Microscopy characterizing of the agglutinate particle shows a “Peanuts” shape, with a length of about 1.2 cm and diameter of about 0.4–0.7 cm. The optical micrograph shows a nearly three dimensional (3D) morphological characteristics. The 3D view of the agglutinate shows that the particle consists of different types of minerals distinguishing by different colors, which makes it easier to see the variety of different components within the agglutinates. It is reported that the agglutinate was formed by gluing tiny rocks, mineral fragments and metal particles in a random fashion to form the highly complex 3D structure[30]. The primary black part of the agglutinate shows that many of its components were encased completely in a dark glassy matrix. The outer surface also shows some bright and yellow parts of mineral fragments. The microstructures of the agglutinate indicate that the particle was formed during the micrometeor strikes. It is explained that when a micrometeor strikes the lunar surface enough heat is generated to fuse together and even melt the underlying soil particles[30–32]. Some of the melt is lost as spray during the initial crater formation. The rest of the impact melt flows into the spaces between the surrounding soil particles encasing them in a glassy matrix as it cools. When a micrometeorite strikes the lunar surface, the intense heat that is generated causes some of the underlying rock or regolith to melt, and as it cools, the molten mixture forms a glass. During this process, the potential volatile components and reduction-generated water gas are released, resulting in gaseous volatiles flowing in the melting material and forming the interior porous structure.
The SEM images of the agglutinate particle are shown in Fig. 1b-e. These figures display the local characteristics of particles by different magnification. A globule was displayed on the lower left corner of the Fig. 1b, which is reported that it was formed by the bubbling of molten glass[31]. As shown in Fig. 1b & 1c, there are myriad of tiny mineral fragments stuck to the outer surface of the particle. By contrast, other tiny fragments on the outer surface of the agglutinate particle have been tightly stuck together by melt flow glass. An open pore was also appeared in the particle, its maximum diameter was about 0.1 mm and the side wall of the pore was smooth and gentle. The feature indicates that the pore was formed by the gas volatilization from the inside out along with its formation due to the micrometeorite impacts. The SEM images of the interior section of the pore and the sidewall were shown in the Fig. 1d & 1e. These figures also confirmed that the sidewall was smooth and gentle, as well as some micropores were formed on the sidewall. These micropores may be also formed due to the gas volatilization with the formation of the agglutinate particle. The SEM images show that the agglutinate contains a large number of holes with complex scales, which indicates that the violent volatilization of gas during the formation of agglutinate particles due to the micrometeorite impacts. However, the identity of the gases responsible for pore formation is not reported for certain, yet many researchers believe that they are a mixture of gases including H, H2O and He implanted by the solar wind.[3, 34]
To understand the interior structure of the agglutinate sample, 3D morphology reconstruction is carried out through X-ray computed tomography (CT) characterizations. Slice images and 3D morphology of the agglutinate particle are shown in Fig. 2. The X-ray computed micro-tomography provides a more accurate impression of the interior structure of this agglutinate particle and how the complex pores interconnect. Images of Fig. 2a-2d show that different pores exist in the glassy matrix from different angles and orientations. The porosity of the sample particle is calculated to be approximately 0.71 based on the CT results. Figure 2e shows the 3D model reconstruction of the agglutinate based on the measured X-ray CT images. The 3D structure shows that the agglutinate particle contains both interior enclosed voids and open pores. The open pores in the agglutinate particle have ellipsoid or spherical shape with smooth edge. These features suggests that the voids are generated by the blister of the components of relatively low boiling temperature[32], and the formation of the agglutinate is a result of vigorous volatilization.
