Design Parameters of a Direct Contact Membrane Distillation and a Case Study of Its Applicability to Low-Grade Waste Energy.

In the design of membrane distillation systems, the effect of different heat transfer coefficient models on the transmembrane flux seems to have been overlooked thus far. Interestingly, the range of discrepancy in the results of the transmembrane flux is wide, especially in the laminar flow region, where MD is often operated. This can be inferred by studying the design and parameters of the direct contact membrane distillation system. In this study, the physical and physiochemical properties that affect the design of MD are comprehensively reviewed, and based on the reviewed parameters, an MD design algorithm is developed. In addition, a cost analysis of the designed MD process for low-grade-energy fluids is conducted. As a result, a total unit product cost of USD 1.59/m3, 2.69/m3, and 15.36/m3 are obtained for the feed velocities of 0.25, 1 and 2.5 m/s, respectively. Among the design parameters, the membrane thickness and velocity are found to be the most influential.

Not from Scratch: Predicting Thermophysical Properties through Model-Based Transfer Learning Using Graph Convolutional Networks.

In this study, a framework for the prediction of thermophysical properties based on transfer learning from existing estimation models is explored. The predictive capabilities of conventional group-contribution methods and traditional machine-learning approaches rely heavily on the availability of experimental datasets and their uncertainty. Through the use of a pretraining scheme, which leverages the knowledge established by other estimation methods, improved prediction models for thermophysical properties can be obtained after fine-tuning networks with more accurate experimental data. As our experiments show, for the case of critical properties of compounds, this pipeline not only improves the performance of the models on commonly found organic structures but can also help these models generalize to less explored areas of chemical space, where experimental data is scarce, such as inorganics and heavier organic compounds. Transfer learning from estimation models data also allows for graph-based deep learning models to create more flexible molecular features over a bigger chemical space, which leads to improved predictive capabilities and can give insights into the relationship between molecular structures and thermophysical properties. The generated molecular features can discriminate behavior discrepancy between isomers without the need of additional parameters. Also, this approach shows better robustness to outliers in experimental datasets.

Comprehensive analysis of a hybrid FO/crystallization/RO process for improving its economic feasibility to seawater desalination.

In this study, the FO/crystallization/RO hybrid process was analyzed comprehensively, including experimentation, modeling, and energy and cost estimation, to examine and improve its feasibility to seawater desalination. A new operating strategy by heating the FO process to 45 °C was suggested, and a detailed process design was conducted. A comparative analysis with the conventional seawater reverse osmosis (SWRO) process was performed in terms of specific energy consumption (SEC) and specific water cost (SWC). The hybrid process can produce fresh water with SWC of 0.6964 $/m3, electrical SEC of 2.71 kWh/m3, and thermal SEC of 14.684 kWh/m3. Compared to the conventional SWRO process (SWC of 0.6890 $/m3 and electrical SEC of 2.674 kWh/m3), the hybrid process can produce water with comparable cost and energy consumption. An economic feasibility study that utilized the waste heat and the developed FO technology was also carried out to investigate future developments of the hybrid process. The SWC can be reduced to 0.6435 $/m3 with free waste heat energy. The permeate water quality of the hybrid process was about half that of the conventional SWRO process on molar basis. The results revealed that the FO/crystallization/RO hybrid process can be utilized as a competitive process for seawater desalination with high recovery and high water quality.

Prediction of parathyroid hormone signalling potency using SVMs.

Parathyroid hormone is the most important endocrine regulator of calcium concentration. Its N-terminal fragment (1-34) has sufficient activity for biological function. Recently, site-directed mutagenesis studies demonstrated that substitutions at several positions within shorter analogues (1-14) can enhance the bioactivity to greater than that of PTH (1-34). However, designing the optimal sequence combination is not simple due to complex combinatorial problems. In this study, support vector machines were introduced to predict the biological activity of modified PTH (1-14) analogues using mono-substituted experimental data and to analyze the key physicochemical properties at each position that correlated with bioactivity. This systematic approach can reduce the time and effort needed to obtain desirable molecules by bench experiments and provide useful information in the design of simpler activating molecules.

On the use of pseudo-spectral method in model reduction and simulation of active dendrites.

Most of dendrites in the central nervous system are now known to have active channels. These active dendrites play important roles not only in signal summation but also in computation. For the simulation of these important active dendrites, the compartment model based on the finite volume or finite difference discretization was mainly adopted. In this paper, we employ the Chebychev pseudo-spectral method well developed in computational physics, and demonstrate that it can achieve a higher precision with the same number of equations than the compartment model. Moreover, it is also shown that the Chebychev pseudo-spectral method converges faster to attain a given precision. Hence, for the simulations of active dendrites, the Chebychev pseudo-spectral method can be an attractive alternative to the compartment model since it leads to a low order model with higher precision or converges faster for a given precision.

Modeling and parameter identification of the simultaneous saccharification-fermentation process for ethanol production.

