Process Analytical Technology (PAT)

Process Analytical Technology (PAT) is a manufacturing methodology for high value chemicals and pharmaceuticals. Critical parameters (CPPs) and key performance indicators (KPIs) of the process are thoroughly understood, well-defined and continually monitored in order to ensure that the pre-defined critical quality attributes (CQA) of the final product are consistently achieved. PAT measures key quality and performance indicators in raw materials, in-process materials and processes in real-time. A well designed PAT-based process is stable, ensuring that the critical parameters and indicators remain within pre-described limits to ensure product quality and process safety. PAT is an important element in a Quality by Design (QbD) system, in which quality is not tested into product, but rather inherent, by-design. PAT is an orthogonalized application of chemical, physical, microbiological, mathematical and risk analysis elements resulting in a complete understanding and control of processes. PAT is applicable to many standard operations performed in chemical processes, including reaction monitoring and crystallization procedures.

Process Analytical Technology (PAT) utilizes a variety of tools, such as spectroscopic and chromatographic compositional analyzers, fixed purpose sensors, automated and statistical data analysis and overall knowledge management methods.  All of these technologies and methods are designed to provide information in a real-time or near real-time manner. Since the purpose of PAT is to improve the safety and quality of manufactured products, it must be incorporated into efforts leading to large-scale production such as early and late stage process development and process scale-up. The application of PAT methodology represents a technological and cultural shift in how a company or an industry approaches development, manufacturing, quality and safety. 

PAT seeks to provide fundamental, science-based insight into those parameters that are key to the stability of a process and resultant product quality. The benefits of incorporating PAT include improved product quality and uniformity, process cost reduction, enhanced process and product safety, eliminates or minimizes product re-work, speeds process cycle, facilitates regulatory acceptance and compliance, and overall creates a robust process that results in manufacturing of product that is right first time.  In process development and scale-up, PAT promotes process understanding by linking the effect of variables to process performance.  Thus, PAT positively influences organizational efficiencies, leads to the development of safe and sustainable processes, speeds the development pipeline, improves product cost through science-based understanding, enables data-driven decision making, and supports regulatory requirements.

There are a number of challenges associated with implementing PAT that encompasses technical, cost and company cultural issues.  With respect to technical challenges, new analyzer technology available in the market must first, be found and then evaluated,  compositional analyzers that produce spectra require development of methods and chemometric models to associate with quality attributes, any new technology will need to be integrated into existing plant equipment and data management systems, control strategies will need to be developed and regulatory approval must be obtained.  Costs of PAT implementation includes acquisition of the new technology, modifications to existing infrastructure, long term maintenance (both hardware and measurement strategy such as quantitative models), as well as training personnel.  As far as company culture, establishing PAT must be considered as a strategic top down goal to ensure smooth implementation without operational barriers.  This means that senior management, the technical staff and production personnel must embrace PAT to ensure success.  Again, a substantial investment in development and training will be necessary.  

PAT has an important role in the implementation of QbD in the pharmaceutical industry providing process knowledge, the effective process design space, and process monitoring to ensure that quality attributes are met and maintained.  Real-time analytics are at the center of this approach, while not relying solely on the classic approach of ofline, post-run QA/QC to ensure quality.  

Real time inline or online analytics encompass technology including:

• Sngle-purpose sensors for pH, O2, temperature and pressure
• Optical probe-based particle size analyzers that provide particle size, particle size distribution and in-situ imaging
• Reaction sampling devices that automatically extract and prep reaction samples for offline analysis
• Compositional monitors such as in-situ FTIR spectroscopy and Raman spectrometers that provide comprehensive reaction understanding and real-time process monitoring

Anchoring these PAT tools are automated chemical reactors, which are used for building process understanding during development and scale-up efforts.  Chemical reactors offer automated and precise control over key process variables, such as pH, temperature, pressure, mixing, dosing, and are used to correlate the effect of these critical variables with process performance.  The combination of automated chemical reactors and in-situ analytics are ideal for obtaining and recording data to support Design of Experiments (DoE) studies to define the effective process design space.  

Contact us to learn more about PAT tools or request an eDemo today.

