- Email: [email protected]
Nanoscale pore structure and mechanical property analysis of coal: An insight combining AFM and SEM images
Fuel 260 (2020) 116352
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Short communication
Nanoscale pore structure and mechanical property analysis of coal: An insight combining AFM and SEM images ⁎
Yong Lia, , Jianghao Yanga, Zhejun Panb, Wangshu Tongc,
T
⁎
a
State Key Laboratory of Coal Resources and Safe Mining and College of Geosciences and Surveying Engineering, China University of Mining and Technology, Beijing 100083, China b CSIRO Energy, Private Bag 10, Clayton South, VIC 3169, Australia c China University of Geosciences, Beijing 100083, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Pore structure AFM SEM Surface roughness Mechanical property
Scanning Electron Microscopy (SEM) and Atomic Force Microscope (AFM), two easily acquired and widely applied image acquisition and analysis methods, have rarely been combined to study the pore structure for unconventional natural gas reservoir rocks. In this work, we present an investigation of nanoscale detection of the pore distribution and mechanical properties of coals using SEM and AFM observations, and conduct quantitative analyses on pore structure distribution, surface roughness and mechanical properties. The morphological characteristics of the coal surface can be revealed by both SEM and AFM methods, and the mechanical parameters of the selected position were obtained under the peakforce quantitative nano-mechanics (PF-QNM) AFM mode, including the Young's modulus, peak force error, deformation, and adhesion forces. By fusing 800 high resolution SEM images into one single image (named as MAPS), the pores morphology and distribution of different scales were acquired. And the studied coal shows different types of cellular pores and gas pores with multiresolution. The mechanical property difference between the matrix and minerals of coal are clearly observed, with the Young’s modulus of organic component around 2 GPa, and that of the minerals generally higher than 10 GPa. The maximum adhesion force values range between 20 and 50 nN. The high values occurred where pores are developed. This work demonstrated that the combination of two dimensional (2D) SEM and three dimensional (3D) AFM results is effective in detection of surface properties, and is of significance in revealing the pore structure and mechanical properties at nanoscale.
1. Introduction
simulation model for gas flow can be reconstructed also by 2D images using stochastic methods, such as simulated annealing method and multiple-point statistics method [13–14]. The SEM observation show that the adjacent micron-sized organic matter and maceral compositions sometimes differ greatly in porosity, which is caused by the innate differences in organic matter deposited as well as by different thermal evolution processes [15,16]. However, SEM cannot fully reflect the fluctuation of the sample surfaces, as the signals received generally shown in gray images, with no information on the surface roughness variation, or mechanical property [3]. AFM is effective in nanometer spatial resolution and force sensing sensitivity detection, which can not only measure the surface topography, but also reflect the mechanical and electrical properties of the tested samples, such as Young's modulus and surface potential [17–20]. As for coal and shale, two important organic rocks containing significant unconventional natural gas resources, their gas adsorption and flow capacity are
Detection of pore structure of unconventional natural gas reservoir rocks is continuously attracting extensive attention, as their nanoscale structure are of strong heterogeneity with no single equipment can fully illustrate the complex composition and structure variations [1–4]. Porescale modelling, e.g. “digital rock” technology, and modern imaging methods are emerging and this requires acquisition of pore structure in high resolution. Experimental methods are widely used including three dimensional (3D) imaging methods of micro- and nano- X-ray computed tomography (CT), Focused ion beam scanning electron microscopy (FIB-SEM) and Helium ion microscope (HIM) [5–8]. Moreover, the two-dimensional (2D) imaging methods, e.g. SEM and Atomic Force Microscope (AFM) are also widely adopted, because of their capability in rapid identification of maceral and mineral composition, pore size distribution and the surface properties [4,9–12]. The numerical
⁎
Corresponding authors. E-mail addresses: [email protected] (Y. Li), [email protected] (W. Tong).
https://doi.org/10.1016/j.fuel.2019.116352 Received 1 May 2019; Received in revised form 28 August 2019; Accepted 4 October 2019 0016-2361/ © 2019 Elsevier Ltd. All rights reserved.
Fuel 260 (2020) 116352
Y. Li, et al.
Fig. 1. Flow chart of the experiment. (a), argon ion polishing; (b), principle of the AFM test; (c), brief introduction to MAPS; (d), image processing procedures.
Fig. 2. Pores occurrence by SEM observations, (a), deformed cellular pores; (b), clay minerals; (c), pores filled by minerals and detrital; (d), MAPS image; (e), fractures in organic matter. (f), macrosporinite; (g), intensively gas pores in collinite.
this paper is 3-fold: (1) combining SEM 2D result and AFM 3D characterizations to study the nano pores morphology and surface characteristics comprehensively; (2) obtaining a large area of ultra-highresolution SEM image by tilting thousands of SEM pictures (MAPS); (3) determining the mechanical property and pores distribution of the same location in a nano scale. The methods can be expanded to study the surface mechanical and pore structure variations under different maturity process and can be widely used in the shale reservoirs evaluation.
closely related to its organic component as its abundance of micro- and nano- pores developed [1,21–24]. Mechanical property is tightly associated coalbed methane and shale gas production. The in-situ stress, bulk modulus, Poisson’ ratio, axial and lateral strain, etc., are all related to the mechanical characteristics, affecting both fracturing efficiency and flow of gas in the reservoir [25]. Organic matter is a major component of coal and shale, hence fragile in nature. Thus, it is difficult to acquire samples of desire size to be experimentally tested for specific parameters [26]. Therefore, investigating the pores distribution and mechanical parameters is important for building robust geological models to access unconventional natural gas resources and enhance gas production efficiency [8]. As a complex issue with multiscale and multi-physics influencing factors, quantitatively and accurately revealing coal surface variation, including pore structure distribution, surface roughness, and its fractal characteristics, et al., has not yet been well understood [16,27]. In this study, we demonstrate the combination of SEM and AFM observations on nanoscale view firstly, providing direct and in situ mechanical characterization of coal matrix and minerals. Then discussion on the pores distribution and surface properties are performed. This approach allows evaluation of compositional and mechanical variations among different macerals (organic matter and minerals) and unravel the surface adsorption and flow capacity variations. The main innovation of
2. Methodology 2.1. Sample preparation and experimental procedure The studied coal samples are collected from the Baode Coal Mines, northeastern Ordos basin, China, where the CBM have been successfully developed [28–30]. The electron beam bombardment on the surface of the sample in SEM may affect the sample to a certain extent. Thus to obtain the image of AFM and SEM in the same horizon, the AFM observation should be conducted first before the SEM experiment. In order to reduce the influence of human influencing factors during the experimental results, the following procedures were considered: (1) reduce the waiting time of the intermediate process; (2) scan more areas for sufficient data; (3) the optical microscope with the instrument was 2
Fuel 260 (2020) 116352
Y. Li, et al.
Fig. 3. SEM and AFM image comparison. (a), SEM image; (b), AFM topography; (c), Young's modulus image; (d), peak force error image; (e), deformation image; (f), adhesion image.
used to observe the samples firstly and select appropriate areas. The AFM instrument used in this experiment was the Dimension Icon AFM from the Bruker Company, Germany. The maximum scanning range was 90 μm × 90 μm × 10 μm, and the resolution is 0.15 nm in the lateral direction and 0.04 nm in the vertical direction. Before AFM scanning, appropriate probes need to be selected and calibrated. The probe AFM Tap525 was adopted, possessing a nominal spring constant k = 200 N/m and a nominal tip radius R = 8 nm, suitable for a studied modulus ranging from 1 GPa up to 10 s GPa. Calibrations for spring constant and deflection sensitivity were conducted using a sapphire
standard provided in the PF-QNM kit. AFM has high requirements for sample roughness. If the surface of the sample fluctuates to a certain degree, the system will not work properly and may even damage the tip of the probe. Therefore, the samples need to be processed before the surface morphology of the samples can be detected. The samples were first cut into cubes of 5 × 5 × 5 mm by precision prototyping machine, and then polished by argon ion polishing to ensure the surface smoothness meeting the experimental requirements. The Argon ion polishing was conducted on a Leica EM TIC 3X, a 3
Fuel 260 (2020) 116352
Y. Li, et al.
Fig. 4. Pore analysis from AFM result, (a) SEM image; (b) 3-D topographic image; (c) (d), Section analysis of AFM (the sections are indicated in red and blue in Fig. 4(b)). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
acceleration voltage was 6 kv and the gun current was 2.2 mA, with the polishing time lasting 4 h. Samples satisfying the flatness requirement can be tested by AFM to obtain the scanning image.
