Batch Approach Research Articles

Assessment of microbial products in the biosorption process of Cu(II) onto aerobic granular sludge: Extracellular polymeric substances contribution and soluble microbial products release

In the present study, the responses of microbial products in the biosorption process of Cu(II) onto aerobic granular sludge were evaluated by using batch and spectroscopic approaches. Batch experimental data showed that extracellular polymeric substances (EPSs) contributed to Cu(II) removal from an aqueous solution, especially when treating low metal concentrations, whereas soluble microbial products (SMPs) were released under the metal stress during biosorption process. A three-dimensional excitation-emission matrix (3D-EEM) identified four main fluorescence peaks in the EPS, i.e., tryptophan protein-like, aromatic protein-like, humic-like and fulvic acid-like substances, and their fluorescence intensities decreased gradually in the presence of Cu(II) during the sorption process. Particularly, tryptophan protein-like substances quenched the Cu(II) binding to a much higher extent through a static quenching process with less than one class of binding sites. According to the synchronous fluorescence spectra, the whole fluorescence intensity of released SMP samples expressed an increased trend with different degrees along with contact time. Two-dimensional correlation spectroscopy (2D-COS) suggested that the fulvic-like fluorescence fraction might be more susceptible to metal exposure than other fractions. The result of molecular weight distribution demonstrated that the SMPs released from the biosorption process differed significantly according to contact time. The result obtained could provide new insights into the responses of microbial products from aerobic granular sludge with heavy metal treatment.

Multimodal Material identification through recursive tactile sensing

Tactile sensing has recently been used in robotics for object identification, grasping, and material identification. Although human tactile sensing is multimodal, existing tactile material recognition approaches use vibration information only. Moreover, material identification through tactile sensing can be solved as an continuous process, yet state of the art approaches use a batch approach where readings are taken for at least one second. This work proposes a recursive multimodal (vibration and thermal) tactile material identification approach. Using the frequency response of the vibration induced by the material and a set of thermal features, we show that it is possible to accurately identify materials in less than half a second. We conducted an exhaustive comparison of our approach with commonly used vibration descriptors and machine learning algorithms for material identification such as k-Nearest Neighbour, Artificial Neural Network and Support Vector Machines. Experimental results show that our approach identifies materials faster than existing techniques and increase the classification accuracy when multiple sensor modalities are used.

Lambda-Based Data Processing Architecture for Two-Level Load Forecasting in Residential Buildings

Building energy management systems (BEMS) have been intensively used to manage the electricity consumption of residential buildings more efficiently. However, the dynamic behavior of the occupants introduces uncertainty problems that affect the performance of the BEMS. To address this uncertainty problem, the BEMS may implement load forecasting as one of the BEMS modules. Load forecasting utilizes historical load data to compute model predictions for a specific time in the future. Recently, smart meters have been introduced to collect electricity consumption data. Smart meters not only capture aggregation data, but also individual data that is more frequently close to real-time. The processing of both smart meter data types for load forecasting can enhance the performance of the BEMS when confronted with uncertainty problems. The collection of smart meter data can be processed using a batch approach for short-term load forecasting, while the real-time smart meter data can be processed for very short-term load forecasting, which adjusts the short-term load forecasting to adapt to the dynamic behavior of the occupants. This approach requires different data processing techniques for aggregation and individual of smart meter data. In this paper, we propose Lambda-based data processing architecture to process the different types of smart meter data and implement the two-level load forecasting approach, which combines short-term and very short-term load forecasting techniques on top of our proposed data processing architecture. The proposed approach is expected to enhance the BEMS to address the uncertainty problem in order to process data in less time. Our experiment showed that the proposed approaches improved the accuracy by 7% compared to a typical BEMS with only one load forecasting technique, and had the lowest computation time when processing the smart meter data.

Targeting Bacteria and Methanogens To Understand the Role of Residual Slurry as an Inoculant in Stored Liquid Dairy Manure.

