text
stringlengths 2
132k
| source
dict |
|---|---|
this way, the entire space can also be represented graphically as a contour plot; this allows the behavior of the factors to be more easily understood and interpreted, as well as allowing predicted response maxima to be easily identified. The response is often also mathematically transformed to give better predictability between two bounds, e.g., log 10 transformed so that a yield prediction cannot exceed 100%. 𝑦 = 𝑏 0 + 𝑏 1 𝑥 1 + 𝑏 2 𝑥 2 + 𝑏 12 𝑥 1 𝑥 2 + 𝑏 11 𝑥 1 2 + 𝑏 22 𝑥 2 2 (1) As described in the empirical model, it is not only the experimental factors that have an effect on the response but also the interactions between these factors; this is an attribute of DoE that is difficult to reproduce through methodologies that utilize human chemical intuition. (81,82) These interaction terms within the model indicate how the experimental factors influence reaction output, when other factors are also changed alongside them. When considering the optimization of some generic reaction while exploring the factors of reagent equivalents and temperature, it may be a significant factor in the modeling of the data to include an interaction term between these factors. In real terms, this interaction could indicate that a higher temperature has a positive influence on the reaction output only at higher reagent equivalents. Similarly, these factors may have an interaction term with themselves, as described by a squared term. For example, a squared temperature model term could indicate that temperature has a larger effect on the response at elevated temperatures, meaning that temperature has a nonlinear effect on the reaction within the explored parameter space. These interaction considerations can typically give a better description of the experimental data, as all synergistic effects between the factors
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
are incorporated into the model. However, all experimental factor and interaction terms can be added or removed from a DoE model depending on whether their contribution to the model is significant. To accurately isolate and determine interaction effects, specific DoE designs can be implemented. Within each design, the experimental factors are split into respective levels: these denote the degree of the experimental factor and are conventionally labeled between the minimum (−1) and maximum (+1). For example, a three-level design that is exploring reaction temperature (10–50 °C) would use the levels −1, 0, and 1, which correspond to 10, 30, and 50 °C, respectively. These levels are defined to ensure that the entire parameter space can be explored regardless of the factor range. Each experimental design may explore different levels, which also depend on the number of factors to be explored, in order to identify specific interaction effects and remove confounding (uncertainty). (83,84) Depending on the practicality of running experiments, it may be necessary to balance model accuracy with experimental measurements, as some designs that feature several factors may require many more experiments at different levels for a marginal increase in predictive power. For example, to estimate all terms for a model containing five experimental factors, a three-level full factorial design would require 246 experiments, while a face-centered central composite design would only require 29 with a minimal reduction in predictive accuracy. It is therefore important for the bench scientist to identify the optimal experimental design for their purposes to avoid conducting unnecessary experimental observations, leading to additional time and material costs. For more detailed information on DoE designs, also refer to Kumar and co-workers. (35) 3.2. Conclusions Chemical reaction optimization using design of experiments can be very powerful when attempting to identify regions of optimal parameter space. The methodology has
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
a relatively low expertise barrier-to-entry, given the advent of DoE software, and provides bench scientists with tools to identify significant experimental variables and model their data. (34) Although there are several options for DoE designs, the statistical knowledge necessary for chemists to select the correct experimental procedure (based on their needs for a given experimental outcome) and analyze the resulting data is low. Because of the relative ease of the technique, DoE can be easily taught to chemistry students and adopted for use in academic laboratories as intuition-based methods are superseded. (1,85,86) Furthermore, the ubiquity of DoE in process laboratories in industry highlights that these statistical methodologies must be taught to chemists; this will help to develop the skillsets of the students and familiarize them with common optimization protocols that they are likely to encounter in future. For further detailed reading on statistical modeling within DoE, refer also to Telford (87) and Severin and co-workers. (88) 4. Kinetic Modeling The use of kinetic modeling, featuring a mechanistic model rather than a statistical one, is also common for reaction understanding and optimization, especially in process laboratories in industry and academia. Kinetic models are constructed from a scientific understanding of the chemical process (89,90) rather than statistical relationships between experimental factors and outcomes. In contrast to DoE, undergraduate courses typically cover kinetic analysis in detail as part of their core physical chemistry modules, covering theories on collision, rate laws, and some basic physical-organic concepts. However, the more practical uses for kinetic analysis (reaction optimization, mechanism elucidation, etc.) are seldom explored by many chemists and are typically reserved for chemical and process engineers. This could be because of large expertise gaps experienced by chemists, or simply because they are unaware of the benefits of kinetic studies for their processes. Because of this distinct
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
knowledge gap, practical kinetic analysis is discussed in this review and how it relates to reaction optimization. When these kinetic models are constructed, they enable scientists to understand and simulate reactions to determine optimal regions of parameter space in silico. The physical modeling of a reaction typically features the rate laws of each individual chemical step and their corresponding rate constants. The main assumption when using a physical model is that the reaction kinetics follow the law of mass action; this states that the rate of an elementary reaction is directly proportional to the product of the concentrations of the reactants, raised to the power of their stoichiometric coefficients. (91) Written simply, this relates the order of a chemical species within an elementary step to the number of molecules reacting within that step. For the reaction shown in eq 2 (where x1–3 are chemical species and α/β/γ are their stoichiometric coefficients), based on the law of mass action, the reaction rate can be described by eq 3 and therefore eq 4. The kinetic rate constant, k, therefore determines the speed of the reaction as well as the reactant concentrations. 𝛼 𝑥 1 + 𝛽 𝑥 2 → 𝛾 𝑥 3 (2) rate ∝ [ 𝑥 1 ] 𝛼 [ 𝑥 2 ] 𝛽 (3) rate = 𝑘 [ 𝑥 1 ] 𝛼 [ 𝑥 2 ] 𝛽 (4) The law of mass action is applicable in almost all cases and is typically only inappropriate when concentrations of particular substrates are very low. More commonly encountered are very fast elementary reactions that occur in some processes, where it is more appropriate to describe an entire reaction with an observed rate rather than the combination of its individual elementary parts. This leads to circumstances where chemical species that are reported to have
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
a zero-order, second-order or even noninteger-order dependence, as it is much more practical to describe the physical model in this way. (92−97) The physical models generated from kinetic analysis contain ingrained chemical information that, unlike their empirical model counterparts, can be used to extrapolate reaction predictions outside of previously conducted experimental constraints. (98,99) The determination of reaction orders and rate laws within a model is used to optimize outputs but also to increase overall chemical understanding. As a result of impurity formation upon scale-up, Ashworth and co-workers studied the kinetics of the alkylation of the indolphenol, 16, with the chloropyrrolidine, 17, to form cediranib, 18, which is a pharmaceutical treatment for solid tumors (Scheme 4). (100) The authors found that overall second-order kinetics, consistent with a direct nucleophilic substitution between the anion of 16 and 17, were not observed in their experiments. Instead, overall first-order kinetics suggested an initial slow step to form another species. Further experimentation confirmed that 17 slowly reacted to form the azetidinium ion, 19, which then reacted quickly with 16 to obtain the desired product. This mechanistic understanding was achieved as a direct result of kinetic experiments, which led to further optimization of the overall process (solvent selection and base equivalents) for an increase in overall reaction yield. Scheme 4 Scheme 4. Alkylation of the Indolphenol, 16, with the Chloropyrrolidine, 17, to Form the Desired Cediranib Product, 18a aThis reaction was found to proceed via the azetidinium intermediate, 19, as a result of kinetic modelling. (100) As reaction progression can be described by the rate laws in eq 4, predictions of reactant/product concentrations can be made for any collection of reaction variables such as reaction time, reagent equivalents, and temperature. This allows response surfaces to be plotted in the same manner as with DoE studies, allowing
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
the visual determination of high-interest experimental parameters for process optimization. One example of this application is in the continuous-flow aqueous reduction of 4-nitrophenol, 20, to 4-aminophenol, 21, using gold nanoparticles (AuNPs) by Chamberlain and co-workers (Scheme 5). (101) This kinetic study related the surface area of AuNPs and flow residence time to reaction conversion, highlighting the optimal reaction conditions for the scale-up of the pharmaceutical building block, 4-aminophenol, as shown in Figure 5. Scheme 5 Scheme 5. Aqueous Reduction of 4-Nitrophenol, 20, to 4-Aminophenol, 21, Using Gold Nanoparticles (AuNPs) and NaBH4 (101) Figure 5 Figure 5. Kinetics-derived response surface for the conversion of 4-nitrophenol, 20, to 4-aminophenol, 21, when exploring the variables of residence time and AuNP surface area per liter. (101) 4.1. Conventional Approaches When fitting a physical model of rate laws to a chemical process, the reaction order and rate constants must be experimentally determined. For a mechanistic model that incorporates reaction temperature, activation energies must also be identified. Conventionally, mathematical transformations to concentration–time data are applied to identify both the reaction order and rate constants in one plot; this can then be repeated at multiple temperatures to obtain the activation energies for the reaction. (102−106)Figure 6 shows the most common, classical data transformations to obtain rate law information from kinetic experiments. (107) For each case, a linear fit to the transformed data indicates that the reaction order is correct, while the gradient of the fit gives information on the rate constant. For a unimolecular reaction, a zero-order reaction is confirmed with a linear fit to the concentration–time data (a) while a first-order reaction can be confirmed via a log-transformed plot of the data (b). For bimolecular reactions, if both reactants are the same, then an overall second-order reaction can be confirmed simply from the plotting of the
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
reciprocal of the concentration data (c). However, when both reactants are not the same, more complex plotting can confirm an overall second-order reaction (d). Additional modeling, such as the conventional modeling of enzymatic (and sometimes other catalytic) reactions, known as Michaelis–Menten kinetics, can also be applied; for more information refer to a recent review by Johnson. (108) Figure 6 Figure 6. Common conventional kinetic analysis techniques for the determination of: (a) unimolecular zero-order kinetics, (b) unimolecular first-order kinetics, (c) bimolecular second-order kinetics between the same reactants, and (d) bimolecular second-order kinetics between different reactants. These techniques are still employed regularly today in both teaching and research laboratories. For example, a study by Etua and co-workers (109) confirmed second-order reaction kinetics of a benzaldehyde oxidation, and work by Guégan and co-workers (110) confirmed first-order reaction kinetics in an epoxybutane polymerization; both cases are recent examples of how these conventional methodologies are still used. As can be seen from Figure 6d, the mathematics required to extract kinetic information grows much more complex as the reaction deviates further from very simple chemistry. Therefore, these conventional approaches become less appropriate when studying processes that are multistep, contain multiple reaction pathways, and have impurity formation as the mathematics becomes more inaccessible. In these complex cases, it is more common to use modern kinetic analysis techniques depending on the chemistry, such as reaction progress kinetic analysis, (111) or kinetic fitting software, such as Compunetics. (101,112) 4.2. Modern Techniques In every chemical case, regardless of the complexity of the process, it is possible to write coded solutions for kinetic analysis using rate constant solvers (using differential equation counterparts to eq 4). (113−115) Although this is a useful strategy that is often employed by engineers, (116−120) it is rarely conducted by chemists, as there is a high expertise
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
barrier to access this form of kinetic analysis, namely, coding and mathematical knowledge. Software solutions have been developed to aid in this kinetic model fitting (Compunetics, (121) Berkeley Madonna, (122) DynoChem, (123) COPASI (124)) that often require minimal coding expertise, which have been adopted by process chemists but have still had relatively low uptake within the wider chemistry community. One example of fitting a kinetic model using rate constant solvers was shown in our lab for the determination of the overall first-order reaction of alanine methyl-ester (Al-Me), 22, and 9-bromo-9-phenylfluorene (PfBr), 23, to form the protected amino acid (Pf-Al-Me), 24, as shown in Scheme 6. (92) This approach computationally identified the “best-fit” parameters for the rate constants and activation energies, with the fit to the experimental data shown in the combined plot in Figure 7. Scheme 6 Scheme 6. Reaction of Al-Me, 22, with PfBr, 23, to Form the Protected Amino Acid Pf-Al-Me, 24 (92) Figure 7 Figure 7. Kinetic profiles for three kinetic experiments at 30 °C, 35 and 40 °C, where red plots indicate PfBr concentrations and blue plots indicate Pf-Al-Me concentrations. At 30 °C: blue solid squares = experimental data, ─ = kinetic fit. At 35 °C: blue solid triangles = experimental data, - - - = kinetic fit. At 40 °C: blue solid circles = experimental data, ······ = kinetic fit. (92) Another common modern kinetic analysis technique used for the determination of physical models, particularly for catalytic processes, is reaction progress kinetic analysis (RPKA). (111) This methodology was pioneered by Blackmond and represents a systematic experimental procedure for kinetic analysis through sequential experimental/analytical steps involving reaction rate/time data. Rather than coding kinetic fitting solutions, this methodology features various plotting techniques and uses qualitative visual confirmations of overlaying graphs to determine catalyst/reactant orders, catalyst deactivation, and
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
product inhibition. Therefore, precise kinetic parameters cannot be elucidated, but the plots required are simple to construct and easy to interpret, which allows easy determination of this kinetic information. Because of its systematic approach to analysis, RPKA has been widely adopted in process chemistry settings and reported in several applications. (125−130) One recent example by Niemeyer and co-workers used RPKA in the regio- and stereoselective reduction of 2-phenylquinoline, 25, with the Hantzsch ester, 26, using the macrocyclic catalyst, 27, to yield the tetrahydroquinoline, 28, as shown in Scheme 7. (131) This work highlighted, using the RPKA methodology, that the process was first-order with respect to both substrates and the catalyst, and that there was no observed catalyst deactivation or product inhibition over time. Each piece of information gained from this study can thereby help in further process development, both for reaction condition optimization and scale-up suitability. Scheme 7 Scheme 7. Stereoselective Reduction of 2-Phenylquinoline, 25, to Yield the Tetrahydroquinoline, 28, Using the Hantzsch Ester, 26, and the Macrocyclic Catalyst, 27 (131) Variable time normalization analysis (VTNA) is another technique that, alongside RPKA, falls into the dubbed category of “visual kinetic analysis”. (132) VTNA can obtain the same chemical and physical model information as RPKA but does not require rate/time data; instead, VTNA can be used directly with concentration/time data, which requires fewer experiments to obtain and less data manipulation. (133,134) As VTNA requires only simple graphical transformations with easily obtained data, this methodology has also been reported in many process chemistry applications and chemical optimization studies. (135−138) One recent example by Carretero and co-workers showed how VTNA was utilized in the cobalt-catalyzed C–H functionalization of N-benzylpicolinamide, 29, with 4-octyne, 30, to yield the dihydroisoquinoline product, 31, as shown in Scheme 8. (139) This study showed that the reaction exhibited a
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
first-order dependence on the cobalt concentration, a zero-order dependence on the alkyne and a partial negative-order in the benzylamide concentration. The authors suggested that this partial negative-order finding was a result of off-cycle unproductive binding interactions with the catalyst, thereby decreasing the effective concentration of catalyst available. Therefore, for further optimization work and scale-up, this quantification of reaction orders has been crucial in showing that there must be an optimal benzylamide concentration range whereby productivity is maximized. Scheme 8 Scheme 8. Cobalt-Catalyzed C–H Functionalization/Alkyne Annulation Reaction of 29 with 30 to Form the Dihydroisoquinoline Product, 31 (139) 4.3. Outlook Kinetic modeling is a very powerful tool for the optimization of chemical processes and represents a much more systematic approach than traditional OFAT optimization. The mechanistic models used in kinetic analysis also provide chemical insights and scientific understanding that DoE does not but may also be more difficult to interpret and conduct experimentally as time-series data is paramount. Although software solutions have been created to lower the expertise barrier for kinetic analysis, the uptake of kinetic modeling techniques is still relatively low among chemistry researchers as the analysis is typically assigned to physical-organic or engineering colleagues, often unnecessarily as many reactions are simplistic and easy to fit a physical model to. Visual kinetic analysis, although a relatively new methodology, has helped to provide an accessible framework for chemists to obtain semiquantitative model information from their processes without the need for coding or software. More complex techniques that feature parameter estimation for kinetic modeling have not been covered, but interested readers are directed to reports on model-based design of experiments (MBDoE). (140−142) As robust physical models are often useful for optimization and necessary for scale-up, familiarity with these modeling techniques is very important as our laboratories become more interdisciplinary and connected. 5.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Self-Optimization Self-optimization is a modern approach to automating the discovery of optimal reaction conditions for chemical processes which does not require the determination of explicit mechanistic or empirical models. Self-optimization proceeds through iterative cycles of automated reaction execution, quantification, and algorithmic condition selection to efficiently identify optimal reaction conditions to maximize process metrics (yield, selectivity, etc.). Although self-optimization was initially applied to tuning analytical instruments as early as the 1970s, (143) DeMello and co-workers (144) first introduced the concept of self-optimization of chemical reactions in 2007, which led to further adoption by many other research groups in the following years. DeMello and co-workers focused on the synthesis of CdSe quantum dots, but subsequent works have applied self-optimization to a wide range of synthetic organic reactions including oxidation, (145) Diels–Alder, (146) methylation, (147) Paal–Knorr, (148) Suzuki–Miyaura cross-coupling, (149,150) hydrolysis, (151) and C–H activation. (152) Self-optimization is often conducted using automated reactors that can independently execute reactions at a specified set of reaction conditions. Automated analytical instruments then quantify the individual components of a reaction mixture, followed by an algorithmic suggestion of new reaction conditions based on previous data to improve key reaction outcomes. Self-optimization research can therefore often be divided into three subsections that map directly on to these three key aspects: development of automated reactors, development of automated analytical methodologies, and development of optimization algorithms (adapted for a specific chemical problem). One or more of these developments are typically reported in individual works in the literature. These ideas are shown conceptually in Figure 8. Figure 8 Figure 8. Three parts of a self-optimizing reactor are an automated reactor system, an analytical method, and an optimization algorithm. The primary goals of self-optimization are to reduce the number of experiments required to optimize a reaction and the experimental burden on the bench
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
scientist, saving time and money. Additionally, self-optimizing systems offer a more consistent way of generating data than human driven reaction optimization. This means that self-optimization has the potential to not only accelerate reaction optimization but also provide data that will enable predictive modeling in the future. Currently, self-optimization has been adopted by both academia and industry, although work from the former has been published more readily. 5.1. Automated Reactors for Self-Optimization Automated reactors must be able to receive and execute a set of reaction conditions without human intervention. Continuous flow and batch reactors have been applied to self-optimization, and prominent examples from the literature are herein discussed, but two example reactor setups are shown in Figure 9. Figure 9 Figure 9. Examples of automated reactors. (a) A bespoke automated flow reactor equipped with pumps, reactors, and analytical equipment. (153) (b) A commercial robotic liquid handler that can be utilized as an automated batch reactor (see section 6.2 for more details). 5.1.1. Automated Flow Reactors Most self-optimization studies in the literature have employed automated flow reactors. (154,155) Utilizing developments in continuous flow chemistry, these automated flow reactors use pumps to deliver solutions at a desired flow rate to a temperature-controlled reactor. By pumping solutions of starting materials and reagents at varying flow rates, precise stoichiometries and reaction times can be achieved. Automated flow reactors are attractive because a bespoke system can be quickly assembled using commercially available parts, or a complete system can be purchased from several specialized vendors. Building a bespoke system offers flexibility and lower costs, while complete systems enable faster (and often easier) application deployment. (156) An additional advantage of automated flow reactors is that the optimal conditions found in an automated flow reactor can be utilized for medium-to-large scale production (grams to kilograms per week) by either
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
running the reactor for an extended period (157) or numbering-up, (158) as discussed in section 7. However, three major challenges are faced by researchers using automated flow reactors for self-optimization. First, automated flow reactors can consume large amounts of starting material and solvent due to the need to flush the reactor when changing conditions. The standard heuristic is to wait at least two residence times (i.e., twice the amount of time necessary for material entering the reactor to exit) prior to initializing analytical measurements, so much of the reaction material is directed to waste. Second, changing categorical reagent conditions such as catalyst, base, or solvent is nontrivial in a standard automated flow reactor because each pump must be loaded with a particular reagent prior to reaction execution. (159) Third, automated flow reactors suffer from standard issues with flow chemistry, particularly reactor clogging due to precipitation of solids, (160) and limits in the maximum residence time of compact reactors. (161) To overcome challenges of material consumption and changing reagents in automated flow reactors, researchers have developed automated droplet flow reactors. (162) These reactors employ liquid handlers that transfer the individual components for a reaction into a sample loop prior to injecting them as a droplet into the reactor tubing. The droplets can be as small as several hundred microliters, resulting in large reaction material savings when compared with traditional continuous flow experimentation. For example, Jensen and co-workers demonstrated the self-optimization of a C–N cross-coupling reaction in an automated droplet flow reactor varying catalyst and base with a liquid handler (see Figure 10); their reactor consumed less than 200 mg of starting material. (163) Figure 10 Figure 10. Development of a Buchwald C–N cross coupling between p-tolyl triflate (32) and aniline (33) via self-optimization in an automated droplet flow reactor. (163) Three
