Towards automation of chemical process route selection based on data mining

Graphical abstract: Towards automation of chemical process route selection based on data mining

A methodology for chemical routes development and evaluation on the basis of data-mining is presented. A section of the Reaxys database was converted into a network, which was used to plan hypothetical synthesis routes to convert a bio-waste feedstock, limonene, to a bulk intermediate, benzoic acid. The route evaluation considered process conditions and used multiple indicators, including exergy, E-factor, solvent score, reaction reliability and route redox efficiency, in a multi-criteria environmental sustainability evaluation. The proposed methodology is the first route evaluation based on data mining, explicitly using reaction conditions, and is amenable to full automation.

In the field of process and synthetic chemistry ‘clean synthesis’ has become one of the standard criteria for good, commercially viable synthesis routes. As a result synthetic and process chemists must be equipped with adequate methodologies for quantification of ‘cleanness’ or ‘greenness’ of alternative routes at the early phases of the development cycle. These new criteria, and the traditional criteria of cost, security of supply, health and safety (H&S), and risk, provide a balanced picture of sustainability of a future technology. Thus, there are two separate aspects to process chemistry: developing the chemistry and the process, and evaluating the overall process, which must occur in parallel. Evaluation of the proposed routes requires data. As data science rapidly evolves, chemistry will inevitably use more of the new tools of data mining and data analysis to automate the routine tasks, such as evaluation of process metrics. In this paper we show some initial results in automation of process evaluation based on deep data mining of process chemistry and multi-criteria decision making.

The evaluation of greenness is a mature field, with a large number of published and standardised approaches, of which many are adopted by industry. 1 However, all published methods are highly case-specific and rather labour-intensive. In the field of synthetic routes development one of the most exciting new areas is the potential for automation of synthesis planning using data mining.2 What has never been attempted before is to automate route generation and evaluation in a coherent methodology, which would aid process development at the early, data-lean, stages. For this we show how to automatically generate process options using a network representation of a section of Reaxys database,3 followed by their screening using multi-criteria decision making, see Fig. 1. As the methods mature and become commercially available, such integration and automation will produce significant savings of time, and would deliver a far more detailed view of the competing synthesis route options than is generally possible at the early stages of design.

To date, obtaining the data, assembling the network and finding potential synthesis routes can already be carried out in a fully automated fashion. Due to issues around data availability the connection to the analysis of the routes still has to be initiated manually, involving a data curation step. The subsequent analysis and multi-criteria decision making have been largely automated in this study. To our knowledge this is the first example of the analysis of synthesis routes generated from the network representation of Reaxys obtained through datamining, using reaction conditions and process data.

image file: c6gc02482c-f2.tif

Fig. 2 A section of a network of organic chemistry. Dots are species and arrows represent reactions.
  1. D. J. C. Constable, C. Jimenez-Gonzalez and A. Lapkin, in Green Chemistry Metrics, John Wiley & Sons, Ltd, Chichester, UK, 2009, pp. 228–247 
  2. S. Szymkuć, E. P. Gajewska, T. Klucznik, K. Molga, P. Dittwald, M. Startek, M. Bajczyk and B. A. Grzybowski, Angew. Chem., Int. Ed., 2016, 55, 5904–5937 
  3. Reed Elsevier Properties SA, Login – Reaxys Login Page [Internet], 2014 [accessed 2014 Jun 8]. Available from: https://www.reaxys.com/. Reaxys is a trademark, copyright owned by Relex Intellectual properties SA and used under licence.

Towards automation of chemical process route selection based on data mining

*Corresponding authors
aDepartment of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge CB2 3RA, UK
E-mail: aal35@cam.ac.uk
Green Chem., 2017,19, 140-152

DOI: 10.1039/C6GC02482C, http://pubs.rsc.org/en/Content/ArticleLanding/2017/GC/C6GC02482C?utm_medium=email&utm_campaign=pub-GC-vol-19-issue-1&utm_source=toc-alert#!divAbstract

Professor Alexei Lapkin, FRSC

Professor Alexei Lapkin FRSC

Professor of Sustainable Reaction Engineering

Fellow of Wolfson College

Catalytic Reaction Engineering

Sustainable Chemical Technologies

Office Phone: 330141

University of Cambridge
Image result for Department of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge CB2 3RA, UK

Biography:

MChem in Biochemistry, Novosibirsk State University, 1994

PhD in Chemical Engineering, University of Bath, 2000

Boreskov Institute of Catalysis, Novosibirsk, Russia (1994-1997)