Despite many environmental advantages of using alcohol as a fuel, there are still serious questions about its economical feasibility when compared with oil-based fuels. The bioethanol industry needs to be more competitive, and therefore, all stages of its production process must be simple, inexpensive, efficient, and "easy" to control. In recent years, there have been significant improvements in process design, such as in the purification technologies for ethanol dehydration (molecular sieves, pressure swing adsorption, pervaporation, etc.) and in genetic modifications of microbial strains. However, a lot of research effort is still required in optimization and control, where the first step is the development of suitable models of the process, which can be used as a simulated plant, as a soft sensor or as part of the control algorithm. Thus, toward developing good, reliable, and simple but highly predictive models that can be used in the future for optimization and process control applications, in this paper an unstructured and a cybernetic model are proposed and compared for the simultaneous saccharification-fermentation process (SSF) for the production of ethanol from starch by a recombinant Saccharomyces cerevisiae strain. The cybernetic model proposed is a new one that considers the degradation of starch not only into glucose but also into dextrins (reducing sugars) and takes into account the intracellular reactions occurring inside the cells, giving a more detailed description of the process. Furthermore, an identification procedure based on the Metropolis Monte Carlo optimization method coupled with a sensitivity analysis is proposed for the identification of the model's parameters, employing experimental data reported in the literature.

Gliding arc plasma processing of CO2 conversion.

Conversion of carbon dioxide (CO2) using gliding arc plasma was performed. The research was done to investigate the effect of variation of total gas flow rates and addition of auxiliary gases--N2, O2, air, water--to the CO2 conversion process. This system shows higher power efficiency than other nonthermal plasma methods. Experiment results indicate the conversion of CO2 reaches 18% at total gas flow rate of 0.8 L/min and produces CO and O2 as the main gaseous products. Among auxiliary gases, only N2 gives positive effect on CO2 conversion and the power efficiency at N2 concentration of 95% and total gas flow rate of 2 L/min increases about three times compared to pure CO2 process.

Lys296 and Arg299 residues in the C-terminus of MD-ACO1 are essential for a 1-aminocyclopropane-1-carboxylate oxidase enzyme activity.

The 1-aminocyclopropane-1-carboxylate (ACC) oxidase catalyzes the last step in the biosynthesis of ethylene from ACC in higher plants. The complex structure of ACC oxidase/Fe(2+)/H(2)O derived from Petunia hybrida has recently been established by X-ray crystallography and it provides a vast structural information for ACC oxidase. Our mutagenesis study shows that both Lys296 and Arg299 residues in the C-terminal helix play important roles in enzyme activity. Both K296R and R299K mutant proteins retain only 30-15% of their enzyme activities with respect to that of the wild-type, implying that the positive charges of C-terminal residues are involved in enzymatic reaction. Furthermore, the sequence alignment of ACC oxidases from 24 different species indicates an existence of the exclusively conserved motif (Lys296-Glu301) especially in the C-terminus. The structure model based on our findings suggests that the positive-charged surface in the C-terminal helix of the ACC oxidase could be a major stabilizer in the spatial arrangement of reactants and that the positive-charge network between the active site and C-terminus is critical for ACC oxidase activity.

The active site and substrate-binding mode of 1-aminocyclopropane-1-carboxylate oxidase determined by site-directed mutagenesis and comparative modelling studies.

The active site and substrate-binding mode of MD-ACO1 (Malus domestica Borkh. 1-aminocyclopropane-1-carboxylate oxidase) have been determined using site-directed mutagenesis and comparative modelling methods. The MD-ACO1 protein folds into a compact jelly-roll motif comprised of eight a-helices, 12 b-strands and several long loops. The active site is well defined as a wide cleft near the C-terminus. The co-substrate ascorbate is located in cofactor Fe2+-binding pocket, the so-called '2-His-1-carboxylate facial triad'. In addition, our results reveal that Arg244 and Ser246 are involved in generating the reaction product during enzyme catalysis. The structure agrees well with the biochemical and site-directed mutagenesis results. The three-dimensional structure together with the steady-state kinetics of both the wild-type and mutant MD-ACO1 proteins reveal how the substrate specificity of MD-ACO1 is involved in the catalytic mechanism, providing insights into understanding the fruit ripening process at atomic resolution.

Cybernetic modeling of the cephalosporin C fermentation process by Cephalosporium acremonium.

A cybernetic mathematical model has been developed to describe the production of cephalosporin C. In developing the model, diauxic behavior of substrate consumption, morphological differentiation of cells, and catabolite repression of cephalosporin C production by the preferred substrate, glucose, were considered. The proposed model was tested on the experimental data from the literature and could adequately describe the morphological differentiation of cells, the sequential utilization of carbon sources and the production of cephalosporin C. It could be a useful tool to optimize the production of cephalosporin C by Cephalosporium acremonium in batch, fed-batch or continuous operations.