XEDS analysis carried out on the fragments attached to the outer surface of the agglutinate particle reveals that it is comprised primarily of magnesium (Mg), calcium (Ca), silicon (Si), aluminum (Al) and oxygen (O) elements. A small amount of iron (Fe) and titanium (Ti) elements are also contained in the fragments. As previously reported, agglutinate particles mainly consisted of different rock, mineral, and glass fragments enclosed in impact-melt glass.[27] Thus, the distribution of elements on the outer surface of the agglutinate particle show nonuniform features. The element types and distribution on the outer surface of the agglutinate particle are similar to that of other CE5 particles, whose dominated elements also including O, Si, Ca, Al, etc[32]. The SEDS analysis of an open pore is shown in Fig. 3b-d. The primary components within the open pores include O, Si, Fe and Al elements, the contents and distributions of these elements are significantly different from the adhered mineral fragments (Fig. 3a). Some regions on the sidewalls are dominated by iron and oxygen (Fig. 3c), which shows that there is a certain amount of FeO minerals. Furthermore, elemental iron particles were also existed in the agglutinate particles. These pure Fe particles were formed via the reduction reaction of solar-wind hydrogen element with the iron-mineral soils. Recently, Li et al.[12] reported the disproportionation origin of nanophase Fe particles in the CE5 sample driven by the micrometeorite impacts. Furthermore, it is reported that the agglutinate particles were also mainly induced by the micrometeorite impacts. Thus, some pure Fe particles could be reserved due to the lack of solar wind irradiation and cosmic radiation. As the whole process was dominated by impact events without the contribution from the solar wind, the initial information may be remained within the hole of the agglutinate particle.
Direct investigation of volatiles transport through the agglutinates is challenging considering the irregular morphology. In this work, a small rectangular agglutinate model is cut out from the original agglutinate sample to preserve its own characteristics to the maximum extent and facilitate the numerical study. Figure 4 depicts the geometric model and transport of the volatiles through the irregular pores. To obtain the dependence of local flow information on pressure in the porous agglutinates, the slip flow model is applied to simulate the gas advection flow in the porous structure under relatively high inlet pressure conditions (0.2 bar < Pin<2 bar). The inlet pressure begins at 20000 Pa with reference temperature of 500 K, and the solid wall is described by the slip boundary condition. The flow field is initialized with P0 = 0 and u0 = 0. Transient study has been conducted to assess the establishment of pressure balance. Figure 5a illustrates the pressure variation at the top surface of the agglutinate for the inlet pressure of 2×105 Pa. It is found that the pressure balances approximately to 1.4×103 Pa within 0.1 ps, which implies that the gas escapes very quickly from the reservoir and the volatiles become thinner and thinner. To evaluate the continuum gas transport property in the porous structure, the gaseous mass flow rate with applied pressure drop is plotted in Fig. 5b. The mass flow rate can be quantitatively calculated by continuity equation in Eq. (1).
where u1 is the superficial velocity at the inlet of the porous medium, r the fluid density, Ac the total cross-section area of flow field, and f the sample porosity. The superficial velocity can further be computed from pressure gradient by Darcy’s Law, namely:
where K is the viscous permeability of the porous agglutinate, m is the gas dynamic viscosity. In the advection flow regime under the continuum theory framework, the produced volatiles escape through the porous agglutinate on the order of 10− 6~10− 4 kg/s for different pressure conditions and the flow rate increases nonlinearly with the gas pressure drop. This is caused by the pressure dependent gas density and slippage condition in the porous structure. Given the limited number of gaseous products and high vacuum condition in the practical Moon regolith, the gas flow is bound to be quite fast and the upstream pressure decreases very quickly until the continuum theory fails and Knudsen diffusion dominates. High vacuum condition accelerates the escape of gas and makes the Moon surface even barren.