Process Analytical Technology (PAT) provides key information for a wide range of applications, including:

• Crystallization and Precipitation and Recrystallization
• Catalyzed Reactions
• Biocatalysis
• Downstream Processing in Biotechnology
• Oligonucleotides Synthesis
• Organometallic Synthesis and Chemistry
• Chemical Reactions Kinetics
• Flow Chemistry
• Chemical Process Safety
• Green Chemistry
• Chemical Synthesis
• Hydroformylations
• Hydrogenation Reactions
• Lithiation and Organolithium Reactions
• Fluorinations and Fluorine Chemistry
• Suzuki and Related Cross-Coupling Reactions
• Grignard Reactions
• Alkylation Reactions and Friedel-Crafts Alkylations Reactions
• Polymerization Reactions
• Quality by Design (QbD)
• Raman Spectroscopy

Modernize Synthesis -  Synthesis workstations provide precise reactor control, automate the capture and digitalization of reaction and process information, and ensure reproducibility between users, experiments, departments and sites. Synthesis workstations serve as superior platforms for PAT technology implementation and DoE methods, help elucidate critical process parameters, and support scale-up since mixing conditions and process geometries can closely match manufacturing environments.

Transform Organizational Productivity –  Use technology, facilities and scientists’ time far more effectively through automation and digital capture of reaction and process information.  Data-rich experiments can be executed 24/7, with samples analyzed automatically at critical points in the process, with results easily shared within the organization, eliminating repetition, lost data and fostering discussion.

Provide Insight For Every Reaction – In-situ, real-time measurement of chemistry provides invaluable process understanding by aiding in the elucidation of reaction kinetics, mechanisms, and the effect of reaction parameters on performance.  Identify and track key reaction intermediates, and define sources for unwanted, toxic, or hazardous by products.  In-depth, data-based reaction understanding leads to stable, predictable processes that are green and safe. Personnel can make actionable decisions in real-time, or later to meet tech transfer or filing requirements. 

Take Control of Crystallization - Using inline tools to obtain process understanding of particle systems, scientists can create fit-for-purpose crystal systems.  Real-time  data enables control of supersaturation, nucleation, agglomerization, particle size, particle count, particle size distribution and crystal morphology for rapidly-changing crystallization processes.  Understanding and improving particle quality in development results in predictable downstream separations, formulations and product quality. 

Sustain a Culture of Safety - Effective deployment of technology enables knowledge-based risk assessment of processes and ensures operation within desired process design space.  Understand and control  the parameters that effect process stability and safety (thermodynamics, accumulation etc) to mitigate risk at all stages of development.  In situ probes alert if process crosses threshold values.   Automated experimentation and in situ sampling minimizes exposure of personnel to hazardous chemicals and potentially hazardous reaction conditions.  PAT and automation aids in eliminating unsafe procedures and processes from the workplace.

Deliver from Lab to Plant - Designing processes for manufacturability starts early in the development cycle and ensures consistency across scales.  PAT enables the development and transferability of parameters defining the process design space, with which personnel responsible for scale-up and manufacturing use to ensure product critical quality attributes.  Monitoring critical parameters in production ensures control strategies are in force and enable corrective action, if necessary.

Jurica, J. A., & McMullen, J. P. (2021). Automation Technologies to Enable Data-Rich Experimentation: Beyond Design of Experiments for Process Modeling in Late-Stage Process Development. Organic Process Research & Development, 25(2), 282–291.

In order to more completely characterize reactions, while mitigating the effect of potentially competing objectives in later stages of pharmaceutical development, the authors present a compelling case for using data rich experimentation (DRE). DRE takes advantage of available technologies that provide extensive, real-time analytical data that are used in conjunction with modeling tools to more fully define reactions and processes. Since reactions often proceed non-linearly, increased extent of time-referenced analytical data provides a more accurate view of reaction progress. They comment that benchtop automated lab reactors (ALR) used in conjunction with automated sampling devices are highly useful for providing chromatographic time-series data.  Whereas ALRs provide optimized control over key reaction variables, automated samplers extract, quench and dilute reaction samples for analysis by chromatographic instrumentation. 