Table 1 Analysis of pore diameter and bearing analysis of coal in Fig. 6. Sections
Long axis (μm)
Short axis (μm)
Average pore size (μm)
Pore depth (μm)
Pore area (μm2)
① ② ③ ④ ⑤ ⑥ ⑦
0.770 1.003 0.902 0.795 1.059 0.799 0.432
0.728 0.709 0.340 0.765 1.018 0.542 0.313
0.745 0.861 0.674 0.785 1.032 0.682 0.386
0.291 0.260 0.182 0.166 0.153 0.178 0.171
0.240 0.267 0.181 0.099 0.130 0.105 0.038
2.2. Acquisition of high-resolution images for large areas As the imaging range of AFM is small (15 μm × 15 μm), it is difficult to obtain the same horizon as the SEM image at one time. Thus a large enough area is needed to ensure the accurate correspondence between the AFM image and the SEM images. In this study, acquisition of high resolution images from large areas under the MAPS mode was conducted on a Helios NanoLab 660. The scanning range of the MAPS image was 900 μm × 900 μm. The basic testing principles of the experiment is to scan thousands (800 in this study) of ultra-high resolution images of the same size in the selected area, and to mosaic small
three-ion beam cutter manufactured by Leica. The argon ion produced under accelerated voltage bombards the sample surface at high speed and denudes the sample layer by layer to achieve polishing effect. The
Table 2 Surface roughness, Young’s modulus, adhesion forces of different observed positions. Position
Ra (nm)
Rq (nm)
Rsk
Rku
Maximum modulus (GPa)
Minimum modulus (GPa)
Average modulus (GPa)
Maximum adhesion force (nN)
Minimum adhesion force (nN)
Average adhesion force (nN)
TL-1 TL-2 TL-3 TA-1 TA-2
50.0 50.9 42.9 26.4 7.9
61.1 64.1 54.9 33.0 9.7
−0.19 −0.36 0.49 −0.29 −0.17
−0.49 0.45 1.07 0.14 −0.22
33.75 40.48 30.19 20.77 39.16
3.08 5.95 2.32 0.83 0.84
10.61 11.43 10.66 12.74 11.19
41.76 57.15 73.27 60.81 137.40
8.79 0.73 0.53 17.68 19.10
22.41 31.36 30.82 36.97 50.80
Note: Ra , average roughness; Rq, root mean square roughness; Rsk, Surface skewness; Rku, Surface kurtosis. 4
Fuel 260 (2020) 116352
Y. Li, et al.
Fig. 5. Young's modulus information obtained by AFM. (a), minerals occurrence (blue part) of samples TA-1; (b), Young’s modulus of samples TA-1; (c). histogram of Young's modulus distribution of samples TA-1; (d), minerals occurrence (blue part) of samples TL-1; (e), Young’s modulus of samples TL-1; (f). histogram of Young's modulus distribution of samples TL-1; (g), minerals occurrence (blue part) of samples TA-2; (h), Young’s modulus of samples TA-2; (i), histogram of Young's modulus distribution of samples TA-2. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
structure and mechanical properties of the material surface [33–35]. The imaging principle is to apply a sinusoidal wave, which driving frequency is much lower than the resonance frequency, of the tip in Z direction of the scanner. The force curve is recorded at every pixel during the scanning and the peak force curve is used as a feedback signal of the imaging. Thus the detailed information of the interaction between the tip and the sample at every pixel can be obtained (Fig. 1-c). The spring constant, tip radius and deflection sensitivity of the tip are corrected by standard method, and the mechanical information distribution of Young's modulus, surface adhesion force and energy dissipation can be obtained at the same time as with the surface morphology variation. The DMT (Derjaguin-Muller-Toporov) model was adopted to calculate Young's modulus [36],
images into a ultra-high resolution, ultra-large area two-dimensional backscatter electronic image (Fig. 1-c). Thus, a more comprehensive and intuitive image can be observed, which avoids the randomness of location selection of traditional observation methods to a certain extent. In addition, the multi-scale pore structure can be presented simultaneously in one image, providing a way for quantitative and qualitative analysis of um scale intergranular pore, micro-fracture and also nano-scale pore. 2.3. Acquisition of mechanical parameters by AFM AFM is a scanning force microscopy characterized by the interaction between probe and sample [31]. The basic working principle is that one end of the elastic cantilever beam is fixed while the other end has a tiny needle tip. When the needle tip scans the sample, the interaction force between the needle tip and the sample (either attractive or repulsive) will cause the cantilever beam to deform (Fig. 1-b). A laser beam emitted from the laser source irradiates the back of the cantilever beam, and the cantilever beam reflects the laser beam to a photoelectric detector [32]. Then the information of the surface morphology or other surface properties of the sample can be obtained. The PeakForce Quantitative Nano-Mechanics (PF-QNM) model was used in this study. The model was developed from tapping mode, and the peak force was used as feedback signal to scan the sample quickly. Meanwhile, the morphology and mechanical properties of the sample can be obtained, which provide an important way to accurately measure the molecular
Ftip =
4 ∗ E Rd3 + Fadh 3
(1)
where Ftip is the force on the tip, Fadh is the adhesion force, R is the tip radius, d is the sample deformation and E* is the reduced modulus, being defined as, −1
2 1 − vtip ⎡ 1 ⎤ Es = (1 − vs2) ⎢ ∗ − Etip ⎥ ⎣E ⎦
(2)
where vs and Es are the Poisson’s ratio and the modulus of the sample; vtip and Etip are the Poisson’s ratio and the modulus of the tip. 5
Fuel 260 (2020) 116352
Y. Li, et al.
Fig. 6. SEM and AFM topography and adhesion images at the same location. (a)(d), SEM images; (b)(e), AFM topography images; (c)(f), AFM adhesion force images.
2.4. Image processing
zmean =
Scanned images were analyzed using Nanoscope Analysis software, version 1.9. The topography images shown in this work were subject to a line-wise 1st order flattening and low-pass filtering, thus Young’s modulus, and deformation images were rendered from the raw data as provided by the NanoScope software. The adhesion force is reflected by the difference between the lowest withdraw part of the force curve and the baseline (Fig. 1-b). The baseline of some points is uneven due to the influence of roughness and other factors, with few of the measured adhesion forces being negative. Thus, “baseline correction” in Nanoscope Analysis Software was used to achieve the correct adhesion force values. There are many methods to measure the surface properties by AFM, such as section analysis [37], grain analysis and bearing analysis. The section analysis reveals the vertical distance and surface roughness along the section fluctuation, and analyses the section condition of the scanning area, thus measuring the diameter and depth of the pore. At the same time, particle analysis is a fast and efficient method in determining pore size distribution. Bearing analysis characterizes the surface nature and pore/grain distribution of different heights of the scanned area [38].
Rq =
Ny
(4)
i= 1 j= 1
1 Nx Ny
Nx
Ny
∑i=1 ∑j=1 (z (i,j)-zmean )2
(5)
where the Nx and Ny are the number of scanning points on the x-axis and y-axis, respectively; Z (i, j) is the height of the (i, j) measuring point; and Z mean is the mean height of all measuring points. Surface skewness (Rsk) gives the information on the symmetry of the distribution of the surface data [43]. If the value equals zero, the distribution is symmetrical. If Rsk > 0, the asymmetry is positive, which reflects that there are many peaks on the surface. If Rsk < 0, the asymmetry is negative, which indicates that there are many deep depressions on the surface. The equation is as follows:
R sk =
1 Nx Ny
N
N
∑i =x 1 ∑ j =y 1 (z(i,j) - zmean )3 R 3q
(6)
Surface kurtosis (Rku), characterizes the sample surface excess and is always proposed in relation to the skewness [44]. Kurtosis for normal distribution is equal to zero. Positive kurtosis occurs when the surface data are more concentrated at the mean value than in normal distribution and negative, when they are less concentrated. The equation is as follows:
By using Amplitude parameters, 3D surface roughness parameters are used to describe the surface morphology characteristics [39]. Amplitude parameters is one of the most important features of surface topography, which characterizes three aspects of surface height: statistical properties, extreme characterizes, and height distribution shapes [40,41]. The average roughness (Ra) shows the average distance of the datum from surface [42]: Nx
Nx
∑ ∑ z (i, j)
Another roughness parameter is the root mean square roughness (Rq), characterizing the variability of the surface topography [41], defined as
2.5. Surface roughness
1 Ra = Nx Ny
1 Nx Ny
Rku =
1 Nx Ny
N
N
∑i =x1 ∑ j =y 1 (z (i, j ) − z mean ) 4 Rq4
−3 (7)
3. Results and discussion
Ny
∑ ∑ |z (i, j) − zmean| i= 1 j= 1
3.1. Pore size distribution
(3)
The studied coal samples are high volatile bituminous coal with
where 6
Fuel 260 (2020) 116352
Y. Li, et al.
The Young's modulus of coal is of great significance in controlling the stabilization of coal and also the formation and extension of hydraulic fractures. The traditional testing methods of macro-mechanical parameters such as uniaxial or triaxial rock mechanics experiments requires specialized core diameter and integrity [42], and can only be conducted in where large coal cores drilling is possible. Furthur, the tested cores cannot be carried out to other mechanical tests after the loading failure, and the repeatability of the experiment is poor. In the AFM test, the sample size is small and can be easily acquired even though in where highly deformed coals developed. And the test can be widely applied for detailed information of different layers or different maceral composition samples. In order to quantify the mechanical properties of different regions in the mechanical images, the Young's modulus of minerals and coal matrix can be distinguished by the threshold method [9]. The blue part shown in the Fig. 5 is the identified mineral area, with the Young's modulus of each sample measured is shown in Table 2. Compared with the traditional experimental method, the Young's modulus of coal matrix measured by AFM is relatively high showing a 10.52 GPa compared with 2 GPa [48], which may be caused by the absence of macro-cracks and the influence of minerals. In the future, if a three dimensional coal core is established with the organic matter, mineral and pores distribution, the correlation between AFM and traditional method can be acquired. The Young's modulus distribution can be detailedly illustrated in the measured picutures and also in the histogram distribution (Fig. 5). The Young’s modulus variation is not unified in different observed areas, with the Fig. 5-c is a normal distribution and the Fig. 5-f and i show diffrences due to the mineral and matrix content difference. The Fig. 5-f is a positively skewed distribution, with large area ratio of the matrix, thus the curve shape is shifted toward the low modulus direction. Fig. 5i shows a negative skewness distribution with the highest peak on the right side of the whole, revaling higher modulus content is relatively high. It can be seen from Fig. 5-g that the content of clay minerals is less, and the area fraction of clay minerals is 0.99%, so the distribution of Fig. 5-i is more concentrated than that of Fig. 5-c.