Microbial communities in residual slurry left after removal of stored liquid dairy manure have been presumed to increase methane emission during new storage, but these microbes have not been studied. While actual manure storage tanks are filled gradually, pilot- and farm-scale studies on methane emissions from such systems often use a batch approach. In this study, six pilot-scale outdoor storage tanks with (10% and 20%) and without residual slurry were filled (gradually or in batch) with fresh dairy manure, and methane and methanogenic and bacterial communities were studied during 120 days of storage. Regardless of filling type, increased residual slurry levels resulted in higher abundance of methanogens and bacteria after 65 days of storage. However, stronger correlation between methanogen abundance and methane flux was observed in gradually filled tanks. Despite some variations in the diversity of methanogens or bacteria with the presence of residual slurry, core phylotypes were not impacted. In all samples, the phylum Firmicutes predominated (∼57 to 70%) bacteria: >90% were members of ClostridiaMethanocorpusculum dominated (∼57 to 88%) archaeal phylotypes, while Methanosarcina gradually increased with storage time. During peak flux of methane, Methanosarcina was the major player in methane production. The results suggest that increased levels of residual slurry have little impact on the dominant methanogenic or bacterial phylotypes, but large population sizes of these organisms may result in increased methane flux during the initial phases of storage.IMPORTANCE Methane is the major greenhouse gas emitted from stored liquid dairy manure. Residual slurry left after removal of stored manure from tanks has been implicated in increasing methane emissions in new storages, and well-adapted microbial communities in it are the drivers of the increase. Linking methane flux to the abundance, diversity, and activity of microbial communities in stored slurries with different levels of residual slurry can help to improve the mitigation strategy. Mesoscale and lab-scale studies conducted so far on methane flux from manure storage systems used batch-filled tanks, while the actual condition in many farms involves gradual filling. Hence, this study provides important information toward determining levels of residual slurry that result in significant reduction of well-adapted microbial communities prior to storage, thereby reducing methane emissions from manure storage tanks filled under farm conditions.

Sorption of uranyl ions on TiO2: Effects of pH, contact time, ionic strength, temperature and HA

Sorption of U(VI) onto TiO2 as functions of pH, ionic strength, contact time, soil humic acid (SHA), solid-to-liquid ratio and temperature was studied under ambient conditions using batch and spectroscopic approaches. The sorption of U(VI) on TiO2 was significantly dependent on pH and ionic strength. The presence of SHA slightly enhanced the sorption of U(VI) on TiO2 below pH4.0, while it inhibited U(VI) sorption in the higher pH range. U(VI) sorption on TiO2 was favored at high temperatures, and the sorption process was estimated to be endothermic and spontaneous. Reduction of U(VI) to lower valent species was confirmed by X-ray photo-electron spectroscopy analysis. It is very interesting to find that U(VI) sorption on TiO2 was promoted in solutions with higher back-ground electrolyte concentrations. In the presence of U(VI), higher back-ground electrolyte made more TiO2 particles aggregate through (001) facets, leading more (101) facets to be exposed. Therefore, the reduction of U(VI) was enhanced by the exposed (101) facets and more U(VI) removal was observed.

Detection of transient signals based on the tricepstrum

In this paper, the problem of detecting transient signals of unknown waveforms and arrival times embedded in white Gaussian noise is addressed. The use of the cepstrum coefficients of the fourth order correlations of the transient signal for forming a detection statistic is demonstrated. It is considered both a batch and adaptive approach for the detection of the signal which is assumed to satisfy a linear constant coefficient difference equation. The adaptive approach is a least squares realization based on Q–R decomposition of the fourth order statistics matrix involved in the computation of the cepstrum coefficients. Their performance is compared in terms of probability of detection and probability of false alarm with the conventional energy detector by means of Monte-Carlo simulations and improved performance is demonstrated. It is shown that the adaptive approach allows for detection of short length transients which are of unknown arrival times using a single data record even before the whole amount of data becomes available. The Q–R decomposition offers good numerical properties when fourth order statistics are involved and in contrast to the energy detector the proposed receiver has false alarm rate independent of the noise if it is of zero mean and independent, identically, distributed (i.i.d).