ALRs with in situ analytical devices are often used to support Design of Experiment (DOE) methods in order to define optimum reaction conditions. The authors comment that statistical tools such as analysis of variance (ANOVA) and response surface modeling, used to define the relationship between an input factor and reaction response, are typically used to address a single time point and may not be as effective for modelling non-linear reaction profiles.  In order to more effectively use all the data points provided by in-situ analysis, the authors explored other modeling tools, choosing dynamic response surface methodology (DRSM) to take full advantage of the data rich experiments.  They commented that DRSM modeling provides information that helps to address time-related optimization issues associated with reaction yield and stability across scales. 

The authors applied DRE to fully characterize a cyclization that forms a pyridone intermediate compound in the doravirine process.  To obtain the experimental data, an EasySampler 1210 automated sampler was used in conjunction with an EasyMax 102 ALR equipped with 100 ml pressure vessels.  The ALR provided superb control of reaction variables to support DoE investigations.  For HPLC analysis, the EasySampler automatically quenched and diluted samples extracted from the pressure vessels at preset times during the course of the reaction.  The DOE featured a 2⁴ full factorial design to assess 4 parameters required a total of 20 experiments.  For each experiment, twelve reaction samples were collected at equal time intervals over a 22 hour period.   Using the DRSM 2.6c method, a model was generated for each response with particular interest in the model trends for pyridone, diene and two impurity compounds.  They also modeled time-dependent competing conditions and trade-offs necessary for achieving both a high yield and reaction stability using a Pareto front solution.  They found that an acceptable stability time coincided with maximum pyridone yield and that yields were only slightly lessened if longer stability was required.

Yang, C., Feng, H., & Stone, K. (2021). Characterization of Propionyl Phosphate Hydrolysis Kinetics by Data-Rich Experiments and In-Line Process Analytical Technology, Organic Process Research & Development, 25(3), 507–515.  doi=10.1021/acs.oprd.0c00451&ref=pdf

The authors commented that enzymatic phosphorylation using propionyl phosphate (PrP) as the phosphate donor is used in the synthesis of an important API.   Since there was limited information available regarding the reaction mechanism and stability of PrP, they decided to investigate these critical parameters.  This is particularly important since PrP hydrolysis competes with the desired enzymatic phosphorylation, and understanding PrP degradation kinetics is key in enabling successful reaction scale up.  For data-rich profiling of the temperature-dependent hydrolysis, they commented that the reaction is not amenable to measurement with off-line analysis methods such as HPLC.  For this reason, they turned to in situ FTIR, which had been previously demonstrated to successfully profile hydrolysis reactions.   To obtain concentration data from the raw FTIR measurements, the authors found that iPLS multivariate analysis provided superior results when compared to univariate models.  The entire wavelength range from 1454 cm-1 to 1010 cm−1 was used to construct models for the PrP mono and phosphate species, and ³¹P NMR analysis on seven reaction samples was used to calibrate the large FTIR dataset.

The real-time FTIR measurements enabled the authors to apply the Repeated Temperature Scanning (RTS) method to study the hydrolysis kinetics of PrP.  This consisted of tracking the concentration of the PrP mono and free phosphate through five temperature scanning cycles, each from 5°C to 38°C in the first part of the cycle and from 38°C to 5°C in the second half of the cycle.  At the end of these cycles, PrP mono was mostly hydrolyzed to the phosphate.  The PrP mono/free phosphate concentrations and reaction temperature profiles were fit to a first-order kinetic model in order to calculate the activation energy (Ea) and rate constant (kref).  To determine the optimum sampling of this data for estimating the kinetic parameters, various numbers of data points were tried in the regression of the DynoChem model.  The authors reported two key kinetic parameters for the PrP hydrolysis.  The activation energy at near neutral pH was found to be 107.2 kJ/mol, and apparent rate constant at 33 °C was 0.0721 h−1.  Further, the authors used Dynochem modeling to simulate reaction performance and to aid in developing process control strategies. The modeling indicated that at 0 °C over 24 hours, just 1% PrP mono was degraded, whereas nearly all the PrP mono degraded at 40 °C.  They commented that based on the modeling, when PrP mono is added to a vessel in the scale up process, appropriate temperature control (<4 °C ) is required to minimize hydrolysis prior to adding other reactants to initiate the enzymatic phosphorylation. 