University of Bath, Department of Chemical Engineering, Research Officer (1997-2000)

University of Bath, Department of Chemical Engineering, Lecturer-SL-Reader (2000-2009)

University of Warwick, School of Engineering, Professor of Engineering (2009-2013)

Research Interests

Reaction Engineering group

Our group is developing cleaner manufacturing processes within chemical and chemistry using industries. We are mainly focusing on liquid- and multi-phase catalytic and biochemical processes. Within the group we have pursued projects on developing functional materials for catalysts, adsorbents and reactors, design of multi-functional intensive reactors, modelling of reaction kinetics and integrated processes, linking reaction kinetics with computational fluid dynamics (CFD) and linking process modelling with life cycle assessment (LCA), integration of reactions and separation.

Public funding:

The group is currently involved in an EU project ‘RECOBA’ (http://www.spire2030.eu/recoba/), in which our group collaborates with Materials and Electronic Engineering at Cambridge to work on innovative measurement techniques for monitoring processes under reaction conditions.

We are involved in the EPSRC project on developing novel routes to platform and functional molecules from waste terpenes, led by University of Bath.

We are involved in “Dial a Molecule 2” network funded by EPSRC.

Keywords

  • Reaction Engineering
  • flow
  • sustainability
  • heterogeneous catalysis
  • catalysis

Key Publications

J. Zakrzhewski, A.P. Smalley, M. Kabeshov, A. Lapkin, M. Gaunt, Continuous flow synthesis and derivatization of aziridines via palladium-catalyzed C(sp3)-H activation, Angew. Chem. Int. Ed., 55 (2016) 8878-8883.

P. Yaseneva, P. Hodgson, J. Zakrzewski, S. Falss, R.E. Meadows, A.A. Lapkin, Continuous flow Buchwald-Hartwig amination of a pharmaceutical intermediate, React. Chem. Eng., 1 (2016) 229-238.

P. Yaseneva, D. Plaza, X. Fan, K. Loponov, A. Lapkin, Synthesis of the antimalarial API artemether in a flow reactor, Catal. Today, 239 (2015) 90-96.

N. Peremezhney, E. Hines, A. Lapkin, C. Connaughton, Combining Gaussian processes, mutual information and a generic algorithm for multi-targeted optimisation of expensive-to-evaluate functions, Engineering Optimisation, 46 (2014) 1593-1607.

P. Yaseneva, C.F. Marti, E. Palomares, X. Fan, T. Morgan,P.S. Perez, M. Ronning, F. Huang,T. Yuranova, L. Kiwi-Minsker, S. Derrouiche, A.A. Lapkin, Efficient reduction of bromates using carbon nanofibre supported catalysts: experimental and a comparative life cycle assessment study, Chem. Eng. J., 248 (2014) 230-241

K.N. Loponov, J. Lopes, M. Barlog, E.V. Astrova, A.V. Malkov, A.A. Lapkin, Optimization of a Scalable Photochemical Reactor for Reactions with Singlet Oxygen, Org.Process Res.Dev., 18 (2014) 1443-1454.

X. Fan, V. Sans, P. Yaseneva, D. Plaza, J.M.J. Williams, A.A. Lapkin, Facile Stoichiometric Reductions in Flow: an Example of Artemisinin, Org.Process Res.Dev., 16 (2012) 1039-1042.

M.V. Sotenko, M. Rebros, V.S. Sans, K.N. Loponov, M.G. Davidson, G. Stephens, A.A. Lapkin, Tandem transformation of glycerol to esters, J. Biotechnol., 162 (2012) 390-397.

A.A. Lapkin, A. Voutchkova, P. Anastas, A conceptual framework for description of complexity in intensive chemical processes, Chem. Eng. Processing. Process intensification, 50 (2011) 1027-1034.

Lapkin, A., Peters, M., Greiner, L., Chemat, S., Leonhard, K., Liauw, M. A. and Leitner, W., Screening of new solvents for artemisinin extraction process using ab-initio methodology, Green Chem., 12 (2010) 241-251.

Lapkin, A. A. and Plucinski, P. K., Engineering factors for efficient flow processes in chemical industries, in Chemical reactions and processes under flow conditions, pp. 1- 43, Eds: Luis, S. V. and Garcia-Verdugo, E., Royal Society of Chemistry, Cambridge, 2010.

Iwan, A., Stephenson, H., Ketchie, W. C. and Lapkin, A. A., High temperature sequestration of CO2 using lithium zirconates, Chem. Eng. J., 146 (2009) 249-258.

