We describe an improved Bayesian inference methodology to characterize photovoltaic materials by matching charge carrier simulations to spectroscopy data. A “parallel tempering” scheme is introduced, which efficiently and reliably locates the global maximum in the complex multimodal distributions that are characteristic of cadmium telluride (CdTe) films. Our results show that the standard carrier transport model cannot explain the observed decay of time-resolved photoluminescence (TRPL) data from CdTe films and that there is carrier trapping within low-lying defect states. This inference has been confirmed by temperature-dependent TRPL and time-resolved emission spectroscopy (TRES). Our work shows that Bayesian inference can discriminate between plausible physics models, as well as determine parameter values for a given model. Finally, we have combined TRPL with time-resolved terahertz spectroscopy (TRTS) to describe the dynamics on nanosecond to microsecond time scales. These results show that sample degradation can be detected by its effect on surface recombination.

We describe a microfluidic device to extract DNA from a cell lysate, without the need for centrifuges, magnetic beads, or gels. Instead, separation is driven by transverse migration of DNA, which occurs when a polyelectrolyte solution flowing through a microfluidic channel is subjected to an electric field. The coupling of the weak shearing with the axial electric field is highly selective for long, flexible, charged molecules, of which DNA is the sole example in a typical cell lysate. As a result of migration to the walls, DNA is held near the channel inlet by electrophoresis (there is no flow near the channel walls), while the remaining components are eluted by the much larger (at least 10-fold) convective flow. We have demonstrated the feasibility of the device by recovering up to 40 ng of purified DNA in less than 30 min from 10 μL of Escherichia coli lysate. Gel electrophoresis indicates minimal additional fragmentation during purification, up to the maximum length recorded by the gel (60 kbp). Electropherograms were also obtained for purified mammalian DNA, using a Femto Pulse system (fragment lengths up to 165 kbp). Extracted samples show strong amplification by PCR, while the original lysate does not. Mixtures of λ-DNA and BSA were used to determine the extent of the separation of DNA from a physiological concentration of proteins (30 mg/mL). The protein concentration in the extract (0.3 to 0.5 ng/µL) was reduced by five orders of magnitude from the initial mixture.

Dissolution of fractured and porous media introduces a positive feedback between fluid flow and reactant transport, leading to the emergence of pronounced, fingerlike channels. We investigate the formation of these structures using a microfluidic Hele-Shaw cell with a soluble bottom. Our experiments show that the shape of dissolution fingers is invariant and reveals itself over time as the fingers extend into the system. By combining reactive-transport theory and conformal mapping techniques we derive these invariant forms. We relate these results to natural dissolution fingers in karst landscapes, and illustrate how to determine the groundwater flow rate responsible for their formation based on the finger shape.

Stalagmites are isolated columns of calcium carbonate growing on a cave floor; their growth is driven by the constant dripping of supersaturated solutions from the roof of the cave. In this paper, we derive a closed-form expression for the shape of a steadily growing stalagmite. Our analysis gives rise to three distinct shapes, all of them observable in nature, with the shape characterized by a single dimensionless parameter. Transitions between different shapes occur at a specific value of this parameter, with additional selection rules determining the shape and size of stalagmites evolving under specific cave conditions. Our theory shows that the stalagmite shape influences the 13C isotope shifts, which are an important source of paleoclimatic information.

Granite-hosted gold deposits, a major component of global gold resources, exhibit complex geochemical evolution and diverse mineralization types. Understanding the mechanisms of gold precipitation and enrichment in these systems is crucial for mineral resources exploration and extraction. Alteration processes such as potassic alteration, sericitization and pyritization significantly influence gold mobility and concentration. However, their roles in fluid-rock interactions and coupled physical-chemical dynamics remain insufficiently understood due to the complexity of mineralogical and environmental factors, as well as limited experimental and modeling data.This study employs PFLOTRAN-based reactive transport modeling to explore the geochemical mechanisms of gold precipitation and enrichment, using the Sanshandao gold deposit in Jiaodong Peninsula, China, as a case study. Integrating geological data, hydrothermal fluid dynamics, and thermodynamic of chemical reaction networks, the model explores the influence of alteration minerals, including K-feldspar, sericite and pyrite, on fluid composition, gold solubility and precipitation. It evaluates the effects of critical parameters such as temperature, pressure and PH on the stability and solubility of gold-bearing complexes, revealing the advantageous conditions for gold precipitation.Alteration minerals affect hydrothermal fluid properties, such as PH and redox potential, which govern gold precipitation. For example, sericitization decreases fluid PH and enhances gold solubility, while pyritization facilitates adsorption, promoting localized gold enrichment. This underscores the importance of fluid-rock interactions and geochemical conditions in controlling gold transport and enrichment.This study offers a framework for understanding the physical-chemical mechanisms of gold mineralization in granite-hosted gold deposits. The use of reactive transport modeling provides insights how alteration processes and fluid-rock interactions shape ore-forming mechanism similar to geological settings.Keywords: Granite-hosted gold deposits; Reactive transport modeling; Fluid-rock interaction; Gold precipitation and enrichment; Ore-forming mechanism