The authors concluded by stating that data rich experimentation using the modified RTS method and real-time, in situ PAT reaction monitoring can provide the information required to produce quantifiable reaction kinetics and process understanding in a single well-designed experimental run. 

Gao, Y., Zhang, T., Ma, Y., Xue, F., Gao, Z. Hou, B. & Gong, J. (2021). Application of PAT-Based Feedback Control Approaches in Pharmaceutical Crystallization. Crystals, 11, 221. https://doi.org/10.3390/cryst11030221

In their review article, the authors comment that controlling crystallizations regulates polymorphs, crystal shape, size and size distribution in the crystalline product, and the key to control is understanding crystal nucleation and growth.  They state that Process Analytical Technology is important to provide the necessary parameters for data-driven control of crystallization processes. The article summarizes the on-line monitoring technology and model-free feedback control that has been applied to investigate various crystallization processes resulting in improved particle size distribution, polymorph control and product quality.

The authors discuss in detail several different model-free control strategies that have been developed by a number of different investigators that use real-time PAT.  These include:

• Supersaturation control (SSC)/concentration feedback control (CFC) for cooling and dissolution of crystals on lab and manufacturing scales using FTIR-ATR and UV/Vis – ATR
• Direct nucleation control (DNC) based on particle count in solution via ParticleTrack with FBRM technology
• Polymorph concentration control (PCC) applying in-solution, Raman-based polymorph measurement
• Image analysis direct nucleation control (IA-DNC) to monitor particles in solution
• SSC-DNC combined with the mass-count (MC) method is performed using ATR-FTIR and FBRM.
• Active polymorphic feedback control (APFC) using Raman and ATR-UV/Vis spectroscopy in combination.

Shvedova, K., Wu, W., Sayin, R., Casati, F., Halkude, B.S., Hermant, P., Shen, D.E., Ramnath, A., Su, Q., Born, S.C., Takizawa, B., Chattopadhyay, S., O’Connor, T.F., Yang, X., Ramanujam, S. & Mascia, S. (2020). Design and Commercialization of an End-to-End Continuous Pharmaceutical Production Process: A Pilot Plant Case Study, Organic Process Research & Development, 24, 12, 2874–2889

The authors comment that pharmaceuticals have typically been made using batch operations, which have inherent technical inefficiencies, and there is considerable interest in Integrated continuous manufacturing (ICM).  ICM uses a series of integrated operations to streamline production.  ICM systems use control systems that are model-based and are equipped with various PAT capabilities.  In their current work, the authors describe an ICM system with six main subfunctions including feeding, dissolution and clarification by-pass, reactive crystallization, filtration and resuspension, drying, extrusion-molding-coating and solvent recovery. 

To measure critical quality and material attributes, PAT probes were installed to provide real-time testing.  PAT in the reactive crystallizer includes ParticleTrack (FBRM) to measure chord length distribution and ReactIR to measure reactant concentration and yield.  These two technologies are also applied in the resuspension unit to determine API crystal CLD and reactant/solvent content in the  slurry.  In the latter case, if out-of-specification  material is detected by the ReactIR probe, the slurry is sent to waste.  Other PATs in the ICM include near-IR probes to measure residual solvent in the API and to determine the content uniformity of the API in the polymer melt. Raman probes are used to determine crystal form/crystallinity in two different locations and a laser diffraction system is used to measure API particle size distribution.  

The authors comment that real-time PAT monitoring in conjunction with the integrated system control is supportive of a Real Time Release and Quality by Design strategy.   This approach minimizes the need for postprocess quality testing, supports meeting regulatory requirements and significantly decreases inventory levels and lead time.  