vitrinite reflectance of 0.80%. The results of coal show that organic matter is the dominant composition of the samples, with widely distributed pores, fractures and minerals. Different kind of pores are developed as well as laminated minerals being observed (Fig. 2-d). The compacted cellular pores are well developed, preserving the original cell morphology, part of the pores is saturated with clay minerals (Fig. 2-b and c). Gas pores are also observed, indicating the coal has been into the gas generation stage (Fig. 2-g). The pores are well developed in the coal matrix, and different of fractures can also be observed, cutting the matrix (Fig. 2-e) or being preserved during the maceral compaction, e.g. macrosporinite compaction (Fig. 2-f). Minerals occurred as isolated grains or saturating pores (Fig. 2-g), and the intercrystallite or intracrystalline pores are also observed (Fig. 2-b). The same view position of the SEM and AFM pictures were also acquired, with the matrix, mineral and pores distribution can be clearly recognized (Fig. 3-a and b). The surface morphological characteristics, in a 3D view, is also displayed by the topographic map (Fig. 4-b). The morphology and distribution of various types of pores in coal can also be directly revealed by the AFM, which provides an important reference for the geometric study of coal pore structure [9]. In order to describe the pore and sample surface more accurately, surface variation profile of the studied samples was conducted (Fig. 4-c and d), with the pore size and pore depth are obtained. The cross-section shows that the coal surface is of strong irregularity. The maximum pore diameter in the cross-section is 2.706 μm, with a pore depth of 168.061 nm, and the minimum pore diameter is 412 nm with a depth of 43.054 nm. And the slit pores shown in the Fig. 3-d is as much as 294.954 nm deep. Several sections were conducted to calculate the average pore size and diameter, with their long axis, short axis, pore depth and pore area can be acquired (Table 1). 3.2. Surface roughness Surface roughness is of significance on the macroscopic contact angle measurement of coal which plays a vital role in measurement of contact angle, and various processes like moistening, spreading, and wetting [45,46]. The 3D surface roughness parameter can provide more comprehensive spatial surface features of the sample surface, which has better statistical significance and is an important indicator for characterizing the surface topography (Fig. 4-b). The measurement results of the sample roughness in the experiment are shown in Table 2. The Rq values of the tested samples are higher than the Ra values. The Rq values are more sensitive to the roughness variation, and small fluctuations in surface height will lead to great variations [47]. The Ra values range from 7.9 to 118.2 nm, with an average value of 49.3 nm, and the Rq values increase from 9.7 to 238.0 nm, averaging of 76.8 nm. The Rsk values range from −0.36 to 0.49 (averaging of −0.09), which indicates slightly unsymmetrical negative distribution of the surface points. And the Rku values of the samples vary from −0.43 to 1.07 (averaging of 0.19), indicating wide variations of the surface roughness.
3.3.2. Adhesion forces The second mechanical property acquired in the mapping is the adhesion force, illustrated as the minimum force in Fig. 6 and Fig. 1. The source of the adhesion force can be any attractive force between the tip and sample. In air, Van der Waals, electrostatics, and forces due to the formation of a capillary meniscus can all contribute to the relative strengths shown as parameters like Hammaker constants, surface charges, and hydrophilicity [49–52]. Adhesion force can reflect the surface energy of the sample [53]. The surface morphology and roughness have a great influence on the adhesion force, which is high in the pore area and low in the matrix (Fig. 6). When the porous surface of coal is formed, at least one side of the carbon atoms on the surface of coal is empty, resulting in unbalanced forces. This unbalance gives the carbon atoms on the surface of coal an additional energy, which can be expressed as surface energy. Surface energy can directly reflect the adsorption amount per unit area of coal surface. Thus the pores areas with complicated surface variations would be of higher adsorption capacity to gases. The Fig. 6 is a good comparison of the results. The plane porosity of the Fig. 6-a is 17.50%, roughness Ra of 118.2 nm and Rq of 238.0 nm, while the Fig. 6d showing a plane porosity of 4.30%, Ra 42.9 nm and Rq of 54.9 nm. The adhesion force of the two samples is as much as 0 nN to 473.26 nN, averaging of 79.47 nN; and the sample with low porosity and roughness is of 0 nN to 16.40nN, averaging of 4.47 nN. The adhesion force of each sample measured by AFM is shown in Table 2. The distribution of adhesion force is different from that of Young's modulus as the adhesion force is influenced by many factors, including the surface energy, micro-morphology and chemical composition of interface. The effect of roughness on the adhesion force is that
3.3. Nano mechanical characteristics 3.3.1. Young’s modulus The Young’s modulus (Fig. 3-c) and deformation (Fig. 3-d) distribution was acquired by the AFM. The Young's modulus variation of minerals is quite different from that of the coal matrix, with the minerals values is as much as 21 GPa while the matrix is only around 2 GPa. Part of the areas where pores developed show high Young's modulus, indicating that this part of the pores is filled with minerals. The deformation range after the coal surface was recorded (Fig. 3-d). Comparing the deformation image with the SEM pictures, it can be clearly seen that the location of the higher deformation ranges occurred in where pores developed, and the minerals show quite low deformation range. Both the deformation image and Young's modulus image reflect the same mechanical properties of samples. 7
Fuel 260 (2020) 116352
Y. Li, et al.
the actual contact area of the probe at a higher point is much smaller than the apparent area. If the contact area between the probe and the sample decreases, the adhesion decreases. This also explains that the adhesion force of the sample is high where more pores are developed. For coal, the location of high adhesion force is more conducive to gas adsorption, and the adhesive force also has a great influence on coal flotation as it influences the particles adhesion [54].