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
continuous variables (residence time, base equivalents, and temperature) and two categorical variables (catalyst and base) were varied to maximize the yield of 4-(p-tolyl)morpholine (34). In the chart, each column contains data for a different catalyst and base combination, and the experiments in each column are shown left to right in the order they were selected by the optimization algorithm. Additionally, the color bar shows experiment selection order. A variety of technologies have been developed to address other standard issues with flow chemistry. Methods to prevent clogging include continuous stirred tank reactors (CSTRs) that can facilitate slurries (164) and reactors with baffles to improve mixing and reduce precipitation. (165) To overcome challenges with limited residence time in flow reactors, researchers have applied oscillatory droplet flow reactors that move droplets back and forth inside a fixed length tubing until the desired reaction time is achieved. (166−168) These oscillatory systems could, in theory, enable very long reaction times (hours), although in practice they have only been used for reactions with shorter reaction times (minutes). Overall, the field of flow chemistry offers a powerful set of tools for building automated reactors for self-optimization. By either building or buying automated flow reactors, research groups and industry can quickly access the basic laboratory tools needed for building self-optimizing systems. However, particular care must be taken for solid handling and the screening of categorical parameters (such as solvent, catalyst etc.) as vital modifications to existing equipment may be necessary. 5.1.2. Automated Batch Reactors Recent work has shown that automated liquid handling robots, historically deployed for high-throughput screening (HTS) in biological applications, can also be used in combination with high-throughput screening microplates to create automated batch reactors for self-optimization. (15,149) Using these liquid handling instruments, it is possible to screen many categorical variables, such as catalysts, ligands and
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
solvents. Hein and co-workers conducted self-optimization of a stereoselective Suzuki coupling reaction using a ChemSpeed liquid handling robot, varying the ligand, several stoichiometries, and temperature to maximize the formation of the E-product and minimize the Z-product, as shown in Figure 11. (149) This self-optimization campaign led to more than a 2-fold increase in yield of the E-product (30% nominal to 73% optimized) and a notable increase in the E/Z ratio (1.5:1 nominal to 2.5:1 optimized) within 161 experiments. Figure 11 Figure 11. Development of a stereoselective Suzuki coupling between sulfonate 35-E and boronic acid 36 to form 37-E and 37-Z via self-optimization in an automated batch reactor. In 161 experiments, the yield of 37-E was improved from 30% nominal to 70%, and the E/Z ratio from 1.5:1 to 2.5:1. AY represents assay yield. (149) As automated liquid handling robots are common among high-throughput experimentation (HTE) and process development groups in industry, automated batch reactors represent a significant opportunity for industrial adoption of self-optimization. Indeed, the few publicly available self-optimization studies that have been executed using automated batch reactors were from industrial-affiliated groups. HTE is further discussed, in detail, in section 6.1. 5.1.3. Analytical Techniques The most common analytical techniques used in self-optimization are chromatographic, either high performance liquid chromatography (HPLC) or ultrahigh performance liquid chromatography (UHPLC). (154) This trend is likely due to the ubiquity of HPLC and UHPLC instruments in chemical synthesis laboratories and the relative ease with which they can give quantitative data and be integrated into self-optimization systems. For automated flow reactors, HPLC sampling is achieved using a switching valve located at the outlet of the reactor, which can alternate between directing reaction material to waste during changes between conditions and sending aliquots to the HPLC instrument for analysis; this is referred to as online sampling, as
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
material is removed from the flow path for analysis. (169) This sampling is typically initialized based on the residence time of the reaction or, in the case of droplet flow reactors, an in-line UV cell. In addition to HPLC, gas chromatography (GC) has also been reported in the literature, but this is much less common. (152) In addition to online HPLC, there have been many reports on developing in-line analytical techniques, such as NMR, IR, UV, and FTIR. In-line NMR has been used to identify both known and novel products synthesized in self-optimizing reactors. (170−172) Additionally, in-line IR and FTIR has been used independently and in combination with NMR and HPLC for reaction quantification. (170) The advantage of in-line analytical techniques is they offer fast data feedback, enabling more rapid optimization. However, because the analysis is conducted on the crude reaction mixture, these complex mixtures can suffer from overlapping peaks which often makes quantification difficult. The supporting information of the review by Rincon and co-workers has a list of flow chemistry self-optimization studies and the on-line/in-line analytical techniques used. (154) 5.2. Optimization Algorithms In self-optimization, an efficient optimization algorithm is required to select new reaction conditions based on previous results. Chemical reactions can be viewed as mathematical functions that receive reaction conditions as input values and output reaction outcomes (e.g., product yield, selectivity, etc.). (173) This functional view of chemical reactions makes it clear why optimization algorithms, which find the optimal values of mathematical functions, can be used to optimize chemical processes. Optimization algorithms iteratively evaluate the output of the function at different input values until a maximum or, if desired, a minimum, is reached. In the case of chemical processes, these iterative evaluations correspond with intelligently suggested experiments to execute in the laboratory until a set of reaction conditions
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
are achieved that give the optimal desired output. 5.2.1. Local Optimization vs Global Optimization The two main classes of optimization algorithms are local and global optimization algorithms. Local optimization algorithms are designed to find the optimal values of a function closest to an initial guess. Therefore, if there is one optimal value, local optimization algorithms will likely find it, but if there are multiple optima, the success of a local optimization method is highly dependent on the initial guess. Example chemical applications of local optimization algorithms include the steepest descent algorithm, which chooses reaction conditions based on the most favorable gradient (direction) in design space to explore, (145) and the simplex algorithm, which explores the design space based on geometric transformations to exploit perceived favorable areas. (143,147) The challenge with local optimization algorithms is their dependence on the reaction conditions used to initialize the algorithm; if there are multiple regions of chemical space with local optima, the algorithm could potentially fail to find the best overall reaction conditions for the transformation, as illustrated in Figure 12. Figure 12 Figure 12. Example of a local optimization algorithm failing to find the global maximum of a function with multiple local optima, i.e., the algorithm finds a maximum peak but not the highest peak. X and Y are hypothetical experimental variables (e.g., temperature, reaction time), and the objective is the value that must be maximized (e.g., yield). Unfilled circles are function evaluations (i.e., experiments). Red indicates local maxima function value, while blue indicates local minima. Global optimization algorithms can identify the best value of a function independent of the initial guess but may require more experiments to obtain. Krishnadasan and several other researchers were the first to apply global optimization to self-optimization. (47,144,145) They used the Stable Noisy Branch and Fit (SNOBFIT), which,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
as its name suggests, relies on sequential branching and fitting steps. The algorithm begins by subdividing the optimization domain into boxes with one data point each (i.e., branching) and subsequently builds full quadratic models for each box and its nearest neighbors (i.e., fitting). (174) SNOBFIT achieves global optimization by sampling evenly from the complete reaction condition input space. This has been shown to result in a larger number of experiments than local optimization in exchange for a higher likelihood of finding the optimum for difficult problems (i.e., challenging nonlinear optimization tasks). (145) SNOBFIT also allows researchers to conduct experiments in batches (i.e., the algorithm makes multiple experimental requests at one time). More recently, researchers have started to apply Bayesian optimization within chemistry optimization problems. (15,175−178) This class of optimization algorithm is a subset of Bayesian statistics, which uses probability to express certainty in future outcomes based on past observations. In the case of reaction optimization, when an experimental iteration is completed, a probabilistic model is trained to predict its reaction outcome (e.g., yield) given the reaction conditions. Then, the Bayesian optimization algorithm chooses experiments that balance further exploration of chemical space and exploitation of the known best performing conditions. (179) Bayesian optimization offers a principled and efficient way to apply global optimization to chemical reactions that is more intuitive and often more effective than other global optimizers. 5.2.2. Categorical Variable Optimization Each of the aforementioned algorithms only utilize continuous input variables (e.g., temperature, concentration, residence time), but chemistry problems often have categorical variables also (e.g., solvent, ligand, etc.). To address this algorithmic limitation, Jensen presented a branch-and-fit algorithm that eliminates possible values for categorical variables (e.g., particular solvents, ligands, etc.) that lead to poor reaction performance. (150,159,180,181) The algorithm fits predefined mathematical models for each combination of categorical variables and,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
once the best categorical combination has been identified, it suggests further experiments to improve the fit of the model. However, this approach requires users to specify a kinetic model a priori, which may be difficult when a full reaction mechanism is not known, as discussed in section 4. Furthermore, because no model is built to describe the relationship between categorical variables, insights cannot be easily drawn about the relationship between different catalysts and bases. This could explain the poor performance seen when this algorithm was used in the aforementioned example of C–N cross coupling (see Figure 10). Alternatively, it is possible to use optimization algorithms that inherently work with categorical variables. In recent studies, Bayesian optimization algorithms were adapted to automatically learn the relationship between categorical variables from experimental data. (177,178) These algorithms tended to perform slightly better in finding optimal reaction conditions than the aforementioned strategies that could not learn a relationship between categorical variables and may be a large research area of interest in future. Categorical variables can be explored more easily by quantifying them with various “continuous variable” chemical descriptors, as highlighted in section 6.2. However, more recently, work by Bourne and co-workers (177,182) showed the effectiveness of their mixed variable multiobjective optimization (MVMOO) algorithm in optimizing categorical/continuous variables without chemical descriptors. In their respective works, they show the use of novel distance metrics based upon Gower similarities that reduce this necessity in both simulated and experimental case studies. 5.2.3. Multiobjective Optimization Optimization problems in chemistry often involve trade-offs between multiple competing objectives, such as balancing high process yields with low costs, so optimization algorithms need to be able to consider and weight these objectives to find optimal solutions. For example, Jensen and co-workers optimized catalyst turnover number (ratio of the rate of product formation to catalyst
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
usage) with the constraint that yield must be greater than 90%. (150) This constraint was implemented to prevent the yield being maximized by simply adding higher loadings of an expensive catalyst. In other cases, researchers have optimized a weighted function of multiple objectives. (144,151,183) Both of these methods require the scientist to make an a priori judgment about the trade-offs between competing objectives and their respective importance. This prior judgment can be limiting and results in a single optimum point identified, whereas there are likely multiple optimal solutions depending on the weightings of the individual objectives. An alternative approach to considering these trade-offs is the use of multiobjective optimization algorithms. These algorithms explore the full set of trade-offs between multiple objectives. (175,184) These algorithms construct a Pareto front, which is defined as a set of points in which an improvement in one objective would result in a detriment to another. By presenting a spectrum of trade-offs, this multiobjective approach allows the scientist to select one optimal point with other softer constraints considered (e.g., budget constraints, manufacturability, downstream separation efficiency, etc.). Due to the need to optimize multiple variables and explore globally, multiobjective Bayesian optimization algorithms can be slower than gradient descent or single objective optimization and require further experimentation. Figure 13 shows an example of the multiobjective Bayesian optimization algorithm TSEMO (184) applied to a N-benzylation reaction. (175) The flow rate of α-methylbenzylamine 38, ratio of benzyl bromide 39 to 38, solvent flow rate, and temperature were modified to maximize production of 40, while minimizing formation of 41. After 20 experiments designed by Latin hypercube sampling (LHS), (185) a balanced random sampling technique, TSEMO quickly identified 58 further experimental conditions on or near the Pareto front. The Pareto front indicated that a 60 kg m–3 h–1 increase in space-time yield
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
(STY) would correspond with an approximate 10% increase in impurity yield, which the scientist can then consider based on the needs of the process. Figure 13 Figure 13. Multiobjective self-optimization of the N-benzylation of N-benzylation of α-methylbenzylamine 38 with benzyl bromide 39. (175) TSEMO (184) was used to maximize space–time yield of the desired 2° amine 40 and minimize production of the percent impurity of 3° amine 41. After 20 experiments designed by Latin hypercube sampling (LHS), (185) TSEMO quickly identified experiments on or near the Pareto front. 5.2.4. Benchmarking of Optimization Algorithms The self-optimization reports explored thus far primarily focus on single experimental case studies, implementing bespoke optimization strategies; this makes it difficult to compare the performance of optimization algorithms objectively. Therefore, recent work has aimed to develop chemically relevant optimization simulations of reactions, or benchmarks, so that these algorithms can be compared without the time and expense of laboratory experiments. (186−188) Benchmarking rarely finds algorithms that will work in all situations, but they can help filter out poorer-performing algorithms and compare the effect of small changes to algorithms for potential benefits. Furthermore, benchmarking studies can act as postverification of algorithms developed initially on real experiments, as it is possible to run large numbers of repeated simulations to understand the average behavior of these algorithmic techniques. 5.3. Future Directions Self-optimization has the potential to significantly accelerate reaction development by enabling autonomous optimizations of reaction conditions. However, currently, as highlighted by Hein and co-workers, automated reactors often require significant human intervention and adjustment to achieve high quality results. (156) Therefore, there are still open research questions tackling how to develop highly flexible automated reactors that can adapt to a wide range of chemistry without significant customization. There are also further necessities when conducting reactions that must be addressed, particularly steps
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
that are easy for humans but more complex for machines, such as phase separations, extractions, crystallizations, etc. Additionally, self-optimization has shown promise in the automated optimization of single reactions, but there is a wealth of reaction data available that current algorithms are unable to utilize. Very recent work has used transfer learning techniques to accelerate optimization by leveraging data from similar reaction optimization campaigns, but this has mainly been demonstrated in silico, (189,190) with one active learning example from Lapkin and co-workers with the focus of pH adjustment. (191) There is also a significant opportunity for benchmarking these algorithms on both in silico and real-life experimental case studies to see how well they generalize to all classes of reactions. 6. Data-Driven Optimization 6.1. High-Throughput Experimentation High-throughput experimentation (HTE) involves running multiple reactions in parallel, which is a useful technique for quickly exploring chemical space in a systematic and standardized manner. (192−200) Traditionally, this process has been employed in the pharmaceutical industry for the parallel synthesis/assessment of chemical compounds on a variety of scales. This includes small, focused arrays of compounds for exploring structure–activity relationships (SAR) around a hit compound of interest, and the synthesis of thousands of compounds to populate high-throughput screening (HTS) libraries. In recent academic settings, HTE has also shown utility in the discovery of new chemical reactions, as this technique is well-suited for the discovery of unexpected reagent combinations enabled by the large number of reactions that can be run in tandem. (201−205) For the purposes of this review, this section will focus on the application of HTE to accelerate the optimization of synthetic organic reactions. Herein, methods employed for the simultaneous exploration of chemical space using multifactorial optimization will be discussed, touching on the strengths and weaknesses of the different methods currently available to HTE practitioners.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
6.1.1. Multifactorial Optimization With HTE, a large proportion of chemical optimization space can be examined at once in an “all vs all” manner, where one exhaustive (full factorial) screen may help to identify optimal conditions much faster and more efficiently than performing reactions individually (e.g., using OFAT). However, reaction optimization using a HTE methodology requires time-intensive reaction design from the outset, as several categorical variables (e.g., catalyst, solvent, base, etc.) may be varied at once. (192) Commonly a fractional factorial approach is employed where a subset of variables are screened in a matrix array with all categorical variables compared against each other. Although this approach can be time-consuming to design and analyze, it is generally more cost-efficient than OFAT (see section 2) or other iterative optimization methods due to the miniaturized scale requiring less reaction material. An example HTE workflow is shown in Figure 14, where a chemical process is explored with each possible variation of 12 catalysts, 4 bases, and 2 solvents. HTE is a powerful resource that has seen widespread use in the pharmaceutical industry. It is employed in different areas of chemical development throughout the drug discovery process, from early stage preclinical development up to process optimization for clinical development, in-human testing, and subsequent market release. (193) Figure 14 Figure 14. Schematic of a typical HTE workflow where a particular chemical process is optimized with respect to 12 catalysts, 4 bases, and 2 solvents. 6.1.2. Reaction Miniaturization During the lead optimization phase of a drug discovery campaign, elaborated compounds (leads) produced in the development of a clinical candidate can be very time and resource intensive to make. High-throughput optimization techniques can be useful to optimize potentially intractable reactions such as transition-metal catalyzed cross-couplings of highly functionalized drug-like compounds, however, these elaborate pharmaceutical intermediates often cost more gram-for-gram
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
than precious transition-metal catalysts that may be employed in their synthesis. By downscaling chemical reactions into a miniaturized format, one can increase the number of data points gained during a HTE screen without the associated impact on material cost. For decades, biological assays have been routinely performed in plate-based formats on micro- or nanoliter scales in a parallelized high-throughput manner. (206) This miniaturization approach not only allows for a greater number of experiments to be run in a material sparing fashion, but is also highly appropriate for automation. Hence, a number of robotic platforms have been developed by vendors (e.g., Tecan, Hamilton, Beckman Coulter Echo, SPTLabtech Mosquito) to facilitate the execution of multiple experiments at once in a Society for Biomolecular Screening (SBS) footprint microtiter plate (MTP). This data capture workflow was further streamlined with the advent of analytical systems equipped with autosamplers that can sample directly from the same footprint MTPs. The standardization of these plates to a universal footprint ensures interchangeability between different robotic systems, and this has undoubtedly had an enormous impact on the rate with which experiments can be performed and analyzed within the biological sciences. Organic chemistry has been slow to adopt MTPs as, unlike biological experiments which are run under aqueous conditions at physiological temperature (37 °C), synthetic organic chemistry experiments employ a wide range of temperatures (cryogenic to elevated, e.g., −78 to 150 °C) and differing polarity solvents, some of which are incompatible with plastic MTPs. Accordingly, when chemists perform reactions in parallel, they tend to miniaturize their chemistry from round-bottomed flasks (RBFs) to 96-well plates and seldom downscale further to 1536-well MTPs. Table 1 explores some of the chemical limitations which need to be taken into consideration when miniaturizing chemical reactions into these plate formats. Table 2 explores how translating chemistry from
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
common synthetic apparatus such as round-bottomed flasks into plate-based formats means that chemical synthesis can be economized, thereby increasing the density of information obtained from the same amount of material. Table 1. Chemical Reaction Limitations with Relation to Current Miniaturization Technologiesa a RBF, round-bottomed flask; MTP, microtiter plate; PFA, perfluoroalkoxy alkane; COC, cyclic olefin copolymer; PP, polypropylene; DCM, dichloromethane. *With specialized glassware or additional apparatus attached. ‡Vessel or equipment dependent. §Depending on amount required. ◇Reagent dependent. Table 2. Physical and Pragmatic Reaction Constraints Using Miniaturization Technologiesa a RBF: round-bottomed flask; MTP: microtiter plate. 6.1.3. Platforms for HTE Optimization Platforms used for automated HTE can be broadly characterized into the following three formats: (1) plate-based reactors where individual reactions are either performed directly in the plate wells or within a glass vial insert, (2) flow-based platforms employing continuous (31,207) or radial (208) flow techniques, or (3) microfluidic (and droplet) reactors. (204) These platforms have been utilized for a number of different applications including miniaturized synthesis and reaction discovery, but the examples shown herein are those which have been applied to reaction optimization only. (201−203,209) 6.1.3.1. Microscale Plate-Based Optimization Microscale plate-based chemistry involving 96 individual glass vials housed within a metal block (Tables 1 and 2) is one of the most routinely used methods for HTE reaction optimization. The apparatus required is relatively low cost, and liquids can be rapidly dosed using multichannel displacement pipettes and solid reagents (either as pure material or as ChemBeads) (210−213) can be weighed out manually with small spatulas or using 3D-printed scoops, (196,210) which can somewhat streamline the meticulous and demanding process. Unlike other nanoscale approaches in 1536 MTP, these 96-well reactors now allow an experimentalist to operate on a scale which is compatible with solid handling of chemical substances. Although the addition of solids to
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