Constable, D. J. C., Jimenez-Gonzalez, C. and Lapkin A., ‘Process metrics’, in Green chemistry metrics: measuring and monitoring sustainable processes, pp.  228- 247, Eds.: Lapkin, A. and Constable, D. J. C., Wiley-Blackwell, Chichester, 2008.

L.Torrente-Murciano, A.Lapkin, D.V. Bavykin, F.C. Walsh, K. Wilson, Highly selective Pd/titanate nanotubes catalysts for the double bond migration reaction, J. Catal., 245 (2007) 270-276.

A. Lapkin, P. Plucinski, Comparative assessment of technologies for extraction of artemisinin, J. Natural Prod., 69 (2006) 1653-1664.

D.V. Bavykin, A.A. Lapkin, S.T. Kolaczkowski, P.K. Plucinski, Selective oxidation of alcohols in a continuous multifunctional reactor: ruthenium oxide catalysed oxidation of benzyl alcohol, Applied Catal. A: General, 288 (2005) 165-174.

Image result for A. A. Lapkin

////////automation, chemical process,  route selection, data mining

A robust and recyclable polyurea-encapsulated copper(I) chloride for one-pot ring-opening/Huisgen cycloaddition/CO2 capture in water

Green Chemistry International

Green Chem., 2016, 18,6357-6366
DOI: 10.1039/C6GC01956K, Paper
Yun Chen, Wei-Qiang Zhang, Bin-Xun Yu, Yu-Ming Zhao, Zi-Wei Gao, Ya-Jun Jian, Li-Wen Xu
One-pot ring-opening/Huisgen cycloaddition reactions combined with CO2 capture were carried out successfully in the presence of polyurea-encapsulated CuCl.
A robust and recyclable polyurea-encapsulated copper(I) chloride for one-pot ring-opening/Huisgen cycloaddition/CO2 capture in water
Yun Chen,a Wei-Qiang Zhang,a Bin-Xun Yu,a Yu-Ming Zhao,a Zi-Wei Gao,*a Ya-Jun Jiana and Li-Wen Xu*ab
*Corresponding authors
aKey Laboratory of Applied Surface and Colloid Chemistry, Ministry of Education (MOE) and School of Chemistry and Chemical Engineering, Shaanxi Normal University, Xi’an 710062, P. R. China
E-mail: liwenxu@hznu.edu.cn, zwgao@snnu.edu.cn
bKey Laboratory of Organosilicon Chemistry and Material Technology of Ministry of Education, Hangzhou Normal University, No 1378, Wenyi West Road, Science Park of HZNU, Hangzhou 311121, P. R. China
Green Chem., 2016,18…

View original post 88 more words

Moving along the pyridine ring — amphoteros

A couple of weeks ago, I finished teaching the synthesis of pyridine and its derivatives in my 4th year synthesis class. Whenever I present this material, I can’t help but appreciate the value of N-oxidation. While there are N-oxides of other heterocycles (thiazole N-oxide stands out for its interesting properties), nowhere else do I feel […]

via Moving along the pyridine ring — amphoteros

Acetylcholine Chloride

New Drug Approvals

Acetylcholine Chloride

2-acetyloxyethyl(trimethyl)azanium;chloride

60-31-1

Molecular Formula: C7H16ClNO2
Molecular Weight: 181.66 g/mol

Acetylcholine chloride is obtained as white or off-white hygroscopic crystals, or as a crystalline powder. The salt is odorless, or nearly odorless, and is a very deliquescent powder. Acetylcholine bromide is obtained as deliquescent crystals, or as a white crystalline powder. The substance is hydrolyzed by hot water and alkali

Image result for acetylcholine chloride

Acetylcholine is an organic chemical that functions in the brain and body of many types of animals, including humans, as a neurotransmitter—a chemical released by nerve cells to send signals to other cells. Its name is derived from its chemical structure: it is an ester of acetic acid and choline. Parts in the body that use or are affected by acetylcholine are referred to as cholinergic. Substances that interfere with acetylcholine activity are called anticholinergics.

Acetylcholine is the neurotransmitter used…

View original post 590 more words

BOSENTAN PRECURSOR

New Drug Approvals

N,N′-(6,6′-(2,2-Dimethyl-1,3-dioxolane-4,5-diyl)bis- (methylene)bis(oxy)bis(5-(2-methoxy phenoxy)-2,2′-bipyrimidine-6,4-diyl))bis(4-tert-butylbenzenesulfonamide)

Mp: 72−74 °C.