Competitive adsorption of U(VI) and Co(II) on montmorillonite: A batch and spectroscopic approach

Abstract The competitive adsorption of U(VI) and Co(II) on montmorillonite (Mt) was investigated by batch and spectroscopic techniques. The batch experiments indicated that competitive adsorption of U(VI) and Co(II) on Mt decreased with increasing ionic strength at low pH conditions, whereas no effect of ionic strength was observed at high pH, indicating that the outer-sphere and inner-sphere surface complexation dominated the competitive adsorption of U(VI) and Co(II) at low and high pH, respectively. In the binary-solute systems, Co(II) slightly inhibited U(VI) adsorption at low pH, whereas no effect of Co(II) on U(VI) adsorption was observed at high pH. The adsorption kinetics of U(VI) and Co(II) on Mt at single-solute systems were both well described by pseudo-second-order kinetic model due to the higher correlation coefficient (R2 > 0.9999). The adsorption isotherms of U(VI) on Mt can be satisfactorily fitted by Freundlich model, while Langmuir model gave the better fits for Co(II) adsorption. The maximum adsorption capacities of U(VI) and Co(II) on Mt calculated from Langmuir model were 4.57 and 6.92 mg·g−1 at pH 5.0 and 323 K, respectively. According to XPS analysis, hydroxyl groups of Mt were responsible for the competitive adsorption of U(VI) and Co(II). The thermodynamic parameters suggested that both the adsorption of U(VI) and Co(II) on Mt were spontaneous and endothermic processes. These observations are crucial for understanding the fate and transport of radionuclides on the clay mineral in the natural environment.

Accelerating the SVD bi-diagonalization of a batch of small matrices using GPUs

The acceleration of many small-sized linear algebra problems has become extremely challenging for current many-core architectures, and in particular GPUs. Standard interfaces have been proposed for some of these problems, called batched problems, so that they get targeted for optimization and used in a standard way in applications, calling them directly from highly optimized, standard numerical libraries, like (batched) BLAS and LAPACK. While most of the developments have been for one-sided factorizations and solvers, many important applications – from big data analytics to information retrieval, low-rank approximations for solvers and preconditioners – require two-sided factorizations, and most notably the SVD factorization. To address these needs and the parallelization challenges related to them, we developed a number of new batched computing techniques and designed batched Basic Linear Algebra Subroutines (BLAS) routines, and in particular the Level-2 BLAS GEMV and the Level-3 BLAS GEMM routines, to solve them. We propose a device functions-based methodology and big-tile setting techniques in our batched BLAS design. The different optimization techniques result in many software versions that must be tuned, for which we adopt an auto-tuning strategy to automatically derive the optimized instances of the routines. We illustrate our batched BLAS approach to optimize batched SVD bi-diagonalization progressively on GPUs. The progression is illustrated on an NVIDIA K40c GPU, and also, ported and presented on AMD Fiji Nano GPU, using AMD's Heterogeneous–Compute Interface for Portability (HIP) C++ runtime API. We demonstrate achieving 80% of the theoretically achievable peak performance for the overall algorithm, and significant acceleration of the Level-2 BLAS GEMV and Level-3 BLAS GEMM needed compared to vendor-optimized libraries on GPUs and multicore CPUs. The optimization techniques in this paper are applicable to the other two-sided factorizations as well.

Measuring Plant-Available Mg, Ca, and K Pools in the Soil—An Isotopic Dilution Assay

In many forest ecosystems, plant-available pools of Mg, Ca, and K are assumed to be stored in the soil as exchangeable cations adsorbed on the cation exchange complex (exchangeable pools). However, between soil minerals and exchangeable cations exists a gradient of Mg, Ca, and K storage forms that have not been fully characterized and may play an important role in plant nutrition and biogeochemical cycles. We hypothesize that sources of Mg, Ca, and K in the soil other than the conventionally measured exchangeable pools are plant-available on very short time scales (<1 day). In the present study, we developed and applied an isotopic dilution technique using the stable isotopes 26Mg, 44Ca, and 41K to trace and quantify the pools of Mg, Ca, and K (isotopically exchangeable pools) in the soil of a hardwood forest that contribute directly to equilibrium processes between the soil water and the soil. We characterize the equilibrium between the soil and soil solution using both a batch approach and a flow-through approach in order (i) to develop and determine the best routine method to measure the isotopically exchangeable pools and (ii) to further the characterization of the forms of storage of Mg, Ca, and K in the isotopically exchangeable pools. We first show that the flow-through reactor approach (equilibrium in unsaturated soil columns) is the most adequate to measure the isotopically exchangeable pools with the fewest equilibrium disturbances. We then show that isotopically exchangeable pools of Mg, Ca, and K are greater than traditionally measured exchangeable pools. The isotopically exchangeable pools of Mg, Ca, and K are mainly composed of traditionally measured exchangeable pools (88.8–98.5% for Mg, 74.7–97.7% for Ca, and 68.7–77.1% for K) but are also composed of pools extracted with the Tamm reagent (oxalic acid, pH 3) and nitric acid (1 mol·L–1): 1.5–11.2% for Mg, 2.3–25.3% for Ca, and 22.9–31.3% for K. Storage forms of Mg, Ca, and K in the isotopically exchangeable pool could include chelation with soil organic matter, retention on soil aluminum and iron oxides and hydroxides through phosphate and/or organic acid bridges and site-specific adsorption. The isotopic dilution method is a relevant tool to quantify the plant-available pools of Mg, Ca, and K on short time scales (source and sink pools) and is a very promising approach to characterize and quantify the processes responsible for the depletion and/or replenishment of these pools over longer time scales.