a microscale reactor can be more time-consuming than the addition of liquids or stock solutions, the ability to use solids is important as it broadens the variety of chemical reactions that can be performed, as not all chemical reagents can be effectively dispensed as stock solutions due to issues such as heterogeneity and chemical instability. Another key benefit of working on microscale in 96-well reaction blocks is that the reactors themselves are suitable for a wider range of reactions than can be performed in plastic MTP reactors, the glass vial inserts in the 96-well plates exhibit good chemical tolerance akin to the traditional round bottomed flask, furthermore, they can be heated or irradiated, and the contents of the vials shaken on an orbital shaker or magnetically stirred with the addition or small magnetic fleas. Although automation is not necessary for reaction implementation using this setup (unlike the 1536-well MTP approach), the SBS footprint of these reactors means that they are compatible with a variety of different platforms for automated liquid and solid handling (214,215) to streamline reaction implementation and optimization. The flexibility and accessibility of this 96-well approach has led to widespread usage with multiple literature examples for HTE optimization (216,217) of a variety of reaction classes with selected examples shown in Table 3. Table 3. Examples of Different Reaction Classes Optimized Using the 96-Well Glass Inert/Metal Reaction Block Approach (218−226) 6.1.3.2. Nanoscale Plate-Based Optimization Nanoscale 1536-well MTP chemistry, recently referred to as ultraHTE, (198) is a commonly reported HTE method which can facilitate the collection of thousands of data points within a multiparameter optimization. (227−229) Due to the minute volumes of material employed in this approach, access to specialized liquid handling robotics is required. Examples include SPT Labtech’s Mosquito positive displacement multichannel pipetting system, which can dispense and aspirate
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
volumes in the range of 25 nL to 1.2 μL, or Beckman’s Echo liquid handler, which can dispense volumes in the range of 2.5 nL to 5 μL using Echo Acoustic Technology. These platforms are increasingly used for nanoscale (or even picoscale) synthesis of pharmaceutically relevant compounds with examples of also incorporating nanoscale biological screening in a “direct-to-biology” approach. (230−239) The Mosquito/1536-well MTP platform pioneered by Merck Research Laboratories is prevalently reported in the literature for nanoscale optimization. Although this approach has limitations related to the breadth of chemical reaction types that can be performed in 1536-well MTPs (Table 1), ultraHTE has been successfully employed on multiple occasions for the multiparameter optimization of a variety of different reaction types from the medicinal chemistry toolbox. (17,240−242) These reactions include Suzuki cross-couplings, reductive aminations, N-alkylations, nucleophilic aromatic substitutions (SNAr), etc., as well as transition-metal catalyzed couplings on pharmaceutically relevant molecules such as the Pd-catalyzed Buchwald–Hartwig amination, metallophotoredox C–N and C–O bond formations, and C–H functionalizations to furnish C(sp2)–C(sp3) bonds (Table 4). Table 4. Examples of Different Reaction Classes Optimized Using the Mosquito/1536 MTP UltraHTE Platforma (228,229,243−247) a ‡, in duplicate; *, in quadruplicate. 6.1.3.3. HTE Plate Analysis When using HTE to optimize chemical reactions, it is important to not only consider the practicalities of how to perform multiple reactions in tandem, but also how to analyze and deconvolute the reaction outcome in a similarly high-throughput manner. Without careful consideration of the overall experimental and analytical workflow, a bottleneck can occur. (248,249) In the past decade, several ground-breaking developments in the analytical sciences have occurred which now permit the ultrafast analysis of high-throughput reaction screening at increasingly impressive speeds. For example, techniques like matrix-assisted laser desorption/ionization (MALDI), (243,250,251) desorption electrospray ionization (DESI), (252−254) and acoustic ejection MS (AE-MS) (233,255−257) have been reported for
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
the high-throughput data acquisition of 1536 reactions in under 10 min and requiring only nanolitre volumes of crude reaction mixtures. These techniques are rapid, and in some cases the analysis can be performed directly from a 1536-well MTP, however, the equipment required can be expensive. Other options for rapid analysis which can be performed on standard UHPLC hardware interfaced with a single-quadrupole mass spectrometer is the Multiple Injections in a Single Experimental Run (MISER) technique developed by Merck. (227,258) This flow-injection analytical method injects multiple samples back-to-back with limited or no chromatographic separation and uses single-ion monitoring (SIM) to detect analytes. With inexpensive UHPLC-MS equipment, run times are reported to be as low as 10 s per sample, resulting in the data acquisition of 1536 samples possible in around six hours. The main drawback of this technique compared to the ultrafast MS methods is that reformatting from a 1536-well MTP to four 384-well MTPs (Figure 15) is required as currently no commercial LC-MS autosamplers can accommodate these plates. Furthermore, when compared to methods like AE-MS, ion suppression from common diluents (like DMSO) can be problematic. (248) Figure 15 Figure 15. Routine HPLC-MS kit equipped with a single-quadruple MS can be used for analysis of ultraHTE reactions direct from 384-well MTP. Two techniques can be employed, either the multiple injections in a single experimental run (MISER) method or a more traditional LCMS/UV method, however, there is a trade-off between speed and the level of quantification that can be achieved. Although the development of these rapid analytical techniques has had a notable impact on the speed at which analytical data from HTE reactions can be acquired, there remains a trade-off between (1) speed of acquisition, (2) sensitivity of the technique employed, (3) the amount of structural information gained from the technique, and
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
(4) the level of quantification that can be achieved. With ultrafast techniques like MALDI, DESI, AE-MS, etc., quantitative data can be acquired providing an internal standard is added to the analytical samples prior to analysis, an authentic sample of the analytes can be obtained, and a calibration curve run to compare measured analyte response and concentration to henceforth determine yield. This calibration is necessary as ionization responses of different analytes can be nonlinear and difficult to compare reliably. However, even if authentic samples of the analytes can be sought, other phenomena like ion suppression can add additional complications to the level of accuracy of yield assessment that can be realistically achieved with MS-based techniques. More conventional LC-MS methods which chromatographically separate analytes and use UV absorbance (e.g., at 210 nm) to determine analyte concentration tend not to suffer from the same issues as MS-based techniques (e.g., nonlinear responses at increased concentration and ion suppression). However, these methods require longer run times (cf MISER) and may require optimization of the elution gradients to ensure peak separation. Thus, although this technique can afford increased confidence in the level of quantitation that can be obtained, this comes at the cost of prolonged analysis times for both acquisition of data and deconvolution of reaction outcome (Figure 15). 6.1.3.4. HTE and Continuous Flow Flow chemistry is a technique that is highly applicable for optimization methods such as self-optimization, as discussed in section 5. Commercially available equipment such as the Vaportec R-series or systems from ThalesNano or Uniqsis, combine an autosampler for reagent selection and following injection into the flow reactor, in-line analytical equipment can also be coupled for in-line analysis. These setups are well-suited to automated synthesis, however, there are currently no off-the-shelf continuous flow setups which can rival the throughput of plate based HTE,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
and consequently there is a scarcity of reports of HTE used in continuous flow. More often continuous flow is used for the scaling up or continuous variable optimization of HTE hits identified in plates, (228,259,260) however, before this can be accomplished, reoptimization is often required as chemistry seldom translates directly from batch or plate to flow. (228,230) Pfizer has recently reported a bespoke flow platform for nanoscale HTE based on modified HPLC equipment, as shown in Figure 16. (207) The autosampler which would normally house analytical samples is repurposed to hold up to 192 source vials containing stock solutions of reagents, which is a significant increase in capacity compared to existing commercial platforms. The reaction segment is assembled in ∼45 s, where 1 μL of each reagent is aspirated and then the whole slug is injected into a flowing solvent stream and delivered into the reactor with approximately one-minute residence time. The reaction mixture is then directed into a 96-well plate fraction collector and subsequently analyzed via LCMS. This system has the benefit of not only containing several switching valves which can permit selection of a several solvents but can also direct the reaction slug to one of two LCMS machines for analysis. This platform enabled the execution and analysis of around 1500 reactions in a 24 h period and was also subsequently employed for the optimization of a photoredox catalyzed decarboxylative minisci-type C–H arylation to afford bicyclo[1.1.1]pentane containing compounds, which are medicinally relevant and traditionally challenging to synthesize. (261) Figure 16 Figure 16. Platform for automated nanomole-scale reaction screening and micromole-scale synthesis in flow developed by Pfizer and reported for the optimization of a Suzuki reaction. (207) A workflow recently reported by Bourne and co-workers also shows the use of a bespoke flow reactor platform for library synthesis and
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
optimization. (262) This work utilizes “stopped-flow” experiments and machine learning models to map chemical reactivity and synthesize diversity-oriented libraries. This led to a system that can predict optimal synthesis conditions with 92% accuracy and a dramatic increase in the success rate of initial library screens for reactivity while achieving a 90% reduction in reagent consumption when compared with continuous flow. 6.1.4. HTE for Data Set Generation Data driven optimization involves the use and analysis of relevant data to guide decision making in pursuit of the global optima for a given process. In practice, the success of this endeavor is highly dependent on the quality of data utilized, a factor that is increasingly important for autonomous reaction optimization systems using algorithmic, generative, and machine-learning (ML) derived models. One of the biggest challenges currently faced in this area is the acquisition of abundant and high-quality data sets. The availability of reaction data accessible to cheminformaticians has dramatically increased as online databases such as SciFinder-CAS, Reaxys, and the United States Patent and Trademarks Office (USPTO) have become accessible, as discussed in section 6.2. However, there are limitations to the condition of data obtained from these databases, as they are collated from numerous sources and usually extracted using text mining software which can lead to poorly standardized data and noisy data sets. Automated HTE offers the potential to acquire high-quality data sets which are well-standardized and most importantly, include negative data points, however, there is currently a scarcity of these data sets which are publicly available. Although data sets compiled from mining patent repositories and published reaction literature are extensive and cover a wide-ranging number of reaction classes, there are limitations to the quality of data as it is merged from many different laboratories using different methods and different equipment for quantification. Deviations in
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
reaction outcomes of the same reaction might arise from human error in the laboratory, transcription, or limitations of the text mining software, varying analytical methods and different reaction scales (mg to kg scale). Moreover, it must be noted that generally literature data is biased toward higher yields as low yielding outcomes are often not reported. In detail, NMR yields are usually calibrated to an internal standard within the reaction sample or probe, although isolated yields are considered the “gold standard,” these can be limited by sample recovery success which is highly dependent on the purification techniques used. UV- (e.g., HPLC-MS diode array responses) or mass-based techniques such as HPLC-MS or GC-MS can also give varying degrees of quantification as UV of MS responses can vary between compounds and purely quantitative results are only achieved by comparing the mass or UV response to an analytical calibration curve. By contrast, HTE offers an opportunity to generate data with a high degree of standardization, reactions tend to be performed and analyzed on the same scale with the same equipment using the same stock solutions. The power of this approach is apparent when comparing the variance in yield prediction models trained on HTE-derived data for specific reactions compared to a much larger data set from USPTO as highlighted by Reymond and co-workers. (263) 6.1.5. HTE Outlook The past decade has seen a shift in the application of HTE from the generation of chemical samples for HTS libraries to conditions screening and optimization of chemical reactions by taking advantage of the ability of HTE to rapidly survey a wide area of chemical space using a small amount of material. Recent advances in HTE and the translation of nanoliter robotics like the SPT Mosquito and Beckman Echo, originally designed for biological assays and now being applied
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
to parallelize organic synthesis, has fueled a rise in reports of ultraHTE where hundreds or thousands of chemical reactions can be executed in parallel on a miniaturized scale. Although limitations still exist with regards to the types of chemistry that are amenable to these plate-based formats, further development to increase the flexibility and generality of chemistry which can be performed in this fashion will be a benefit to the field of reaction optimization by HTE. Furthermore, as HTE plates are sealed and heated as a block, flexibility in reaction times, reagent equivalents, and temperature controls are limited as the entire plate is subjected to the same conditions. This means that although promising categorical variables can be identified as “hits” for promising potential optimal conditions in reaction optimization, many continuous variables must then be optimized upon scale-up to more traditional batch or flow laboratory scales. As discussed herein, another advantage of translating organic chemistry to a miniaturized HTE configuration is the ability to rapidly generate large amounts of standardized data in a relatively short time frame creating information density for a desired reaction or class of substrates. Currently, there are few examples of utilizing HTE-generated data to train ML models and the increased availability of more publicly available data sets would be particularly useful to the machine learning in the chemistry community. In this regard, challenges exist around the extent of quantitative data which can be obtained using an ultraHTE approach; currently there is a trade-off between speed and quantitation, and there are also constraints around the number of different analytes that can be analyzed in a quantitative manner in tandem. Recently, there has been significant progress in the field of analytical chemistry and robotics, which have directly facilitated the renaissance of HTE in the field of organic synthesis and further
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
developments are sure to modernize the area further and increase the number of data sets available to data scientists. 6.2. Data Mining, Machine Learning, and Optimization Benchmarking Machine learning (ML) has already revolutionized various areas, such as image recognition, (264) natural language processing, (265) and autonomous driving. (266) Within the field of organic chemistry, ML also represents an emerging tool, particularly for prediction tasks such as retrosynthesis, optimal reaction conditions, or reaction outcomes. It is then also possible to use these predictions from ML to influence starting points and process bounds for real-world optimizations, whether in self-optimization, HTE, or otherwise. The prediction of reaction outcomes at specific reaction conditions or direct prediction of reaction conditions are relevant and particularly attractive for reaction optimization. As shown in Figure 17, this problem is subdivided into data collection and model training. Once a data set is extracted from a high-throughput experiment or a reaction database, a chosen ML model is trained to predict reaction conditions or outcomes. Typical inputs for ML models include continuous reaction parameters such as reaction time, temperature, or reagent equivalents and categorical parameters involving the choice of catalyst, reactants, or bases. In the case of reaction outcome prediction, outputs are typically the targets of reaction optimization such as yield, conversion, or enantiomeric excess (ee). The predictive performance of these trained models is subsequently evaluated on unseen test data. Thus, one main goal is to develop modeling strategies that capture the correlation between reactants, reagents, and chemical reactivity to avoid brute-force laboratory screening which can be wasteful (particularly without the use of HTE equipment). Choosing the best molecular parametrization is a key aspect of achieving that goal. Figure 17 Figure 17. Schematic of the data collection and model training steps of using machine learning for reaction optimization from public or
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
private reaction databases. 6.2.1. Molecular Parameterization Molecules must be translated to a machine-readable, typically numerical, format that can be used as an input for ML models, prior to their use. We refer to this translation process as molecular parametrization as it aims to capture relevant molecular properties for a particular reaction. For different chemical transformations, different properties can influence reaction outcomes such as steric hindrance of a functional group or electronegativity of neighboring atoms. Moreover, the parametrization strategy should also be chosen to allow optimal compatibility with the ML model used, as prediction performance will depend on the compatibility between input format and ML model. The baseline parametrization method for representing chemical inputs is one-hot encoding (OHE). A one (1) or a zero (0) represent the presence or absence of specified reaction components respectively: no chemical information is encoded. This approach has been shown to be effective for a variety of chemical tasks, including yield prediction, but cannot extrapolate to new parts of chemical space. (229) Extended-connectivity fingerprints (ECFP) is a parametrization method that captures atom types, neighboring connectivity relationships, bond types, and represents the outcome in a machine-readable one-dimensional bit-vector. Circular fingerprints (e.g., Morgan fingerprints) are generated by (1) assigning identifiers to each atom in the molecule, (2) updating each atom’s identifiers depending on the neighboring atoms, (3) removing duplicates, and (4) compressing the data into a vector of set length, e.g., 1024 bit (a number of zeros and ones). (267) One of the advantages of these fingerprint based methods is that they are considered cheap features for modeling; their generation does not require a vast amount of computing power/time. Yet, their ability to explicitly capture molecular properties (e.g., sterics, electronics) of molecules is limited. Typically, models that use fingerprints develop knowledge in an indirect manner, such as an
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
implicit understanding of electronegativity associated with different halides, for example. (268) A much more comprehensive parametrization approach is calculating molecular descriptors using density functional theory (DFT). DFT can be used to determine the ground/excited state of molecules and thus offer fundamental insights into geometric and electronic properties. (229) As a result, DFT can be used to calculate descriptors that quantify the specific chemical properties of the given set of ligands such as the bulkiness of a molecule or electronegativity of atoms within a molecule. (269) However, DFT calculations for large libraries are often more time-consuming than actually running the corresponding reactions in a high-throughput screening format. More recently, parametrization work has utilized neural networks to achieve the tailored nature of DFT descriptors without the computational expense. This work is divided into two approaches: natural language processing models and graph neural networks. The former leverages recent advances in language models such as transformers, (270) where results can be achieved by training a model to predict the next word in a sentence across a wide variety of texts. Because chemistry can be represented as a language in the form of simplified molecular-input line-entry system (SMILES), (271) a language model can be trained to predict the next atom in a molecule when given only a portion of the molecule, thereby saving computational expense. (272) Because the model must understand a significant volume of chemistry to be able to predict a SMILES string, its numerical output can be used as a “learned fingerprint” for other prediction tasks. (273) Furthermore, the learned fingerprint can be tuned for each downstream task such as yield prediction using standard neural network training. Alternatively, graph neural networks represent a molecule as an interconnected network of atoms and bonds. These networks can be trained to produce a “learned fingerprint” for
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
prediction tasks. One of the most widely used forms of graph neural networks in chemistry are message passing neural networks (MPNNs), which learn relationships between neighboring atoms through iterative “messages” passed along bonds. (274,275) MPNNs have been extended to generate fingerprints for reactions, with state of the art results. (276) An overview of these techniques is shown in Table 5. Table 5. Overview of the Commonly Used Molecular Parameterization Techniques for Modelling Chemical Data 6.2.2. Prediction of Chemical Reaction Yields from High-Throughput Experiments Reaction outcome prediction has primarily been carried out on data obtained via HTE or similar techniques for generating consistent data sets. Doyle and co-workers trained a random forest (RF) algorithm on HTE data of a Buchwald–Hartwig reaction, aiming to generate an automatic feature generation algorithm. They demonstrated success using their approach, which was trained on 5% of data and outperformed linear regression trained on 70% of data. (243) Subsequently, Hirst and co-workers continued the work by Ahneman by using another machine learning technique, support vector machines (SVM), in which they demonstrated improved prediction performance. (277) Glorius and co-workers successfully boosted this prediction performance using a concatenation of fingerprints. (278) Doyle and co-workers then used a combination of fingerprint based and DFT based descriptors for the prediction of reaction performance for a deoxyfluorination HTE data set, thereby guiding the search toward high yielding conditions. (279) Overall, due to the consistency of HTE generated data sets, good results could be achieved with regard to the predictive performance of the applied ML models. 6.2.3. Reaction Databases While most of this covered literature used data solely generated via HTE or flow chemical platforms for training ML models to prediction chemical reactivity, an increasing effort is made in developing predictive models with chemical data mined from online databases. Reaxys, a commercial database
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
by Elsevier, and the United States Patent and Trademark Office (USPTO) reaction database are the two most used sources for data-mined applications. The Reaxys database is proprietary and contains information on more than 56 million reactions from over 16 000 journals. In contrast, the USPTO database is accessible to the public and contains chemical reactivity information obtained from over nine million patent applications. (280,281) Additionally, Pistachio represents a commercially available data set, based on USPTO data, electronic lab notebook (ELN) data, and information obtained from journals or other patent literature containing 13.3 million reactions. (282) Most recently, the open reaction database was created to build a standard format and open-access location for reaction data, which represents a large shift in fair data accessibility. (283)Figure 18 illustrates the information overlap between several of these reaction databases. (284) Figure 18 Figure 18. Illustration of the overlap of chemical reaction databases (Reaxys, Pistachio, USPTO, and a subset of AstraZeneca ELN13). (284) 6.2.4. Data Preprocessing To build predictive models based on reaction databases, one of the most significant challenges is extracting data into a standard format amenable to machine learning. The database providers mentioned perform some amount of preprocessing to extract the data into a common format. However, the databases often do not capture important details of the procedure such as the reaction temperature, workup protocol, or analytical results. Furthermore, the same reagent can be represented in different ways across the database (e.g., metal catalysts can be represented with their ligands or as separate components). To overcome this issue with data quality, recent work has developed deep learning models that can extract text descriptions of synthetic procedures in a standard format. (285−287) Subsequently, the data must then be filtered, in particular, as noted by Varnek and co-workers, there are many duplicate reactions in databases.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