Agosti, A., Panzeri, S., Gassa, F., Magnani, M., Forni, G., Quaroni, M., Feliciani, L. & Bertolini, G. (2020). Continuous Safety Improvements to Avoid Runaway Reactions: The Case of a Chloro-Thiadiazole Intermediate Synthesis toward Timolol, Organic Process Research & Development, 24, 1032−1042. https://doi.org/10.1021/acs.oprd.0c00048

The authors comment that constantly changing regulatory and safety requirements require re-examining processes that have been running for years.  For this reason, they decided to examine the process for timolol, and specifically the synthesis of a key morpholine adduct intermediate.  They stated that the current process has some safety concerns, specifically exothermic reaction steps and that the reaction is run neat (no solvent other than the morpholine reactant).   Thus, a risk assessment for the reaction of 3,4- dichloro-1,2,5-thiadiazole (DCTDA) to morpholine was considered in the event of a cooling failure, which might result in significant decomposition.  Therefore they investigated the thermal stability of the reagents and products using Differential Scanning Calorimetry to better define the risk level.  Further, they investigated the reaction to identify at what point that a loss of cooling would cause the reaction temperature to rise and trigger decomposition.  This was performed on  100ml scale using an EasyMax HFCal in order to measure the accumulated heat and calculate the maximum temperature of the synthesis reaction (MTSR). The reaction was found to be highly exothermic and the amount of heat generated, combined with increasing concentration as the morpholine solvent is removed as it converts to morpholine HCl, presents significant safety risks for this process. Also, a larger scale reaction performed using an Optimax HFCal system showed that the stirring rate for this reaction had to be quite high to avoid solids from settling and acting as a heat sink, further adding to the seriousness of the thermal issues.  

From this calorimetric investigation, the authors decided to modify the process to reduce safety risks.  The intended goals of the modification were to reduce the overall heat of the reaction, avoid an unwanted bis-morpholine HCL by-product that forms at high temperature, and to minimize the amount of morpholine equivalents required.  Using the Optimax calorimeter, they found that performing the reaction in an inert solvent (rather than neat) and reversing the order of reagent addition resulted in a reaction with lower MTSR and slower decomposition kinetics.  Furthermore, the bis morpholine HCL impurity was reduced by an order of magnitude.  Thus, they achieved their goals of a safer process and higher quality product by a thorough re-evaluation of the process with these calorimetric tools.   

Dance, E.X.Z., Crawford, M., Moment, A., Brunskill, A., Wabuyele, B., (2020). Kinetics, Thermodynamics, and Scale-Up of an Azeotropic Drying Process: Mapping Rapid Phase Conversion with Process Analytical Technology. Org. Process Res. Dev. 24, 1665−1674 https://doi.org/10.1021/acs.oprd.0c00275

The authors comment that distillation processes with multiple solid-state phases and changing liquid phase compositions are often difficult to understand and scale-up.  Even though the distillation process might be efficient, it may be inadequately reproducible due to a lack of information, and therefore abandoned.  In this current work, the authors reporting the development of an effective distillation process using in-line FTIR spectroscopy, in-line Raman spectroscopy, off line analytical measurements, all in combination with process modeling.  They undertook a detailed investigation of the thermodynamics and kinetics of the distillation process, which resulted in the development of an effective control strategy that was applied in manufacturing.

They explain that when the intermediate compound, 2′-C-methyluridine, is crystallized in water, a dihydrate forms, and it is necessary to remove the water, since it detrimental to further steps in the overall synthesis.  They formed the anhydrous compound by azeotropic distillation in acetonitrile, which was then telescoped into the following, water-sensitive reaction. The system complexity increased when a hemihydrate from of the molecule was discovered.  In order to better understand the overall system in more detail, in situ FTIR and in situ Raman spectroscopy were applied.  In situ FTIR (ReactIR) was used to monitor the water content in the system in real-time and Raman was used to analyze the solid-state form.  With the information gained from these tools, the researchers were able to construct a process phase map and identify the kinetics of the dihydrate, hemihydrate and anhydrate transformations.  With the thermodynamic and kinetic understanding achieved and documented, the authors were able to successfully transfer the distillation process from the gram scale in the lab to production of hundreds of kilograms of the anhydrous intermediate at another facility.  

EasyMax and OptiMax automated chemical reactors are synthesis workstations for fundamental parameter optimization.  These fully automated systems are ideal for the development and scale-up of chemical reactions. Both automated systems offer precise control over process parameters. 