Quantifying rates of charge carrier recombination is a crucial step in developing solar cells with high power conversion efficiencies. However, recovering characteristic parameters such as carrier mobility, doping concentration, and recombination rate constants is hindered by the interplay of the carrier dynamics with multiple recombination mechanisms. Interpretation of optoelectronic measurements, such as time-resolved photoluminescence (TRPL), usually relies on analytically tractable simplifications to the underlying physics models for carrier mobility and recombination, which sacrifices some of the information content of the measurement. We have recently shown that, by incorporating simulations of the complete carrier physics into a Bayesian analysis, previously unrecoverable material parameters, such as carrier mobility and the doping level, can be determined from TRPL measurements. Unfortunately, the large number of simulations required by a random sampling of the parameter space necessitated access to high-performance computing resources, limiting the usefulness of the approach. Here, we introduce an importance sampling algorithm (Metropolis Monte Carlo), which reduces the computational requirements by 2–3 orders of magnitude, rendering the Bayesian inference tractable on desktop computers. These developments affirm the utility of a simulation-driven analysis of optical characterization measurements, and make a physics-informed Bayesian inference available to all semiconductor researchers.

The recovery of characteristic absorber parameters such as the carrier mobility, doping concentration, and rate constants of each carrier recombination mechanism from optical characterization measurements has historically been hindered by the complexity of the carrier dynamics. This necessitates the use of simplified, analytically solvable physics models that sacrifice some of the potential information content of the measurement data. In this work, we introduce a desktop-scale Markov Chain Monte Carlo (MCMC) sampler that utilizes simulation of the full carrier physics to recover material parameters with increased accuracy or which were previously inaccessible. From a “power scan” consisting of time-resolved photoluminescence (TRPL) data at varying excitation intensity, we recover the ambipolar carrier mobility, the doping concentration, and rate constants for Auger and bimolecular radiative recombination. We also obtain an effective lifetime for defect-assisted nonradiative recombination, which can be decomposed further into bulk and surface recombination components by introducing additional data from multiple material sample thicknesses. These results reaffirm the potential for simulation-driven statistical analyzers to greatly expand the utility of optical characterization measurements, though herein the need for expensive computational resources is not required.

A rotating disk is the canonical experiment for measuring surface reaction rates in geochemical and electrochemical systems. Using the similarity solution for laminar flow around an infinite disk, the mass transfer coefficient can be simply related to the intrinsic reaction rate at the surface. However, measurements of mass transfer rates use a finite-size disk within a larger container of solution; here the flow is no longer strictly laminar, but there must always be some recirculation. Our interest was initially in the assumption of a uniform radial concentration field, how this breaks down near the perimeter of the disk, and what effect that might have on the measured mass transfer rates. However, our numerical simulations suggest that the flow around a finite-size disk becomes time dependent at Reynolds number ( $Re$ ) below 1000, which is much smaller than the typical values in mass-transfer measurements ( ${Re} \sim 10^4$ ). We observe the formation of coherent structures in the flow, which suggest a non-uniform mass transfer at the disk surface. The rotating-disk flow follows a similar sequence of instabilities to the Taylor–Couette flow: a centrifugal instability leading an axisymmetric, time-invariant flow, followed by a Hopf bifurcation to a time-periodic flow. To minimise the possibility that our results are a numerical artefact, we have also simulated the instability in the stationary boundary layer of a rotor–stator flow, comparing with self-similar solutions at low ${Re}$ and with spectral methods near the critical Reynolds number.

Upscaling methods are frequently used to derive transport equations at the macroscopic scale from more fundamental equations formulated at the pore scale. These methods typically give a suitable structure for the macroscopic equations and can also provide explicit expressions for the constitutive parameters, such as permeability and dispersion coefficients. Introducing chemical reactions complicates upscaling in at least two important ways. First, the interplay between chemical reactions and transport processes introduces a new length scale, which can be much smaller than the convective or dispersive length scales. A small reactive length scale breaks one of the key assumptions in upscaling; that there is a significant separation in length between the pore-scale and macro-scale processes. The second complication is that if reactions take place at mineral surfaces (dissolution or precipitation) then the pore space itself is evolving in time. In this paper we suggest ways in which these difficulties can be approached, based on analysis of pore-scale simulation data. First, we noticed that the concentration field in successive unit cells has an almost identical spatial variation, with a single scaling factor for each unit cell that is proportional to the incoming reactant flux. Using pore-scale simulations to determine the mass transfer coefficient in a few unit cells, we can calculate the concentration field in the whole domain, even when dissolution is entirely transport limited. Second, we have noticed a time-dependent mapping of the grain shapes from different unit cells. From these observations, we can deduce constitutive relations where the only time-varying parameter is the porosity. We show that a model based on these ideas can quantitatively account for the pore-scale simulation data.