(288) This is often caused by scientists using the same procedure for a standard reaction as is reported in the literature. Therefore, a filtering process often includes removing duplicate reactions, discarding reactions with missing key reactants or reagents (e.g., a Suzuki reaction should always have an organohalide and boronic acid), and excluding reactions without a numerical yield. This filtering process can often result in less than the 30% of the original extracted data being utilized for machine learning. For example, Reymond and co-workers created a data set of Buchwald couplings based on data extracted from several databases and, after filtering, only 15% of the original reaction records remained. (289) 6.2.5. Machine Learning for Reaction Condition Prediction Upon data set extraction from the literature, machine learning models can then be used to predict reaction conditions directly given a set of reactants. The first examples of such a model were developed by Jensen and co-workers, who used a feed forward neural network to directly predict reaction conditions given the difference in the ECFP fingerprints of the products and reactants, as shown in Figure 19. (290) Their neural network architecture was designed to reflect a chemist’s intuition. Often, catalysts are selected followed by solvents, reagents, and temperature, so Jensen and co-workers’s network first predicted catalysts and then conditioned each further prediction (solvents, reagents, and temperature) on the prior ones. On unseen reactions, the neural network’s top-3 predictions included those used in the literature with 50% accuracy. Furthermore, the correct catalyst was selected with over 93% accuracy. Figure 19 Figure 19. Demonstration of the architecture used by Jensen and co-workers for predicting reaction conditions. (290) Approaches that directly predict reaction conditions do not account for the yield of the reaction; instead, these techniques aim to predict the conditions used most often in the literature.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
The downside of this approach is that optimal reaction conditions for a particular transformation often require going beyond the standard conditions used for a transformation. Therefore, reaction condition prediction models might be used to give suggestions of starting points for further optimization rather than predicting optimal conditions directly. (291) In addition to predicting reaction conditions, machine learning models can be trained to predict the likely products of a reaction given the reactants. A variety of models have been trained for this purpose, ranging from natural language processing models (e.g., transformers) (292) to custom neural network architectures. (293) Furthermore, machine learning models can be used for retrosynthesis, which has been reviewed extensively elsewhere. (294) 6.2.6. Future Directions Thus far, ML has been applied to yield prediction and reaction condition prediction, both of which have potential use for reaction optimization. Scientists can use yield prediction models to narrow the potential optimal sets of conditions for a reaction without brute-force screening. Recent work by Grzybowski, Burke, and co-workers showed this by highlighting the use of ML and closed-loop optimization to identify general high-performing reaction conditions for a Suzuki–Miyaura coupling. (295) These ML models can therefore be combined with the optimization techniques described in section 5 to automate the identification of optimal reaction conditions with viable starting points. (15) Currently, ML models have been very successful in reaction outcome prediction on HTE data sets for a single reactant pair. We foresee that leveraging a broad selection of data from targeted experiments could aid general prediction of key reaction outcomes for specific classes of reactions. Reymond and co-workers attempted to build a general yield prediction model based on the USPTO data set, but their model has low predictive quality due to the sparse nature of the USPTO yield data. (263) Therefore, more high-quality data sets
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
with reaction outcomes recorded and further model development are needed to create general ML models for reaction optimization. If successful, this research could be transformative in reaction optimization and transition the field to more direct predictions of optimal reaction conditions. For a recent review of in-depth modeling of HTE data sets, also refer to Jensen and co-workers. (296) 7. Scale-Up and Manufacture Lab-scale reaction optimization studies focus on improvements in reaction yield and purity, optimization of reaction cost and greenness, and development of the optimal workup/separation approach, whereas subsequent process design steps (necessary for scale-up) need to address a different set of requirements. (297) The ability to deliver commercially relevant quantities of product, with an acceptable investment, operational and environmental cost, as well as process safety, are the focus of this stage of process design. Furthermore, continuous fulfilment of quality critical parameters is investigated at this step, fundamental for pharmaceutical and fine chemicals industries. (298) Development of processes meeting these criteria is defined as scale-up and involves determination of the critical scale-dependent factors that would affect the choices of the most functional reactor architecture, process conditions, and separations steps. These are the next steps from bench-scale reaction optimization and understanding how previously obtained optimal conditions may change is crucial to the successful scale-up of chemical processes. Scale-up implies the attainment of significantly larger product throughputs compared to laboratory reaction discovery or reaction development studies. This is traditionally realized by deploying significantly larger reactors with very different gradients of temperature, pressure, and reactant concentrations to the small-scale processes. (299) Therefore, it is often very difficult to reproduce at scale exactly the same conditions obtained when using small-scale equipment and the most promising operating conditions determined during lab-scale optimization studies do not necessarily yield an optimal large-scale process. As a result, many
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
initially promising discoveries and reaction routes may not achieve broad deployment in industrial production. (300,301) While large companies can invest vast resources into solving scale-up related challenges, (302) smaller research organizations such as start-ups or academia might not be in place to allocate such resources, limiting the technology-readiness level of developed reactions. (297) Determining the optimal reaction conditions for the scaled-up reaction involves careful consideration of how the processing conditions change in space and time and how these changes relate to the time scale of the molecular events driving the reaction. Elucidating these phenomena requires simultaneous insights into the kinetics, heat transfer, and mass transfer happening inside of the reactor. In contrast to exhaustive scale-up guidelines formulated for chemical engineers and process chemist experts, (303−305) this section discusses scale-up on a conceptual level, providing insight for chemists working on molecular discovery and benchtop optimization. We anticipate that the consideration of scale-up challenges and complexity in the early stages of process optimization can help to guide the laboratory studies toward the achievement of metrics meaningful for large-scale plants and consequently accelerate the process of launching new molecules and products to market. The following subsections briefly discuss scale-up considerations to be addressed during benchtop experiments, as well as which phenomena change during the transition to larger reactors and how to quantify these changes. Furthermore, we introduce two strategies used to develop large-volume processes: scale-up and numbering-up, as well as guide the user toward a choice of applicable equipment for each case. 7.1. Scale-Up Considerations within Reaction Optimization In most chemical applications, the design of a large-throughput process involves a multifold increase in the size of the reactor vessels that were used for kinetic or optimization studies in the laboratory environment. The processing environment is likely to drastically change; this is because of
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
significant reductions in surface area/volume ratios that imposes limitations on heat transfer rate, sensitivity to mixing, and different time of addition and removal of products. (306)Figure 20 summarizes the typical time ranges of characteristic mixing, heat transfer, and liquid space time (reactor volume divided per volumetric flow rate) for different reactors used in academia and industry (shake flask, flow reactors, microreactors, stirred-tank reactors). During the transition to the large-scale reactors, mixing and heat transfer become several orders of magnitude slower. While this imposes little or no consequences for slow reactions, which can be scaled-up in a relatively straightforward manner, (303) for fast reactions the extended time necessary to mix phases in a large-reactor results in limited chemical availability of each reactant leading to suboptimal outputs. Figure 20 Figure 20. Laboratory vs industrial scale: comparison of characteristic times (in seconds) for mixing, heat transfer and liquid space time observed in reactors used in benchtop optimization and large-scale industrial reactors. (306,307) Therefore, finding the optimal reaction conditions for a scaled-up process cannot rely on a simple linear increase of the reactant feed and processing time from the optimal lab benchmark. From a process design perspective, the availability of the intrinsic kinetic model can significantly accelerate process development, (308) and the ability to generate such models with automated flow methods has been highlighted as a potential paradigm shift in the process systems engineering community. (162) Notably, such models should be obtained with no mass/heat/mixing limitations. To design kinetic determination experiments unhindered by these effects, Jensen and co-workers (309) provided a simplified chart-based method for the evaluation of mixing and dispersion in small-scale flow systems, while detailed insights into overcoming heat transfer limitation during optimization in flow reactors were discussed by Mase and co-workers. (310) Although kinetic models are a powerful tool for accelerating
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
process scale-up, in some cases, the experimental effort necessary to derive these models is extensive, making it infeasible to realize in the fast-paced process development environment. In such cases, Stitt and Simmons (303) recommend determination from benchtop optimization, at a minimum, the following information: (i) reaction network and the significant byproducts, (ii) if the reaction kinetics follow simple power-laws, (iii) sensitivity of the reaction selectivity to mass transfer effects, and (iv) heat evolution and the potential for reaction runaway. Practical guidelines toward the use of flow chemistry setup for heat measurements were proposed by Meier and co-workers, (311) while Bourne and co-workers (312) discussed best practices toward continuous-flow aided kinetic analysis. Other, case-specific strategies for extracting meaningful information for process scale-up from optimization studies are summarized in the reviews of industrial practices in large-scale deployment of novel reactive routes, (313) such as photoredox catalysis, (314,315) electrochemistry, (316) C–H activation, (317) reductive coupling, (318) and flow chemistry. (319) Another aspect to address within optimization studies is the preferred production mode for the scaled-up process, as the rationale for its choice may differ between early benchtop experiments and large-scale production, as broadly described by Trout and co-workers. (320) In a batch process, feedstocks are supplied all at once to the reactor, and the treatment of the subsequent load of feed materials starts only when the previous batch is fully processed and removed. Batch processing is frequently preferred due to its simplicity and is particularly convenient in pharmaceutical processes that require frequent cleaning of used equipment, or even removal and disposal of single-use reactors between batches (e.g., biomanufacturing lines (321)). However, the design simplicity comes at an extra cost of slow processing, increased energy consumption for start-up and shutdown of each batch, significant environmental impact from cleaning solvents, and more challenging process parameters
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
control. Continuous processing, in contrast, allows for a significant cost and environmental footprint reduction (if solvents are recycled (322,323)) by faster conversion and increased productivity, reduced down-time, and improved quality by facilitating continuous monitoring of critical parameters. From the perspective of the chemist, it is important to be aware of the preferred processing mode and consider process limitations related to each mode within the design of lab-scale experiments. Within the pharmaceutical and fine chemical industries, the ability to fulfill quality-critical parameters is one of the key goals of process development, but these aspects are typically not investigated before moving toward scale-up studies. Most recent industrial practices, described by Tsai and co-workers, (324) describe how to systematically include quality considerations across all phases, including reaction optimization. 7.1.1. Classical Approaches Toward Scale-Up After transferring the data from benchtop optimization to process development, models of scaled-up reactors are obtained to account for engineering phenomena, e.g., characterization of heat and mass transfer environments. If detailed kinetic models are available, the engineer can simulate the operation of a large-scale process and determine a new optimal set of conditions. Gatica and co-workers provided a detailed overview of the industrial practices for the development of design equations to model such processes, (325) whereas a review by Patterson (326) addresses the challenges in modeling of mixing- and temperature-sensitive chemical reactions at larger scales. Other reports from the literature also provide insightful examples of computational fluid dynamics (CFD) use for modeling of mixing phenomena. (327,328) Scale-up engineers seek to quantify the influence of heat and mass transfer, as well as other physical properties, on reaction progression. This also includes the interactions between each physicochemical property. This problem is frequently simplified by dimensional analysis that enables scientists to develop relations among physical quantities (e.g., velocity, viscosity, surface tension) using their
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
dimensions expressed in base units (combination of, e.g., meter, kilogram, second). Detailed instructions on the derivation and use of dimensional analysis equations, followed by practical examples of calculations for stirred-tank reactors, were provided by Zlokarnik (329,330) and Wild and co-workers. (331) These equations enable an engineer to disregard the parameters that are not critical for the given reaction, and further develop the nonlinear relationships guiding scale-up. If necessary, experiments to determine the significance of these variables for successful scale-up can be designed. The experiments can be planned for a partial similarity or complete similarity approach, and advantages and limitations of both approaches are described by Kind and co-workers (332) and illustrated by the case study of a scale-up of competitive chemical reactions. Another aspect addressed at the scale-up stage is the choice of a reactor. While stirred-tank or tubular reactors are by far the most deployed reactors in the industrial setting, development of the process intensification field (333) yields a wide variety of novel reactors offering designs that can drastically enhance heat and mass transfer as well as increase the controllability of reactants and product spatial concentrations distributions. The choice of the reactor also includes insights into construction materials and their interactions with the reaction medium, compliance with the regulatory requirements for the given final product, and the lead manufacturing time. Guidelines provided by Moran and Henkel (334) compare the reactors typically used in industry; the toolbox proposed by Roberge and co-workers includes both the choice of processing mode and the reactor, (335) while the contribution of Lindeque (336) specifically addresses reactor considerations in biocatalytic production of pharmaceuticals compounds; Dautzenberg and Mukherjee (337) discuss the choice and deployment of multifunctional reactors; West and co-workers (338) focus on scalable autothermal reactors. Methods for selection between a wider range of novel, process
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
intensification reactors were proposed by Commenge and Falk, (306) and an overview of different options was presented in an open-source database published by Gorak and co-workers. (339) Selected reactor architectures and the optimal conditions determined for a scaled-up process are frequently verified on a scale of pilot or mini-plant operation, whereas operability and stability are tested in pilot plants. Industrial practice proves that this intermediate step in process development has the substantial potential to improve process understanding and consequently product quality and operational safety, however, it also increases time-to-market and overall project costs. Design of such plants was described in detail by Whalley. (340) Examples of scale-up toward a pilot/manufacturing plant capacity are available in the literature, including a hydrazine condensation study by Lane and co-workers and the edaravone synthesis by Sun and co-workers. (341,342) 7.1.2. Alternative Approaches: Numbering Up The discussed strategies for process scale-up include both extensive experimental insights on laboratory scale, simultaneous modeling of numerous phenomena, and potential pilot plant testing. Despite the availability of the methodological tools described above, it still may not be possible to reproduce the selectivity achieved in benchtop tests in the scale-up setting, particularly if the optimization studies were conducted in microreactors that offer excellent heat and mass transfer characteristics. In such cases, increases in the plant throughput can be achieved instead by the numbering-up approach, which involves the simultaneous use of hundreds to thousands of reactors of the same or similar scale to the ones used in process optimization studies. (158) This enables one to achieve exactly the same conditions as found in the lab, improves safety by better temperature control and drastically reduces scale-up time. Possibilities of the accelerated process development by numbering-up of microreactors are covered in detail by Kockmann and co-workers (343) and Roberge and co-workers. (344) However,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
numbering-up is associated with high investment cost: numerous reactors are required instead of a single stirred-tank reactor, along with multiple process control devices. Detailed insights into the economics of numbering-up were described by Weber and Snowden-Swan, (345) and the potential to intensify the process can be evaluated by the process intensification score proposed by van der Meer and co-workers. (346,347) Capital costs considerations mean that this approach can be economically justified only for high-end products, and for cases where a drastic improvement of output (e.g., reaction yield) has been demonstrated. Interestingly, an analysis conducted by the Process Development Team from Lonza (348) reveals that 50% of reactions in the fine chemical/pharmaceutical industry could benefit from the deployment of microreactor technology. Lowe and co-workers (349) proposed a benchmarking method allowing for a similar evaluation across different industries. Notably, as the economic factors are limiting the deployment of microreactors, their market availability is also lower, and the extended lead time could be a key limitation for decision makers that require fast and large product delivery. By more frequent adoption of novel process optimization methods that involve the use of microreactors, we anticipate the unveiling of more case studies where excellent optima are achievable solely by this method, and hence, we can expect accelerated adoption of more efficient process intensification technologies. 7.2. Safety Considerations For both scaling- and numbering-up approaches, in batch or continuous mode, there are several safety concerns to be addressed. Large volumes of flammable solvent are almost always required, meaning that precautions must be taken to minimize the risk of fire from static electricity build up or exposed flame. However, the greatest concern with organic processes is thermal runaway. (350) It is important to consider these physical limitations when running optimization campaigns, as algorithms or experimental designs may suggest experiments
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
that are out of safe operating bounds: this includes reactions that are suggested at unsafe temperatures or concentrations. Thermal runaway occurs when the rate of an exothermic reaction is accelerated by increases in temperature. In the worst case, this rate acceleration can lead to secondary decomposition reactions that are more energetic and hazardous than the primary synthesis. While reactions commonly used in fine chemicals such as the Suzuki–Miyaura cross coupling are often not thought to be dangerously exothermic at the bench, they can have significant exotherms that could lead to thermal runaway at scale. (351) The exotherms of reactions need to be understood via calorimetry studies in the lab, so that control strategies can be implemented during scale-up. Reaction calorimetry enables the measurement of the heat produced by a reaction over time. Often, two parameters are determined: the adiabatic temperature rise (ΔTadia) and the maximum temperature of synthesis reaction (MTSR). ΔTadia is the temperature rise in a specific volume for a specific reaction when all the heat of reaction is delivered to an adiabatic system (i.e., no heat is transferred between the system and its surroundings). MTSR is defined as the maximum temperature an adiabatic reactor would reach if cooling failed. In other words, MSTR is the sum of the reaction temperature Tp and ΔTadia. Together, ΔTadia and MTSR can give insights of the worst-case scenario for a particular reaction and reactor. For example, Wang and co-workers conducted calorimetry studies of copper-mediated fluorinations of bromopyridines and found that the MTSR was above the decomposition of the chosen reaction solvent, DMSO. (352) Because decomposing DMSO is an explosion hazard, (353,354) control strategies such as reduced speed dosing or switching to an alternative solvent needed to be implemented. Readers interested in learning more about the safety considerations for reaction development are referred
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
to the excellent reviews by Yang on safety aspects of DMSO (354) and Pd-catalyzed reactions, (350) as well as the textbook by Stoessel on “Thermal Safety of Chemical Processes”. (355) 8. Conclusion In this review, we have outlined several modern techniques that are utilized for chemical reaction optimization to serve as an accessible reference for interested bench scientists. We have also given discussions on their relationship to further stages of process development, namely scale-up. There are many distinct methodologies that can be used to obtain optimal reactions conditions for desired outcomes (reaction yield, selectivity, E-factor, etc.), and there are trade-offs to consider for research organizations when deciding which to implement. These decisions must balance the costs associated with each technique (equipment costs, training costs, time investments) with the deliverables that they hope to achieve and their associated accuracy and reliability. As automated equipment is becoming more ubiquitous and user-friendly, this is one possible solution to unify several fields (chemistry, process engineering, computer science) to conduct reaction optimization techniques in novel, systematic ways to maximize process outputs. This could be in conducting automated DoE or kinetic studies, screening every combination of reaction variables in HTE or utilizing self-optimization, or more! Researchers may willingly run intuition-driven experimentation, even when knowing of more effective techniques, as the novelty of their research may focus on other aspects of chemistry (such as reaction discovery, substrate scopes, etc.) rather than complete optimization of their processes. However, as modern laboratories are becoming more diversified in skillsets and more interdisciplinary research is conducted, familiarity with these techniques must be embraced and undergraduate/postgraduate courses will undoubtedly reflect this more in coming years. It is easy to envision an evolving chemistry course with practical modules in DoE, HTE, and more, as the skillsets of chemists diversify beyond traditional synthesis to
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
meet the needs of the modern laboratory. It is the hope that this timely review will prove the accessibility of these optimization techniques and help to encourage inspired chemists to incorporate them into their workflows. Author Information Corresponding Authors Connor J. Taylor - Astex Pharmaceuticals, 436 Cambridge Science Park, Milton Road, Cambridge CB4 0QA, U.K.; Innovation Centre in Digital Molecular Technologies, Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, U.K.; Email: C.Taylor@compunetics.net Alexei A. Lapkin - Innovation Centre in Digital Molecular Technologies, Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, U.K.; Email: aal35@cam.ac.uk Authors Alexander Pomberger - Innovation Centre in Digital Molecular Technologies, Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, U.K. Kobi C. Felton - Department of Chemical Engineering and Biotechnology, University of Cambridge, Philippa Fawcett Drive, Cambridge CB3 0AS, U.K. Rachel Grainger - Astex Pharmaceuticals, 436 Cambridge Science Park, Milton Road, Cambridge CB4 0QA, U.K.; Magda Barecka - Chemical Engineering Department, Northeastern University, 360 Huntington Avenue, Boston, Massachusetts 02115, United States; Chemistry and Chemical Biology Department, Northeastern University, 360 Huntington Avenue, Boston, Massachusetts 02115, United States; Cambridge Centre for Advanced Research and Education in Singapore, 1 Create Way, 138602 Singapore; Thomas W. Chamberlain - Institute of Process Research and Development, School of Chemistry and School of Chemical and Process Engineering, University of Leeds, Leeds LS2 9JT, U.K.; Richard A. Bourne - Institute of Process Research and Development, School of Chemistry and School of Chemical and Process Engineering, University of Leeds, Leeds LS2 9JT, U.K.; Christopher N. Johnson - Astex Pharmaceuticals, 436 Cambridge Science Park, Milton Road, Cambridge CB4 0QA, U.K.; Author Contributions CRediT: Connor J. Taylor conceptualization, formal analysis, investigation, methodology, project administration, visualization, writing-original draft, writing-review & editing; Alexander Pomberger data