• Manage all Critical Process Parameters (CPPs) including reagent dosing rate, temperature, mixing and pH.  The systems are ideal platforms for conducting DoE studies and data-rich experiments. 
• Fully integrate in-situ analytics including ReactIR (FTIR), ReactRaman (Raman), Particle Track (FBRM), EasyViewer (in-situ image analysis) and EasySampler (automated, in-situ reactor sampling) when offline analysis is required.) through a single software package, iControl. 
• iC software enables automated execution and control of all experiments and peripheral equipment as well as full data capture and record keeping

View a live eDemo from your work or home office on your schedule.

ReactIR enables scientists to measure reaction progress and dynamics in real-time, providing highly specific information about kinetics, mechanism, pathways, and the influence of reaction variables on performance.  Using ReactIR, directly trend reactants, reagents, intermediates, products and by-products as they change during the course of the reaction.  ReactIR provides critical information to scientists as they research, develop and scale-up chemical syntheses and chemical processes.

• Monitor multiple product quality attributes in-situ.
• Eliminate analysis delays and irreproducibility associated with sampling for offline analysis
• Quantify the impact of processing steps on product yield and process efficiency
• Supports ICH regulatory guidelines through data-rich experimentation and digitalization
• Measures batch or continuous flow operations
• Measure solutes directly in solution to support optimizing crystallization processes

ReactRaman is a probe-based, Raman spectroscopy instrument that  monitors chemical reactions and crystallization processes in situ, revealing mechanisms and kinetics in single or multiphase systems.  ReactRaman is part of an integrated family of products, which includes ReactIR™ and EasyMax ®. Designed specifically for chemical and process development, these tools are combined across the powerful iC software platform to provide unparalleled process understanding.

• Raman and FTIR are complementary methods and depending on the chemistry investigated, each has its own strengths. ReactRaman and ReactIR form a powerful molecular spectroscopy suite for the study of chemical and biological processes.  
• When used in conjunction with ReactIR, both solution phase (FTIR) and solid phase (Raman) are measured in slurries, providing a thorough understanding of the effect of parameters on supersaturation, nucleation, crystal growth, morphology, etc.

ParticleTrack using Focused Beam Reflectance Measurement (FBRM)

Particle size and count directly impact performance in multiphase processes including crystallization, emulsification and flocculation. By monitoring particle size and count in real time scientists can understand, optimize and scale-up processes confidently using evidence based methods.

• Track in-situ changing particle size and count in real time. Uses Chord Length Distribution measurements to obtain Particle Size Distribution (PSD).
• Characterize crystallization  across multiple particle types, morphologies and particle mechanisms related to process parameters
• Characterize conditions which affect nucleation, agglomeration, polymorphism and other system attributes
• Use ParticleTrack for solids analysis in slurries along with  ReactIR for solute-solution analysis to fully define crystallization systems

ParticleTrack with Focused Beam Reflectance Measurement (FBRM) technology is a probe-based particle analyzer that is inserted directly into a vessel to track changing particle size and count in real time. Uses Chord Length Distribution measurements to obtain Particle Size Distribution (PSD). 

EasyViewer is an inline particle size analysis tool based on high-resolution microscope images and verifiable image analysis. Together with image analysis based particle size measurements, EasyViewer provides rapid and early understanding of a how particles and droplets change. This can be leveraged to improve yield, purity, filtration and quality in smaller volumes and faster than ever before. Without sampling or diluting, researchers can visualize crystals, particles and droplets in an outstanding level of detail.

• Particle accumulation, size distribution and morphology are obtained from time-resolved inline images of the process allowing populations to be trended over time.
• Particles and crystal structures are monitored continuously, as experimental conditions vary as a function of process parameters
• Track particles in batch or continuous flow chemistries in real time with no sample removal required
• Characterize crystallization conditions which affect nucleation, precipitation, polymorphism and other system attributes

The value of Process Analytical Technology (PAT) data is enhanced with the advanced modeling of Dynochem and Reaction Lab. In experiments designed to develop kinetic parameters and insight into mechanisms, PAT data provides accurate rate constants and activation energies for use in models. Using this information, Reaction Lab simultaneously models the effect of variables such as temperature, pressure, reactant loading, product and by-product concentrations to provide a best set of reaction conditions. The response of the reaction to varying specific parameters and conditions can be observed, and response surfaces generated that give insight into yield/impurity tradeoffs as well as overall robustness of the reaction. Dynochem enables scale-up of reactions by modeling the effect of variables, including temperature, pressure, mixing rates and heat transfer to quickly identify appropriate process conditions that ensure safe, sustainable operation and meet yield/impurity targets. Particularly important in scaleup and production, Dynochem aids in managing risk through the development of design space models for optimized operating conditions and achieving QbD objectives. The synergy of data-rich PAT experiments and advanced modeling software enhances reaction and process understanding. 