pore characteristics in coal and coal-bearing shale. Mar. Pet. Geol. 2019;105:331–7. [11] Liu X, Song D, He X, Wang Z, Zeng M, Deng K. Nanopore structure of deep-burial coals explored by AFM. Fuel 2019;246:9–17. [12] Liu X, Song D, He X, Wang Z, Zeng M, Wang L. Quantitative analysis of coal nanopore characteristics using atomic force microscopy. Powder Technol. 2019;346:332–40. [13] Keehm Y, Mukerji T, Nur A. Permeability prediction from thin sections: 3D reconstruction and Lattice-Boltzmann flow simulation. Geophys. Res. Lett. 2004;31(4):1668. [14] Wu K, Dijke M, Couples G, Jiang Z, Ma J, Sorbie K, et al. 3D stochastic modelling of heterogeneous porous media–applications to reservoir rocks. Transp. Porous Media 2006;65(3):443–67. [15] Curtis M, Sondergeld C, Ambrose R, Rai C. Microstructural investigation of gas shales in two and three dimensions using nanometer-scale resolution imaging. AAPG Bull. 2012;96(4):665–77. [16] Li Y, Zhang C, Tang D, Gan Q, Niu X, Shen R. Coal pore size distributions controlled by the coalification process: an experimental study of coals from the Junggar, Ordos, and Qinshui basins in China. Fuel 2017;206:352–63. [17] Abedi S, Slim M, Ulm F. Nanomechanics of organic-rich shales: the role of thermal maturity and organic matter content on texture. Acta Geotech. 2016;11:775–87. [18] Foster B. New atomic force microscopy (AFM) approaches life sciences gently, quantitatively, and correlatively. Am. Lab. 2012;4:24–8. [19] Trtik P, Kaufmann J, Volz U. On the use of peak-force tapping atomic force microsco-py for quantification of the local elastic modulus in hardened cement paste. Cem. Concr. Res. 2012;42(1):215–21. [20] He X, Liu X, Song D, Nie B. Effect of microstructure on electrical property of coal surface. Appl. Surf. Sci. 2019;483:713–20. [21] Lee T, Bocquet L, Coasne B. Activated desorption at heterogeneous interfaces and long-time kinetics of hydrocarbon recovery from nanoporous media. Nat. Commun. 2016;7:11890. [22] Ju Y, Huang C, Sun Y, Wan Q, Lu X, Lu S, et al. Nanogeosciences: research history, current status, and development trends. J. Nanosci. Nanotechnol. 2017;17(9):5930–65. [23] Zhu H, Ju Y, Qi Y, Huang C, Zhang L. Impact of tectonism on pore type and pore structure evolution in organic-rich shale: Implications for gas storage and migration pathways in naturally deformed rocks. Fuel 2018;228:272–89. [24] Cai J, Lin D, Singh H, Wei W, Zhou S. Shale gas transport model in 3D fractal porous media with variable pore sizes. Mar. Pet. Geol. 2018;98:437–47. [25] Tan L, Ren T, Yang X, He X. A numerical simulation study on mechanical behaviour of coal with bedding planes under coupled static and dynamic load. Int. J. Min. Sci. Technol. 2018;28:791–7. [26] Kumar H, Mishra S, Mishra K. Experimental Evaluation of Geo-mechanical properties of Coal using Sonic Wave Velocity. International Conference on Advances in Agricultural, Biological & Environmental Sciences July 22-23, 2015 London (UK). [27] Li Y, Tang D, Elsworth D, Xu H. Characterization of coalbed methane reservoirs at multiple length scales: a cross-section from southeastern Ordos Basin, China. Energ. Fuel 2014;28:5587–95. [28] Li Y, Wang Z, Gan Q, Niu X, Xu W. Paleoenvironmental conditions and organic matter accumulation in Upper Paleozoic organic-rich rocks in the east margin of the Ordos Basin, China. Fuel 2019;252:172–87. [29] Li Y, Tang D, Fang Y, Xu H, Meng Y. Distribution and genesis of stable carbon isotope in coalbed methane from the east margin of Ordos Basin. Sci China Earth Sci 2014;57(8):1741–8. [30] Li Y, Yang J, Pan Z, Meng S, Wang K, Niu X. Unconventional natural gas accumulations in stacked deposits: a discussion of Upper Paleozoic coal-bearing strata in the east margin of the Ordos Basin. China Acta Geol Sin (English Edition) 2019;93(1):111–29. [31] Binning G, Quate C, Gerber C. Atomic force microscopy. Phys Rev Lett 1986;56(9):930–3. [32] Patel A, Kranz C. (Multi) functional atomic force microscopy imaging. Annu Rev Anal Chem 2018;11:329–50. [33] Cappella B, Dietler G. Force-distance curves by atomic force microscopy. Surf Sci Rep 1999;34(1–3):5–104. [34] Wang W, Guo Z, Sun J, Li Z. Nano-mechanical characterization of disassembling amyloid fibrils using the Peak Force QNM method. Biopolymers 2017;107(2):61–9. [35] Young T, Monclus M, Burnett T, Broughton W, Ogin S, Smith P. The use of the PeakForce TM quantitative nanomechanical mapping AFM-based method for highresolution Young’s modulus measurement of polymers. Meas Sci Technol 2011;22(12). 125703. [36] Derjaguin B, Muller V, Toporov Y. Effect of contact deformations on the adhesion of particles. J Colloid Interface Sci 1975;53(2):314–26. [37] Bruening F, Cohen A. Measuring surface properties and oxidation of coal macerals using the atomic force microscope. Coal Geol 2005;63:195–204. [38] Yao S, Jiao K, Zhang K, Hu W, Ding H, Li M, et al. An atomic force microscopy study of coal nanopore structure. Chinese Sci Bull 2011;56:2706–12. [39] Pancewicz T, Mruk I. Holographic contouring for determination of three dimensional description of surface roughness. Wear 1996;199(1):127–31. [40] Li C, Dong S. The parameters and methods of characterizing 3D surface microtopography. J Astronaut Metrol Measure 1999;19(6):33–43. [41] Gadelmawla E, Koura M, Maksoud T, Elewa I, Soliman H. Roughness parameters. J Mater Process Technol 2002;123(1):133–45. [42] Abouelatta O, Madl J. Surface roughness prediction based on cutting parameters and tool vibrations in turning operations. J Mater Process Technol 2001;118(1–3):269–77. [43] Carvalho A, Maugeri F, Silva V, Hernandez A, Palacio L, Pradanos P. AFM analysis of the surface of nanoporous membranes: application to the nanofiltration of
4. Conclusions Observations combining AFM and SEM tests were conducted to reveal the pore structure and mechanical properties of coal. The following conclusions can be drawn: (1) Combing AFM and SEM observations in a nanoscale is effective to reveal pore structure and mechanical properties in 2D and 3D dimensions. The SEM provides information of pore size distribution, and matrix and mineral compositions, and the AFM presents pore structure, surface roughness and mechanical properties. (2) MAPS provide information of coal surfaces in a larger observation size with high resolution degrees, which is of significance in revealing pores distribution and maceral composition at different scales. The combination of SEM equipped with energy spectrum and AFM tests in different areas would be useful in revealing geochemical and geomechanical information of coal and shale reservoirs. (3) The PF-QNM mode of AFM quantitatively illustrates the surface roughness, Young’s modulus and adhesion force variations. The coal matrix and minerals show huge difference in mechanical properties, and the influence of mineral distribution in hydraulic fracturing should be carefully considered. (4) The quantitative information of Young's modulus and adhesion force of coal samples can be obtained by AFM. The adhesion force which reflects the surface energy information is affected by the surface morphology. The measurement of adhesion force is conducive to the study of coal adsorption capacity and coal flotation. Acknowledgments This study was supported by the National Natural Science Foundation of China (Grant No. 41702171), National Major Science and Technology Projects of China (No.2016ZX05066001-002). We are grateful to the reviewers for their valuable advice during their busy schedule. References [1] Clarkson C, Solano N, Bustin R, Bustin A, Chalmers G, He L, et al. Pore structure characterization of North American shale gas reservoirs using USANS/SANS, gas adsorption, and mercury intrusion. Fuel 2013;103:606–16. [2] Wu T, Zhang D. Impact of adsorption on gas transport in nanopores. Sci Rep 2016;6:23629. [3] Yang J, Hatcherian J, Hackley P, Pomerantz A. Nanoscale geochemical and geomechanical characterization of organic matter in shale. Nat Commun 2017;8(1):2179. [4] Li Y, Wang Z, Pan Z, Niu X, Yu Y, Meng S. Pore structure and its fractal dimensions of transitional shale: A cross section from east margin of the Ordos Basin, China. Fuel 2019;241:417–31. [5] Chen C, Hu D, Westacott D, Loveless D. Nanometer-scale characterization of microscopic pores in shale kerogen by image analysis and pore-scale modeling. Geochem. Geophys. Geosyst. 2013;14(10):4066–75. [6] Blunt M, Bijeljic B, Dong H, Gharbi O, Iglauer S, Mostaghimi P, et al. Pore-scale imaging and modelling. Adv. Water Resour. 2013;51:197–216. [7] Kelly S, El-Sobky H, Torres-Verdín C, Balhoff M. Assessing the utility of FIB-SEM images for shale digital rock physics. Adv. Water Resour. 2016;95:302–16. [8] Wu T, Li X, Zhao J, Zhang D. Multiscale pore structure and its effect on gas transport in organic-rich shale. Water Resour. Res. 2017;53:5438–50. [9] Zhao S, Li Y, Wang Y, Ma Z, Huang X. Quantitative study on coal and shale pore structure and surface roughness based on atomic force microscopy and image processing. Fuel 2019;244:78–90. [10] Liu X, Nie B, Wang W, Wang Z, Zhang L. The use of AFM in quantitative analysis of
8
Fuel 260 (2020) 116352
Y. Li, et al.
[49] Tiwari A, Gladwell G. Nanomechanical analysis of high performance materials. Springer; 2013. p. 43–4. [50] Weisenhorn A, Hansma P, Albrecht T, Quate C. Forces in atomic force microscopy in air and water. Appl Phys Lett 1989;54:2651–3. [51] Dokukin M, Sokolov I. Quantitative mapping of the elastic modulus of soft materials with Harmoni X and peak force QNM AFM modes. Langmuir 2012;28(46):16060–7. [52] Rezende C, Lee L, Galembeck F. Surface mechanical properties of thin polymer films investigated by AFM in pulsed force mode. Langmuir 2009;25(17):9938–46. [53] Bastian S, Mikhail S, Wael A, Maryam H, Maxim F, Jan B, et al. Quantifying mineral surface energy by scanning force microscopy. J Colloid Interface Sci 2016;472:237–46. [54] Tri M, Tsuyoshi H, Nur L, Lina N, Sudarsono S. Surface physicochemical properties of semi-anthracitic coal from Painan-Sumatra during air oxidation. Int J Coal Sci Technol 2018;5(2):156–66.
potassium clavulanate. J. Mater Sci 2011;46(10):3356–69. [44] Li W, Wang C, Shi Z, Wei Y, Zhou H, Deng K. The description of shale reservoir pore structure based on method of moments estimation. PLoS ONE 2016;11(3):0151631. [45] Kang J, Zhang B, Wang J, Wu J, Wang L, Yin B, et al. Mechanical testing of anthracite to assess its surface energy and temperature dependence. Fuel 2019;239:76–86. [46] Han B, Liang S, Wang B, Zheng J, Xie X, Xiao K, et al. Simultaneous determination of surface energy and roughness of dense membranes by a modified contact angle method. Colloids Surf A 2019;562:370–6. [47] Liu J, Jiang X, Huang X, Wu S. Morphological characterization of super fine pulverized coal particle. Part 2. AFM investigation of single coal particle. Fuel 2010;89(12):3884–91. [48] Domke J, Radmacher M. Measuring the elastic properties of thin polymer films with the atomic force microscope. Langmuir 1998;14:3320–5.