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
curation, formal analysis, investigation, writing-original draft, writing-review & editing; Kobi C. Felton formal analysis, investigation, methodology, writing-original draft, writing-review & editing; Rachel Grainger formal analysis, investigation, methodology, writing-original draft, writing-review & editing; Magda Barecka formal analysis, investigation, methodology, writing-original draft, writing-review & editing; Thomas W. Chamberlain funding acquisition, resources, supervision, writing-review & editing; Richard A. Bourne funding acquisition, resources, supervision, writing-review & editing; Christopher Norbert Johnson funding acquisition, resources, supervision, writing-review & editing; Alexei A. Lapkin funding acquisition, resources, supervision, writing-review & editing. Notes The authors declare no competing financial interest. Biographies Connor J. Taylor is currently a Sustaining Innovation Postdoctoral Research Associate at Astex Pharmaceuticals in collaboration with the University of Cambridge, working on autonomous reactor platforms. He completed his integrated B.Sc. MChem Chemistry degree at the University of Leeds (2017), and his Chemical and Process Engineering Ph.D. under the supervision of Prof. Richard Bourne and Dr. Thomas Chamberlain (2021). During his Ph.D., he founded the process optimization company, Compunetics, in partnership with the University of Leeds and maintains a strong interest in optimization, reaction engineering and digital chemistry. Alexander Pomberger received his Bachelor’s degree in Chemistry from the University of Vienna (2016) and his Master’s degree in Technical Chemistry from the Vienna University of Technology (2019). During his studies, he joined Prof. Jensen at MIT to conduct research in the area of continuous flow chemistry and reactor development in 2018. In 2020, he joined Prof. Lapkin at the University of Cambridge to pursue his Ph.D. studies, where he encompasses the application of machine learning for chemical reaction optimization and prediction. Kobi Felton is a fourth year Ph.D. student at the University of Cambridge working at the intersection of data science and chemical engineering working under the supervision of Alexei Lapkin. He obtained a Bachelor’s in Chemical Engineering from
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
the NC State University and a MPhil Research from the University of Cambridge. Rachel Grainger received her MChem in Chemistry for Drug Discovery (University of Bath, 2010), and her Ph.D. under the supervision of Prof. Igor Larrosa (QMUL, 2014). She then moved to Cambridge (UK) and worked as a postdoc with Prof. Matthew Gaunt (University of Cambridge) before moving to Astex, where she started her industrial career as a Sustaining Innovation postdoc. Currently, she is a Senior Research Associate at Astex, and she has extensive experience in automated high-throughput experimentation and has helped establish Astex’s automation capabilities within the Medicinal Chemistry Department. Magda H. Barecka is an Assistant Professor at Northeastern University in Boston, Department of Chemical Engineering and Chemistry and Chemical Biology. Her group focuses on the discovery and scale-up of intensified chemical manufacturing technologies, mainly electrochemical CO2 utilization methods. Dr Barecka completed postdoctoral training at the University of Cambridge, Research Centre for Research and Education in Singapore (CARES). She developed a new approach for CO2 utilization (Carbon Capture On-Site Recycling), as well as coauthored three patents. She holds a Ph.D. degree from TU Dortmund (Germany) and, as a part of her Ph.D. thesis, developed a methodology supporting implementation of intensified technologies in chemical manufacturing. Thomas W. Chamberlain is an Associate Professor within the Institute of Process Research and Development at the University of Leeds, applying his understanding of nanomaterials to both fundamental and applied research challenges. His interdisciplinary research group exploits automated chemical synthesis platforms, machine learning approaches, and on- and off-line characterization to explore and optimize chemical reactions to form chemical products, including; heterogeneous catalysts, metal nanoparticles, quantum dots, polymers, pharmaceutical, and agrochemicals. He completed an MSci in Chemistry at the University of Nottingham in 2005. He was then awarded a University Interdisciplinary award to study for
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
a Ph.D. with Professors Andrei Khlobystov and Neil Champness in Chemistry and Peter Beton in the School of Physics, working on the synthesis of novel functional fullerene molecules and the subsequent formation of fullerene/carbon nanotube peapod structures. He received his Ph.D. in 2009 and then joined the Nottingham Nanocarbon group as a postdoctoral research associate studying the use of supramolecular forces. Richard A. Bourne is currently a Professor of Digital Chemical Manufacturing at the University of Leeds based at the Institute of Process Research and Development. His group works on automated flow systems, combining online analysis, feedback control, and evolutionary algorithms to provide process understanding and optimization. Richard is a Royal Academy of Engineering Research Chair working on Digital Discovery and Manufacturing of Pharmaceuticals. He completed a BSC MChem at the University of Nottingham (2004), then his Ph.D. under the supervision of Prof. Martyn Poliakoff (2008). Christopher N. Johnson obtained an MA degree at the University of Cambridge and a Ph.D. at the University of Bristol. In 2008, after several years working as a medicinal chemist at GSK and legacy companies, he moved to Astex, where he has worked on fragment-based approaches applied to multiple target classes including PPI antagonists, with a particular focus on the oncology and neurodegeneration disease areas. Dr. Johnson is a coinventor or coauthor of over 100 patents and scientific publications. Alexei A. Lapkin studied Biochemistry at Novosibirsk State University and then completed his Ph.D. in Catalytic Reaction Engineering at the University of Bath. He began his academic career at the University of Bath, focusing on catalytic reaction engineering and process intensification, then moved to the University of Warwick, shifting attention more on sustainability of chemical processes. Since 2013, he has held the position of Professor of Sustainable Reaction Engineering at the University of Cambridge. Acknowledgments
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
C.J.T. is a Sustaining Innovation Postdoctoral Research Associate at Astex Pharmaceuticals and thanks Astex Pharmaceuticals for funding, as well as his Astex colleagues Mark Wade, Gianni Chessari, and David Rees for their support. K.C.F. has received Ph.D. funding from the Marshall Scholarship, Cambridge Trust, and BASF SE. A.P. is grateful for receiving funding from BASF SE. This work is supported by European Regional Development Fund (ERDF) via the Innovation Centre in Digital Molecular Technologies and Engineering and Physical Sciences Research Council (EPSRC) via project EP/S024220/1 “Centre for Doctoral Training in Automated Chemical Synthesis Enabled by Digital Molecular Technologies”. References This article references 355 other publications. 1Taylor, C. J.; Baker, A.; Chapman, M. R.; Reynolds, W. R.; Jolley, K. E.; Clemens, G.; Smith, G. E.; Blacker, A. J.; Chamberlain, T. W.; Christie, S. D.Flow Chemistry for Process Optimisation Using Design of Experiments. J. Flow. Chem. 2021, 11, 75– 86, DOI: 10.1007/s41981-020-00135-0 Google Scholar 2Tsay, C.; Pattison, R. C.; Piana, M. R.; Baldea, M. A Survey of Optimal Process Design Capabilities and Practices in the Chemical and Petrochemical Industries. Comput. Chem. Eng. 2018, 112, 180– 189, DOI: 10.1016/j.compchemeng.2018.01.012 Google Scholar 3Andraos, J. Designing a Green Organic Chemistry Lecture Course. In Green Organic Chemistry in Lecture and Laboratory, 1st ed.; CRC Press: Boca Raton, FL, 2012; pp 29– 68. Google Scholar 4Brain, C. T.; Steer, J. T. An Improved Procedure for the Synthesis of Benzimidazoles, Using Palladium-Catalyzed Aryl-Amination Chemistry. J. Org. Chem. 2003, 68, 6814– 6816, DOI: 10.1021/jo034824l Google Scholar 5Wang, H.; Jung, H.; Song, F.; Zhu, S.; Bai, Z.; Chen, D.; He, G.; Chang, S.; Chen, G. Nitrene-Mediated Intermolecular N–N Coupling for Efficient Synthesis of Hydrazides. Nat. Chem. 2021, 13, 378– 385, DOI: 10.1038/s41557-021-00650-0 Google Scholar 6Kuang, Y.; Cao, H.; Tang, H.; Chew, J.; Chen, W.; Shi, X.; Wu, J. Visible Light
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Driven Deuteration of Formyl C–H and Hydridic C (Sp 3)–H Bonds in Feedstock Chemicals and Pharmaceutical Molecules. Chem. Sci. 2020, 11, 8912– 8918, DOI: 10.1039/D0SC02661A Google Scholar 7Derosa, J.; van der Puyl, V. A.; Tran, V. T.; Liu, M.; Engle, K. M. Directed Nickel-Catalyzed 1, 2-Dialkylation of Alkenyl Carbonyl Compounds. Chem. Sci. 2018, 9, 5278– 5283, DOI: 10.1039/C8SC01735B Google Scholar 8Kim, I.; Im, H.; Lee, H.; Hong, S. N-Heterocyclic Carbene-Catalyzed Deaminative Cross-Coupling of Aldehydes with Katritzky Pyridinium Salts. Chem. Sci. 2020, 11, 3192– 3197, DOI: 10.1039/D0SC00225A Google Scholar 9Roque, J. B.; Kuroda, Y.; Göttemann, L. T.; Sarpong, R. Deconstructive Diversification of Cyclic Amines. Nature. 2018, 564, 244– 248, DOI: 10.1038/s41586-018-0700-3 Google Scholar 10Zhu, S.; Chen, Y.-H.; Wang, Y.-B.; Yu, P.; Li, S.-Y.; Xiang, S.-H.; Wang, J.-Q.; Xiao, J.; Tan, B. Organocatalytic Atroposelective Construction of Axially Chiral Arylquinones. Nat. Commun. 2019, 10, 4268, DOI: 10.1038/s41467-019-12269-4 Google Scholar 11Derosa, J.; Kleinmans, R.; Tran, V. T.; Karunananda, M. K.; Wisniewski, S. R.; Eastgate, M. D.; Engle, K. M. Nickel-Catalyzed 1, 2-Diarylation of Simple Alkenyl Amides. J. Am. Chem. Soc.. 2018, 140, 17878– 17883, DOI: 10.1021/jacs.8b11942 Google Scholar 12Derosa, J.; Tran, V. T.; Boulous, M. N.; Chen, J. S.; Engle, K. M. Nickel-Catalyzed Β, Γ-Dicarbofunctionalization of Alkenyl Carbonyl Compounds Via Conjunctive Cross-Coupling. J. Am. Chem. Soc. 2017, 139, 10657– 10660, DOI: 10.1021/jacs.7b06567 Google Scholar 13Czitrom, V. One-Factor-at-a-Time Versus Designed Experiments. Am. Stat. 1999, 53, 126– 131, DOI: 10.1080/00031305.1999.10474445 Google Scholar 14Nor, N. M.; Mohamed, M. S.; Loh, T. C.; Foo, H. L.; Rahim, R. A.; Tan, J. S.; Mohamad, R. Comparative Analyses on Medium Optimization Using One-Factor-at-a-Time, Response Surface Methodology, and Artificial Neural Network for Lysine–Methionine Biosynthesis by Pediococcus Pentosaceus Rf-1. Biotechnol. Biotechnol. Equip. 2017, 31, 935– 947, DOI: 10.1080/13102818.2017.1335177 Google Scholar 15Shields, B. J.; Stevens, J.; Li, J.; Parasram, M.; Damani, F.; Alvarado,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
J. I. M.; Janey, J. M.; Adams, R. P.; Doyle, A. G. Bayesian Reaction Optimization as a Tool for Chemical Synthesis. Nature. 2021, 590, 89– 96, DOI: 10.1038/s41586-021-03213-y Google Scholar 16Lendrem, D. W.; Lendrem, B. C.; Woods, D.; Rowland-Jones, R.; Burke, M.; Chatfield, M.; Isaacs, J. D.; Owen, M. R. Lost in Space: Design of Experiments and Scientific Exploration in a Hogarth Universe. Drug Discovery Today. 2015, 20, 1365– 1371, DOI: 10.1016/j.drudis.2015.09.015 Google Scholar 17Boström, J.; Brown, D. G.; Young, R. J.; Keserü, G. M. Expanding the Medicinal Chemistry Synthetic Toolbox. Nat. Rev. Drug Discovery. 2018, 17, 709– 727, DOI: 10.1038/nrd.2018.116 Google Scholar 18Murray, P. M.; Bellany, F.; Benhamou, L.; Bučar, D.-K.; Tabor, A. B.; Sheppard, T. D. The Application of Design of Experiments (Doe) Reaction Optimisation and Solvent Selection in the Development of New Synthetic Chemistry. Org. Biomol. Chem. 2016, 14, 2373– 2384, DOI: 10.1039/C5OB01892G Google Scholar 19Lendrem, D.; Owen, M.; Godbert, S. Doe (Design of Experiments) in Development Chemistry: Potential Obstacles. Org. Process Res. Dev. 2001, 5, 324– 327, DOI: 10.1021/op000025i Google Scholar 20Aggarwal, V. K.; Staubitz, A. C.; Owen, M. Optimization of the Mizoroki– Heck Reaction Using Design of Experiment (Doe). Org. Process Res. Dev. 2006, 10, 64– 69, DOI: 10.1021/op058013q Google Scholar 21Peris-Díaz, M. D.; Sentandreu, M. A.; Sentandreu, E. Multiobjective Optimization of Liquid Chromatography–Triple-Quadrupole Mass Spectrometry Analysis of Underivatized Human Urinary Amino Acids through Chemometrics. Anal. Bioanal. Chem. 2018, 410, 4275– 4284, DOI: 10.1007/s00216-018-1083-x Google Scholar 22Sagués, F.; Epstein, I. R. Nonlinear Chemical Dynamics. Dalton Trans. 2003, 1201– 1217, DOI: 10.1039/b210932h Google Scholar 23Owen, M. R.; Luscombe, C.; Lai, L.-W.; Godbert, S.; Crookes, D. L.; Emiabata-Smith, D. Efficiency by Design: Optimisation in Process Research. Org. Process Res. Dev. 2001, 5, 308– 323, DOI: 10.1021/op000024q Google Scholar 24Wahid, Z.; Nadir, N. Improvement of One Factor at
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
a Time through Design of Experiments. World Appl. Sci. J. 2013, 21, 56– 61 Google Scholar 25Abtahi, B.; Tavakol, H. Choline Chloride-Urea Deep Eutectic Solvent as an Efficient Media for the Synthesis of Propargylamines Via Organocuprate Intermediate. Appl. Organomet. Chem. 2020, 34, e5895 DOI: 10.1002/aoc.5895 Google Scholar 26Liu, J.; Ye, W.; Qing, X.; Wang, C. Solvent-Free Dabco-Mediated [3+ 2] Cycloadditions of Donor–Acceptor Cyclopropanes with Aldehydes: Strategy for Synthesis of Fully Substituted Furans. J. Org. Chem. 2016, 81, 7970– 7976, DOI: 10.1021/acs.joc.6b01259 Google Scholar 27Yamaguchi, T.; Kommagalla, Y.; Aihara, Y.; Chatani, N. Cobalt-Catalyzed Chelation Assisted C–H Allylation of Aromatic Amides with Unactivated Olefins. Chem. Commun. 2016, 52, 10129– 10132, DOI: 10.1039/C6CC05330K Google Scholar 28Gupta, S.; Chandna, N.; Dubey, P.; Singh, A. K.; Jain, N. Go–Cu 7 S 4 Catalyzed Ortho-Aminomethylation of Phenol Derivatives with N, N-Dimethylbenzylamines: Site-Selective Oxidative Cdc. Chem. Commun. 2018, 54, 7511– 7514, DOI: 10.1039/C8CC03396J Google Scholar 29Bayana, M.; Blacker, A. J.; Clayton, A. D.; Jolley, K. E.; Labes, R.; Taylor, C. J.; Reynolds, W. A Practical Experiment to Teach Students Continuous Flow and Physico-Chemical Methods: Acetylation of Ethylene Diamine in Liquid Bi-Phase. J. Flow. Chem. 2021, 11, 31– 36, DOI: 10.1007/s41981-020-00114-5 Google Scholar 30Shen, Y.; Borowski, J. E.; Hardy, M. A.; Sarpong, R.; Doyle, A. G.; Cernak, T. Automation and Computer-Assisted Planning for Chemical Synthesis. Nat. Rev. Methods Primers. 2021, 1, 23, DOI: 10.1038/s43586-021-00022-5 Google Scholar 31Hardwick, T.; Ahmed, N. Digitising Chemical Synthesis in Automated and Robotic Flow. Chem. Sci. 2020, 11, 11973– 11988, DOI: 10.1039/D0SC04250A Google Scholar 32Weissman, S. A.; Anderson, N. G. Design of Experiments (Doe) and Process Optimization. A Review of Recent Publications. Org. Process Res. Dev. 2015, 19, 1605– 1633, DOI: 10.1021/op500169m Google Scholar 33Elliott, P.; Billingham, S.; Bi, J.; Zhang, H. Quality by Design for Biopharmaceuticals: A Historical Review and Guide for Implementation.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Pharm. Bioprocess. 2013, 1, 105– 122, DOI: 10.4155/pbp.13.6 Google Scholar 34Tye, H. Application of Statistical ‘Design of Experiments’ Methods in Drug Discovery. Drug Discovery Today. 2004, 9, 485– 491, DOI: 10.1016/S1359-6446(04)03086-7 Google Scholar 35Dhoot, A. S.; Fernandes, G. J.; Naha, A.; Rathnanand, M.; Kumar, L. Design of Experiments in Pharmaceutical Development. Pharm. Chem. J. 2019, 53, 730– 735, DOI: 10.1007/s11094-019-02070-4 Google Scholar 36Denmark, S. E.; Butler, C. R. Vinylation of Aromatic Halides Using Inexpensive Organosilicon Reagents. Illustration of Design of Experiment Protocols. J. Am. Chem. Soc. 2008, 130, 3690– 3704, DOI: 10.1021/ja7100888 Google Scholar 37 MODDE ─ Explore. Improve. Advance!; Sartorius AG, 2022; (accessed 2023-01-20). Google Scholar 38 Statistical Software for Insight-Driven Improvement Seeing is Understanding; JMP Statistical Discovery LLC, 2023 (accessed 2023-01-20). Google Scholar 39 Design Expert, version 13; STAT-EASE, 2023; (accessed 2023-01-20). Google Scholar 40Bjørsvik, H.-R.; Liguori, L.; Minisci, F. High Selectivity in the Oxidation of Mandelic Acid Derivatives and in O-Methylation of Protocatechualdehyde: New Processes for Synthesis of Vanillin, I So-Vanillin, and Heliotropin. Org. Process Res. Dev. 2000, 4, 534– 543, DOI: 10.1021/op0000529 Google Scholar 41Nieuwland, P. J.; Segers, R.; Koch, K.; van Hest, J. C.; Rutjes, F. P. Fast Scale-up Using Microreactors: Pyrrole Synthesis from Micro to Production Scale. Org. Process Res. Dev. 2011, 15, 783– 787, DOI: 10.1021/op100338z Google Scholar 42Delville, M. M.; Nieuwland, P. J.; Janssen, P.; Koch, K.; van Hest, J. C.; Rutjes, F. P. Continuous Flow Azide Formation: Optimization and Scale-Up. Chem. Eng. J. 2011, 167, 556– 559, DOI: 10.1016/j.cej.2010.08.087 Google Scholar 43Cao, B.; Adutwum, L. A.; Oliynyk, A. O.; Luber, E. J.; Olsen, B. C.; Mar, A.; Buriak, J. M. How to Optimize Materials and Devices Via Design of Experiments and Machine Learning: Demonstration Using Organic Photovoltaics. ACS Nano 2018, 12, 7434– 7444, DOI: 10.1021/acsnano.8b04726 Google Scholar 44Eckert,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
T.; Klein, F. C.; Frieler, P.; Thunich, O.; Abetz, V. Experimental Design in Polymer Chemistry─a Guide Towards True Optimization of a Raft Polymerization Using Design of Experiments (Doe). Polymers. 2021, 13, 3147, DOI: 10.3390/polym13183147 Google Scholar 45Fukuda, I. M.; Pinto, C. F. F.; Moreira, C. d. S.; Saviano, A. M.; Lourenço, F. R. Design of Experiments (Doe) Applied to Pharmaceutical and Analytical Quality by Design (Qbd). Braz. J. Pharm. Sci. 2018, 54, e1006, DOI: 10.1590/s2175-97902018000001006 Google Scholar 46Holmes, N.; Akien, G. R.; Savage, R. J.; Stanetty, C.; Baxendale, I. R.; Blacker, A. J.; Taylor, B. A.; Woodward, R. L.; Meadows, R. E.; Bourne, R. A. Online Quantitative Mass Spectrometry for the Rapid Adaptive Optimisation of Automated Flow Reactors. React. Chem. Eng. 2016, 1, 96– 100, DOI: 10.1039/C5RE00083A Google Scholar 47Holmes, N.; Akien, G. R.; Blacker, A. J.; Woodward, R. L.; Meadows, R. E.; Bourne, R. A. Self-Optimisation of the Final Stage in the Synthesis of Egfr Kinase Inhibitor Azd9291 Using an Automated Flow Reactor. React. Chem. Eng. 2016, 1, 366– 371, DOI: 10.1039/C6RE00059B Google Scholar 48Mear, S. J.; Jamison, T. F. Diazotization of S-Sulfonyl-Cysteines. J. Org. Chem. 2019, 84, 15001– 15007, DOI: 10.1021/acs.joc.9b02630 Google Scholar 49Taylor, B.; Fernandez Barrat, C.; Woodward, R. L.; Inglesby, P. A. Synthesis of 2-Oxopropanethioamide for the Manufacture of Lanabecestat: A New Route for Control, Robustness, and Operational Improvements. Org. Process Res. Dev. 2017, 21, 1404– 1412, DOI: 10.1021/acs.oprd.7b00170 Google Scholar 50Raw, S. A.; Taylor, B. A.; Tomasi, S. A Simplified Process for the Manufacture of Azd0530, a Potent Src Kinase Inhibitor. Org. Process Res. Dev. 2011, 15, 688– 692, DOI: 10.1021/op200079g Google Scholar 51Bell, N. L.; Kupper, M.; Cronin, L. Design of Experiments for Optimization of Polyoxometalate Syntheses. Chem. Mater. 2021, 33, 7263– 7271, DOI: 10.1021/acs.chemmater.1c01401 Google Scholar 52Tye, H.; Whittaker, M. Use of a
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Design of Experiments Approach for the Optimisation of a Microwave Assisted Ugi Reaction. Org. Biomol. Chem. 2004, 2, 813– 815, DOI: 10.1039/b400298a Google Scholar 53Hobson, L. A.; Akiti, O.; Deshmukh, S. S.; Harper, S.; Katipally, K.; Lai, C. J.; Livingston, R. C.; Lo, E.; Miller, M. M.; Ramakrishnan, S.Development of a Scaleable Process for the Synthesis of a Next-Generation Statin. Org. Process Res. Dev. 2010, 14, 441– 458, DOI: 10.1021/op100010n Google Scholar 54Hajmowicz, H.; Wisialski, J.; Synoradzki, L. Direct Hydrolysis of O, O′-Dibenzoyltartaric Anhydride-an Alternative Method for Manufacture of the Same Acid: Tartaric Acid and Its O-Acyl Derivatives. Part 9. Org. Process Res. Dev. 2011, 15, 427– 434, DOI: 10.1021/op100315k Google Scholar 55Guercio, G.; Perboni, A.; Tinazzi, F.; Rovatti, L.; Provera, S. The Synthesis of Gv143253a: A Case Study for the Use of Analytical and Statistical Tools to Elucidate the Reaction Mechanism and to Optimize the Process. Org. Process Res. Dev. 2010, 14, 840– 848, DOI: 10.1021/op100097c Google Scholar 56Roemmele, R. C.; Christie, M. A. Development and Scale-up of an Optimized Route to the Peptide Boronic Acid, Cep-18770. Org. Process Res. Dev. 2013, 17, 422– 426, DOI: 10.1021/op400010u Google Scholar 57Moseley, J. D.; Moss, W. O.; Welham, M. J.; Ancell, C. L.; Banister, J.; Bowden, S. A.; Norton, G.; Young, M. J. A New Approach to Rapid Parallel Development of Four Neurokinin Antagonists. Part 2. Synthesis of Zd6021 Cyano Acid. Org. Process Res. Dev. 2003, 7, 58– 66, DOI: 10.1021/op020065h Google Scholar 58Klöpfer, V.; Eckl, R.; Floß, J.; Roth, P. M.; Reiser, O.; Barham, J. P. Catalyst-Free, Scalable Heterocyclic Flow Photocyclopropanation. Green Chem. 2021, 23, 6366– 6372, DOI: 10.1039/D1GC01624E Google Scholar 59Adlington, N. K.; Agnew, L. R.; Campbell, A. D.; Cox, R. J.; Dobson, A.; Barrat, C. F.; Gall, M. A.; Hicks, W.; Howell, G. P.; Jawor-Baczynska, A. Process Design
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
and Optimization in the Pharmaceutical Industry: A Suzuki–Miyaura Procedure for the Synthesis of Savolitinib. J. Org. Chem. 2019, 84, 4735– 4747, DOI: 10.1021/acs.joc.8b02351 Google Scholar 60ten Dam, J.; Kapteijn, F.; Djanashvili, K.; Hanefeld, U. Tuning Selectivity of Pt/Caco3 in Glycerol Hydrogenolysis─a Design of Experiments Approach. Catal. Commun. 2011, 13, 1– 5, DOI: 10.1016/j.catcom.2011.06.007 Google Scholar 61Arunajatesan, V.; McIntosh, S. L.; Cruz, M.; Renneke, R. F.; Chen, B., Rapid Optimization of Hydrogenation Process Variables by Combining Experimental Design and Parallel Reactors. In Catalysis of Organic Reactions; CRC Press, 2005; pp 219– 224. Google Scholar 62Weiberth, F. J.; Gill, H. S.; Jiang, Y.; Lee, G. E.; Lienard, P.; Pemberton, C.; Powers, M. R.; Subotkowski, W.; Tomasik, W.; Vanasse, B. J.; Yu, Y. Process Development of Selectively Benzoylated and Fluorinated Glycosyl Donors. Org. Process Res. Dev. 2010, 14, 623– 631, DOI: 10.1021/op100053k Google Scholar 63Charoo, N. A.; Shamsher, A. A.; Zidan, A. S.; Rahman, Z. Quality by Design Approach for Formulation Development: A Case Study of Dispersible Tablets. Int. J. Pharm. 2012, 423, 167– 178, DOI: 10.1016/j.ijpharm.2011.12.024 Google Scholar 64Naeem, M.; Rahman, N. U.; Khan, J. A.; Sehti, A.; Nawaz, Z. Development and Optimization of Microemulsion Formulation Using Box-Behnken Design for Enhanced Transdermal Delivery of Lornoxicam. Lat. Am. J. Pharm. 2013, 32, 1196– 1204 Google Scholar 65Kumar, P. M.; Ghosh, A. Development and Evaluation of Silver Sulfadiazine Loaded Microsponge Based Gel for Partial Thickness (Second Degree) Burn Wounds. Eur. J. Pharm. Sci. 2017, 96, 243– 254, DOI: 10.1016/j.ejps.2016.09.038 Google Scholar 66Malladi, M.; Jukanti, R. Formulation Development and Evaluation of a Novel Bi-Dependent Clarithromycin Gastroretentive Drug Delivery System Using Box-Behnken Design. J. Drug Delivery Sci. Technol. 2016, 35, 134– 145, DOI: 10.1016/j.jddst.2016.06.003 Google Scholar 67Chavez, P.-F.; Lebrun, P.; Sacré, P.-Y.; De Bleye, C.; Netchacovitch, L.; Cuypers, S.; Mantanus, J.; Motte, H.; Schubert, M.; Evrard,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
B.; HUbert, P.; Ziemons, E. Optimization of a Pharmaceutical Tablet Formulation Based on a Design Space Approach and Using Vibrational Spectroscopy as Pat Tool. Int. J. Pharm. 2015, 486, 13– 20, DOI: 10.1016/j.ijpharm.2015.03.025 Google Scholar 68Rahali, Y.; Saulnier, P.; Benoit, J.-P.; Bensouda, Y. Incorporating Vegetal Oils in Parenteral Nutrition Using Lipid Nanocapsules. J. Drug Delivery Sci. Technol. 2010, 20, 425– 429, DOI: 10.1016/S1773-2247(10)50074-1 Google Scholar 69Karatzas, A.; Politis, S.; Rekkas, D. Development of Rapidly Dissolving Pellets within the Quality by Design Approach. Drug Dev. Ind. Pharm. 2017, 43, 770– 779, DOI: 10.1080/03639045.2016.1220576 Google Scholar 70Kumar, P. M.; Ghosh, A. Development and Evaluation of Metronidazole Loaded Microsponge Based Gel for Superficial Surgical Wound Infections. J. Drug Delivery Sci. Technol. 2015, 30, 15– 29, DOI: 10.1016/j.jddst.2015.09.006 Google Scholar 71Maretti, E.; Rustichelli, C.; Romagnoli, M.; Balducci, A. G.; Buttini, F.; Sacchetti, F.; Leo, E.; Iannuccelli, V. Solid Lipid Nanoparticle Assemblies (Slnas) for an Anti-Tb Inhalation Treatment. A Design of Experiments Approach to Investigate the Influence of Pre-Freezing Conditions on the Powder Respirability. Int. J. Pharm. 2016, 511, 669– 679, DOI: 10.1016/j.ijpharm.2016.07.062 Google Scholar 72Kumar, L.; Reddy, M. S.; Managuli, R. S.; Pai K, G. Full Factorial Design for Optimization, Development and Validation of Hplc Method to Determine Valsartan in Nanoparticles. Saudi Pharm. J. 2015, 23, 549– 555, DOI: 10.1016/j.jsps.2015.02.001 Google Scholar 73Abdelgawad, M. A.; Owis, A. I.; Abdelwahab, N. S. Design and Optimization of a Reversed-Phase Hplc with Diode Array Detection Method for the Determination of Acemetacin and Its Toxic Impurities Using Experimental Design. Sep. Sci. Plus. 2018, 1, 244– 252, DOI: 10.1002/sscp.201700045 Google Scholar 74Naik, S.; Mullick, P.; Mutalik, S. P.; Hegde, A. R.; Lewis, S. A.; Bhat, K.; Rao, B. S. S.; Mutalik, S. Full Factorial Design for Development and Validation of a Stability-Indicating Rp-Hplc Method for the Estimation of Timolol