To investigate the details of gas flow through the porous agglutinate, the velocity distribution in typical pore spaces is examined for Pin=2 bar. The velocity profiles of gaseous volatiles at the cross section of y/Ly=0.11, 0.23, 0.34, 0.42, 0.53, and 0.65 are illustrated in Fig. 6, respectively. It is noted that the velocity field is greatly distorted by the irregular connected pores, which also provide different pathways for the volatiles to travel through the porous agglutinates. By examine the velocity field in Fig. 6, it is found that the gas flows with a mean velocity approximately 2084 m/s in the porous structure. This large velocity is caused by extrusion of the micropores in the porous agglutinate and the suction effect by the high vacuum condition on the Moon. In particular, the gas flow is severely suppressed and significantly accelerated at the narrow throats and the maximum velocity can reach 2480 m/s, which further reduces the volatiles concentration in the moon regolith and the total energy is thus decreased. The gas velocity begins to decrease after it escapes from the agglutinate and enters the vacuum space. Combining with the element distribution, the transport behavior of gaseous volatiles in the porous agglutinate provides more details on the formation of the barren Moon regolith, and offers ideas for the construction of artificial environment of the future Lunar research station.
With the rapid dissipation of gaseous volatiles in the moon regolith, further transport of the remained rarefied gas no longer obeys the continuum hypothesis and the Knudsen diffusion regime starts to dominate. In the Knudsen diffusion regime, the gas molecules are fixed at the bottom layer to mimic the volatiles underneath. Once released, volatile molecules rapidly diffuse through the porous medium and pile up in the throat regions as illustrated in Fig. 7. It is noted that the time of gas transport through the porous sample under Knudsen diffusion regime is on the order of 1 ms, which is much longer than the continuum regime. Gas molecules gather at the bottom pores and diffuse into the sample after initially released. At t = 0.5 ms, the large portion of the gas molecules are still trapped at the bottom inlet pores for the system temperature T = 500 K, and only a very small amount of gas molecules transport through the sample, the transmission probability () measuring the percentage of gas molecules transport through the sample is approximately 0.054. As depicted in Fig. 7, the gas molecules migrate with a mean velocity more than a magnitude slower than that in the continuum regime. Due to the large mean free path of the rarefied volatiles, gas molecules interact frequently with the solid surfaces instead of self-interactions. Thus, the molecule diffusion inside the porous medium will be greatly decelerated due to the irregular scattering process. At t = 1.1 ms, approximately half portion of the gas molecules transport through the porous sample while the diffusion process slows down after t = 1.7 ms. Although the residual gas molecules in the porous agglutinate is less and less after 2 ms, it is difficult to expel them completely.
To evaluate the statistical pattern of the volatile diffusion via the agglutinate, we computed the gas transmission probability variation under different temperature. In Fig. 8a, it is found that the gas transmission probability monotonically increases with temperature. As surrounding temperature increases, the mean kinetic energy of gas molecules also increases according to from statistical thermodynamics, which subsequently increases the probability of gas molecules overcoming the energy barrier of the porous sample. As indicated in Fig. 8a, the transmission probability increases approximately 4 times as temperature increases from 300 K to 1300 K at t = 1 ms. It is noteworthy that the slope of the transmission probability curve is decreasing as temperature increases. This is mainly caused by the surface mitigation as the result of molecule-surface scattering. Generally, the gas molecules diffuse through the porous sample under molecule-wall interaction and self-interaction. As temperature increases, extra kinetic energy will bolster the microscopic interactions and more and more gas molecules get through the porous region. Therefore, the number density of gas molecules in the porous region decreases and gas-wall interaction dominates over self-interaction. The diffusion of gas molecules through the porous sample will be greatly mitigated considering the irregular pore geometry and rarefied gas density. This phenomenon is clearer in Fig. 8b. The curves of transmission probability variation can be clearly divided into three stages by their slopes. At the initial stage, a slowly increases with time and is lower than 0.01. This corresponds to the entrance of the porous agglutinate and gas molecules can hardly get through the sample. Once the gas molecules fully occupy the pore space and get to the outlet boundary, they begin to escape the porous sample and rush into the vacuum space. This corresponding to the middle stage and a sharply increases with time. In the last stage, a gradually increases with time and asymptotes to 1, which represents the slow diffusion process due to the rarefied gas density and gas-wall scattering. For T = 300 K, a increases to 0.9 in 2.9 ms, while it only needs 1.4 ms for T = 1300 K. Nevertheless, it is difficult to get rid of all the gas molecules in the porous region.