9
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Short communication
Nanoscale pore structure and mechanical property analysis of coal: An insight combining AFM and SEM images ⁎
Yong Lia, , Jianghao Yanga, Zhejun Panb, Wangshu Tongc,
T
⁎
a
State Key Laboratory of Coal Resources and Safe Mining and College of Geosciences and Surveying Engineering, China University of Mining and Technology, Beijing 100083, China b CSIRO Energy, Private Bag 10, Clayton South, VIC 3169, Australia c China University of Geosciences, Beijing 100083, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Pore structure AFM SEM Surface roughness Mechanical property
Scanning Electron Microscopy (SEM) and Atomic Force Microscope (AFM), two easily acquired and widely applied image acquisition and analysis methods, have rarely been combined to study the pore structure for unconventional natural gas reservoir rocks. In this work, we present an investigation of nanoscale detection of the pore distribution and mechanical properties of coals using SEM and AFM observations, and conduct quantitative analyses on pore structure distribution, surface roughness and mechanical properties. The morphological characteristics of the coal surface can be revealed by both SEM and AFM methods, and the mechanical parameters of the selected position were obtained under the peakforce quantitative nano-mechanics (PF-QNM) AFM mode, including the Young's modulus, peak force error, deformation, and adhesion forces. By fusing 800 high resolution SEM images into one single image (named as MAPS), the pores morphology and distribution of different scales were acquired. And the studied coal shows different types of cellular pores and gas pores with multiresolution. The mechanical property difference between the matrix and minerals of coal are clearly observed, with the Young’s modulus of organic component around 2 GPa, and that of the minerals generally higher than 10 GPa. The maximum adhesion force values range between 20 and 50 nN. The high values occurred where pores are developed. This work demonstrated that the combination of two dimensional (2D) SEM and three dimensional (3D) AFM results is effective in detection of surface properties, and is of significance in revealing the pore structure and mechanical properties at nanoscale.
1. Introduction
simulation model for gas flow can be reconstructed also by 2D images using stochastic methods, such as simulated annealing method and multiple-point statistics method [13–14]. The SEM observation show that the adjacent micron-sized organic matter and maceral compositions sometimes differ greatly in porosity, which is caused by the innate differences in organic matter deposited as well as by different thermal evolution processes [15,16]. However, SEM cannot fully reflect the fluctuation of the sample surfaces, as the signals received generally shown in gray images, with no information on the surface roughness variation, or mechanical property [3]. AFM is effective in nanometer spatial resolution and force sensing sensitivity detection, which can not only measure the surface topography, but also reflect the mechanical and electrical properties of the tested samples, such as Young's modulus and surface potential [17–20]. As for coal and shale, two important organic rocks containing significant unconventional natural gas resources, their gas adsorption and flow capacity are
Detection of pore structure of unconventional natural gas reservoir rocks is continuously attracting extensive attention, as their nanoscale structure are of strong heterogeneity with no single equipment can fully illustrate the complex composition and structure variations [1–4]. Porescale modelling, e.g. “digital rock” technology, and modern imaging methods are emerging and this requires acquisition of pore structure in high resolution. Experimental methods are widely used including three dimensional (3D) imaging methods of micro- and nano- X-ray computed tomography (CT), Focused ion beam scanning electron microscopy (FIB-SEM) and Helium ion microscope (HIM) [5–8]. Moreover, the two-dimensional (2D) imaging methods, e.g. SEM and Atomic Force Microscope (AFM) are also widely adopted, because of their capability in rapid identification of maceral and mineral composition, pore size distribution and the surface properties [4,9–12]. The numerical
⁎
Corresponding authors. E-mail addresses: [email protected] (Y. Li), [email protected] (W. Tong).
https://doi.org/10.1016/j.fuel.2019.116352 Received 1 May 2019; Received in revised form 28 August 2019; Accepted 4 October 2019 0016-2361/ © 2019 Elsevier Ltd. All rights reserved.
Fuel 260 (2020) 116352
Y. Li, et al.
Fig. 1. Flow chart of the experiment. (a), argon ion polishing; (b), principle of the AFM test; (c), brief introduction to MAPS; (d), image processing procedures.
Fig. 2. Pores occurrence by SEM observations, (a), deformed cellular pores; (b), clay minerals; (c), pores filled by minerals and detrital; (d), MAPS image; (e), fractures in organic matter. (f), macrosporinite; (g), intensively gas pores in collinite.
this paper is 3-fold: (1) combining SEM 2D result and AFM 3D characterizations to study the nano pores morphology and surface characteristics comprehensively; (2) obtaining a large area of ultra-highresolution SEM image by tilting thousands of SEM pictures (MAPS); (3) determining the mechanical property and pores distribution of the same location in a nano scale. The methods can be expanded to study the surface mechanical and pore structure variations under different maturity process and can be widely used in the shale reservoirs evaluation.
closely related to its organic component as its abundance of micro- and nano- pores developed [1,21–24]. Mechanical property is tightly associated coalbed methane and shale gas production. The in-situ stress, bulk modulus, Poisson’ ratio, axial and lateral strain, etc., are all related to the mechanical characteristics, affecting both fracturing efficiency and flow of gas in the reservoir [25]. Organic matter is a major component of coal and shale, hence fragile in nature. Thus, it is difficult to acquire samples of desire size to be experimentally tested for specific parameters [26]. Therefore, investigating the pores distribution and mechanical parameters is important for building robust geological models to access unconventional natural gas resources and enhance gas production efficiency [8]. As a complex issue with multiscale and multi-physics influencing factors, quantitatively and accurately revealing coal surface variation, including pore structure distribution, surface roughness, and its fractal characteristics, et al., has not yet been well understood [16,27]. In this study, we demonstrate the combination of SEM and AFM observations on nanoscale view firstly, providing direct and in situ mechanical characterization of coal matrix and minerals. Then discussion on the pores distribution and surface properties are performed. This approach allows evaluation of compositional and mechanical variations among different macerals (organic matter and minerals) and unravel the surface adsorption and flow capacity variations. The main innovation of
2. Methodology 2.1. Sample preparation and experimental procedure The studied coal samples are collected from the Baode Coal Mines, northeastern Ordos basin, China, where the CBM have been successfully developed [28–30]. The electron beam bombardment on the surface of the sample in SEM may affect the sample to a certain extent. Thus to obtain the image of AFM and SEM in the same horizon, the AFM observation should be conducted first before the SEM experiment. In order to reduce the influence of human influencing factors during the experimental results, the following procedures were considered: (1) reduce the waiting time of the intermediate process; (2) scan more areas for sufficient data; (3) the optical microscope with the instrument was 2
Fuel 260 (2020) 116352
Y. Li, et al.
Fig. 3. SEM and AFM image comparison. (a), SEM image; (b), AFM topography; (c), Young's modulus image; (d), peak force error image; (e), deformation image; (f), adhesion image.
used to observe the samples firstly and select appropriate areas. The AFM instrument used in this experiment was the Dimension Icon AFM from the Bruker Company, Germany. The maximum scanning range was 90 μm × 90 μm × 10 μm, and the resolution is 0.15 nm in the lateral direction and 0.04 nm in the vertical direction. Before AFM scanning, appropriate probes need to be selected and calibrated. The probe AFM Tap525 was adopted, possessing a nominal spring constant k = 200 N/m and a nominal tip radius R = 8 nm, suitable for a studied modulus ranging from 1 GPa up to 10 s GPa. Calibrations for spring constant and deflection sensitivity were conducted using a sapphire
standard provided in the PF-QNM kit. AFM has high requirements for sample roughness. If the surface of the sample fluctuates to a certain degree, the system will not work properly and may even damage the tip of the probe. Therefore, the samples need to be processed before the surface morphology of the samples can be detected. The samples were first cut into cubes of 5 × 5 × 5 mm by precision prototyping machine, and then polished by argon ion polishing to ensure the surface smoothness meeting the experimental requirements. The Argon ion polishing was conducted on a Leica EM TIC 3X, a 3
Fuel 260 (2020) 116352
Y. Li, et al.
Fig. 4. Pore analysis from AFM result, (a) SEM image; (b) 3-D topographic image; (c) (d), Section analysis of AFM (the sections are indicated in red and blue in Fig. 4(b)). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
acceleration voltage was 6 kv and the gun current was 2.2 mA, with the polishing time lasting 4 h. Samples satisfying the flatness requirement can be tested by AFM to obtain the scanning image.