An overview of incremental feature extraction methods based on linear subspaces

With the massive explosion of machine learning in our day-to-day life, incremental and adaptive learning has become a major topic, crucial to keep up-to-date and improve classification models and their corresponding feature extraction processes. This paper presents a categorized overview of incremental feature extraction based on linear subspace methods which aim at incorporating new information to the already acquired knowledge without accessing previous data. Specifically, this paper focuses on those linear dimensionality reduction methods with orthogonal matrix constraints based on global loss function, due to the extensive use of their batch approaches versus other linear alternatives. Thus, we cover the approaches derived from Principal Components Analysis, Linear Discriminative Analysis and Discriminative Common Vector methods. For each basic method, its incremental approaches are differentiated according to the subspace model and matrix decomposition involved in the updating process. Besides this categorization, several updating strategies are distinguished according to the amount of data used to update and to the fact of considering a static or dynamic number of classes. Moreover, the specific role of the size/dimension ratio in each method is considered. Finally, computational complexity, experimental setup and the accuracy rates according to published results are compiled and analyzed, and an empirical evaluation is done to compare the best approach of each kind.

Високопродуктивний пакетний добуток матриць афінних перетворень за допомогою Android NDK та JNI

This work contains description of high-performance batched affine matrices multiplication approach in scope of An-droid NDK and JNI. Approach is based on combination of separate technics and practices. Android Native Development Kit and Java Native Interfaces were used to develop and run high-performance code via Java Virtual Machine. This approach overhead was minimized with input data batching for computation. Compile-time optimization approaches were used to speed-up C/C++/Java based solutions. As a development, analysis and research result precise performance benchmarks were made. Output solution was compared with analogues, which are using limited sets of applied techniques, features and approaches. Charts were built on top of that results. Appropriate decisions were made. Problems in programming 2018; 1: 36-45

Organocatalyzed Decarboxylative Trichloromethylation of Morita-Baylis-Hillman Adducts in Batch and Continuous Flow.

Two protocols for the organocatalyzed decarboxylative trichloromethylation of Morita-Baylis-Hillman (MBH) substrates have been developed. Applying sodium trichloroacetate, as the trichloromethyl anion precursor, in combination with an organocatalyst and acetylated MBH-alcohols, the desired trichloromethylated products were obtained in good yields at room temperature in batch. The method was next extrapolated into a two-step continuous flow protocol, starting directly from the MBH alcohols, in combination with tributylamine acting both as base and catalyst. The flow process proved superior to the batch approach, reducing the reaction time from 16 hours to only 20 minutes, with increased yields for all investigated entries. Two examples were also taken to scale-up in flow producing more than 10 grams of both trichloromethylated targets. Finally, substitution of the organocatalyst to (DHQ)2 PHAL or (DHQD)2 PHAL induced chiral transfer to the generated stereocenter in the reaction attaining selectivities with nearly 90 % ee.

Exploring the relationship between Th(IV) adsorption and the structure alteration of phlogopite