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Maleate in Surfactant-Based Elastic Nano-Vesicular Systems. J. Chromatogr. Sci. 2022, 60, 584– 594, DOI: 10.1093/chromsci/bmab101 Google Scholar 75Dai, S.; Xu, B.; Luo, G.; Li, J.; Xue, Z.; Shi, X.; Qiao, Y. Application of Design of Experiment and Design Space (Doe-Ds) Methodology for the Hplc Separation of Panax Notoginseng Saponins. Open Chem. Eng. J. 2015, 9, 47, DOI: 10.2174/1874123101509010047 Google Scholar 76Patel, N.; Jain, S.; Madan, P.; Lin, S. Application of Design of Experiments for Formulation Development and Mechanistic Evaluation of Iontophoretic Tacrine Hydrochloride Delivery. Drug Dev. Ind. Pharm. 2016, 42, 1894– 1902, DOI: 10.1080/03639045.2016.1181646 Google Scholar 77Badawi, M. A.; El-Khordagui, L. K. A Quality by Design Approach to Optimization of Emulsions for Electrospinning Using Factorial and D-Optimal Designs. Eur. J. Pharm. Sci. 2014, 58, 44– 54, DOI: 10.1016/j.ejps.2014.03.004 Google Scholar 78Shariare, M.; De Matas, M.; York, P.; Shao, Q. The Impact of Material Attributes and Process Parameters on the Micronisation of Lactose Monohydrate. Int. J. Pharm. 2011, 408, 58– 66, DOI: 10.1016/j.ijpharm.2011.01.047 Google Scholar 79Al-Gheethi, A.; Noman, E.; Radin Mohamed, R. M. S.; Ismail, N.; Bin Abdullah, A. H.; Mohd Kassim, A. H. Optimizing of Pharmaceutical Active Compounds Biodegradability in Secondary Effluents by Β-Lactamase from Bacillus Subtilis Using Central Composite Design. J. Hazard. Mater. 2019, 365, 883– 894, DOI: 10.1016/j.jhazmat.2018.11.068 Google Scholar 80Picker, T. S. Digitalization in Laboratories of the Pharmaceutical Industry; Solid State Development Processing of Pharmaceutical Molecules: Salts, Cocrystals, Polymorphism: Wiley, 2021; Vol. 79. Chapter 8. DOI: 10.1002/9783527823048.ch8 Google Scholar 81Csóka, I.; Pallagi, E.; Paál, T. L. Extension of Quality-by-Design Concept to the Early Development Phase of Pharmaceutical R&D Processes. Drug Discovery Today. 2018, 23, 1340– 1343, DOI: 10.1016/j.drudis.2018.03.012 Google Scholar 82Zidan, A. S.; Sammour, O. A.; Hammad, M. A.; Megrab, N. A.; Habib, M. J.; Khan, M. A. Quality by Design: Understanding the Formulation Variables of a
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Cyclosporine a Self-Nanoemulsified Drug Delivery Systems by Box–Behnken Design and Desirability Function. Int. J. Pharm. 2007, 332, 55– 63, DOI: 10.1016/j.ijpharm.2006.09.060 Google Scholar 83Kulpreechanan, N.; Sorasitthiyanukarn, F. N. Preparation and Optimization of Capsaicin-Loaded Chitosan Nanoparticles as Delivery Systems Using Box-Behnken Design. Mater. Sci. Forum. 2020, 998, 277– 282, DOI: 10.4028/www.scientific.net/MSF.998.277 Google Scholar 84Sahu, B. P.; Das, M. K. Optimization of Felodipine Nanosuspensions Using Full Factorial Design. Int. J. PharmTech Res. 2013, 5, 553– 561 Google Scholar 85McAllister, G. D.; Parsons, A. F. Going Green in Process Chemistry: Optimizing an Asymmetric Oxidation Reaction to Synthesize the Antiulcer Drug Esomeprazole. J. Chem. Educ. 2019, 96, 2617– 2621, DOI: 10.1021/acs.jchemed.9b00350 Google Scholar 86St Germain, E. J.; Horowitz, A. S.; Rucco, D.; Rezler, E. M.; Lepore, S. D. Teaching Experiment to Elucidate a Cation– Π Effect in an Alkyne Cycloaddition Reaction and Illustrate Hypothesis-Driven Design of Experiments. J. Chem. Educ. 2017, 94, 240– 243, DOI: 10.1021/acs.jchemed.6b00318 Google Scholar 87Telford, J. K. A Brief Introduction to Design of Experiments. Johns Hopkins APL Technol. Dig. 2007, 27, 224– 232 Google Scholar 88Poroch-Seritan, M.; Gutt, S.; Gutt, G.; Cretescu, I.; Cojocaru, C.; Severin, T. Design of Experiments for Statistical Modeling and Multi-Response Optimization of Nickel Electroplating Process. Chem. Eng. Res. Des. 2011, 89, 136– 147, DOI: 10.1016/j.cherd.2010.05.010 Google Scholar 89Carey, F. A.; Sundberg, R. J. Study and Description of Organic Reaction Mechanisms. In Advanced Organic Chemistry: Part A, Structure and Mechanisms, 5th ed.; Springer, 2000; pp 187– 261. Google Scholar 90Willis, M. J.; von Stosch, M. Inference of Chemical Reaction Networks Using Mixed Integer Linear Programming. Comput. Chem. Eng. 2016, 90, 31– 43, DOI: 10.1016/j.compchemeng.2016.04.019 Google Scholar 91Sayikli, C.; Bagci, E. Z. Limitations of Using Mass Action Kinetics Method in Modeling Biochemical Systems: Illustration for a Second Order Reaction. International Conference on Computational Science and Its Applications.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
2011, 6786, 521– 526, DOI: 10.1007/978-3-642-21934-4_42 Google Scholar 92Taylor, C. J.; Seki, H.; Dannheim, F. M.; Willis, M. J.; Clemens, G.; Taylor, B. A.; Chamberlain, T. W.; Bourne, R. A. An Automated Computational Approach to Kinetic Model Discrimination and Parameter Estimation. React. Chem. Eng. 2021, 6, 1404– 1411, DOI: 10.1039/D1RE00098E Google Scholar 93Ashworth, I. W.; Bush, E.; Chan, L. C.; Cherryman, J.; Cox, B. G.; Muir, J.; Korupoju, S. R.; Keshwan, J. Where Has My Acid Gone? Understanding the Self-Catalyzed Esterification of Maleic Acid in Methanol During Salt Formation. Org. Process Res. Dev. 2012, 16, 1646– 1651, DOI: 10.1021/op3001959 Google Scholar 94Minegishi, S.; Loos, R.; Kobayashi, S.; Mayr, H. Kinetics of the Reactions of Halide Anions with Carbocations: Quantitative Energy Profiles for Sn1 Reactions. J. Am. Chem. Soc. 2005, 127, 2641– 2649, DOI: 10.1021/ja045562n Google Scholar 95Somerville, R. J.; Hale, L. V.; Gómez-Bengoa, E.; Burés, J.; Martin, R. Intermediacy of Ni–Ni Species in Sp2 C–O Bond Cleavage of Aryl Esters: Relevance in Catalytic C–Si Bond Formation. J. Am. Chem. Soc. 2018, 140, 8771– 8780, DOI: 10.1021/jacs.8b04479 Google Scholar 96Shi, L.; Wang, Y.; Yan, B.; Song, W.; Shao, D.; Lu, A.-H. Progress in Selective Oxidative Dehydrogenation of Light Alkanes to Olefins Promoted by Boron Nitride Catalysts. Chem. Commun. 2018, 54, 10936– 10946, DOI: 10.1039/C8CC04604B Google Scholar 97Poisson, P.-A.; Tran, G.; Besnard, C.; Mazet, C. Nickel-Catalyzed Kumada Vinylation of Enol Phosphates: A Comparative Mechanistic Study. ACS Catal. 2021, 11, 15041– 15050, DOI: 10.1021/acscatal.1c04800 Google Scholar 98Zaborenko, N.; Bedore, M. W.; Jamison, T. F.; Jensen, K. F. Kinetic and Scale-up Investigations of Epoxide Aminolysis in Microreactors at High Temperatures and Pressures. Org. Process Res. Dev. 2011, 15, 131– 139, DOI: 10.1021/op100252m Google Scholar 99Tufvesson, P.; Lima-Ramos, J.; Haque, N. A.; Gernaey, K. V.; Woodley, J. M. Advances in the Process Development of Biocatalytic
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Processes. Org. Process Res. Dev. 2013, 17, 1233– 1238, DOI: 10.1021/op4001675 Google Scholar 100Ashworth, I. W.; Chan, L. C.; Cox, B. G.; McFarlane, I. M.; Phillips, A. R. Understanding the Alkylation of a Phenol by 1-(3-Chloropropyl) Pyrrolidine: Evidence for the Intermediacy of an Azetidinium Ion. J. Org. Chem. 2019, 84, 4754– 4762, DOI: 10.1021/acs.joc.8b02458 Google Scholar 101Hall, B. L.; Taylor, C. J.; Labes, R.; Massey, A. F.; Menzel, R.; Bourne, R. A.; Chamberlain, T. W. Autonomous Optimisation of a Nanoparticle Catalysed Reduction Reaction in Continuous Flow. Chem. Commun. 2021, 57, 4926– 4929, DOI: 10.1039/D1CC00859E Google Scholar 102Widstrom, A. L.; Lear, B. J. Structural and Solvent Control over Activation Parameters for a Pair of Retro Diels-Alder Reactions. Sci. Rep. 2019, 9, 18267, DOI: 10.1038/s41598-019-54156-4 Google Scholar 103Laidler, K. J. The Development of the Arrhenius Equation. J. Chem. Educ. 1984, 61, 494, DOI: 10.1021/ed061p494 Google Scholar 104Rode, C.; Vaidya, M.; Chaudhari, R. Synthesis of P-Aminophenol by Catalytic Hydrogenation of Nitrobenzene. Org. Process Res. Dev. 1999, 3, 465– 470, DOI: 10.1021/op990040r Google Scholar 105Bay, S.; Baumeister, T.; Hashmi, A. S. K.; Röder, T. Safe and Fast Flow Synthesis of Functionalized Oxazoles with Molecular Oxygen in a Microstructured Reactor. Org. Process Res. Dev. 2016, 20, 1297– 1304, DOI: 10.1021/acs.oprd.6b00118 Google Scholar 106Moore, J. S.; Smith, C. D.; Jensen, K. F. Kinetics Analysis and Automated Online Screening of Aminocarbonylation of Aryl Halides in Flow. React. Chem. Eng. 2016, 1, 272– 279, DOI: 10.1039/C6RE00007J Google Scholar 107Cox, B. G. Modern Liquid Phase Kinetics; Oxford University Press, 1994: pp 6– 7. Google Scholar 108Johnson, K. A. A Century of Enzyme Kinetic Analysis, 1913 to 2013. FEBS Lett. 2013, 587, 2753– 2766, DOI: 10.1016/j.febslet.2013.07.012 Google Scholar 109Atolaiye, B.; Edward-Ekpu, D.; Etua, E. Kinetics of the Redox Reaction of Benzaldehyde and Dichromate Ions in Aqueous Acidic Medium. J. Appl.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Sci. Environ. Manage. 2019, 23, 1309– 1313, DOI: 10.4314/jasem.v23i7.18 Google Scholar 110Puchelle, V.; Du, H.; Illy, N.; Guégan, P. Polymerization of Epoxide Monomers Promoted by T Bup 4 Phosphazene Base: A Comparative Study of Kinetic Behavior. Polymer Chemistry. 2020, 11, 3585– 3592, DOI: 10.1039/D0PY00437E Google Scholar 111Blackmond, D. G. Reaction Progress Kinetic Analysis: A Powerful Methodology for Mechanistic Studies of Complex Catalytic Reactions. Angew. Chem., Int. Ed. 2005, 44, 4302– 4320, DOI: 10.1002/anie.200462544 Google Scholar 112Schotten, C.; Taylor, C. J.; Bourne, R. A.; Chamberlain, T. W.; Nguyen, B. N.; Kapur, N.; Willans, C. E. Alternating Polarity for Enhanced Electrochemical Synthesis. React. Chem. Eng. 2021, 6, 147– 151, DOI: 10.1039/D0RE00399A Google Scholar 113Bezemer, E.; Rutan, S. C. Multivariate Curve Resolution with Non-Linear Fitting of Kinetic Profiles. Chemom. Intell. Lab. Syst. 2001, 59, 19– 31, DOI: 10.1016/S0169-7439(01)00141-1 Google Scholar 114Bezemer, E.; Rutan, S. C. Analysis of Three-and Four-Way Data Using Multivariate Curve Resolution-Alternating Least Squares with Global Multi-Way Kinetic Fitting. Chemom. Intell. Lab. Syst. 2006, 81, 82– 93, DOI: 10.1016/j.chemolab.2005.10.005 Google Scholar 115Taylor, C. J.; Booth, M.; Manson, J. A.; Willis, M. J.; Clemens, G.; Taylor, B. A.; Chamberlain, T. W.; Bourne, R. A. Rapid, Automated Determination of Reaction Models and Kinetic Parameters. Chem. Eng. J. 2021, 413, 127017, DOI: 10.1016/j.cej.2020.127017 Google Scholar 116Krzelj, V.; Ferreira Liberal, J.; Papaioannou, M.; van der Schaaf, J.; Neira d’Angelo, M. F. Kinetic Model of Xylose Dehydration for a Wide Range of Sulfuric Acid Concentrations. Ind. Eng. Chem. Res. 2020, 59, 11991– 12003, DOI: 10.1021/acs.iecr.0c01197 Google Scholar 117Mandapati, R. N.; Ghodke, P. K. Kinetics of Pyrolysis of Cotton Stalk Using Model-Fitting and Model-Free Methods. Fuel. 2021, 303, 121285, DOI: 10.1016/j.fuel.2021.121285 Google Scholar 118Chaudhuri, A.; Backx, W. G.; Moonen, L. L.; Molenaar, C. W.; Winkenweder, W.; Ljungdahl, T.; van der Schaaf, J. Kinetics and Intensification of Tertiary
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Amine N-Oxidation: Towards a Solventless, Continuous and Sustainable Process. Chem. Eng. J. 2021, 416, 128962, DOI: 10.1016/j.cej.2021.128962 Google Scholar 119Nestler, F.; Müller, V.; Ouda, M.; Hadrich, M.; Schaadt, A.; Bajohr, S.; Kolb, T. A Novel Approach for Kinetic Measurements in Exothermic Fixed Bed Reactors: Advancements in Non-Isothermal Bed Conditions Demonstrated for Methanol Synthesis. React. Chem. Eng. 2021, 6, 1092– 1107, DOI: 10.1039/D1RE00071C Google Scholar 120Chen, Q.; Zheng, J.; Yang, Q.; Dang, Z.; Zhang, L. Effect of Carbon Chain Structure on the Phthalic Acid Esters (Paes) Adsorption Mechanism by Mesoporous Cellulose Biochar. Chem. Eng. J. 2019, 362, 383– 391, DOI: 10.1016/j.cej.2019.01.052 Google Scholar 121 Compunetics; Compunetics, 2023; (accessed 2023-01-20). Google Scholar 122 Berkeley Madonna, version 10; Berkeley Madonna, 2023; (accessed 2023-01-20). Google Scholar 123 DynoChem; Scale-up Systems Ltd, 2023; (Accessed 2023-01-20). Google Scholar 124 COPASI: Biochemical System Simulator; COPASI, 2023; (accessed 2023-01-20). Google Scholar 125Wang, D. M.; Feng, W.; Wu, Y.; Liu, T.; Wang, P. Redox-Neutral Nickel (Ii) Catalysis: Hydroarylation of Unactivated Alkenes with Arylboronic Acids. Angew. Chem., Int. Ed. 2020, 59, 20399– 20404, DOI: 10.1002/anie.202009195 Google Scholar 126Turek, A. K.; Sak, M. H.; Miller, S. J. Kinetic Analysis of a Cysteine-Derived Thiyl-Catalyzed Asymmetric Vinylcyclopropane Cycloaddition Reflects Numerous Attractive Noncovalent Interactions. J. Am. Chem. Soc. 2021, 143, 16173– 16183, DOI: 10.1021/jacs.1c07323 Google Scholar 127Lin, Q.; Diao, T. Mechanism of Ni-Catalyzed Reductive 1, 2-Dicarbofunctionalization of Alkenes. J. Am. Chem. Soc. 2019, 141, 17937– 17948, DOI: 10.1021/jacs.9b10026 Google Scholar 128Shevick, S. L.; Obradors, C.; Shenvi, R. A. Mechanistic Interrogation of Co/Ni-Dual Catalyzed Hydroarylation. J. Am. Chem. Soc. 2018, 140, 12056– 12068, DOI: 10.1021/jacs.8b06458 Google Scholar 129Wendlandt, A. E.; Vangal, P.; Jacobsen, E. N. Quaternary Stereocentres Via an Enantioconvergent Catalytic Sn 1 Reaction. Nature. 2018, 556, 447– 451, DOI: 10.1038/s41586-018-0042-1 Google Scholar 130Tran, G. l.; Shao, W.; Mazet, C. m.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Ni-Catalyzed Enantioselective Intermolecular Hydroamination of Branched 1, 3-Dienes Using Primary Aliphatic Amines. J. Am. Chem. Soc. 2019, 141, 14814– 14822, DOI: 10.1021/jacs.9b07253 Google Scholar 131Jansen, D.; Gramüller, J.; Niemeyer, F.; Schaller, T.; Letzel, M. C.; Grimme, S.; Zhu, H.; Gschwind, R. M.; Niemeyer, J. What Is the Role of Acid–Acid Interactions in Asymmetric Phosphoric Acid Organocatalysis? A Detailed Mechanistic Study Using Interlocked and Non-Interlocked Catalysts. Chem. Sci. 2020, 11, 4381– 4390, DOI: 10.1039/D0SC01026J Google Scholar 132Nielsen, C. D.-T.; Burés, J. Visual Kinetic Analysis. Chem. Sci. 2019, 10, 348– 353, DOI: 10.1039/C8SC04698K Google Scholar 133Burés, J. Variable Time Normalization Analysis: General Graphical Elucidation of Reaction Orders from Concentration Profiles. Angew. Chem., Int. Ed. 2016, 128, 16318– 16321, DOI: 10.1002/ange.201609757 Google Scholar 134Burés, J. A Simple Graphical Method to Determine the Order in Catalyst. Angew. Chem., Int. Ed. 2016, 55, 2028– 2031, DOI: 10.1002/anie.201508983 Google Scholar 135Kawakita, K.; Beaumier, E. P.; Kakiuchi, Y.; Tsurugi, H.; Tonks, I. A.; Mashima, K. Bis (Imido) Vanadium (V)-Catalyzed [2+ 2+ 1] Coupling of Alkynes and Azobenzenes Giving Multisubstituted Pyrroles. J. Am. Chem. Soc. 2019, 141, 4194– 4198, DOI: 10.1021/jacs.8b13390 Google Scholar 136Pearce, A. J.; See, X. Y.; Tonks, I. A. Oxidative Nitrene Transfer from Azides to Alkynes Via Ti (Ii)/Ti (Iv) Redox Catalysis: Formal [2+ 2+ 1] Synthesis of Pyrroles. Chem. Commun. 2018, 54, 6891– 6894, DOI: 10.1039/C8CC02623H Google Scholar 137Nishii, Y.; Ikeda, M.; Hayashi, Y.; Kawauchi, S.; Miura, M. Triptycenyl Sulfide: A Practical and Active Catalyst for Electrophilic Aromatic Halogenation Using N-Halosuccinimides. J. Am. Chem. Soc. 2020, 142, 1621– 1629, DOI: 10.1021/jacs.9b12672 Google Scholar 138Martínez-Carrión, A.; Howlett, M. G.; Alamillo-Ferrer, C.; Clayton, A. D.; Bourne, R. A.; Codina, A.; Vidal-Ferran, A.; Adams, R. W.; Burés, J. Kinetic Treatments for Catalyst Activation and Deactivation Processes Based on Variable Time Normalization Analysis. Angew. Chem., Int. Ed. 2019,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
131, 10295– 10299, DOI: 10.1002/ange.201903878 Google Scholar 139Martínez, Á. M.; Rodríguez, N.; Gómez-Arrayás, R.; Carretero, J. C. Cobalt-Catalyzed Ortho-C– H Functionalization/Alkyne Annulation of Benzylamine Derivatives: Access to Dihydroisoquinolines. Chem. - Eur. J. 2017, 23, 11669– 11676, DOI: 10.1002/chem.201702283 Google Scholar 140Echtermeyer, A.; Amar, Y.; Zakrzewski, J.; Lapkin, A. Self-Optimisation and Model-Based Design of Experiments for Developing a C–H Activation Flow Process. Beilstein J. Org. Chem. 2017, 13, 150– 163, DOI: 10.3762/bjoc.13.18 Google Scholar 141Franceschini, G.; Macchietto, S. Model-Based Design of Experiments for Parameter Precision: State of the Art. Chem. Eng. Sci. 2008, 63, 4846– 4872, DOI: 10.1016/j.ces.2007.11.034 Google Scholar 142Maheshwari, V.; Rangaiah, G. P.; Samavedham, L. Multiobjective Framework for Model-Based Design of Experiments to Improve Parameter Precision and Minimize Parameter Correlation. Ind. Eng. Chem. Res. 2013, 52, 8289– 8304, DOI: 10.1021/ie400133m Google Scholar 143Routh, M. W.; Swartz, P. A.; Denton, M. B. Performance of the Super Modified Simplex. Anal. Chem. 1977, 49, 1422– 1428, DOI: 10.1021/ac50017a033 Google Scholar 144Krishnadasan, S.; Brown, R. J. C.; DeMello, A. J.; DeMello, J. C. Intelligent Routes to the Controlled Synthesis of Nanoparticles. Lab Chip. 2007, 7, 1434, DOI: 10.1039/b711412e Google Scholar 145McMullen, J. P.; Jensen, K. F. An Automated Microfluidic System for Online Optimization in Chemical Synthesis. Org. Process Res. Dev. 2010, 14, 1169– 1176, DOI: 10.1021/op100123e Google Scholar 146McMullen, J. P.; Jensen, K. F. Rapid Determination of Reaction Kinetics with an Automated Microfluidic System. Org. Process Res. Dev. 2011, 15, 398– 407, DOI: 10.1021/op100300p Google Scholar 147Bourne, R. A.; Skilton, R. A.; Parrott, A. J.; Irvine, D. J.; Poliakoff, M. Adaptive Process Optimization for Continuous Methylation of Alcohols in Supercritical Carbon Dioxide. Org. Process Res. Dev. 2011, 15, 932– 938, DOI: 10.1021/op200109t Google Scholar 148Moore, J. S.; Jensen, K. F. Automated Multitrajectory Method for Reaction Optimization in a Microfluidic System Using Online Ir
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Analysis. Org. Process Res. Dev. 2012, 16, 1409– 1415, DOI: 10.1021/op300099x Google Scholar 149Christensen, M.; Yunker, L. P. E.; Adedeji, F.; Hase, F.; Roch, L. M.; Gensch, T.; dos Passos Gomes, G.; Zepel, T.; Sigman, M. S.; Aspuru-Guzik, A.; Hein, J. E. Data-Science Driven Autonomous Process Optimization. Commun. Chem. 2021, 4, 112, DOI: 10.1038/s42004-021-00550-x Google Scholar 150Reizman, B. J.; Wang, Y. M.; Buchwald, S. L.; Jensen, K. F. Suzuki-Miyaura Cross-Coupling Optimization Enabled by Automated Feedback. React. Chem. Eng. 2016, 1, 658– 666, DOI: 10.1039/C6RE00153J Google Scholar 151Fitzpatrick, D. E.; Battilocchio, C.; Ley, S. V. A Novel Internet-Based Reaction Monitoring, Control and Autonomous Self-Optimization Platform for Chemical Synthesis. Org. Process Res. Dev. 2016, 20, 386– 394, DOI: 10.1021/acs.oprd.5b00313 Google Scholar 152Echtermeyer, A.; Amar, Y.; Zakrzewski, J.; Lapkin, A. Self-Optimisation and Model-Based Design of Experiments for Developing a C-H Activation Flow Process. Beilstein J. Org. Chem. 2017, 13, 150– 163, DOI: 10.3762/bjoc.13.18 Google Scholar 153Clayton, A. D.; Schweidtmann, A. M.; Clemens, G.; Manson, J. A.; Taylor, C. J.; Niño, C. G.; Chamberlain, T. W.; Kapur, N.; Blacker, A. J.; Lapkin, A. A.; Bourne, R. A. Automated Self-Optimisation of Multi-Step Reaction and Separation Processes Using Machine Learning. Chem. Eng. J. 2020, 384, 123340, DOI: 10.1016/j.cej.2019.123340 Google Scholar 154Mateos, C.; Nieves-Remacha, M. J.; Rincon, J. A. Automated Platforms for Reaction Self-Optimization in Flow. React. Chem. Eng. 2019, 4, 1536– 1544, DOI: 10.1039/C9RE00116F Google Scholar 155Chatterjee, S.; Guidi, M.; Seeberger, P. H.; Gilmore, K. Automated Radial Synthesis of Organic Molecules. Nature. 2020, 579, 379– 384, DOI: 10.1038/s41586-020-2083-5 Google Scholar 156Christensen, M.; Yunker, L. P. E.; Shiri, P.; Zepel, T.; Prieto, P. L.; Grunert, S.; Bork, F.; Hein, J. E. Automation Isn’t Automatic. Chem. Sci.. 2021, 12, 15473– 15490, DOI: 10.1039/D1SC04588A Google Scholar 157Cole, K. P.; Reizman, B. J.; Hess, M.; Groh, J. M.; Laurila, M.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
E.; Cope, R. F.; Campbell, B. M.; Forst, M. B.; Burt, J. L.; Maloney, T. D.Small-Volume Continuous Manufacturing of Merestinib. Part 1. Process Development and Demonstration. Org. Process Res. Dev. 2019, 23, 858– 869, DOI: 10.1021/acs.oprd.8b00441 Google Scholar 158Dong, Z.; Wen, Z.; Zhao, F.; Kuhn, S.; Noël, T. Scale-up of Micro- and Milli-Reactors: An Overview of Strategies, Design Principles and Applications. Chem. Eng. Sci. 2021, 10, 100097, DOI: 10.1016/j.cesx.2021.100097 Google Scholar 159Reizman, B. J.; Jensen, K. F. Feedback in Flow for Accelerated Reaction Development. Acc. Chem. Res. 2016, 49, 1786– 1796, DOI: 10.1021/acs.accounts.6b00261 Google Scholar 160Chen, Y. Z.; Sabio, J. C.; Hartman, R. L. When Solids Stop Flow Chemistry in Commercial Tubing. J. Flow. Chem. 2015, 5, 166– 171, DOI: 10.1556/1846.2015.00001 Google Scholar 161Abolhasani, M.; Oskooei, A.; Klinkova, A.; Kumacheva, E.; Gunther, A. Shaken, and Stirred: Oscillatory Segmented Flow for Controlled Size-Evolution of Colloidal Nanomaterials. Lab Chip. 2014, 14, 2309– 2318, DOI: 10.1039/C4LC00131A Google Scholar 162Reizman, B. J.; Jensen, K. F. Feedback in Flow for Accelerated Reaction Development. Acc. Chem. Res. 2016, 49, 1786– 1796, DOI: 10.1021/acs.accounts.6b00261 Google Scholar 163Baumgartner, L. M.; Dennis, J. M.; White, N. A.; Buchwald, S. L.; Jensen, K. F. Use of a Droplet Platform to Optimize Pd-Catalyzed C-N Coupling Reactions Promoted by Organic Bases. Org. Process Res. Dev. 2019, 23, 1594– 1601, DOI: 10.1021/acs.oprd.9b00236 Google Scholar 164Pomberger, A.; Mo, Y.; Nandiwale, K. Y.; Schultz, V. L.; Duvadie, R.; Robinson, R. I.; Altinoglu, E. I.; Jensen, K. F. A Continuous Stirred-Tank Reactor (Cstr) Cascade for Handling Solid-Containing Photochemical Reactions. Org. Process Res. Dev. 2019, 23, 2699– 2706, DOI: 10.1021/acs.oprd.9b00378 Google Scholar 165Bianchi, P.; Williams, J. D.; Kappe, C. O. Oscillatory Flow Reactors for Synthetic Chemistry Applications. J. Flow. Chem. 2020, 10, 475– 490, DOI: 10.1007/s41981-020-00105-6 Google Scholar 166Coley, C. W.; Abolhasani, M.; Lin, H. K.; Jensen,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
K. F. Material-Efficient Microfluidic Platform for Exploratory Studies of Visible-Light Photoredox Catalysis. Angew. Chem., Int. Ed. 2017, 56, 9847– 9850, DOI: 10.1002/anie.201705148 Google Scholar 167Hsieh, H. W.; Coley, C. W.; Baumgartner, L. M.; Jensen, K. F.; Robinson, R. I. Photoredox Iridium-Nickel Dual-Catalyzed Decarboxylative Arylation Cross-Coupling: From Batch to Continuous Flow Via Self-Optimizing Segmented Flow Reactor. Org. Process Res. Dev. 2018, 22, 542– 550, DOI: 10.1021/acs.oprd.8b00018 Google Scholar 168Zhu, C.; Raghuvanshi, K.; Coley, C. W.; Mason, D.; Rodgers, J.; Janka, M. E.; Abolhasani, M. Flow Chemistry-Enabled Studies of Rhodium-Catalyzed Hydroformylation Reactions. Chem. Commun. 2018, 54, 8567– 8570, DOI: 10.1039/C8CC04650F Google Scholar 169Hone, C. A.; Kappe, C. O. Towards the Standardization of Flow Chemistry Protocols for Organic Reactions. Chem.: Methods. 2021, 1, 454– 467, DOI: 10.1002/cmtd.202100059 Google Scholar 170Sagmeister, P.; Williams, J. D.; Hone, C. A.; Kappe, C. O. Laboratory of the Future: A Modular Flow Platform with Multiple Integrated Pat Tools for Multistep Reactions. React. Chem. Eng. 2019, 4, 1571– 1578, DOI: 10.1039/C9RE00087A Google Scholar 171Granda, J. M.; Donina, L.; Dragone, V.; Long, D.-L.; Cronin, L. Controlling an Organic Synthesis Robot with Machine Learning to Search for New Reactivity. Nature. 2018, 559, 377– 381, DOI: 10.1038/s41586-018-0307-8 Google Scholar 172Sans, V.; Porwol, L.; Dragone, V.; Cronin, L. A Self Optimizing Synthetic Organic Reactor System Using Real-Time in-Line Nmr Spectroscopy. Chem. Sci. 2015, 6, 1258– 1264, DOI: 10.1039/C4SC03075C Google Scholar 173Clayton, A. D.; Manson, J. A.; Taylor, C. J.; Chamberlain, T. W.; Taylor, B. A.; Clemens, G.; Bourne, R. A. Algorithms for the Self-Optimisation of Chemical Reactions. React. Chem. Eng. 2019, 4, 1545– 1554, DOI: 10.1039/C9RE00209J Google Scholar 174Huyer, W.; Neumaier, A. Snobfit - Stable Noisy Optimization by Branch and Fit. ACM Trans. Math. Softw. 2008, 35, 1– 25, DOI: 10.1145/1377612.1377613 Google Scholar 175Schweidtmann, A. M.; Clayton, A. D.; Holmes, N.; Bradford, E.;
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Bourne, R. A.; Lapkin, A. A. Machine Learning Meets Continuous Flow Chemistry: Automated Optimization Towards the Pareto Front of Multiple Objectives. Chem. Eng. J. 2018, 352, 277– 282, DOI: 10.1016/j.cej.2018.07.031 Google Scholar 176Amar, Y.; Schweidtmann, A. M.; Deutsch, P.; Cao, L.; Lapkin, A. Machine Learning and Molecular Descriptors Enable Rational Solvent Selection in Asymmetric Catalysis. Chem. Sci. 2019, 10, 6697– 6706, DOI: 10.1039/C9SC01844A Google Scholar 177Manson, J. A.; Chamberlain, T. W.; Bourne, R. A. Mvmoo: Mixed Variable Multi-Objective Optimisation. Journal of Global Optimization. 2021, 80, 865– 886, DOI: 10.1007/s10898-021-01052-9 Google Scholar 178Hase, F.; Aldeghi, M.; Hickman, R. J.; Roch, L. M.; Aspuru-Guzik, A. Gryffin: An Algorithm for Bayesian Optimization of Categorical Variables Informed by Expert Knowledge. Appl. Phys. Rev. 2021, 8, 031406, DOI: 10.1063/5.0048164 Google Scholar 179Shahriari, B.; Swersky, K.; Wang, Z. Y.; Adams, R. P.; de Freitas, N. Taking the Human out of the Loop: A Review of Bayesian Optimization. P Ieee. 2016, 104, 148– 175, DOI: 10.1109/JPROC.2015.2494218 Google Scholar 180Baumgartner, L. M.; Coley, C. W.; Reizman, B. J.; Gao, K. W.; Jensen, K. F. Optimum Catalyst Selection over Continuous and Discrete Process Variables with a Single Droplet Microfluidic Reaction Platform. React. Chem. Eng. 2018, 3, 301– 311, DOI: 10.1039/C8RE00032H Google Scholar 181Reizman, B. J. Automated Feedback in Flow for Accelerated Reaction Screening, Optimization, and Kinetic Parameter Estimation. Ph.D. Thesis. Massachusetts Institute of Technology, 2015. Google Scholar 182Kershaw, O. J.; Clayton, A. D.; Manson, J. A.; Barthelme, A.; Pavey, J.; Peach, P.; Mustakis, J.; Howard, R. M.; Chamberlain, T. W.; Warren, N. J.; Bourne, R. A. Machine Learning Directed Multi-Objective Optimization of Mixed Variable Chemical Systems. Chem. Eng. J. 2023, 451, 138443, DOI: 10.1016/j.cej.2022.138443 Google Scholar 183Häse, F.; Roch, L. M.; Aspuru-Guzik, A. Chimera: Enabling Hierarchy Based Multi-Objective Optimization for Self-Driving Laboratories. Chem. Sci. 2018, 9, 7642– 7655, DOI:
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
10.1039/C8SC02239A Google Scholar 184Bradford, E.; Schweidtmann, A. M.; Lapkin, A. Efficient Multiobjective Optimization Employing Gaussian Processes, Spectral Sampling and a Genetic Algorithm. Journal of Global Optimization. 2018, 71, 407– 438, DOI: 10.1007/s10898-018-0609-2 Google Scholar 185McKay, M. D. Latin Hypercube Sampling as a Tool in Uncertainty Analysis of Computer Models. In WSC '92: Proceedings of the 24th Conference on Winter Simulation , 1992; pp 557– 564. DOI: 10.1145/167293.167637 Google Scholar 186Felton, K. C.; Rittig, J. G.; Lapkin, A. A. Summit: Benchmarking Machine Learning Methods for Reaction Optimisation. Chem.: Methods. 2021, 1, 116– 122, DOI: 10.1002/cmtd.202000051 Google Scholar 187Häse, F.; Aldeghi, M.; Hickman, R. J.; Roch, L. M.; Christensen, M.; Liles, E.; Hein, J. E.; Aspuru-Guzik, A. Olympus: A Benchmarking Framework for Noisy Optimization and Experiment Planning. Machine Learning Science and Technology. 2021, 2, 035021, DOI: 10.1088/2632-2153/abedc8 Google Scholar 188Epps, R. W.; Volk, A. A.; Reyes, K. G.; Abolhasani, M. Accelerated Ai Development for Autonomous Materials Synthesis in Flow. Chem. Sci. 2021, 12, 6025– 6036, DOI: 10.1039/D0SC06463G Google Scholar 189Felton, K. C.; Wigh, D.; Lapkin, A. Multi-Task Bayesian Optimization of Chemical Reactions. In Machine Learning for Molecules Workshop at NeurIPS 2020 , 2020. Google Scholar 190Hickman, R. J.; Ruža, J.; Roch, L. M.; Tribukait, H.; Garcia, A. D. Equipping Data-Driven Experiment Planning for Self-Driving Laboratories with Semantic Memory: Case Studies of Transfer Learning in Chemical Reaction Optimization. ChemRxiv , 2022, 2022-jt4sm DOI: 10.26434/chemrxiv-2022-jt4sm . Google Scholar 191Pomberger, A.; Jose, N.; Walz, D.; Meissner, J.; Holze, C.; Kopczynski, M.; Müller-Bischof, P.; Lapkin, A. Automated Ph Adjustment Driven by Robotic Workflows and Active Machine Learning. Chem. Eng. J. 2023, 451, 139099, DOI: 10.1016/j.cej.2022.139099 Google Scholar 192Shevlin, M. Practical High-Throughput Experimentation for Chemists. ACS Med. Chem. Lett. 2017, 8, 601– 607, DOI: 10.1021/acsmedchemlett.7b00165 Google Scholar 193Mennen, S. M.; Alhambra, C.; Allen, C. L.; Barberis, M.;