Table 1 Analysis of pore diameter and bearing analysis of coal in Fig. 6. Sections
Long axis (μm)
Short axis (μm)
Average pore size (μm)
Pore depth (μm)
Pore area (μm2)
① ② ③ ④ ⑤ ⑥ ⑦
0.770 1.003 0.902 0.795 1.059 0.799 0.432
0.728 0.709 0.340 0.765 1.018 0.542 0.313
0.745 0.861 0.674 0.785 1.032 0.682 0.386
0.291 0.260 0.182 0.166 0.153 0.178 0.171
0.240 0.267 0.181 0.099 0.130 0.105 0.038
2.2. Acquisition of high-resolution images for large areas As the imaging range of AFM is small (15 μm × 15 μm), it is difficult to obtain the same horizon as the SEM image at one time. Thus a large enough area is needed to ensure the accurate correspondence between the AFM image and the SEM images. In this study, acquisition of high resolution images from large areas under the MAPS mode was conducted on a Helios NanoLab 660. The scanning range of the MAPS image was 900 μm × 900 μm. The basic testing principles of the experiment is to scan thousands (800 in this study) of ultra-high resolution images of the same size in the selected area, and to mosaic small
three-ion beam cutter manufactured by Leica. The argon ion produced under accelerated voltage bombards the sample surface at high speed and denudes the sample layer by layer to achieve polishing effect. The
Table 2 Surface roughness, Young’s modulus, adhesion forces of different observed positions. Position
Ra (nm)
Rq (nm)
Rsk
Rku
Maximum modulus (GPa)
Minimum modulus (GPa)
Average modulus (GPa)
Maximum adhesion force (nN)
Minimum adhesion force (nN)
Average adhesion force (nN)
TL-1 TL-2 TL-3 TA-1 TA-2
50.0 50.9 42.9 26.4 7.9
61.1 64.1 54.9 33.0 9.7
−0.19 −0.36 0.49 −0.29 −0.17
−0.49 0.45 1.07 0.14 −0.22
33.75 40.48 30.19 20.77 39.16
3.08 5.95 2.32 0.83 0.84
10.61 11.43 10.66 12.74 11.19
41.76 57.15 73.27 60.81 137.40
8.79 0.73 0.53 17.68 19.10
22.41 31.36 30.82 36.97 50.80
Note: Ra , average roughness; Rq, root mean square roughness; Rsk, Surface skewness; Rku, Surface kurtosis. 4
Fuel 260 (2020) 116352
Y. Li, et al.
Fig. 5. Young's modulus information obtained by AFM. (a), minerals occurrence (blue part) of samples TA-1; (b), Young’s modulus of samples TA-1; (c). histogram of Young's modulus distribution of samples TA-1; (d), minerals occurrence (blue part) of samples TL-1; (e), Young’s modulus of samples TL-1; (f). histogram of Young's modulus distribution of samples TL-1; (g), minerals occurrence (blue part) of samples TA-2; (h), Young’s modulus of samples TA-2; (i), histogram of Young's modulus distribution of samples TA-2. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
structure and mechanical properties of the material surface [33–35]. The imaging principle is to apply a sinusoidal wave, which driving frequency is much lower than the resonance frequency, of the tip in Z direction of the scanner. The force curve is recorded at every pixel during the scanning and the peak force curve is used as a feedback signal of the imaging. Thus the detailed information of the interaction between the tip and the sample at every pixel can be obtained (Fig. 1-c). The spring constant, tip radius and deflection sensitivity of the tip are corrected by standard method, and the mechanical information distribution of Young's modulus, surface adhesion force and energy dissipation can be obtained at the same time as with the surface morphology variation. The DMT (Derjaguin-Muller-Toporov) model was adopted to calculate Young's modulus [36],
images into a ultra-high resolution, ultra-large area two-dimensional backscatter electronic image (Fig. 1-c). Thus, a more comprehensive and intuitive image can be observed, which avoids the randomness of location selection of traditional observation methods to a certain extent. In addition, the multi-scale pore structure can be presented simultaneously in one image, providing a way for quantitative and qualitative analysis of um scale intergranular pore, micro-fracture and also nano-scale pore. 2.3. Acquisition of mechanical parameters by AFM AFM is a scanning force microscopy characterized by the interaction between probe and sample [31]. The basic working principle is that one end of the elastic cantilever beam is fixed while the other end has a tiny needle tip. When the needle tip scans the sample, the interaction force between the needle tip and the sample (either attractive or repulsive) will cause the cantilever beam to deform (Fig. 1-b). A laser beam emitted from the laser source irradiates the back of the cantilever beam, and the cantilever beam reflects the laser beam to a photoelectric detector [32]. Then the information of the surface morphology or other surface properties of the sample can be obtained. The PeakForce Quantitative Nano-Mechanics (PF-QNM) model was used in this study. The model was developed from tapping mode, and the peak force was used as feedback signal to scan the sample quickly. Meanwhile, the morphology and mechanical properties of the sample can be obtained, which provide an important way to accurately measure the molecular
Ftip =
4 ∗ E Rd3 + Fadh 3
(1)
where Ftip is the force on the tip, Fadh is the adhesion force, R is the tip radius, d is the sample deformation and E* is the reduced modulus, being defined as, −1
2 1 − vtip ⎡ 1 ⎤ Es = (1 − vs2) ⎢ ∗ − Etip ⎥ ⎣E ⎦
(2)
where vs and Es are the Poisson’s ratio and the modulus of the sample; vtip and Etip are the Poisson’s ratio and the modulus of the tip. 5
Fuel 260 (2020) 116352
Y. Li, et al.
Fig. 6. SEM and AFM topography and adhesion images at the same location. (a)(d), SEM images; (b)(e), AFM topography images; (c)(f), AFM adhesion force images.
2.4. Image processing
zmean =
Scanned images were analyzed using Nanoscope Analysis software, version 1.9. The topography images shown in this work were subject to a line-wise 1st order flattening and low-pass filtering, thus Young’s modulus, and deformation images were rendered from the raw data as provided by the NanoScope software. The adhesion force is reflected by the difference between the lowest withdraw part of the force curve and the baseline (Fig. 1-b). The baseline of some points is uneven due to the influence of roughness and other factors, with few of the measured adhesion forces being negative. Thus, “baseline correction” in Nanoscope Analysis Software was used to achieve the correct adhesion force values. There are many methods to measure the surface properties by AFM, such as section analysis [37], grain analysis and bearing analysis. The section analysis reveals the vertical distance and surface roughness along the section fluctuation, and analyses the section condition of the scanning area, thus measuring the diameter and depth of the pore. At the same time, particle analysis is a fast and efficient method in determining pore size distribution. Bearing analysis characterizes the surface nature and pore/grain distribution of different heights of the scanned area [38].
Rq =
Ny
(4)
i= 1 j= 1
1 Nx Ny
Nx
Ny
∑i=1 ∑j=1 (z (i,j)-zmean )2
(5)
where the Nx and Ny are the number of scanning points on the x-axis and y-axis, respectively; Z (i, j) is the height of the (i, j) measuring point; and Z mean is the mean height of all measuring points. Surface skewness (Rsk) gives the information on the symmetry of the distribution of the surface data [43]. If the value equals zero, the distribution is symmetrical. If Rsk > 0, the asymmetry is positive, which reflects that there are many peaks on the surface. If Rsk < 0, the asymmetry is negative, which indicates that there are many deep depressions on the surface. The equation is as follows:
R sk =
1 Nx Ny
N
N
∑i =x 1 ∑ j =y 1 (z(i,j) - zmean )3 R 3q
(6)
Surface kurtosis (Rku), characterizes the sample surface excess and is always proposed in relation to the skewness [44]. Kurtosis for normal distribution is equal to zero. Positive kurtosis occurs when the surface data are more concentrated at the mean value than in normal distribution and negative, when they are less concentrated. The equation is as follows:
By using Amplitude parameters, 3D surface roughness parameters are used to describe the surface morphology characteristics [39]. Amplitude parameters is one of the most important features of surface topography, which characterizes three aspects of surface height: statistical properties, extreme characterizes, and height distribution shapes [40,41]. The average roughness (Ra) shows the average distance of the datum from surface [42]: Nx
Nx
∑ ∑ z (i, j)
Another roughness parameter is the root mean square roughness (Rq), characterizing the variability of the surface topography [41], defined as
2.5. Surface roughness
1 Ra = Nx Ny
1 Nx Ny
Rku =
1 Nx Ny
N
N
∑i =x1 ∑ j =y 1 (z (i, j ) − z mean ) 4 Rq4
−3 (7)
3. Results and discussion
Ny
∑ ∑ |z (i, j) − zmean| i= 1 j= 1
3.1. Pore size distribution
(3)
The studied coal samples are high volatile bituminous coal with
where 6
Fuel 260 (2020) 116352
Y. Li, et al.
The Young's modulus of coal is of great significance in controlling the stabilization of coal and also the formation and extension of hydraulic fractures. The traditional testing methods of macro-mechanical parameters such as uniaxial or triaxial rock mechanics experiments requires specialized core diameter and integrity [42], and can only be conducted in where large coal cores drilling is possible. Furthur, the tested cores cannot be carried out to other mechanical tests after the loading failure, and the repeatability of the experiment is poor. In the AFM test, the sample size is small and can be easily acquired even though in where highly deformed coals developed. And the test can be widely applied for detailed information of different layers or different maceral composition samples. In order to quantify the mechanical properties of different regions in the mechanical images, the Young's modulus of minerals and coal matrix can be distinguished by the threshold method [9]. The blue part shown in the Fig. 5 is the identified mineral area, with the Young's modulus of each sample measured is shown in Table 2. Compared with the traditional experimental method, the Young's modulus of coal matrix measured by AFM is relatively high showing a 10.52 GPa compared with 2 GPa [48], which may be caused by the absence of macro-cracks and the influence of minerals. In the future, if a three dimensional coal core is established with the organic matter, mineral and pores distribution, the correlation between AFM and traditional method can be acquired. The Young's modulus distribution can be detailedly illustrated in the measured picutures and also in the histogram distribution (Fig. 5). The Young’s modulus variation is not unified in different observed areas, with the Fig. 5-c is a normal distribution and the Fig. 5-f and i show diffrences due to the mineral and matrix content difference. The Fig. 5-f is a positively skewed distribution, with large area ratio of the matrix, thus the curve shape is shifted toward the low modulus direction. Fig. 5i shows a negative skewness distribution with the highest peak on the right side of the whole, revaling higher modulus content is relatively high. It can be seen from Fig. 5-g that the content of clay minerals is less, and the area fraction of clay minerals is 0.99%, so the distribution of Fig. 5-i is more concentrated than that of Fig. 5-c.