Abstract The adsorption of Th(IV) and the structural transformation of phlogopite were explored using batch, extraction, and spectroscopic approaches in this study. The batch experiments showed that ion exchange, surface complexation, and precipitation were predominant for Th(IV) adsorption on phlogopite under the observed pH ranges. XRD patterns showed that phlogopite was indeed transforming into the vermiculite- and/or montmorillonite-like minerals during Th(IV) adsorption, which indicated that the adsorption of Th(IV) could enhance the weathering process of phlogopite. Such weathering process could be attributed to the intercalation of hydrated Th(IV) ions instead of K+ ions locating at the interlayer sites, and then induced the expansion of the interlayer with a d-spacing of ~ 14 A. Higher temperature and initial Th(IV) concentration were benefit for the structural transformation, but the presence of alkaline cations inhibited the expansion process following the order of Rb+ > K+ > Na+ ≈ Li+, which is strictly consistent with their hydrated radii and energy. Fulvic acid (FA) affected the distribution of Th(IV) on phlogopite to a large extent, which induced different adsorption mechanisms for Th(IV) and the alteration process of phlogopite structure. The results clearly showed a positively relationship between the immobilization ability for Th(IV) and alteration degree of phlogopite structure, which is very important to understand the mutual interaction of radionuclides and micaceous minerals in the environment.

A Comparison between Continuous and Batch Processes to Capture Aldehydes and Ketones by Using a Scavenger Resin

A low-cost isoniazid-based scavenger resin was previously prepared and successfully used to capture aldehydes and ketones in solution by using a batch approach. The present work compares batch and continuous processes in relation to the preparation of the scavenger resin and the effectiveness of cyclohexanone scavenging. The batch preparation of the resin proved to be the better choice because the resin produced in the continuous flow suffered deformation on its surface that prevented access to the active sites in the internal porous polymer matrix. Conversely, the continuous-flow approach presented higher capturing yields than the batch. The optimized scavenging process was successfully used to remove formaldehyde present in a solution and the carbonyl compounds present in an ethanolic extract of Cymbopogon citratus.

Biosorption performance evaluation of heavy metal onto aerobic granular sludge-derived biochar in the presence of effluent organic matter via batch and fluorescence approaches

In present study, the biosorption process of Cu(II) onto aerobic granular sludge-derived biochar was evaluated in the absence and presence of effluent organic matter (EfOM) by using batch and fluorescence approaches. It was found that EfOM gave rise to enhancement of Cu(II) removal efficiency onto biochar, and the sorption data were better fitted with pseudo-second order model and Freundlich equation, in despite of the absence and presence of EfOM. According to excitation-emission matrix (EEM), EfOM was mainly comprised by humic-like substances and fulvic-like substances and their intensities were reduced in the addition of biochar and Cu(II) from batch biosorption process. Synchronous fluorescence spectra coupled to two-dimensional correlation spectroscopy (2D-COS) further implied that a successive fluorescence quenching was observed in various EfOM fractions with the increasing Cu(II) concentration. Moreover, fulvic-like fraction was more susceptibility than other fractions for fluorescence quenching of EfOM.

Natural soil mediated photo Fenton-like processes in treatment of pharmaceuticals: Batch and continuous approach

This paper manifests the potential viability of soil as a cost-free catalyst in photo-Fenton-like processes for treating pharmaceuticals at large scale. Naturally available soil without any cost intensive modification was utilized as a catalyst to degrade pharmaceuticals, specifically ornidazole (ORZ) and ofloxacin (OFX). Soil was characterized and found enriched with various iron oxides like hematite, magnetite, goethite, pyrite and wustite, which contributes toward enhanced dissolution of Fe3+ than Fe2+ in the aqueous solution resulting in augmented rate of photo-Fenton reaction. The leached iron concentration in solution was detected during the course of experiments. The degradation of ORZ and OFX was assessed in solar induced batch experiments using H2O2 as oxidant and 95% ORZ and 92% OFX removal was achieved. Elevated efficiencies were achieved due to Fe2+/Fe3+ cycling, producing more hydroxyl radical leading to the existence of homogeneous and heterogeneous reactions simultaneously. The removal efficiency of solar photo-Fenton like process was also compared to photo-Fenton process with different irradiation sources (UV-A and UV-B) and were statistically analysed. Continuous-scale studies were conducted employing soil either in the form of soil beads or as a thin layer spread on the surface of baffled reactor. Soil beads were found to have satisfactory reusability and stability. 84 and 79% degradation of ORZ and OFX was achieved using soil as thin layer while with soil beads 71 and 68% degradation, respectively. HPLC and TOC study confirmed the efficient removal of both the compounds. Toxicity assessment demonstrates the inexistence of toxic intermediates during the reaction.