Quantifying charge-carrier dynamics within a material or device from analysis of optoelectronic measurements is a crucial aspect of improving next-generation solar cells. However, analyses by hand are limited in information yield due to their reliance on simplified physics models. Simulation of full physics models can be computationally expensive. Here we demonstrate a GPU-accelerated machine learning approach via Bayesian parameter estimation for rapid analysis of optoelectronic data. Using time-resolved photoluminescence (TRPL) data of a perovskite absorber as a case study, we demonstrate our ability to estimate carrier mobilities, the doping level, and the radiative recombination rate. Furthermore, while most TRPL analyses are limited to determining an effective minority carrier lifetime, we reliably decompose this recombination lifetime into radiative, bulk nonradiative, and surface nonradiative components by introduction of TRPL data from multiple absorbers of different thicknesses. Our simultaneous collection of these parameters represents a significant increase in the typical information yield from TRPL measurements.

To enhance the accuracy and effectiveness of optoelectronic characterization, Bayesian inference has been applied to the statistical analysis of time-resolved photoluminescence (TRPL) data with large-scale graphics-processing unit (GPU)-based simulations of electron dynamics. A simulated TRPL dataset, derived from a CH3NH3PbI3-xClx perovskite absorber, was used as a case study. From a power scan, Bayesian inference extracts values of the (ambipolar) carrier mobility, free-carrier density, radiative recombination rate, and the overall lifetime of the nonradiative recombination processes. Analysis of the experimental TRPL data yields similar parameter values. The independent contributions of bulk and surface recombination can be distinguished via the introduction of an additional sample thickness. Further experiments separate the front and back surface recombination velocities and can distinguish the electron and hole mobilities. Ultimately, Bayesian inference enables a significant increase in the information yield from TRPL measurements.

We review theoretical and computational research, primarily from the past 10 years, addressing the flow of reactive fluids in porous media. The focus is on systems where chemical reactions at the solid-fluid interface cause dissolution of the surrounding porous matrix, creating nonlinear feedback mechanisms that can often lead to greatly enhanced permeability. We discuss insights into the evolution of geological forms that can be inferred from these feedback mechanisms, as well as some geotechnical applications such as enhanced oil recovery, hydraulic fracturing, and carbon sequestration. Until recently, most practical applications of reactive transport have been based on Darcy-scale modeling, where averaged equations for the flow and reactant transport are solved. We summarize the successes and limitations of volume averaging, which leads to Darcy-scale equations, as an introduction to pore-scale modeling. Pore-scale modeling is computationally intensive but offers new insights as well as tests of averaging theories and pore-network models. We include recent research devoted to validation of pore-scale simulations, particularly the use of visual observations from microfluidic experiments.

Dissolution of porous rocks by reactive fluids is a highly nonlinear process resulting in a variety of dissolution patterns, the character of which depends on physical conditions such as flow rate and reactivity of the fluid. Long, finger-like dissolution channels, “wormholes”, are the main subject of interest in the literature, however, the underlying dynamics of their growth remains unclear.While analyzing the tomography data on wormhole growth. one open question is to define the exact position of the tip of the wormhole. Near the tip the wormhole gradually thins out and the proper resolution of its features is hindered by the finite spatial resolution of the tomographs. In particular, we often observe in the near-tip region several disconnected regions of porosity growth, which - as we hypothesized - are connected by the dissolution channels at subpixel scale. In this study, we show how these features can be better resolved by using numerically calculated flow fields in the reconstructed pore-space.We used 70 micrometers, 16-bit grayscale X-ray computed microtomography (XCMT) time series scans of limestone cores, 14mm in diameter and 25mm in length. Scans were performed during the entire dissolution experiment with an interval of 8 minutes. These scans were further processed using a 3-phase segmentation proposed by Luquot et al.[1], in which grayscale voxels are converted to macro-porosity, micro-porosity and grain phases from their grayscale values. The macro-porous phase is assigned a porosity of 1, while the grain phase is assigned 0. Micro-porous regions are assigned an intermediate value determined by linear interpolation between pore and grain threshold using grayscale values. An OpenFOAM based, Darcy-Brinkman solver, porousFoam, is then used to calculate the flow field in this extracted porosity field.Porosity contours reconstructed from the tomographs show some disconnected porosity growth near the tip region which later become part of the wormhole in subsequent scans. We have used a novel approach by including the micro-porosity phase in pore-space to calculate the flow-fields in the near-tip region. The calculated flow fields clearly show an extended region of focused flow in front of the wormhole tip, which is a manifestation of the presence of a wormhole at the subpixel scale. These results show that micro-porosity plays an important role in dissolution and 3-phase segmentation combined with the flow field calculations is able to capture the sub-resolved dissolution channels.[1] Luquot, L., Rodriguez, O., and Gouze, P.: Experimental characterization of porosity structure and transport property changes in limestone undergoing different dissolution regimes, Transport Porous Med., 101, 507–532, 2014