vitrinite reflectance of 0.80%. The results of coal show that organic matter is the dominant composition of the samples, with widely distributed pores, fractures and minerals. Different kind of pores are developed as well as laminated minerals being observed (Fig. 2-d). The compacted cellular pores are well developed, preserving the original cell morphology, part of the pores is saturated with clay minerals (Fig. 2-b and c). Gas pores are also observed, indicating the coal has been into the gas generation stage (Fig. 2-g). The pores are well developed in the coal matrix, and different of fractures can also be observed, cutting the matrix (Fig. 2-e) or being preserved during the maceral compaction, e.g. macrosporinite compaction (Fig. 2-f). Minerals occurred as isolated grains or saturating pores (Fig. 2-g), and the intercrystallite or intracrystalline pores are also observed (Fig. 2-b). The same view position of the SEM and AFM pictures were also acquired, with the matrix, mineral and pores distribution can be clearly recognized (Fig. 3-a and b). The surface morphological characteristics, in a 3D view, is also displayed by the topographic map (Fig. 4-b). The morphology and distribution of various types of pores in coal can also be directly revealed by the AFM, which provides an important reference for the geometric study of coal pore structure [9]. In order to describe the pore and sample surface more accurately, surface variation profile of the studied samples was conducted (Fig. 4-c and d), with the pore size and pore depth are obtained. The cross-section shows that the coal surface is of strong irregularity. The maximum pore diameter in the cross-section is 2.706 μm, with a pore depth of 168.061 nm, and the minimum pore diameter is 412 nm with a depth of 43.054 nm. And the slit pores shown in the Fig. 3-d is as much as 294.954 nm deep. Several sections were conducted to calculate the average pore size and diameter, with their long axis, short axis, pore depth and pore area can be acquired (Table 1). 3.2. Surface roughness Surface roughness is of significance on the macroscopic contact angle measurement of coal which plays a vital role in measurement of contact angle, and various processes like moistening, spreading, and wetting [45,46]. The 3D surface roughness parameter can provide more comprehensive spatial surface features of the sample surface, which has better statistical significance and is an important indicator for characterizing the surface topography (Fig. 4-b). The measurement results of the sample roughness in the experiment are shown in Table 2. The Rq values of the tested samples are higher than the Ra values. The Rq values are more sensitive to the roughness variation, and small fluctuations in surface height will lead to great variations [47]. The Ra values range from 7.9 to 118.2 nm, with an average value of 49.3 nm, and the Rq values increase from 9.7 to 238.0 nm, averaging of 76.8 nm. The Rsk values range from −0.36 to 0.49 (averaging of −0.09), which indicates slightly unsymmetrical negative distribution of the surface points. And the Rku values of the samples vary from −0.43 to 1.07 (averaging of 0.19), indicating wide variations of the surface roughness.
3.3.2. Adhesion forces The second mechanical property acquired in the mapping is the adhesion force, illustrated as the minimum force in Fig. 6 and Fig. 1. The source of the adhesion force can be any attractive force between the tip and sample. In air, Van der Waals, electrostatics, and forces due to the formation of a capillary meniscus can all contribute to the relative strengths shown as parameters like Hammaker constants, surface charges, and hydrophilicity [49–52]. Adhesion force can reflect the surface energy of the sample [53]. The surface morphology and roughness have a great influence on the adhesion force, which is high in the pore area and low in the matrix (Fig. 6). When the porous surface of coal is formed, at least one side of the carbon atoms on the surface of coal is empty, resulting in unbalanced forces. This unbalance gives the carbon atoms on the surface of coal an additional energy, which can be expressed as surface energy. Surface energy can directly reflect the adsorption amount per unit area of coal surface. Thus the pores areas with complicated surface variations would be of higher adsorption capacity to gases. The Fig. 6 is a good comparison of the results. The plane porosity of the Fig. 6-a is 17.50%, roughness Ra of 118.2 nm and Rq of 238.0 nm, while the Fig. 6d showing a plane porosity of 4.30%, Ra 42.9 nm and Rq of 54.9 nm. The adhesion force of the two samples is as much as 0 nN to 473.26 nN, averaging of 79.47 nN; and the sample with low porosity and roughness is of 0 nN to 16.40nN, averaging of 4.47 nN. The adhesion force of each sample measured by AFM is shown in Table 2. The distribution of adhesion force is different from that of Young's modulus as the adhesion force is influenced by many factors, including the surface energy, micro-morphology and chemical composition of interface. The effect of roughness on the adhesion force is that
3.3. Nano mechanical characteristics 3.3.1. Young’s modulus The Young’s modulus (Fig. 3-c) and deformation (Fig. 3-d) distribution was acquired by the AFM. The Young's modulus variation of minerals is quite different from that of the coal matrix, with the minerals values is as much as 21 GPa while the matrix is only around 2 GPa. Part of the areas where pores developed show high Young's modulus, indicating that this part of the pores is filled with minerals. The deformation range after the coal surface was recorded (Fig. 3-d). Comparing the deformation image with the SEM pictures, it can be clearly seen that the location of the higher deformation ranges occurred in where pores developed, and the minerals show quite low deformation range. Both the deformation image and Young's modulus image reflect the same mechanical properties of samples. 7
Fuel 260 (2020) 116352
Y. Li, et al.
the actual contact area of the probe at a higher point is much smaller than the apparent area. If the contact area between the probe and the sample decreases, the adhesion decreases. This also explains that the adhesion force of the sample is high where more pores are developed. For coal, the location of high adhesion force is more conducive to gas adsorption, and the adhesive force also has a great influence on coal flotation as it influences the particles adhesion [54].
pore characteristics in coal and coal-bearing shale. Mar. Pet. Geol. 2019;105:331–7. [11] Liu X, Song D, He X, Wang Z, Zeng M, Deng K. Nanopore structure of deep-burial coals explored by AFM. Fuel 2019;246:9–17. [12] Liu X, Song D, He X, Wang Z, Zeng M, Wang L. Quantitative analysis of coal nanopore characteristics using atomic force microscopy. Powder Technol. 2019;346:332–40. [13] Keehm Y, Mukerji T, Nur A. Permeability prediction from thin sections: 3D reconstruction and Lattice-Boltzmann flow simulation. Geophys. Res. Lett. 2004;31(4):1668. [14] Wu K, Dijke M, Couples G, Jiang Z, Ma J, Sorbie K, et al. 3D stochastic modelling of heterogeneous porous media–applications to reservoir rocks. Transp. Porous Media 2006;65(3):443–67. [15] Curtis M, Sondergeld C, Ambrose R, Rai C. Microstructural investigation of gas shales in two and three dimensions using nanometer-scale resolution imaging. AAPG Bull. 2012;96(4):665–77. [16] Li Y, Zhang C, Tang D, Gan Q, Niu X, Shen R. Coal pore size distributions controlled by the coalification process: an experimental study of coals from the Junggar, Ordos, and Qinshui basins in China. Fuel 2017;206:352–63. [17] Abedi S, Slim M, Ulm F. Nanomechanics of organic-rich shales: the role of thermal maturity and organic matter content on texture. Acta Geotech. 2016;11:775–87. [18] Foster B. New atomic force microscopy (AFM) approaches life sciences gently, quantitatively, and correlatively. Am. Lab. 2012;4:24–8. [19] Trtik P, Kaufmann J, Volz U. On the use of peak-force tapping atomic force microsco-py for quantification of the local elastic modulus in hardened cement paste. Cem. Concr. Res. 2012;42(1):215–21. [20] He X, Liu X, Song D, Nie B. Effect of microstructure on electrical property of coal surface. Appl. Surf. Sci. 2019;483:713–20. [21] Lee T, Bocquet L, Coasne B. Activated desorption at heterogeneous interfaces and long-time kinetics of hydrocarbon recovery from nanoporous media. Nat. Commun. 2016;7:11890. [22] Ju Y, Huang C, Sun Y, Wan Q, Lu X, Lu S, et al. Nanogeosciences: research history, current status, and development trends. J. Nanosci. Nanotechnol. 