Value assessment method for expansion planning of generators and transmission networks: a non-iterative approach

In the context of efficient generation expansion planning (GEP) and transmission expansion planning (TEP), value assessment method (VAM) is the critical topic to discuss. Presently, two well-known VAMs, min-cut max-flow (MCMF) and load curtailment strategy (LCS), are used for GEP and TEP. MCMF does not follow electrical laws and is unable to calculate congestion cost (CC) and re-dispatch cost (RDC). LCS calculates both, but in iterative way, thus takes a long time to provide solution. In the constrained network, multiple quantities like demand/energy not served (D/ENS) and generation not served (GNS), wheeling loss (WL), CC and RDC are existing together and thus have to be calculated together to encounter the loss in all aspects. Existing methods show limitations in this regard and do not calculate all above described quantities simultaneously. Thus, in this paper, a non-iterative VAM (NVAM) is presented based on electrical laws, which calculates value of the present and the planned systems by incorporating all system quantities of D/ENS, GNS, WL, CC and RDC together. Due to non-iterative batch approach, it is quite faster compared to the above-mentioned traditional VAMs, i.e., MCMF and LCS. Furthermore, comparative results on IEEE-5 bus and IEEE-24 bus power systems show its higher efficiency. The MATLAB code of the introduced NVAM is provided in “Appendix” for further development by the researchers.

Learning Power Spectrum Maps From Quantized Power Measurements

Power spectral density (PSD) maps providing the distribution of RF power across space and frequency are constructed using power measurements collected by a network of low-cost sensors. By introducing linear compression and quantization to a small number of bits, sensor measurements can be communicated to the fusion center with minimal bandwidth requirements. Strengths of data- and model-driven approaches are combined to develop estimators capable of incorporating multiple forms of spectral and propagation prior information while fitting the rapid variations of shadow fading across space. To this end, novel nonparametric and semiparametric formulations are investigated. It is shown that PSD maps can be obtained using support vector machine-type solvers. In addition to batch approaches, an online algorithm attuned to real-time operation is developed. Numerical tests assess the performance of the novel algorithms.

Enhanced Crystallization of Lysozyme Mediated by the Aggregation of Inorganic Seed Particles

We show that aggregation plays a major role in seeded growth of protein crystals. The seeded batch approach provides the opportunity to set the starting conditions for protein crystallization by adding a defined amount of well-characterized seed particles. The experimental observations for tetragonal hen egg-white lysozyme (LSZ) confirm the concept of the oriented aggregation of larger building blocks to form a protein crystal. It was shown that the aggregation of the seed particles/bioconjugates is advantageous for the product quality in terms of larger and more defined LSZ crystals and in terms of accelerated reaction kinetics. We present a population balance (PB) model for the seeded batch crystallization of LSZ considering the aggregation of growth units to form protein crystals. For the modeling of crystal growth, evolving particle size distributions (PSDs) of agglomerating LSZ molecules were measured by dynamic light scattering (DLS). Moreover, the aggregation of seed particles in LSZ solutions unde...

Continuous Flow Synthesis of Anisotropic Cadmium Selenide and Zinc Selenide Nanoparticles

AbstractAnisotropic semiconductor nanoparticles find use in various applications ranging from electronics to photocatalysis and biolabeling. Batch synthesis methods typically used for their synthesis are often hampered by slow mixing, slow heating/cooling, and lack of batch‐to‐batch reproducibility, especially when scaling up. The modular continuous flow reactor reported here overcomes some of these challenges. It enables air‐sensitive syntheses at temperatures as high as 750 °C, supports rapid heating and cooling times (≈1 s or less), and enables syntheses that involve reagents that are viscous or even solid at room temperature. For validation we pursued two systems: the synthesis of (i) CdSe nanorods and bipods, and of (ii) ZnSe nanorods. Nanoparticles with low variance in quantum confined dimension (width) −16 % for CdSe and 11 % for ZnSe were obtained. For comparison, the same products were also synthesized using two batch approaches, hot‐injection and heat‐up, under similar conditions. The batch products were less uniform: 30 % variance in quantum confined dimension. Furthermore, the lack of precise temperature control in the batch processes resulted in CdSe nanorods with irregular‐shaped, jagged branches whereas the continuous process produced CdSe nanorods with uniform, straight branches. The modular continuous flow reactor design is suited for scale up, allowing working flow rates as high as ≈10 mL min−1, which translates into a production rate of ≈158 g day−1 for CdSe.