1 H NMR (400 MHz, CDCl3): δ 1.25 (6H, s), 1.29 (18H, s), 3.84−3.90 (4H, m), 4.27−4.31 (2H, m), 6.84−6.87 (3H, t), 6.97−7.00 (2H, dd), 7.09−7.13 (3H, t), 7.43−7.45 (10H, m), 9.0−9.01 (4H, d), 8.43 (2H, br s);

13C NMR (100 MHz, CDCl3): δ 25.88, 30.02, 34.10, 55.01, 61.53, 77.36, 108.43, 111.4, 118.73, 120.4, 124.09, 124.34, 126.67, 127.38, 128.35, 135.30, 138.25, 144.74, 148.62, 150.99, 156.07, 156.71, 160.56;

MS: m/z 1142.2 (M + H);

Elem. Anal: Found: C 59.87, H 5.20, N 12.38; Calcd for C57H60N10O12S2: C 59.99, H 5.30, N 12.27

Abstract Image

A new and efficient synthetic process for the synthesis of an endothelin receptor antagonist, bosentan monohydrate, involves the coupling of ptert-butyl-N-(6-chloro-5-(2-methoxy phenoxy)-2,2′-bipyrimidin-4-yl)benzenesulfonamide (7) with (2,2-dimethyl-1,3-dioxolane-4,5-diyl)dimethanol (14) as a key step. This new process provides desired bosentan monohydrate (1) with better quality and…

View original post 52 more words

Heck–Matsuda Reaction in Flow

Abstract Image

Product 3 was obtained as a mixture of diastereomers (58:42). The NMR data are consistent with literature precedent.20a

Major diastereomer: 1H NMR (300 MHz, CDCl3) δ (ppm) 7.25-7.28 (m, 2H), 7.14-7.17 (m, 2H), 5.14 (dd, 1H, J = 2.5, 5.8 Hz), 4.29 (t, 1H, J = 8.3 Hz), 3.79 (dd, 1H, J = 6.9, 8.4 Hz), 3.54-3.62 (m, 1H), 3.38 (s, 3H), 2.32 (dd, 1H, J = 7.7, 12.9 Hz), 2.04 (ddd, 1H, J = 5.1, 9.3, 13.1 Hz);

Minor diastereomer: 1H NMR (300 MHz, CDCl3) δ 7.25-7.28 (m, 4H), 5.16 (d, 1H, J = 4.4 Hz), 4.17 (t, 1H, J = 8.1 Hz), 3.72 (dd, 1H, J = 8.5, 9.7 Hz), 3.42 (s, 3H), 3.32-3.36 (m, 1H), 2.59 (ddd, 1H, J = 5.5, 10.3, 13.7 Hz), 1.91 (ddd, 1H, J = 2.4, 7.7, 10.2 Hz);

13C NMR (75 MHz, CDCl3) δ (ppm) 141.4, 140.0, 132.4, 132.3, 129.1, 128.7, 128.7, 128.5, 105.7, 105.4, 73.7, 73.0, 54.9, 54.7, 43.6, 42.1, 41.4, 41.1.

(20) (a) Oliveira, C. C.; Angnes, R. A.; Correia, C. R. D. J. Org. Chem. 2013, 78, 4373. (b) Oliveira, C. C.; Pfaltz, A.; Correia, C. R. D. Angew. Chem. Int. Ed. 2015, 54, 14036.

The optimization of a palladium-catalyzed Heck–Matsuda reaction using an optimization algorithm is presented. We modified and implemented the Nelder–Mead method in order to perform constrained optimizations in a multidimensional space. We illustrated the power of our modified algorithm through the optimization of a multivariable reaction involving the arylation of a deactivated olefin with an arenediazonium salt. The great flexibility of our optimization method allows to fine-tune experimental conditions according to three different objective functions: maximum yield, highest throughput, and lowest production cost. The beneficial properties of flow reactors associated with the power of intelligent algorithms for the fine-tuning of experimental parameters allowed the reaction to proceed in astonishingly simple conditions unable to promote the coupling through traditional batch chemistry.

/////////

BioAssay Express: models mixing assays & compounds — Cheminformatics 2.0

The latest experimental feature of the BioAssay Express project involves taking all of the curated assays (3500 so far) and their corresponding compounds from PubChem (hundreds of thousands of unique structures) and feeding them all into one giant Bayesian model. Rather than the usual approach of modelling compound ⟹ activity separately for each assay, this approach takes advantage […]

via BioAssay Express: models mixing assays & compounds — Cheminformatics 2.0