2017;17(9):5930–65. [23] Zhu H, Ju Y, Qi Y, Huang C, Zhang L. Impact of tectonism on pore type and pore structure evolution in organic-rich shale: Implications for gas storage and migration pathways in naturally deformed rocks. Fuel 2018;228:272–89. [24] Cai J, Lin D, Singh H, Wei W, Zhou S. Shale gas transport model in 3D fractal porous media with variable pore sizes. Mar. Pet. Geol. 2018;98:437–47. [25] Tan L, Ren T, Yang X, He X. A numerical simulation study on mechanical behaviour of coal with bedding planes under coupled static and dynamic load. Int. J. Min. Sci. Technol. 2018;28:791–7. [26] Kumar H, Mishra S, Mishra K. Experimental Evaluation of Geo-mechanical properties of Coal using Sonic Wave Velocity. International Conference on Advances in Agricultural, Biological & Environmental Sciences July 22-23, 2015 London (UK). [27] Li Y, Tang D, Elsworth D, Xu H. Characterization of coalbed methane reservoirs at multiple length scales: a cross-section from southeastern Ordos Basin, China. Energ. Fuel 2014;28:5587–95. [28] Li Y, Wang Z, Gan Q, Niu X, Xu W. Paleoenvironmental conditions and organic matter accumulation in Upper Paleozoic organic-rich rocks in the east margin of the Ordos Basin, China. Fuel 2019;252:172–87. [29] Li Y, Tang D, Fang Y, Xu H, Meng Y. Distribution and genesis of stable carbon isotope in coalbed methane from the east margin of Ordos Basin. Sci China Earth Sci 2014;57(8):1741–8. [30] Li Y, Yang J, Pan Z, Meng S, Wang K, Niu X. Unconventional natural gas accumulations in stacked deposits: a discussion of Upper Paleozoic coal-bearing strata in the east margin of the Ordos Basin. China Acta Geol Sin (English Edition) 2019;93(1):111–29. [31] Binning G, Quate C, Gerber C. Atomic force microscopy. Phys Rev Lett 1986;56(9):930–3. [32] Patel A, Kranz C. (Multi) functional atomic force microscopy imaging. Annu Rev Anal Chem 2018;11:329–50. [33] Cappella B, Dietler G. Force-distance curves by atomic force microscopy. Surf Sci Rep 1999;34(1–3):5–104. [34] Wang W, Guo Z, Sun J, Li Z. Nano-mechanical characterization of disassembling amyloid fibrils using the Peak Force QNM method. Biopolymers 2017;107(2):61–9. [35] Young T, Monclus M, Burnett T, Broughton W, Ogin S, Smith P. The use of the PeakForce TM quantitative nanomechanical mapping AFM-based method for highresolution Young’s modulus measurement of polymers. Meas Sci Technol 2011;22(12). 125703. [36] Derjaguin B, Muller V, Toporov Y. Effect of contact deformations on the adhesion of particles. J Colloid Interface Sci 1975;53(2):314–26. [37] Bruening F, Cohen A. Measuring surface properties and oxidation of coal macerals using the atomic force microscope. Coal Geol 2005;63:195–204. [38] Yao S, Jiao K, Zhang K, Hu W, Ding H, Li M, et al. An atomic force microscopy study of coal nanopore structure. Chinese Sci Bull 2011;56:2706–12. [39] Pancewicz T, Mruk I. Holographic contouring for determination of three dimensional description of surface roughness. Wear 1996;199(1):127–31. [40] Li C, Dong S. The parameters and methods of characterizing 3D surface microtopography. J Astronaut Metrol Measure 1999;19(6):33–43. [41] Gadelmawla E, Koura M, Maksoud T, Elewa I, Soliman H. Roughness parameters. J Mater Process Technol 2002;123(1):133–45. [42] Abouelatta O, Madl J. Surface roughness prediction based on cutting parameters and tool vibrations in turning operations. J Mater Process Technol 2001;118(1–3):269–77. [43] Carvalho A, Maugeri F, Silva V, Hernandez A, Palacio L, Pradanos P. AFM analysis of the surface of nanoporous membranes: application to the nanofiltration of
4. Conclusions Observations combining AFM and SEM tests were conducted to reveal the pore structure and mechanical properties of coal. The following conclusions can be drawn: (1) Combing AFM and SEM observations in a nanoscale is effective to reveal pore structure and mechanical properties in 2D and 3D dimensions. The SEM provides information of pore size distribution, and matrix and mineral compositions, and the AFM presents pore structure, surface roughness and mechanical properties. (2) MAPS provide information of coal surfaces in a larger observation size with high resolution degrees, which is of significance in revealing pores distribution and maceral composition at different scales. The combination of SEM equipped with energy spectrum and AFM tests in different areas would be useful in revealing geochemical and geomechanical information of coal and shale reservoirs. (3) The PF-QNM mode of AFM quantitatively illustrates the surface roughness, Young’s modulus and adhesion force variations. The coal matrix and minerals show huge difference in mechanical properties, and the influence of mineral distribution in hydraulic fracturing should be carefully considered. (4) The quantitative information of Young's modulus and adhesion force of coal samples can be obtained by AFM. The adhesion force which reflects the surface energy information is affected by the surface morphology. The measurement of adhesion force is conducive to the study of coal adsorption capacity and coal flotation. Acknowledgments This study was supported by the National Natural Science Foundation of China (Grant No. 41702171), National Major Science and Technology Projects of China (No.2016ZX05066001-002). We are grateful to the reviewers for their valuable advice during their busy schedule. References [1] Clarkson C, Solano N, Bustin R, Bustin A, Chalmers G, He L, et al. Pore structure characterization of North American shale gas reservoirs using USANS/SANS, gas adsorption, and mercury intrusion. Fuel 2013;103:606–16. [2] Wu T, Zhang D. Impact of adsorption on gas transport in nanopores. Sci Rep 2016;6:23629. [3] Yang J, Hatcherian J, Hackley P, Pomerantz A. Nanoscale geochemical and geomechanical characterization of organic matter in shale. Nat Commun 2017;8(1):2179. [4] Li Y, Wang Z, Pan Z, Niu X, Yu Y, Meng S. Pore structure and its fractal dimensions of transitional shale: A cross section from east margin of the Ordos Basin, China. Fuel 2019;241:417–31. [5] Chen C, Hu D, Westacott D, Loveless D. Nanometer-scale characterization of microscopic pores in shale kerogen by image analysis and pore-scale modeling. Geochem. Geophys. Geosyst. 2013;14(10):4066–75. [6] Blunt M, Bijeljic B, Dong H, Gharbi O, Iglauer S, Mostaghimi P, et al. Pore-scale imaging and modelling. Adv. Water Resour. 2013;51:197–216. [7] Kelly S, El-Sobky H, Torres-Verdín C, Balhoff M. Assessing the utility of FIB-SEM images for shale digital rock physics. Adv. Water Resour. 2016;95:302–16. [8] Wu T, Li X, Zhao J, Zhang D. Multiscale pore structure and its effect on gas transport in organic-rich shale. Water Resour. Res. 2017;53:5438–50. [9] Zhao S, Li Y, Wang Y, Ma Z, Huang X. Quantitative study on coal and shale pore structure and surface roughness based on atomic force microscopy and image processing. Fuel 2019;244:78–90. [10] Liu X, Nie B, Wang W, Wang Z, Zhang L. The use of AFM in quantitative analysis of
8
Fuel 260 (2020) 116352
Y. Li, et al.
[49] Tiwari A, Gladwell G. Nanomechanical analysis of high performance materials. Springer; 2013. p. 43–4. [50] Weisenhorn A, Hansma P, Albrecht T, Quate C. Forces in atomic force microscopy in air and water. Appl Phys Lett 1989;54:2651–3. [51] Dokukin M, Sokolov I. Quantitative mapping of the elastic modulus of soft materials with Harmoni X and peak force QNM AFM modes. Langmuir 2012;28(46):16060–7. [52] Rezende C, Lee L, Galembeck F. Surface mechanical properties of thin polymer films investigated by AFM in pulsed force mode. Langmuir 2009;25(17):9938–46. [53] Bastian S, Mikhail S, Wael A, Maryam H, Maxim F, Jan B, et al. Quantifying mineral surface energy by scanning force microscopy. J Colloid Interface Sci 2016;472:237–46. [54] Tri M, Tsuyoshi H, Nur L, Lina N, Sudarsono S. Surface physicochemical properties of semi-anthracitic coal from Painan-Sumatra during air oxidation. Int J Coal Sci Technol 2018;5(2):156–66.
potassium clavulanate. J. Mater Sci 2011;46(10):3356–69. [44] Li W, Wang C, Shi Z, Wei Y, Zhou H, Deng K. The description of shale reservoir pore structure based on method of moments estimation. PLoS ONE 2016;11(3):0151631. [45] Kang J, Zhang B, Wang J, Wu J, Wang L, Yin B, et al. Mechanical testing of anthracite to assess its surface energy and temperature dependence. Fuel 2019;239:76–86. [46] Han B, Liang S, Wang B, Zheng J, Xie X, Xiao K, et al. Simultaneous determination of surface energy and roughness of dense membranes by a modified contact angle method. Colloids Surf A 2019;562:370–6. [47] Liu J, Jiang X, Huang X, Wu S. Morphological characterization of super fine pulverized coal particle. Part 2. AFM investigation of single coal particle. Fuel 2010;89(12):3884–91. [48] Domke J, Radmacher M. Measuring the elastic properties of thin polymer films with the atomic force microscope. Langmuir 1998;14:3320–5.
9