2022
DOI: 10.1039/d2sc04041g
|View full text |Cite
|
Sign up to set email alerts
|

A broadly applicable quantitative relative reactivity model for nucleophilic aromatic substitution (SNAr) using simple descriptors

Abstract: We report a multivariate linear regression model able to make accurate predictions for the relative rate and regioselectivity of nucleophilic aromatic substitution (SNAr) reactions based on the electrophile structure. This...

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

1
20
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
7
1

Relationship

0
8

Authors

Journals

citations
Cited by 24 publications
(21 citation statements)
references
References 135 publications
1
20
0
Order By: Relevance
“…For aryl nuclei containing electron‐withdrawing groups, including sulphonyl, polyhalogens, nitro, cyano, and trifluoromethyl groups, studies on the S N Ar of fluorine by various nucleophilic substituents have been reported. [ 18,19 ] Using these electron‐withdrawing groups increases the speed of the S N Ar reaction process and produces excellent yields. Solvents, including hexamethylphosphoramide, dimethyl sulfoxide, dimethylacetamide, and dimethylformamide, are often used for S N Ar reactions because of their polarity and aprotic nature.…”
Section: Introductionmentioning
confidence: 99%
“…For aryl nuclei containing electron‐withdrawing groups, including sulphonyl, polyhalogens, nitro, cyano, and trifluoromethyl groups, studies on the S N Ar of fluorine by various nucleophilic substituents have been reported. [ 18,19 ] Using these electron‐withdrawing groups increases the speed of the S N Ar reaction process and produces excellent yields. Solvents, including hexamethylphosphoramide, dimethyl sulfoxide, dimethylacetamide, and dimethylformamide, are often used for S N Ar reactions because of their polarity and aprotic nature.…”
Section: Introductionmentioning
confidence: 99%
“…Tested on Pfizer’s in-house dataset and USPTO public data, the complete workflow achieves around 95% accuracy on both datasets. The model developed by Jorner et al achieved an accuracy of 86% under the same metrics, and another recent linear regression model by Lu and co-workers reached 91% accuracy. While direct comparison is difficult since they were tested on different data sets, our workflow outperforms the two cited work mainly because of two points.…”
Section: Discussionmentioning
confidence: 78%
“…Moreover, confident predictions dominate the first-step ML predictions. Compared to the reported models, , only a minor portion of the reactions will trigger DFT modeling (compared to the conventional TST modeling which computes all reacting pairs, the proposed workflow only takes 23 and 36% of computing resources, estimated on Pfizer and USPTO datasets, respectively), which allows us to screen more candidates or save computational resources for other tasks.…”
Section: Discussionmentioning
confidence: 99%
“…Moreover, confident predictions dominate the first-step machine learning predictions. Compared to reported models 13,63 , only a minor portion of the reactions will trigger DFT modeling, which allows us to screen more candidates or save computational resources for other tasks.…”
Section: Discussionmentioning
confidence: 99%