Extended Data Fig. 1: Partial scope and quality of optical yield estimates. | Nature

Extended Data Fig. 1: Partial scope and quality of optical yield estimates.

From: Robot-assisted mapping of chemical reaction hyperspaces and networks

Extended Data Fig. 1

a, Examples of reactions compatible (green background), borderline (yellow) and incompatible (pink) with yield determination using UV-Vis and spectral unmixing. The compatible reactions include but are not limited to Claisen–Schmidt condensation 1, Ullmann-type coupling 2, Suzuki–Miyaura cross-coupling 3, Cu(I)-catalysed alkyne-azide cycloaddition 4, Friedel–Crafts acylation 5, imine condensation 6, Glaser-type coupling 7 or Beckmann rearrangement 8. The Diels–Alder reaction 9 shown on yellow background is borderline in terms of the quality of unmixing (the overlap of the UV-Vis signals from the product and from dimethyl acetylenedicarboxylate prevents accurate fitting, leading to an overestimation of the yield). For the Diels–Alder reaction 10 shown on red background, the band of the product (λmax < 286 nm) is obscured by the toluene solvent (λmax = 286 nm). Other reactions unsuitable for the method are, for example, SN2 reaction of alkyl bromide 11 (not absorbing above 220 nm) or condensation of dichlorodiaminobenzene with phenanthrenequinone 12 (poor solubility of the product). b, R2 = 0.96 correlation between the optically determined yields versus the yields of the same reactions determined by conventional means (HPLC for E1 from Fig. 3 and Ugi-type MCR from Fig. 4; 1H NMR for Claisen–Schmidt condensation 1 and imine condensation 6; the weights of isolated compounds for Cu(I)-catalysed alkyne-azide cycloaddition 4 and Beckmann rearrangement 8). For numerical values of each data point, see Supplementary Tables 57 in Supplementary Information Sections 4.3 and 4.4.

Back to article page