We apply conformal mapping to find the evolving shapes of a dissolving cylinder in a potential flow. Similar equations can be used to describe melting in a flowing liquid phase. Results are compared with microfluidic experiments and numerical simulations. Shapes predicted by conformal mapping agree almost perfectly with experimental observations, after a modest (20 %) rescaling of the time. Finite-volume simulations show that the differences with experiment are connected to the underlying assumptions of the analytical model: potential flow and diffusion-limited dissolution. Approximate solutions of the equations describing the evolution of the shape of the undissolved solid can be derived from a Laurent expansion of the mapping function from the unit circle. Asymptotic expressions for the evolution of the area of the disk and the shift in its centre of mass have been derived at low and high Peclet number. Analytic approximations to the leading-order Laurent coefficients provide additional insight into the mechanisms underlying pore-scale dissolution.

We report separation of genomic DNA (48 kbp) from bovine serum albumin (BSA) by the electro-hydrodynamic coupling between a pressure-driven flow and a parallel electric field. Electro-hydrodynamic extraction exploits this coupling to trap DNA molecules at the entrance of a microfluidic contraction channel, while allowing proteins and salts to be flushed from the device. Samples (10 μL) containing λ-DNA (1 ng) and BSA (0.3 mg) were injected directly into the device and convected to the contraction channel entrance by a flowing buffer solution. The DNA remains trapped in this region essentially indefinitely, while proteins and salts are eluted. The effectiveness of the concept has been assessed by fluorescence measurements of DNA and BSA concentrations. Electro-hydrodynamic extraction in a single-stage device was found to enhance the concentration of DNA 40-fold, while reducing the BSA concentration by four orders of magnitude. The relative concentrations of DNA to BSA at the contraction channel entrance can be as large as 1.5 : 1, corresponding to an A260/280 ratio of 1.9. The maximum yield of DNA from a salt-free solution is 50%, while salted (150 mM) solutions have a lower yield (38%).

In this work we have investigated numerically the formation of channelised dissolution patterns, termed “wormholes”, using initial pore geometries generated from tomographic images of limestone cores. We have employed an OpenFOAM-based Darcy-scale numerical solver, porousFoam, which combines a Darcy/Darcy-Brinkman flow solver and a reactive transport solver in an evolving pore space. Simulated geometries, of both final and intermediate steps, are compared to dissolution experiments on samples the initial pore geometry is generated from, with the same acid concentration and flow rate applied.The initial condition of porosity distribution is set from X-Ray Computed Microtomography (XCMT) images via three phase segmentation into macroporosity, microporosity, and grain regions. Porosity values for microporous regions are set using linear interpolation between pore and grain grayscale values [1]. The inlet boundary conditions of flow rate and acid concentration are set as in the dissolution experiment. To test the effect of the permeability-porosity constitutive relationship we have investigated several options including power laws of varying exponent, and the Carman-Kozeny relation. We have also analyzed the impact of using Darcy versus Darcy-Brinkman flow solvers. Despite a qualitatively similar appearance to experimental results, the simulated wormholes are usually significantly thicker than their experimental counterparts, a fact noted by other researchers as well [2]. We comment on possible reasons for this discrepancy and on the limitations of Darcy-scale solvers in general. Additionally, we find that higher exponents in the power law makes the numerical dissolution very sensitive to grayscale threshold values as a small variation in this value changes the path of the wormhole.[1] Luquot, L., Rodriguez, O., and Gouze, P.: Experimental characterization of porosity structure and transport property changes in limestone undergoing different dissolution regimes, Transport Porous Med., 101, 507–532, 2014.[2] Yue Hao, Megan Smith, Yelena Sholokhova, Susan Carroll, CO2-induced dissolution of low permeability carbonates. Part II: Numerical modeling of experiments, Advances in Water Resources, 62, 388-408, 2013