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Berritt, S.; Brandt, T. A.; Campbell, A. D.; Castañón, J.; Cherney, A. H.; Christensen, M.The Evolution of High-Throughput Experimentation in Pharmaceutical Development and Perspectives on the Future. Org. Process Res. Dev. 2019, 23, 1213– 1242, DOI: 10.1021/acs.oprd.9b00140 Google Scholar 194Krska, S. W.; DiRocco, D. A.; Dreher, S. D.; Shevlin, M. The Evolution of Chemical High-Throughput Experimentation to Address Challenging Problems in Pharmaceutical Synthesis. Acc. Chem. Res. 2017, 50, 2976– 2985, DOI: 10.1021/acs.accounts.7b00428 Google Scholar 195Isbrandt, E. S.; Sullivan, R. J.; Newman, S. G. High Throughput Strategies for the Discovery and Optimization of Catalytic Reactions. Angew. Chem., Int. Ed. 2019, 58, 7180– 7191, DOI: 10.1002/anie.201812534 Google Scholar 196Cook, A.; Clément, R.; Newman, S. G. Reaction Screening in Multiwell Plates: High-Throughput Optimization of a Buchwald–Hartwig Amination. Nat. Prot. 2021, 16, 1152– 1169, DOI: 10.1038/s41596-020-00452-7 Google Scholar 197Cernak, T.; Gesmundo, N. J.; Dykstra, K.; Yu, Y.; Wu, Z.; Shi, Z.-C.; Vachal, P.; Sperbeck, D.; He, S.; Murphy, B. A.Microscale High-Throughput Experimentation as an Enabling Technology in Drug Discovery: Application in the Discovery of (Piperidinyl)Pyridinyl-1h-Benzimidazole Diacylglycerol Acyltransferase 1 Inhibitors. J. Med. Chem. 2017, 60, 3594– 3605, DOI: 10.1021/acs.jmedchem.6b01543 Google Scholar 198Mahjour, B.; Shen, Y.; Cernak, T. Ultrahigh-Throughput Experimentation for Information-Rich Chemical Synthesis. Acc. Chem. Res. 2021, 54, 2337– 2346, DOI: 10.1021/acs.accounts.1c00119 Google Scholar 199Gaunt, M. J.; Janey, J. M.; Schultz, D. M.; Cernak, T. Myths of High-Throughput Experimentation and Automation in Chemistry. Chem. 2021, 7, 2259– 2260, DOI: 10.1016/j.chempr.2021.08.012 Google Scholar 200Maier, W. F. Early Years of High-Throughput Experimentation and Combinatorial Approaches in Catalysis and Materials Science. ACS Comb. Sci. 2019, 21, 437– 444, DOI: 10.1021/acscombsci.8b00189 Google Scholar 201Robbins, D. W.; Hartwig, J. F. A Simple, Multidimensional Approach to High-Throughput Discovery of Catalytic Reactions. Science. 2011, 333, 1423– 1427, DOI: 10.1126/science.1207922 Google Scholar 202McNally, A.; Prier, C. K.; MacMillan, D. W. C. Discovery of
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
an Α-Amino C–H Arylation Reaction Using the Strategy of Accelerated Serendipity. Science. 2011, 334, 1114– 1117, DOI: 10.1126/science.1213920 Google Scholar 203Troshin, K.; Hartwig, J. F. Snap Deconvolution: An Informatics Approach to High-Throughput Discovery of Catalytic Reactions. Science. 2017, 357, 175– 181, DOI: 10.1126/science.aan1568 Google Scholar 204Sun, A. C.; Steyer, D. J.; Allen, A. R.; Payne, E. M.; Kennedy, R. T.; Stephenson, C. R. J. A Droplet Microfluidic Platform for High-Throughput Photochemical Reaction Discovery. Nat. Commun. 2020, 11, 6202, DOI: 10.1038/s41467-020-19926-z Google Scholar 205Wu, L.; Wan, Q.; Nie, W.; Hao, Y.; Feng, G.; Chen, M.; Chen, S. High-Throughput Nano-Electrostatic-Spray Ionization/Photoreaction Mass Spectrometric Platform for the Discovery of Visible-Light-Activated Photocatalytic Reactions in the Picomole Scale. Anal. Chem. 2021, 93, 14560– 14567, DOI: 10.1021/acs.analchem.1c03639 Google Scholar 206Wong, H.; Cernak, T. Reaction miniaturization in eco-friendly solvents. Curr. Opin. Green Sust. Chem. 2018, 11, 91– 98, DOI: 10.1016/j.cogsc.2018.06.001 Google Scholar 207Perera, D.; Tucker, J. W.; Brahmbhatt, S.; Helal, C. J.; Chong, A.; Farrell, W.; Richardson, P.; Sach, N. W. A Platform for Automated Nanomole-Scale Reaction Screening and Micromole-Scale Synthesis in Flow. Science 2018, 359, 429– 434, DOI: 10.1126/science.aap9112 Google Scholar 208Guidi, M.; Moon, S.; Anghileri, L.; Cambié, D.; Seeberger, P. H.; Gilmore, K. Combining Radial and Continuous Flow Synthesis to Optimize and Scale-up the Production of Medicines. React. Chem. Eng. 2021, 6, 220– 224, DOI: 10.1039/D0RE00445F Google Scholar 209Buglioni, L.; Raymenants, F.; Slattery, A.; Zondag, S. D.; Noël, T. Technological Innovations in Photochemistry for Organic Synthesis: Flow Chemistry, High-Throughput Experimentation, Scale-up, and Photoelectrochemistry. Chem. Rev. 2022, 122, 2752– 2906, DOI: 10.1021/acs.chemrev.1c00332 Google Scholar 210Tu, N. P.; Dombrowski, A. W.; Goshu, G. M.; Vasudevan, A.; Djuric, S. W.; Wang, Y. High-Throughput Reaction Screening with Nanomoles of Solid Reagents Coated on Glass Beads. Angew. Chem., Int. Ed. 2019, 58, 7987– 7991, DOI: 10.1002/anie.201900536 Google Scholar 211Martin, M.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
C.; Goshu, G. M.; Hartnell, J. R.; Morris, C. D.; Wang, Y.; Tu, N. P. Versatile Methods to Dispense Submilligram Quantities of Solids Using Chemical-Coated Beads for High-Throughput Experimentation. Org. Process Res. Dev. 2019, 23, 1900– 1907, DOI: 10.1021/acs.oprd.9b00213 Google Scholar 212Kang, K.; Loud, N. L.; DiBenedetto, T. A.; Weix, D. J. A General, Multimetallic Cross-Ullmann Biheteroaryl Synthesis from Heteroaryl Halides and Heteroaryl Triflates. J. Am. Chem. Soc. 2021, 143, 21484– 21491, DOI: 10.1021/jacs.1c10907 Google Scholar 213Aguirre, A. L.; Loud, N. L.; Johnson, K. A.; Weix, D. J.; Wang, Y. Chembead Enabled High-Throughput Cross-Electrophile Coupling Eeveals a New Complementary Ligand. Chem. - Eur. J. 2021, 27, 12981– 12986, DOI: 10.1002/chem.202102347 Google Scholar 214Bahr, M. N.; Damon, D. B.; Yates, S. D.; Chin, A. S.; Christopher, J. D.; Cromer, S.; Perrotto, N.; Quiroz, J.; Rosso, V. Collaborative Evaluation of Commercially Available Automated Powder Dispensing Platforms for High-Throughput Experimentation in Pharmaceutical Applications. Org. Process Res. Dev. 2018, 22, 1500– 1508, DOI: 10.1021/acs.oprd.8b00259 Google Scholar 215Bahr, M. N.; Morris, M. A.; Tu, N. P.; Nandkeolyar, A. Recent Advances in High-Throughput Automated Powder Dispensing Platforms for Pharmaceutical Applications. Org. Process Res. Dev. 2020, 24, 2752– 2761, DOI: 10.1021/acs.oprd.0c00411 Google Scholar 216Prieto Kullmer, C. N.; Kautzky, J. A.; Krska, S. W.; Nowak, T.; Dreher, S. D.; MacMillan, D. W. Accelerating Reaction Generality and Mechanistic Insight through Additive Mapping. Science. 2022, 376, 532– 539, DOI: 10.1126/science.abn1885 Google Scholar 217Qi, N.; Wismer, M. K.; Conway, D. V.; Krska, S. W.; Dreher, S. D.; Lin, S. Development of a High Intensity Parallel Photoreactor for High Throughput Screening. React. Chem. Eng. 2022, 7, 354– 360, DOI: 10.1039/D1RE00317H Google Scholar 218Buitrago Santanilla, A.; Christensen, M.; Campeau, L.-C.; Davies, I. W.; Dreher, S. D. P2et Phosphazene: A Mild, Functional Group Tolerant Base for Soluble, Room Temperature Pd-Catalyzed C–N, C–O, and C–C
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Cross-Coupling Reactions. Org. Lett. 2015, 17, 3370– 3373, DOI: 10.1021/acs.orglett.5b01648 Google Scholar 219Huff, C. A.; Cohen, R. D.; Dykstra, K. D.; Streckfuss, E.; DiRocco, D. A.; Krska, S. W. Photoredox-Catalyzed Hydroxymethylation of Heteroaromatic Bases. J. Org. Chem. 2016, 81, 6980– 6987, DOI: 10.1021/acs.joc.6b00811 Google Scholar 220DiRocco, D. A.; Dykstra, K.; Krska, S.; Vachal, P.; Conway, D. V.; Tudge, M. Late-Stage Functionalization of Biologically Active Heterocycles through Photoredox Catalysis. Angew. Chem., Int. Ed. 2014, 53, 4802– 4806, DOI: 10.1002/anie.201402023 Google Scholar 221Primer, D. N.; Molander, G. A. Enabling the Cross-Coupling of Tertiary Organoboron Nucleophiles through Radical-Mediated Alkyl Transfer. J. Am. Chem. Soc. 2017, 139, 9847– 9850, DOI: 10.1021/jacs.7b06288 Google Scholar 222Brocklehurst, C. E.; Gallou, F.; Hartwieg, J. C. D.; Palmieri, M.; Rufle, D. Microtiter Plate (Mtp) Reaction Screening and Optimization of Surfactant Chemistry: Examples of Suzuki–Miyaura and Buchwald–Hartwig Cross-Couplings in Water. Org. Process Res. Dev. 2018, 22, 1453– 1457, DOI: 10.1021/acs.oprd.8b00200 Google Scholar 223Zarate, C.; Ardolino, M.; Morriello, G. J.; Logan, K. M.; Kaplan, W. P.; Torres, L.; Li, D.; Chen, M.; Li, H.; Su, J.Development of Scalable Routes to 1-Bicyclo [1.1. 1] Pentylpyrazoles. Org. Process Res. Dev. 2021, 25, 642– 647, DOI: 10.1021/acs.oprd.0c00446 Google Scholar 224Fier, P. S.; Maloney, K. M. Reagent Design and Ligand Evolution for the Development of a Mild Copper-Catalyzed Hydroxylation Reaction. Org. Lett. 2017, 19, 3033– 3036, DOI: 10.1021/acs.orglett.7b01403 Google Scholar 225Nicastri, M. C.; Lehnherr, D.; Lam, Y.-h.; DiRocco, D. A.; Rovis, T. Synthesis of Sterically Hindered Primary Amines by Concurrent Tandem Photoredox Catalysis. J. Am. Chem. Soc. 2020, 142, 987– 998, DOI: 10.1021/jacs.9b10871 Google Scholar 226Corcoran, E. B.; Pirnot, M. T.; Lin, S.; Dreher, S. D.; DiRocco, D. A.; Davies, I. W.; Buchwald, S. L.; MacMillan, D. W. Aryl Amination Using Ligand-Free Ni (Ii) Salts and Photoredox Catalysis. Science. 2016, 353, 279– 283, DOI: 10.1126/science.aag0209
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Google Scholar 227Buitrago Santanilla, A.; Regalado, E. L.; Pereira, T.; Shevlin, M.; Bateman, K.; Campeau, L.-C.; Schneeweis, J.; Berritt, S.; Shi, Z.-C.; Nantermet, P.Nanomole-Scale High-Throughput Chemistry for the Synthesis of Complex Molecules. Science. 2015, 347, 49– 53, DOI: 10.1126/science.1259203 Google Scholar 228Grainger, R.; Heightman, T. D.; Ley, S. V.; Lima, F.; Johnson, C. N. Enabling Synthesis in Fragment-Based Drug Discovery by Reactivity Mapping: Photoredox-Mediated Cross-Dehydrogenative Heteroarylation of Cyclic Amines. Chem. Sci. 2019, 10, 2264– 2271, DOI: 10.1039/C8SC04789H Google Scholar 229Pomberger, A.; Pedrina McCarthy, A. A.; Khan, A.; Sung, S.; Taylor, C.; Gaunt, M.; Colwell, L.; Walz, D.; Lapkin, A. The Effect of Chemical Representation on Active Machine Learning Towards Closed-Loop Optimization. React. Chem. Eng. 2022, 7, 1368, DOI: 10.1039/D2RE00008C Google Scholar 230González-Esguevillas, M.; Fernández, D. F.; Rincón, J. A.; Barberis, M.; de Frutos, O.; Mateos, C.; García-Cerrada, S.; Agejas, J.; MacMillan, D. W. C. Rapid Optimization of Photoredox Reactions for Continuous-Flow Systems Using Microscale Batch Technology. ACS Cent. Sci. 2021, 7, 1126– 1134, DOI: 10.1021/acscentsci.1c00303 Google Scholar 231Weis, E.; Johansson, M.; Korsgren, P.; Martín-Matute, B.; Johansson, M. J. Merging Directed C–H Activations with High-Throughput Experimentation: Development of Iridium-Catalyzed C–H Aminations Applicable to Late-Stage Functionalization. JACS Au. 2022, 2, 906– 916, DOI: 10.1021/jacsau.2c00039 Google Scholar 232Shaabani, S.; Xu, R.; Ahmadianmoghaddam, M.; Gao, L.; Stahorsky, M.; Olechno, J.; Ellson, R.; Kossenjans, M.; Helan, V.; Dömling, A. Automated and Accelerated Synthesis of Indole Derivatives on a Nano-Scale. Green Chem. 2019, 21, 225– 232, DOI: 10.1039/C8GC03039A Google Scholar 233Wang, Y.; Shaabani, S.; Ahmadianmoghaddam, M.; Gao, L.; Xu, R.; Kurpiewska, K.; Kalinowska-Tluscik, J.; Olechno, J.; Ellson, R.; Kossenjans, M.Acoustic Droplet Ejection Enabled Automated Reaction Scouting. ACS Cent. Sci. 2019, 5, 451– 457, DOI: 10.1021/acscentsci.8b00782 Google Scholar 234Osipyan, A.; Shaabani, S.; Warmerdam, R.; Shishkina, S. V.; Boltz, H.; Dömling, A. Automated, Accelerated Nanoscale Synthesis of Iminopyrrolidines.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Angew. Chem., Int. Ed. 2020, 59, 12423– 12427, DOI: 10.1002/anie.202000887 Google Scholar 235Sutanto, F.; Shaabani, S.; Oerlemans, R.; Eris, D.; Patil, P.; Hadian, M.; Wang, M.; Sharpe, M. E.; Groves, M. R.; Dömling, A. Combining High-Throughput Synthesis and High-Throughput Protein Crystallography for Accelerated Hit Identification. Angew. Chem., Int. Ed. 2021, 60, 18231– 18239, DOI: 10.1002/anie.202105584 Google Scholar 236Hadian, M.; Shaabani, S.; Patil, P.; Shishkina, S. V.; Böltz, H.; Dömling, A. Sustainability by Design: Automated Nanoscale 2,3,4-Trisubstituted Quinazoline Diversity. Green Chem. 2020, 22, 2459– 2467, DOI: 10.1039/D0GC00363H Google Scholar 237Gao, K.; Shaabani, S.; Xu, R.; Zarganes-Tzitzikas, T.; Gao, L.; Ahmadianmoghaddam, M.; Groves, M. R.; Dömling, A. Nanoscale, Automated, High Throughput Synthesis and Screening for the Accelerated Discovery of Protein Modifiers. RSC Med. Chem. 2021, 12, 809– 818, DOI: 10.1039/D1MD00087J Google Scholar 238Sangouard, G.; Zorzi, A.; Wu, Y.; Ehret, E.; Schüttel, M.; Kale, S.; Díaz-Perlas, C.; Vesin, J.; Bortoli Chapalay, J.; Turcatti, G.Picomole-Scale Synthesis and Screening of Macrocyclic Compound Libraries by Acoustic Liquid Transfer. Angew. Chem., Int. Ed. 2021, 60, 21702– 21707, DOI: 10.1002/anie.202107815 Google Scholar 239Gesmundo, N. J.; Sauvagnat, B.; Curran, P. J.; Richards, M. P.; Andrews, C. L.; Dandliker, P. J.; Cernak, T. Nanoscale Synthesis and Affinity Ranking. Nature. 2018, 557, 228– 232, DOI: 10.1038/s41586-018-0056-8 Google Scholar 240Schiffer, Z. J.; Chung, M.; Steinberg, K.; Manthiram, K. Selective Electrochemical Reductive Amination of Benzaldehyde at Heterogeneous Metal Surfaces. Chem. Catal. 2023, 100500, DOI: 10.1016/j.checat.2022.100500 Google Scholar 241Brown, D. G.; Boström, J. Analysis of Past and Present Synthetic Methodologies on Medicinal Chemistry: Where Have All the New Reactions Gone?. J. Med. Chem. 2016, 59, 4443– 4458, DOI: 10.1021/acs.jmedchem.5b01409 Google Scholar 242Dombrowski, A. W.; Aguirre, A. L.; Shrestha, A.; Sarris, K. A.; Wang, Y. The Chosen Few: Parallel Library Reaction Methodologies for Drug Discovery. J. Org. Chem. 2022, 87, 1880– 1897, DOI: 10.1021/acs.joc.1c01427 Google
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Scholar 243Ahneman, D. T.; Estrada, J. G.; Lin, S.; Dreher, S. D.; Doyle, A. G. Predicting Reaction Performance in C–N Cross-Coupling Using Machine Learning. Science. 2018, 360, 186– 190, DOI: 10.1126/science.aar5169 Google Scholar 244Nielsen, M. K.; Ahneman, D. T.; Riera, O.; Doyle, A. G. Deoxyfluorination with Sulfonyl Fluorides: Navigating Reaction Space with Machine Learning. J. Am. Chem. Soc. 2018, 140, 5004– 5008, DOI: 10.1021/jacs.8b01523 Google Scholar 245Lin, S.; Dikler, S.; Blincoe, W. D.; Ferguson, R. D.; Sheridan, R. P.; Peng, Z.; Conway, D. V.; Zawatzky, K.; Wang, H.; Cernak, T.Mapping the Dark Space of Chemical Reactions with Extended Nanomole Synthesis and Maldi-Tof Ms. Science. 2018, 361, eaar6236 DOI: 10.1126/science.aar6236 Google Scholar 246Lee, H.; Boyer, N. C.; Deng, Q.; Kim, H.-Y.; Sawyer, T. K.; Sciammetta, N. Photoredox Ni-Catalyzed Peptide C (Sp 2)–O Cross-Coupling: From Intermolecular Reactions to Side Chain-to-Tail Macrocyclization. Chem. Sci. 2019, 10, 5073– 5078, DOI: 10.1039/C9SC00694J Google Scholar 247Gesmundo, N.; Dykstra, K.; Douthwaite, J.; Mahjour, B.; Ferguson, R.; Dreher, S.; Sauvagnat, B.; Sauri, J.; Cernak, T. Miniaturization of Popular Reactions from the Medicinal Chemists’ Toolbox for Ultrahigh-Throughput Experimentation. ChemRxiv 2022, chemrxiv-2022-x6msn. DOI: 10.26434/chemrxiv-2022-x6msn Google Scholar 248Grainger, R.; Whibley, S. A Perspective on the Analytical Challenges Encountered in High-Throughput Experimentation. Org. Process Res. Dev. 2021, 25, 354– 364, DOI: 10.1021/acs.oprd.0c00463 Google Scholar 249Welch, C. J. High Throughput Analysis Enables High Throughput Experimentation in Pharmaceutical Process Research. React. Chem. Eng. 2019, 4, 1895– 1911, DOI: 10.1039/C9RE00234K Google Scholar 250Krenkel, H.; Hartmane, E.; Piras, C.; Brown, J.; Morris, M.; Cramer, R. Advancing Liquid Atmospheric Pressure Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry toward Ultrahigh-Throughput Analysis. Anal. Chem. 2020, 92, 2931– 2936, DOI: 10.1021/acs.analchem.9b05202 Google Scholar 251Blincoe, W. D.; Lin, S.; Dreher, S. D.; Sheng, H. Practical Guide on Maldi-Tof Ms Method Development for High Throughput Profiling of Pharmaceutically Relevant, Small Molecule Chemical Reactions. Tetrahedron.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
2020, 76, 131434, DOI: 10.1016/j.tet.2020.131434 Google Scholar 252Logsdon, D. L.; Li, Y.; Paschoal Sobreira, T. J.; Ferreira, C. R.; Thompson, D. H.; Cooks, R. G. High-Throughput Screening of Reductive Amination Reactions Using Desorption Electrospray Ionization Mass Spectrometry. Org. Process Res. Dev. 2020, 24, 1647– 1657, DOI: 10.1021/acs.oprd.0c00230 Google Scholar 253Sawicki, J. W.; Bogdan, A. R.; Searle, P. A.; Talaty, N.; Djuric, S. W. Rapid Analytical Characterization of High-Throughput Chemistry Screens Utilizing Desorption Electrospray Ionization Mass Spectrometry. React. Chem. Eng. 2019, 4, 1589– 1594, DOI: 10.1039/C9RE00054B Google Scholar 254Wleklinski, M.; Loren, B. P.; Ferreira, C. R.; Jaman, Z.; Avramova, L.; Sobreira, T. J. P.; Thompson, D. H.; Cooks, R. G. High Throughput Reaction Screening Using Desorption Electrospray Ionization Mass Spectrometry. Chem. Sci. 2018, 9, 1647– 1653, DOI: 10.1039/C7SC04606E Google Scholar 255DiRico, K. J.; Hua, W.; Liu, C.; Tucker, J. W.; Ratnayake, A. S.; Flanagan, M. E.; Troutman, M. D.; Noe, M. C.; Zhang, H. Ultra-High-Throughput Acoustic Droplet Ejection-Open Port Interface-Mass Spectrometry for Parallel Medicinal Chemistry. ACS Med. Chem. Lett. 2020, 11, 1101– 1110, DOI: 10.1021/acsmedchemlett.0c00066 Google Scholar 256Sinclair, I.; Bachman, M.; Addison, D.; Rohman, M.; Murray, D. C.; Davies, G.; Mouchet, E.; Tonge, M. E.; Stearns, R. G.; Ghislain, L.Acoustic Mist Ionization Platform for Direct and Contactless Ultrahigh-Throughput Mass Spectrometry Analysis of Liquid Samples. Anal. Chem. 2019, 91, 3790– 3794, DOI: 10.1021/acs.analchem.9b00142 Google Scholar 257Liu, C. Acoustic Ejection Mass Spectrometry: Fundamentals and Applications in High-Throughput Drug Discovery. Expert Opin. Drug Discovery. 2022, 17, 775– 787, DOI: 10.1080/17460441.2022.2084069 Google Scholar 258Welch, C. J.; Gong, X.; Schafer, W.; Pratt, E. C.; Brkovic, T.; Pirzada, Z.; Cuff, J. F.; Kosjek, B. Miser Chromatography (Multiple Injections in a Single Experimental Run): The Chromatogram Is the Graph. Tetrahedron: Asymmetry. 2010, 21, 1674– 1681, DOI: 10.1016/j.tetasy.2010.05.029 Google Scholar 259Stevens, J. M.; Simmons, E. M.; Tan, Y.; Borovika,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
A.; Fan, J.; Forest, R. V.; Geng, P.; Guerrero, C. A.; Lou, S.; Skliar, D.Leveraging High-Throughput Experimentation to Drive Pharmaceutical Route Invention: A Four-Step Commercial Synthesis of Branebrutinib (Bms-986195). Org. Process Res. Dev. 2022, 26, 1174– 1183, DOI: 10.1021/acs.oprd.1c00443 Google Scholar 260Carter, H. L.; Connor, A. W.; Hart, R.; McCabe, J.; McIntyre, A. C.; McMillan, A. E.; Monks, N. R.; Mullen, A. K.; Ronson, T. O.; Steven, A.Rapid Route Design of Azd7594. React. Chem. Eng. 2019, 4, 1658– 1673, DOI: 10.1039/C9RE00118B Google Scholar 261Mousseau, J. J.; Perry, M. A.; Bundesmann, M. W.; Chinigo, G. M.; Choi, C.; Gallego, G.; Hicklin, R. W.; Hoy, S.; Limburg, D. C.; Sach, N. W.Automated Nanomole-Scale Reaction Screening toward Benzoate Bioisosteres: A Photocatalyzed Approach to Highly Elaborated Bicyclo[1.1.1]Pentanes. ACS Catal. 2022, 12, 600– 606, DOI: 10.1021/acscatal.1c05076 Google Scholar 262Avila, C.; Cassani, C.; Kogej, T.; Mazuela, J.; Sarda, S.; Clayton, A. D.; Kossenjans, M.; Green, C. P.; Bourne, R. A. Automated Stopped-Flow Library Synthesis for Rapid Optimisation and Machine Learning Directed Experimentation. Chem. Sci. 2022, 13, 12087– 12099, DOI: 10.1039/D2SC03016K Google Scholar 263Schwaller, P.; Vaucher, A. C.; Laino, T.; Reymond, J.-L. Prediction of Chemical Reaction Yields Using Deep Learning. Machine learning: science and technology. 2021, 2, 015016, DOI: 10.1088/2632-2153/abc81d Google Scholar 264Zoph, B.; Vasudevan, V.; Shlens, J.; Le, Q. V. Learning Transferable Architectures for Scalable Image Recognition. In ArXiv , 2018, arXiv:1707.07012. DOI: 10.48550/arXiv.1707.07012 Google Scholar 265Goldberg, Y. Neural Network Methods for Natural Language Processing. Synthesis lectures on human language technologies. 2017, 10, 1– 309, DOI: 10.1007/978-3-031-02165-7_1 Google Scholar 266Dogan, U.; Edelbrunner, J.; Iossifidis, I. Autonomous Driving: A Comparison of Machine Learning Techniques by Means of the Prediction of Lane Change Behavior. In 2011 IEEE International Conference on Robotics and Biomimetics, ROBIO 2011, Karon Beach, Thailand, December 7-11, 2011 2011. 1837 1843 Google Scholar 267Rogers, D.;
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Hahn, M. Extended-Connectivity Fingerprints. J. Chem. Inf. Model. 2010, 50, 742– 754, DOI: 10.1021/ci100050t Google Scholar 268Eyke, N. S.; Green, W. H.; Jensen, K. F. Iterative Experimental Design Based on Active Machine Learning Reduces the Experimental Burden Associated with Reaction Screening. React. Chem. Eng. 2020, 5, 1963– 1972, DOI: 10.1039/D0RE00232A Google Scholar 269Brethomé, A. V.; Fletcher, S. P.; Paton, R. S. Conformational Effects on Physical-Organic Descriptors: The Case of Sterimol Steric Parameters. ACS Catal. 2019, 9, 2313– 2323, DOI: 10.1021/acscatal.8b04043 Google Scholar 270Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, L.; Polosukhin, I. Attention Is All You Need. Adv. Neural Inform. Process. Syst. 2017, 30. Google Scholar 271Weininger, D. Smiles, a Chemical Language and Information-System.1. Introduction to Methodology and Encoding Rules. J. Chem. Inf. Model. 1988, 28, 31– 36, DOI: 10.1021/ci00057a005 Google Scholar 272Schwaller, P.; Laino, T.; Gaudin, T.; Bolgar, P.; Hunter, C. A.; Bekas, C.; Lee, A. A. Molecular Transformer: A Model for Uncertainty-Calibrated Chemical Reaction Prediction. ACS Cent. Sci. 2019, 5, 1572– 1583, DOI: 10.1021/acscentsci.9b00576 Google Scholar 273Schwaller, P.; Hoover, B.; Reymond, J. L.; Strobelt, H.; Laino, T. Extraction of Organic Chemistry Grammar from Unsupervised Learning of Chemical Reactions. Sci. Adv. 2021, 7, eabe4166 DOI: 10.1126/sciadv.abe4166 Google Scholar 274Gilmer, J.; Schoenholz, S. S.; Riley, P. F.; Vinyals, O.; Dahl, G. E. Neural Message Passing for Quantum Chemistry. Proc. Machine Learn. Res. 2017, 70, 1263– 1272 Google Scholar 275Yang, K.; Swanson, K.; Jin, W. G.; Coley, C.; Eiden, P.; Gao, H.; Guzman-Perez, A.; Hopper, T.; Kelley, B.; Mathea, M.Analyzing Learned Molecular Representations for Property Prediction (Vol 59, Pg 3370, 2019). J. Chem. Inf. Model. 2019, 59, 5304– 5305, DOI: 10.1021/acs.jcim.9b01076 Google Scholar 276Heid, E.; Green, W. H. Machine Learning of Reaction Properties Via Learned Representations of the Condensed Graph of Reaction. J. Chem.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Inf. Model. 2022, 62, 2101– 2110, DOI: 10.1021/acs.jcim.1c00975 Google Scholar 277Haywood, A.; Redshaw, J.; Hanson-Heine, M.; Taylor, A.; Brown, A.; Mason, A.; Gaertner, T.; Hirst, J. Kernel Methods for Predicting Yields of Chemical Reactions. J. Chem. Inf. Model. 2022, 62, 2077, DOI: 10.1021/acs.jcim.1c00699 Google Scholar 278Sandfort, F.; Strieth-Kalthoff, F.; Kühnemund, M.; Beecks, C.; Glorius, F. A Structure-Based Platform for Predicting Chemical Reactivity. Chem 2020, 6, 1379– 1390, DOI: 10.1016/j.chempr.2020.02.017 Google Scholar 279Żurański, A. M.; Martinez Alvarado, J. I.; Shields, B. J.; Doyle, A. G. Predicting Reaction Yields Via Supervised Learning. Acc. Chem. Res. 2021, 54, 1856– 1865, DOI: 10.1021/acs.accounts.0c00770 Google Scholar 280Lowe, D. M. Extraction of Chemical Structures and Reactions from the Literature Doctoral Thesis. University of Cambridge, 2012. DOI: 10.17863/CAM.16293 Google Scholar 281Apodaca, R. L. The NextMove Patent Reaction Dataset; Depth-First , 2019; (accessed 2023-01-20). Google Scholar 282Mayfield, J.; O'Boyle, N.; Sayle, R. Pistachio - Search and Faceting of Large Reaction Databases ACS Fall 2017 2017. Google Scholar 283Kearnes, S. M.; Maser, M. R.; Wleklinski, M.; Kast, A.; Doyle, A. G.; Dreher, S. D.; Hawkins, J. M.; Jensen, K. F.; Coley, C. W. The Open Reaction Database. J. Am. Chem. Soc. 2021, 143, 18820– 18826, DOI: 10.1021/jacs.1c09820 Google Scholar 284Thakkar, A.; Kogej, T.; Reymond, J.-L.; Engkvist, O.; Bjerrum, E. J. Datasets and Their Influence on the Development of Computer Assisted Synthesis Planning Tools in the Pharmaceutical Domain. Chem. Sci. 2020, 11, 154– 168, DOI: 10.1039/C9SC04944D Google Scholar 285Vaucher, A. C.; Zipoli, F.; Geluykens, J.; Nair, V. H.; Schwaller, P.; Laino, T. Automated Extraction of Chemical Synthesis Actions from Experimental Procedures. Nat. Commun. 2020, 11, 3601, DOI: 10.1038/s41467-020-17266-6 Google Scholar 286Vaucher, A. C.; Schwaller, P.; Geluykens, J.; Nair, V. H.; Iuliano, A.; Laino, T. Inferring Experimental Procedures from Text-Based Representations of Chemical Reactions. Nat. Commun. 2021, 12, 2573, DOI:
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
10.1038/s41467-021-22951-1 Google Scholar 287Guo, J.; Ibanez-Lopez, A. S.; Gao, H. Y.; Quach, V.; Coley, C. W.; Jensen, K. F.; Barzilay, R. Automated Chemical Reaction Extraction from Scientific Literature. J. Chem. Inf. Model. 2021, 61, 4124– 4124, DOI: 10.1021/acs.jcim.1c00834 Google Scholar 288Gimadiev, T. R.; Lin, A.; Afonina, V. A.; Batyrshin, D.; Nugmanov, R. I.; Akhmetshin, T.; Sidorov, P.; Duybankova, N.; Verhoeven, J.; Wegner, J.Reaction Data Curation I: Chemical Structures and Transformations Standardization. Mol. Inf. 2021, 40, 2100119, DOI: 10.1002/minf.202100119 Google Scholar 289Fitzner, M.; Wuitschik, G.; Koller, R. J.; Adam, J. M.; Schindler, T.; Reymond, J. L. What Can Reaction Databases Teach Us About Buchwald-Hartwig Cross-Couplings?. Chem. Sci. 2020, 11, 13085– 13093, DOI: 10.1039/D0SC04074F Google Scholar 290Gao, H.; Struble, T. J.; Coley, C. W.; Wang, Y.; Green, W. H.; Jensen, K. F. Using Machine Learning to Predict Suitable Conditions for Organic Reactions. ACS Cent. Sci. 2018, 4, 1465– 1476, DOI: 10.1021/acscentsci.8b00357 Google Scholar 291Coley, C. W.; Thomas, D. A.; Lummiss, J. A. M.; Jaworski, J. N.; Breen, C. P.; Schultz, V.; Hart, T.; Fishman, J. S.; Rogers, L.; Gao, H. Y.A Robotic Platform for Flow Synthesis of Organic Compounds Informed by Ai Planning. Science. 2019, 365, eaax1566, DOI: 10.1126/science.aax1566 Google Scholar 292Schwaller, P.; Gaudin, T.; Lányi, D.; Bekas, C.; Laino, T. Found in Translation”: Predicting Outcomes of Complex Organic Chemistry Reactions Using Neural Sequence-to-Sequence Models. Chem. Sci. 2018, 9, 6091– 6098, DOI: 10.1039/C8SC02339E Google Scholar 293Lei, T.; Jin, W. G.; Barzilay, R.; Jaakkola, T. Deriving Neural Architectures from Sequence and Graph Kernels. Proc. Machine Learn. Res. 2017, 70, 2024– 2033 Google Scholar 294Coley, C. W.; Green, W. H.; Jensen, K. F. Machine Learning in Computer-Aided Synthesis Planning. Acc. Chem. Res. 2018, 51, 1281– 1289, DOI: 10.1021/acs.accounts.8b00087 Google Scholar 295Angello, N. H.; Rathore, V.; Beker, W.; Wołos, A.; Jira, E. R.; Roszak, R.;