Below is a selection of recent publications where Process Analytical Technology (PAT) was applied.

• Hutchinson, G., Alamillo-Ferrer, C., Burés, J. (2021). Mechanistically Guided Design of an Efficient and Enantioselective Aminocatalytic α-Chlorination of Aldehydes J. Am. Chem. Soc. 143, 18, 6805–6809. https://doi.org/10.1021/jacs.1c02997
• Gao, Y., Zhang, T., Ma, Y., Xue, F., Gao, Z. Hou, B. & Gong, J. (2021). Application of PAT-Based Feedback Control Approaches in Pharmaceutical Crystallization. Crystals, 11, 221. https://doi.org/10.3390/cryst11030221
• Allsop, G. L., Carey, J. S., Joshi, S., Leong, P., & Mirata, M. A. (2021). Process Development toward a Pro-Drug of R-Baclofen. Organic Process Research & Development, 25(1), 136–147. https://doi.org/10.1021/acs.oprd.0c00491
• Wu, X., Ding, G., Lu, W., Yang, L., Wang, J., Zhang, Y., Xie, X. & Zhang, Z. (2021). Nickel-Catalyzed Hydrosilylation of Terminal Alkenes with Primary Silanes via Electrophilic Silicon–Hydrogen Bond Activation. Organic Letters, 23(4), 1434–1439.
• Hosoya, M., Shiino, G. & Tsuno, N. (2021). A Practical Transferring Method from Batch to Flow Synthesis of Dipeptides via Acid Chloride Assisted by Simulation of the Reaction Rate., https://doi.org/10.1246/cl.210103
• Jurica, J. A., & McMullen, J. P. (2021). Automation Technologies to Enable Data-Rich Experimentation: Beyond Design of Experiments for Process Modeling in Late-Stage Process Development. Organic Process Research & Development, 25(2), 282–291.
• Burns, M., Perkins, D., Chan, L.C., Pilling, M.J., Jawor-Baczynska, A., Mullen, A.K., Steven, A., Wimsey, C., Elmekawy, A., Lamacraft, A., Dobson, B.C., McMillan, A.E., Hose, D.R.J., Inglesby, P.A., Raw, .A and Jones, M.F. (2021). Route Design to Manufacture: Synthesis of the Heterocyclic Fragment of AZD5718 Using a Non-cryogenic Lithiation-Alkoxycarbonylation Reaction, Org. Process Res. Dev. 25, 858−870. https://doi.org/10.1021/acs.oprd.0c00533
• Daniel A. Strassfeld, D.A., Algera, R.F., Wickens, Z.K. and Jacobsen, E.N. (2021). A Case Study in Catalyst Generality: Simultaneous, Highly-Enantioselective Brønsted- and Lewis-Acid Mechanisms in Hydrogen-Bond-Donor Catalyzed Oxetane Openings J. Am. Chem. Soc., https://doi.org/10.1021/jacs.1c03992  
• Smith, J.P., Obligacion, J.V., Dance, Z.E.X., Lomont, J.P., Ralbovsky, N.M., Bu, X. and Mann, B.F. (2021). Investigation of Lithium Acetyl Phosphate Synthesis Using Process Analytical Technology, Organic Process Research & Development, 25, 6, 1402-1413. DOI: 10.1021/acs.oprd.1c00091
• Trampuž, M., Teslić, D. and Likozar,B., Crystal-size distribution-based dynamic process modelling, optimization, and scaling for seeded batch cooling crystallization of Active Pharmaceutical Ingredients (API)(2021). Chem. Eng. Res. Design, 165, 254-269.
• Ren, R., Huang, P., Zhao, W., Li, T., Liu, M.and Wu, Y., (2021). A New ternary organometallic Pd(II)/Fe(III)/Ru(III) self-assembly monolayer: the essential ensemble synergistic for improving catalytic activity, RSC Adv.,11,1250-1260. DOI: 10.1039/D0RA09347E
• Payne, J., Cronin, J., Haer, M., Krouse, J., Prosperi, W., Drolet-Vives, K., Lieve, M., Soika, M., Balmer,M. and Kirkitadze, M. (2021). In-line monitoring of surfactant clearance in viral vaccine downstream processing, Comp. Struct. Biotech. J., 19, 1829-1837. https://doi.org/10.1016/j.csbj.2021.03.030
• Hosoya, M., Shiino, G. and Tsuno, N. (2021), A Practical Transferring Method from Batch to Flow Synthesis of Dipeptides via Acid Chloride Assisted by Simulation of the Reaction Rate. Chem. Letters, 50(6), 1254-1258. https://doi.org/10.1246/cl.210103
• Tang, W., Quan, Y., Gong, J., Wang, J., Yin, Q and Li, T. (2021). Form selection of concomitant polymorphs: A case study informed by crystallization kinetics modeling, AIChE, J., 67(4),  https://doi.org/10.1002/aic.17129
• Bosits, M.H., Szalay, Z., Pataki, H., Marosi, G. and Ádám Demeter (2021). Development of a Continuous Crystallization Process of the Spironolactone Hydrate Form with a Turbidity-Based Level Control Method, Org. Process Res. Dev., 25(4), 760–768. https://doi.org/10.1021/acs.oprd.0c00409
• Xu, J., Song, Y., He, J., Dong, S., Lin, L. and Feng, X. (2021). Asymmetric Catalytic Vinylogous Addition Reactions Initiated by Meinwald Rearrangement of Vinyl Epoxides, Angew. Chem., 60(26), 14521-14527. https://doi.org/10.1002/anie.202102054.
• An, Q., Wang, L., Bi,S., Zhao, W., Wei, D., Li, T., Liu,M. and Wu, Y. (2021).  Sandwich structured aryl-diimine Pd (II)/Co (II) monolayer—Fabrication, catalytic performance, synergistic effect and mechanism investigation. Mol. Catal., 501, 111359. https://doi.org/10.1016/j.mcat.2020.111359
• Acevedo, D., Wu, W-L., Yang, X., Pavurala, N., Mohammad, A. and O’Connor, T. (2021) Evaluation of focused beam reflectance measurement (FBRM) for monitoring and predicting the crystal size of carbamazepine in crystallization processes, CrystEngComm, 2021,23, 972-985. https://doi.org/10.1039/D0CE01388A.
• Johnson, C., Dabral, S., Rudolf, P., Licht, U., Hashmi, A.S.K. and Schaub, T. (2021). Liquid-liquid-phase Synthesis of exo-Vinylene Carbonates from Primary Propargylic Alcohols: Catalyst Design and Recycling. ChemCatChem, 13, 353–361. https://doi.org/10.1002/cctc.202001551
• Duprez, J., Kalbfleisch, K., Deshmukh, S., Payne, J., Haer, M., Williams, w., Durowoju, I. and Kirkitadze, M. (2021). Structure and compositional analysis of aluminum oxyhydroxide adsorbed pertussis vaccine, Comp. Struct. Biotech. J., 19, 439-447. https://doi.org/10.1016/j.csbj.2020.12.023
• Yang, C., Feng, H., & Stone, K. (2021). Characterization of Propionyl Phosphate Hydrolysis Kinetics by Data-Rich Experiments and In-Line Process Analytical Technology, Organic Process Research & Development, 25(3), 507–515.  doi=10.1021/acs.oprd.0c00451&ref=pdf
• Millward , M.J., Ellis, E., Ward, J.W., Clayden, J. (2021). Hydantoin-bridged medium ring scaffolds by migratory insertion of urea-tethered nitrile anions into aromatic C–N bonds, Chem.Sci., 12, 2091-2096. https://doi.org/10.1039/D0SC06188C