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Wu, T. C.; Schroeder, C. M.; Aspuru-Guzik, A.; Grzybowski, B. A.; Burke, M. D. Closed-Loop Optimization of General Reaction Conditions for Heteroaryl Suzuki-Miyaura Coupling. Science. 2022, 378, 399– 405, DOI: 10.1126/science.adc8743 Google Scholar 296Xu, J.; Kalyani, D.; Struble, T. J.; Dreher, S. D.; Krska, S.; Buchwald, S. L.; Jensen, K. Roadmap to Pharmaceutically Relevant Reactivity Models Leveraging High-Throughput Experimentation. ChemRxiv 2022 chemrxiv-2022-x694w. DOI: 10.26434/chemrxiv-2022-x694w Google Scholar 297Kulp, C. Route Scouting, Process Optimization, and Scale-up Considerations: Drug Development Firms and Cros. Chim. Oggi. 2018, 36, 33– 34 Google Scholar 298Yu, L. X.; Amidon, G.; Khan, M. A.; Hoag, S. W.; Polli, J.; Raju, G. K.; Woodcock, J. Understanding Pharmaceutical Quality by Design. AAPS Journal. 2014, 16, 771– 783, DOI: 10.1208/s12248-014-9598-3 Google Scholar 299Bumphenkiattikul, P.; Limtrakul, S.; Vatanatham, T.; Khongprom, P.; Ramachandran, P. A. Heat Transfer Effect in Scaling-up a Fluidized Bed Reactor for Propylene Polymerization. RSC Adv. 2018, 8, 28293– 28312, DOI: 10.1039/C8RA04834G Google Scholar 300Rossetti, I.; Compagnoni, M. Chemical Reaction Engineering, Process Design and Scale-up Issues at the Frontier of Synthesis: Flow Chemistry. Chem. Eng. J. 2016, 296, 56– 70, DOI: 10.1016/j.cej.2016.02.119 Google Scholar 301McMullen, J. P.; Marton, C. H.; Sherry, B. D.; Spencer, G.; Kukura, J.; Eyke, N. S. Development and Scale-up of a Continuous Reaction for Production of an Active Pharmaceutical Ingredient Intermediate. Org. Process Res. Dev. 2018, 22, 1208– 1213, DOI: 10.1021/acs.oprd.8b00192 Google Scholar 302Dugger, R. W.; Ragan, J. A.; Ripin, D. H. B. Survey of Gmp Bulk Reactions Run in a Research Facility between 1985 and 2002. Org. Process Res. Dev. 2005, 9, 253– 258, DOI: 10.1021/op050021j Google Scholar 303Stitt, E. H.; Simmons, M. J. H. Scale-up of Chemical Reactions. In Process Understanding.; Wiley, 2011; pp 155– 198. DOI: 10.1002/9783527637140.ch6 Google Scholar 304Levin, M. Pharmaceutical Process Scale-Up, 3rd ed.; CRC Press, 2011. Google Scholar 305Leng, R.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
From Bench to Plant: Scale up Specialty Chemical Processes Directly. Chem. Eng. Prog. 2004, 100, 37– 44 Google Scholar 306Commenge, J. M.; Falk, L. Methodological Framework for Choice of Intensified Equipment and Development of Innovative Technologies. Chem. Eng. Proc. 2014, 84, 109– 127, DOI: 10.1016/j.cep.2014.03.001 Google Scholar 307Tan, R. K.; Eberhard, W.; Buchs, J. Measurement and Characterization of Mixing Time in Shake Flasks. Chem. Eng. Sci. 2011, 66, 440– 447, DOI: 10.1016/j.ces.2010.11.001 Google Scholar 308Hoffmann, W., Use of Models to Enhance Process Understanding. In Process Understanding; Wiley, 2011; pp 127– 153. DOI: 10.1002/9783527637140.ch5 Google Scholar 309Nagy, K. D.; Shen, B.; Jamison, T. F.; Jensen, K. F. Mixing and Dispersion in Small-Scale Flow Systems. Org. Process Res. Dev. 2012, 16, 976– 981, DOI: 10.1021/op200349f Google Scholar 310Vamosi, P.; Matsuo, K.; Masuda, T.; Sato, K.; Narumi, T.; Takeda, K.; Mase, N. Rapid Optimization of Reaction Conditions Based on Comprehensive Reaction Analysis Using a Continuous Flow Microwave Reactor. Chem. Rec. 2019, 19, 77– 84, DOI: 10.1002/tcr.201800048 Google Scholar 311Moser, M.; Georg, A. G.; Steinemann, F. L.; Rutti, D. P.; Meier, D. M. Continuous Milli-Scale Reaction Calorimeter for Direct Scale-up of Flow Chemistry. J. Flow. Chem. 2021, 11, 691– 699, DOI: 10.1007/s41981-021-00204-y Google Scholar 312Taylor, C. J.; Manson, J. A.; Clemens, G.; Taylor, B. A.; Chamberlain, T. W.; Bourne, R. A. Modern Advancements in Continuous-Flow Aided Kinetic Analysis. React. Chem. Eng. 2022, 7, 1037– 1046, DOI: 10.1039/D1RE00467K Google Scholar 313Lovato, K.; Fier, P. S.; Maloney, K. M. The Application of Modern Reactions in Large-Scale Synthesis. Nat. Rev. Chem. 2021, 5, 546– 563, DOI: 10.1038/s41570-021-00288-z Google Scholar 314Douglas, J. J.; Albright, H.; Sevrin, M. J.; Cole, K. P.; Stephenson, C. R. J. A Visible-Light-Mediated Radical Smiles Rearrangement and Its Application to the Synthesis of a Difluoro-Substituted Spirocyclic Orl-1 Antagonist. Angew. Chem., Int. Ed. 2015, 54,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
14898– 14902, DOI: 10.1002/anie.201507369 Google Scholar 315Yayla, H. G.; Peng, F.; Mangion, I. K.; McLaughlin, M.; Campeau, L. C.; Davies, I. W.; DiRocco, D. A.; Knowles, R. R. Discovery and Mechanistic Study of a Photocatalytic Indoline Dehydrogenation for the Synthesis of Elbasvir. Chem. Sci. 2016, 7, 2066– 2073, DOI: 10.1039/C5SC03350K Google Scholar 316Leech, M. C.; Garcia, A. D.; Petti, A.; Dobbs, A. P.; Lam, K. Organic Electrosynthesis: From Academia to Industry. React. Chem. Eng. 2020, 5, 977– 990, DOI: 10.1039/D0RE00064G Google Scholar 317Ouellet, S. G.; Roy, A.; Molinaro, C.; Angelaud, R.; Marcoux, J. F.; O’Shea, P. D.; Davies, I. W. Preparative Scale Synthesis of the Biaryl Core of Anacetrapib Via a Ruthenium-Catalyzed Direct Arylation Reaction: Unexpected Effect of Solvent Impurity on the Arylation Reaction. J. Org. Chem. 2011, 76, 1436– 1439, DOI: 10.1021/jo1018574 Google Scholar 318Watanabe, E.; Chen, Y. D.; May, O.; Ley, S. V. A Practical Method for Continuous Production of Sp3-Rich Compounds from (Hetero)Aryl Halides and Redox-Active Esters. Chem. - Eur. J. 2020, 26, 186– 191, DOI: 10.1002/chem.201905048 Google Scholar 319Broom, T.; Hughes, M.; Szczepankiewicz, B. G.; Ace, K.; Hagger, B.; Lacking, G.; Chima, R.; Marchbank, G.; Alford, G.; Evans, P.The Synthesis of Bromomethyltrifluoroborates through Continuous Flow Chemistry. Org. Process Res. Dev. 2014, 18, 1354– 1359, DOI: 10.1021/op400090a Google Scholar 320Schaber, S. D.; Gerogiorgis, D. I.; Ramachandran, R.; Evans, J. M. B.; Barton, P. I.; Trout, B. L. Economic Analysis of Integrated Continuous and Batch Pharmaceutical Manufacturing: A Case Study. Ind. Eng. Chem. Res. 2011, 50, 10083– 10092, DOI: 10.1021/ie2006752 Google Scholar 321Pollard, D. J.; Pralong, A. Single-Use Technology Implementation for Biologics and Vaccines Production. Biopharmaceutical Processing: Development, Design, and Implementation of Manufacturing Processes. 2018, 721– 740, DOI: 10.1016/B978-0-08-100623-8.00035-9 Google Scholar 322Lapkin, A.; Jacob, P. M.; Yaseneva, P.; Gordon, C.; Peace, A. Assessment of Sustainability within Holistic Process
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Design. Wiley Ser. Renew Reso. 2015, 37– 50, DOI: 10.1002/9781118933916.ch3 Google Scholar 323Lapkin, A. A.; Yaseneva, P. Life Cycle Assessment of Flow Chemistry Processes. Sustainable Flow Chemistry: Methods and Applications; Wiley-VCH, 2017; pp 249– 276 Google Scholar 324Demmon, S.; Bhargava, S.; Ciolek, D.; Halley, J.; Jaya, N.; Joubert, M. K.; Koepf, E.; Smith, P.; Trexler-Schmidt, M.; Tsai, P. A Cross-Industry Forum on Benchmarking Critical Quality Attribute Identification and Linkage to Process Characterization Studies. Biologicals. 2020, 67, 9– 20, DOI: 10.1016/j.biologicals.2020.06.008 Google Scholar 325Hlavacek, V.; Puszynski, J. A.; Viljoen, H. J.; Gatica, J. E.. Model Reactors and Their Design Equations. In Ullmann's Encyclopedia of Industrial Chemistry; Wiley-VCH, 2000. DOI: 10.1002/14356007.b04_121 Google Scholar 326Patterson, G. Modeling and Scale-up of Mixing- and Temperature-Sensitive Chemical Reactions. Ind. Eng. Chem. Res. 2005, 44, 5325– 5341, DOI: 10.1021/ie049161s Google Scholar 327Marchisio, D. L.; Barresi, A. A. Cfd Simulation of Mixing and Reaction: The Relevance of the Micro-Mixing Model. Chem. Eng. Sci. 2003, 58, 3579– 3587, DOI: 10.1016/S0009-2509(03)00264-1 Google Scholar 328Rigopoulos, S.; Jones, A. A Hybrid Cfd - Reaction Engineering Framework for Multiphase Reactor Modelling: Basic Concept and Application to Bubble Column Reactors. Chem. Eng. Sci. 2003, 58, 3077– 3089, DOI: 10.1016/S0009-2509(03)00179-9 Google Scholar 329Zlokarnik, M. Scale-up of Processes Using Material Systems with Variable Physical Properties. Chem. Biochem. Eng. Q. 2001, 15, 43– 47 Google Scholar 330Zlokarnik, M. Dimensional Analysis and Scale-up in Theory and Industrial Application. J. Liposome. Res. 2001, 11, 269– 307, DOI: 10.1081/LPR-100108610 Google Scholar 331Froment, G. F.; Bischoff, K. B.; Wild, J. D. Chemical Reactor Analysis and Design, 3rd ed.;Wiley, 2011. Google Scholar 332Rehage, H.; Bartsch, M.; Kind, M. A New Scale-up Method for Competitive Chemical Model Reactions Based on Complete Similarity. Chem. Eng. J. 2020, 400, 125763, DOI: 10.1016/j.cej.2020.125763 Google Scholar 333Stankiewicz, A. I.; Moulijn, J. A. Process Intensification: Transforming Chemical Engineering.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Chem. Eng. Prog. 2000, 96, 22– 34 Google Scholar 334Moran, S.; Henkel, K.-D., Reactor Types and Their Industrial Applications. In Ullmann's Encyclopedia of Industrial Chemistry; Wiley, 2016; pp 1– 49. DOI: 10.1002/14356007.b04_087.pub2 Google Scholar 335Plouffe, P.; Macchi, A.; Roberge, D. M. From Batch to Continuous Chemical Synthesis-a Toolbox Approach. Org. Process Res. Dev. 2014, 18, 1286– 1294, DOI: 10.1021/op5001918 Google Scholar 336Lindeque, R. M.; Woodley, J. M. Reactor Selection for Effective Continuous Biocatalytic Production of Pharmaceuticals. Catalysts. 2019, 9, 262, DOI: 10.3390/catal9030262 Google Scholar 337Dautzenberg, F. M.; Mukherjee, M. Process Intensification Using Multifunctional Reactors. Chem. Eng. Sci. 2001, 56, 251– 267, DOI: 10.1016/S0009-2509(00)00228-1 Google Scholar 338Balakotaiah, V.; Sun, Z.; Gu, T.; West, D. H. Scaling Relations for Autothermal Operation of Catalytic Reactors. Ind. Eng. Chem. Res. 2021, 60, 6565– 6582, DOI: 10.1021/acs.iecr.0c05594 Google Scholar 339Barecka, M. H.; Skiborowski, M.; Gorak, A. A Novel Approach for Process Retrofitting through Process Intensification: Ethylene Oxide Case Study. Chem. Eng. Res. Des. 2017, 123, 295– 316, DOI: 10.1016/j.cherd.2017.05.014 Google Scholar 340Whalley, A. Understand Pilot-Plant Design Specifications. Chem. Eng. Prog. 2016, 112, 25– 28 Google Scholar 341Zhu, R. H.; Reddy, R.; Ding, M.; Xu, M.; Deng, C. Y.; Tadayon, S.; Li, H.; Depew, K.; Lane, B. Development and Scale-up of a Continuous Manufacturing Process for a Hydrazine Condensation Reaction. Org. Process Res. Dev. 2021, 25, 199– 205, DOI: 10.1021/acs.oprd.0c00424 Google Scholar 342Zhou, S. H.; Hong, Q. X.; Mei, W. L.; He, Y.; Wu, C. J.; Sun, T. M. Scale-up of a Continuous Manufacturing Process of Edaravone. Org. Process Res. Dev. 2021, 25, 2146– 2153, DOI: 10.1021/acs.oprd.1c00228 Google Scholar 343Roberge, D. M.; Gottsponer, M.; Eyholzer, M.; Kockmann, N. Industrial Design, Scale-up, and Use of Microreactors. Chim. Oggi. 2009, 27, 8– 11 Google Scholar 344Kockmann, N.; Gottsponer, M.; Roberge, D. M. Scale-up Concept of Single-Channel Microreactors from
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Process Development to Industrial Production. Chem. Eng. J. 2011, 167, 718– 726, DOI: 10.1016/j.cej.2010.08.089 Google Scholar 345Weber, R.; Snowden-Swan, L. The Economics of Numbering-up a Chemical Process Enterprise. J. Adv. Manuf. Process. 2019, 1, e10011 DOI: 10.1002/amp2.10011 Google Scholar 346Fernandez Rivas, D.; Castro-Hernandez, E.; Villanueva Perales, A. L.; van der Meer, W. Evaluation Method for Process Intensification Alternatives. Chem. Eng. Proc. 2018, 123, 221– 232, DOI: 10.1016/j.cep.2017.08.013 Google Scholar 347Fernandez Rivas, D.; Cintas, P. On an Intensification Factor for Green Chemistry and Engineering: The Value of an Operationally Simple Decision-Making Tool in Process Assessment. Sustainable Chem. Pharm. 2022, 27, 100651, DOI: 10.1016/j.scp.2022.100651 Google Scholar 348Roberge, D. M.; Ducry, L.; Bieler, N.; Cretton, P.; Zimmermann, B. Microreactor Technology: A Revolution for the Fine Chemical and Pharmaceutical Industries?. Chem. Eng. Technol. 2005, 28, 318– 323, DOI: 10.1002/ceat.200407128 Google Scholar 349Pennemann, H.; Watts, P.; Haswell, S. J.; Hessel, V.; Lowe, H. Benchmarking of Microreactor Applications. Org. Process Res. Dev. 2004, 8, 422– 439, DOI: 10.1021/op0341770 Google Scholar 350Yang, Q.; Babij, N. R.; Good, S. Potential Safety Hazards Associated with Pd-Catalyzed Cross-Coupling Reactions. Org. Process Res. Dev. 2019, 23, 2608– 2626, DOI: 10.1021/acs.oprd.9b00377 Google Scholar 351Yang, Q.; Canturk, B.; Gray, K.; McCusker, E.; Sheng, M.; Li, F. Z. Evaluation of Potential Safety Hazards Associated with the Suzuki-Miyaura Cross-Coupling of Aryl Bromides with Vinylboron Species. Org. Process Res. Dev. 2018, 22, 351– 359, DOI: 10.1021/acs.oprd.8b00001 Google Scholar 352Yang, Q.; Cabrera, P. J.; Li, X. Y.; Sheng, M.; Wang, N. X. Safety Evaluation of the Copper-Mediated Cross-Coupling of 2-Bromopyridines with Ethyl Bromodifluoroacetate. Org. Process Res. Dev. 2018, 22, 1441– 1447, DOI: 10.1021/acs.oprd.8b00261 Google Scholar 353Wang, Z.; Richter, S. M.; Bellettini, J. R.; Pu, Y. M.; Hill, D. R. Safe Scale-up of Pharmaceutical Manufacturing Processes with Dimethyl Sulfoxide as the Solvent and a Reactant or a Byproduct.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Org. Process Res. Dev. 2014, 18, 1836– 1842, DOI: 10.1021/op500260n Google Scholar 354Yang, Q.; Sheng, M.; Li, X. Y.; Tucker, C.; Vásquez Cespedes, S. V.; Webb, N. J.; Whiteker, G. T.; Yu, J. Potential Explosion Hazards Associated with the Autocatalytic Thermal Decomposition of Dimethyl Sulfoxide and Its Mixtures. Org. Process Res. Dev. 2020, 24, 916– 939, DOI: 10.1021/acs.oprd.0c00159 Google Scholar 355Stoessel, F. Thermal Safety of Chemical Processes; Wiley, 2021. Google Scholar Cited By This article is cited by 197 publications. Brendan J. Wall, Mason T. Koeritz, Levi M. Stanley, Brett VanVeller. A Practical Start-Up Guide for Synthetic Chemists to Implement Design of Experiments (DoE). ACS Catalysis 2025, Article ASAP. Bernd Schaefer, (BASF SE)Shashank Shekhar, (AbbVie)James Murray (Amgen). Excellence in Industrial Organic Synthesis 2024. The Journal of Organic Chemistry 2025, 90 (15) , 5081-5082. Bernd Schaefer, (BASF SE)Shashank Shekhar, (AbbVie)James Murray (Amgen). Excellence in Industrial Organic Synthesis 2024. Organic Process Research & Development 2025, 29 (3) , 601-602. Matthew J. Takle, Linden Schrecker, Benjamin J. Deadman, Joachim Dickhaut, Andy Wieja, Klaus Hellgardt, King Kuok Mimi Hii. Flash Thermal Racemization of Chiral Amine in Continuous Flow: An Exploration of Reaction Space Using DoE and Multivariate Transient Flow. Organic Process Research & Development 2025, 29 (2) , 545-554. Janis M. Zakis, Rebeka A. Lipina, Sharon Bell, Simon R. Williams, Maurus Mathis, Magnus J. Johansson, Joanna Wencel-Delord, Tomas Smejkal. High-Throughput Enabled Iridium-Catalyzed C–H Borylation Platform for Late-Stage Functionalization. ACS Catalysis 2025, 15 (4) , 3525-3534. Saweta Garg, Pankaj Singla, Sarbjeet Kaur, Francesco Canfarotta, Eirini Velliou, James A. Dawson, Nikil Kapur, Nicholas J. Warren, Shoba Amarnath, Marloes Peeters. Future Perspectives on the Automation and Biocompatibility of Molecularly Imprinted Polymers for Healthcare Applications. Macromolecules 2025, 58 (3) , 1157-1168. Oliver Weismantel, Lakshani J. Weerarathna, Tanja Junkers. Autonomous Size-Targeting for Block Copolymer
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Nanoparticles. ACS Applied Polymer Materials 2025, 7 (2) , 938-947. Charitomeni Angeli, Sara Atienza-Sanz, Simon Schröder, Annika Hein, Yongxin Li, Alexander Argyrou, Angelina Osipyan, Henrik Terholsen, Sandy Schmidt. Recent Developments and Challenges in the Enzymatic Formation of Nitrogen–Nitrogen Bonds. ACS Catalysis 2025, 15 (1) , 310-342. Dong-Zhi Li, Xue-Qing Gong. Challenges with Literature-Derived Data in Machine Learning for Yield Prediction: A Case Study on Pd-Catalyzed Carbonylation Reactions. The Journal of Physical Chemistry A 2024, 128 (48) , 10423-10430. Elaine Reichert Raguram, Jakob C. Dahl, Klavs F. Jensen, Stephen L. Buchwald. Kinetic Modeling Enables Understanding of Off-Cycle Processes in Pd-Catalyzed Amination of Five-Membered Heteroaryl Halides. Journal of the American Chemical Society 2024, 146 (48) , 33035-33047. Rodolfo I. Teixeira, Michael Andresini, Renzo Luisi, Brahim Benyahia. Computer-Aided Retrosynthesis for Greener and Optimal Total Synthesis of a Helicase-Primase Inhibitor Active Pharmaceutical Ingredient. JACS Au 2024, 4 (11) , 4263-4272. Nahiane Pipaón Fernández, Odhran Cruise, Sarah E. F. Easton, Justin M. Kaplan, John L. Woodard, Damian P. Hruszkewycz, David C. Leitch. Direct Heterocycle C–H Alkenylation via Dual Catalysis Using a Palladacycle Precatalyst: Multifactor Optimization and Scope Exploration Enabled by High-Throughput Experimentation. The Journal of Organic Chemistry 2024, 89 (22) , 16145-16160. Jayalakshmi Sridhar, Galina Goloverda. Reaction Optimization Experiment for Undergraduate Capstone Organic Chemistry Laboratory Course. Journal of Chemical Education 2024, 101 (11) , 4680-4685. Alexander S. Shved, Blake E. Ocampo, Elena S. Burlova, Casey L. Olen, N. Ian Rinehart, Scott E. Denmark. molli: A General Purpose Python Toolkit for Combinatorial Small Molecule Library Generation, Manipulation, and Feature Extraction. Journal of Chemical Information and Modeling 2024, 64 (21) , 8083-8090. Sarah L. Boyall, Holly Clarke, Thomas Dixon, Robert W. M. Davidson, Kevin Leslie, Graeme Clemens, Frans L. Muller, Adam D. Clayton, Richard A. Bourne, Thomas W. Chamberlain.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Automated Optimization of a Multistep, Multiphase Continuous Flow Process for Pharmaceutical Synthesis. ACS Sustainable Chemistry & Engineering 2024, 12 (41) , 15125-15133. Gary Tom, Stefan P. Schmid, Sterling G. Baird, Yang Cao, Kourosh Darvish, Han Hao, Stanley Lo, Sergio Pablo-García, Ella M. Rajaonson, Marta Skreta, Naruki Yoshikawa, Samantha Corapi, Gun Deniz Akkoc, Felix Strieth-Kalthoff, Martin Seifrid, Alán Aspuru-Guzik. Self-Driving Laboratories for Chemistry and Materials Science. Chemical Reviews 2024, 124 (16) , 9633-9732. Yao Xu, Yuechen Gao, Min Wang, Xi Zhu. Human-Guided Metaverse Synthesis for Quantum Dots: Advancing Nanomaterial Research through Augmented Artificial Intelligence. ACS Applied Materials & Interfaces 2024, 16 (34) , 45207-45213. Claudio Lehmann, Kevin Eckey, Maria Viehoff, Christoph Greve, Thorsten Röder. Autonomous Online Optimization in Flash Chemistry Using Online Mass Spectrometry. Organic Process Research & Development 2024, 28 (8) , 3108-3118. Pauline Bianchi, Jean-Christophe M. Monbaliu. New Opportunities for Organic Synthesis with Superheated Flow Chemistry. Accounts of Chemical Research 2024, 57 (15) , 2207-2218. Hanyu Zhang, Yuan Zhou, Zhichao Zhang, Huaicheng Sun, Ziqi Pan, Minjie Mou, Wei Zhang, Qing Ye, Tingjun Hou, Honglin Li, Chang-Yu Hsieh, Feng Zhu. Large Language Model-Based Natural Language Encoding Could Be All You Need for Drug Biomedical Association Prediction. Analytical Chemistry 2024, 96 (30) , 12395-12403. Haripriya Thalla, Varshini Uma Jayaraman, Maheshkumar Uppada, Vishnuvardhan Reddy Eda, Saikat Sen, Rakeshwar Bandichhor, Srinivas Oruganti. Supervised Machine-Learning Algorithm using Low Data Sets: Flow Chemistry Optimization of the Key Urea Moiety Construction in Larotrectinib. Organic Process Research & Development 2024, 28 (7) , 2552-2560. A. Filipa Almeida, Sofia Branco, Luisa C. R. Carvalho, Andre Raposo Moreira Dias, Emília P. T. Leitão, Rui M. S. Loureiro, Susana D. Lucas, Ricardo F. Mendonça, Rudi Oliveira, Inês L. D. Rocha, Joao Sardinha, Saúl Silva, Luís M. S. Sobral, Nuno M. T. Lourenço, Pedro
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
C. Valente. Benchmarking Strategies of Sustainable Process Chemistry Development: Human-Based, Machine Learning, and Quantum Mechanics. Organic Process Research & Development 2024, 28 (7) , 2885-2895. Florian L. Wagner, Peter Sagmeister, Thomas G. Tampone, Vidhyadhar Manee, Dauzhan Yerkozhanov, Frederic G. Buono, Jason D. Williams, C. Oliver Kappe. Self-Optimizing Flow Reactions for Sustainability: An Experimental Bayesian Optimization Study. ACS Sustainable Chemistry & Engineering 2024, 12 (26) , 10002-10010. Danny Long, Richard G. Finke, Wolfgang Bangerth. Methodology for the Computational Optimization of Nanoparticle Synthesis and Particle Size Distribution, Using Ir(0)n Nanoparticles as an Example System. ACS Applied Nano Materials 2024, 7 (12) , 14090-14101. Jiapeng Guo, Guihua Luo, Kejie Chai, Weike Su, An Su. Continuous Flow Synthesis of N,O-Dimethyl-N′-nitroisourea Monitored by Inline Fourier Transform Infrared Spectroscopy: Bayesian Optimization and Kinetic Modeling. Industrial & Engineering Chemistry Research 2024, 63 (23) , 10162-10171. Aravind Senthil Vel, Kouakou E. Konan, Daniel Cortés-Borda, François-Xavier Felpin. Enhancing Optimization of Mixed Variables on a Robotic Flow Platform: Integrating Statistical Filtering with Nelder–Mead and Bayesian Methods. Organic Process Research & Development 2024, 28 (5) , 1597-1606. Felix Strieth-Kalthoff, Sara Szymkuć, Karol Molga, Alán Aspuru-Guzik, Frank Glorius, Bartosz A. Grzybowski. Artificial Intelligence for Retrosynthetic Planning Needs Both Data and Expert Knowledge. Journal of the American Chemical Society 2024, 146 (16) , 11005-11017. E. William Webb, Kevin Cheng, Wade P. Winton, Brandon J.C. Klein, Gregory D. Bowden, Mami Horikawa, S. Wendy Liu, Jay S. Wright, Stefan Verhoog, Dipannita Kalyani, Michael Wismer, Shane W. Krska, Melanie S. Sanford, Peter J.H. Scott. Development of High-Throughput Experimentation Approaches for Rapid Radiochemical Exploration. Journal of the American Chemical Society 2024, 146 (15) , 10581-10590. Sandip Bapu Khatal, Siddhartha K. Purkayastha, Ankur K. Guha, Srinu Tothadi, Sanjay Pratihar. Enhancing Precatalyst Performance and Robustness through Aromaticity: Insights from Iridaheteroaromatics. The Journal
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
of Organic Chemistry 2024, 89 (4) , 2480-2493. Casey L. Olen, Andrew F. Zahrt, Sean W. Reilly, Danielle Schultz, Khateeta Emerson, David Candito, Xiao Wang, Neil A. Strotman, Scott E. Denmark. Chemoinformatic Catalyst Selection Methods for the Optimization of Copper–Bis(oxazoline)-Mediated, Asymmetric, Vinylogous Mukaiyama Aldol Reactions. ACS Catalysis 2024, 14 (4) , 2642-2655. Wensheng Wang, Mingxin Li, Cong Xu. Selective Scaling-out of a Serpentine Micromixer: Scaling Criteria. Industrial & Engineering Chemistry Research 2024, 63 (5) , 2374-2383. Ramon Arora, Mark Lautens. Photoexcited Nickel-Catalyzed Carbohalogenation. ACS Catalysis 2024, 14 (3) , 1970-1975. Jingru Lu, David C. Leitch. Organopalladium Catalysis as a Proving Ground for Data-Rich Approaches to Reaction Development and Quantitative Predictions. ACS Catalysis 2023, 13 (24) , 15691-15707. Devin M. Makey, Roger C. Diehl, Yue Xin, Bridget E. Murray, Dwight R. Stoll, Brandon T. Ruotolo, James P. Grinias, Alison R. H. Narayan, Veronica Lopez-Carillo, Michaela Stark, Philipp Johnen, Robert T. Kennedy. High-Throughput Liquid Chromatographic Analysis Using a Segmented Flow Injector with a 1 s Cycle Time. Analytical Chemistry 2023, 95 (46) , 17028-17036. Nung Siong Lai, Yi Shen Tew, Xialin Zhong, Jun Yin, Jiali Li, Binhang Yan, Xiaonan Wang. Artificial Intelligence (AI) Workflow for Catalyst Design and Optimization. Industrial & Engineering Chemistry Research 2023, 62 (43) , 17835-17848. Joshua Schrier, Alexander J. Norquist, Tonio Buonassisi, Jakoah Brgoch. In Pursuit of the Exceptional: Research Directions for Machine Learning in Chemical and Materials Science. Journal of the American Chemical Society 2023, 145 (40) , 21699-21716. Elena Braconi, Edouard Godineau. Bayesian Optimization as a Sustainable Strategy for Early-Stage Process Development? A Case Study of Cu-Catalyzed C–N Coupling of Sterically Hindered Pyrazines. ACS Sustainable Chemistry & Engineering 2023, 11 (28) , 10545-10554. Eunjae Shim, Ambuj Tewari, Tim Cernak, Paul M. Zimmerman. Machine Learning Strategies for Reaction
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Development: Toward the Low-Data Limit. Journal of Chemical Information and Modeling 2023, 63 (12) , 3659-3668. Jason D. Williams, C. Oliver Kappe. Self-Optimizing Flow Reactors Get a Boost by Multitasking. ACS Central Science 2023, 9 (5) , 864-866. Tianyi Wu, Sina Kheiri, Riley J. Hickman, Huachen Tao, Tony C. Wu, Zhi-Bo Yang, Xin Ge, Wei Zhang, Milad Abolhasani, Kun Liu, Alan Aspuru-Guzik, Eugenia Kumacheva. Self-driving lab for the photochemical synthesis of plasmonic nanoparticles with targeted structural and optical properties. Nature Communications 2025, 16 (1) Anushka Satpathy, Koel Mukherjee, Vinod Kumar Nigam. Improved production of thermostable pectinase from novel Bacillus siamensis (BIOSMNF45) using computational-conventional approach. Scientific Reports 2025, 15 (1) Pinank Patel, Divya Adalja, Nikunj Mashru, Pradeep Jangir, Arpita, Reena Jangid, Gulothungan G, Mohammad Khishe. Multi objective elk herd optimization for efficient structural design. Scientific Reports 2025, 15 (1) Youngchun Kwon, Hyunjeong Jeon, Joonhyuk Choi, Youn-Suk Choi, Seokho Kang. Enhancing chemical reaction search through contrastive representation learning and human-in-the-loop. Journal of Cheminformatics 2025, 17 (1) Hailong Wang, Lijun Li, Fubin Yang, Dongming Li, Shuming Wu, Wei Yan, Guodong Sun, Jie Tan, Yishao Li, Hua Yang, Zhongqing Wang. Process development and optimization of apalutamide synthesis aided by the Design of Experiments (DoE). Tetrahedron 2025, 184 , 134725. Jiapeng Guo, Kejie Chai, Guihua Luo, Weike Su, An Su. Prior knowledge-based multi-round multi-objective Bayesian optimization: continuous flow synthesis and scale-up of O-methylisourea. Chemical Engineering and Processing - Process Intensification 2025, 215 , 110376. Emmanuel Chimeh Ezeako, Barine Innocent Nwiloh, Malachy Chigozie Odo, Vincent E. Ozougwu. Harnessing synthetic biology for sustainable industrial innovation: Advances, challenges, and future direction. Biochemical Engineering Journal 2025, 221 , 109777. Vikas Kumar Singh, Abesh Chatterjee, Payal Maiti, Pankaj Parmar, Subhrajit Mukherjee, B.C. Meikap. Pyrolysis of used lubricating oil using industrial waste-based
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.