Skip to main content
NCBI home page
ACS Medicinal Chemistry Letters logoLink to ACS Medicinal Chemistry Letters
. 2023 Mar 16;14(4):521–529. doi: 10.1021/acsmedchemlett.2c00538

ChemBeads-Enabled Photoredox High-Throughput Experimentation Platform to Improve C(sp2)–C(sp3) Decarboxylative Couplings

Nathan J Gesmundo 1,*, Noah P Tu 1, Kathy A Sarris 1, Ying Wang 1
PMCID: PMC10108395  PMID: 37077401

Abstract

graphic file with name ml2c00538_0007.jpg

Enthusiasm surrounding nickel/photoredox C(sp2)–C(sp3) cross-couplings is very high; however, these methods are sometimes challenged by complex drug-like substrates in discovery chemistry. In our hands this has been especially true of the decarboxylative coupling, which has lagged behind other photoredox couplings in internal adoption and success. Herein, the development of a photoredox high-throughput experimentation platform to optimize challenging C(sp2)–C(sp3) decarboxylative couplings is described. Chemical-coated glass beads (ChemBeads) and a novel parallel bead dispenser are used to expedite the high-throughput experimentation process and identify improved coupling conditions. In this report, photoredox high-throughput experimentation is utilized to dramatically improve low-yielding decarboxylative C(sp2)–C(sp3) couplings, and libraries, using conditions not previously identified in the literature.

Keywords: High-throughput experimentation, chemical-coated glass beads, ChemBeads, photoredox catalysis, C(sp2)−C(sp3) cross-coupling, decarboxylative coupling, parallel library synthesis

High-throughput experimentation (HTE) enables one to rapidly evaluate multiple reaction conditions in parallel for a challenging transformation, making it a powerful tool for reaction optimization.1,2 The technologies utilized in HTE allow one to miniaturize reactions, down to the micromole or nanomole scale, to increase reaction throughput and stretch materials further.3,4 Within pharmaceutical drug discovery, HTE can be highly impactful, as medicinal chemists often work with complex functionalized cores, meaning literature reaction conditions may not work as well as expected, or at all. Medicinal chemists also often operate under pressing timelines, so technology like HTE, which makes analogue synthesis more efficient, ultimately enables more structure–activity relationship (SAR) questions to be answered in less time. Even using HTE to identify a “hit”, reaction conditions which prepare isolable amounts of material regardless of yield, can help a team answer key SAR questions. In the context of pharmaceutically relevant transformations, HTE has been demonstrated in the optimization of challenging palladium-catalyzed Buchwald–Hartwig aminations and Suzuki cross-couplings,4−6 SNAr,3 C–H alkylation,7 photoredox aminations,8 and C(sp2)–C(sp3) couplings.9,10

Photocatalysis has had a profound impact on how we think about preparing molecules in drug discovery,11,12 and photoredox C(sp2)–C(sp3) cross-couplings specifically have emerged as attractive methods to directly form C(sp2)–C(sp3) bonds.13−17 These methods rely on dual nickel/photoredox catalysis and form the C(sp2)–C(sp3) bond from aryl halides and readily available redox-active alkyl coupling partners (C(sp3)-centered radical precursors) such as carboxylic acids,18 potassium alkyl trifluoroborate salts (BF3K),19 or other reagents20−24 (Figure 1). In discovery chemistry these methods are attractive because they allow direct access to alkyl-substituted analogues, which possess increased fraction sp3 (Fsp3) compared to biaryl counterparts. Increasing Fsp3 can improve properties such as solubility and crystallinity while also broadening accessible chemical space.25,26 We recently undertook a systematic evaluation of seven C(sp2)–C(sp3) cross-coupling methods to better understand the strengths and limitations of each method, and to facilitate use of these methods internally.16 Three of seven methods were nickel/photoredox reactions: the BF3K coupling (using alkyl BF3K salts), the cross-electrophile coupling (CEC, using alkyl bromides), and the decarboxylative coupling (using alkyl carboxylic acids) (Figure 1). Using library synthesis, we found that the BF3K and CEC methods worked very well across a broad range of primary and secondary alkyl coupling partners, while tertiary alkyl partners were limitations. The decarboxylative coupling, on the other hand, was not as successful in our evaluation (Figure 1A). The average library yield for primary carboxylic acids was just 3% (compared to 21% and 18% for the BF3K and CEC methods respectively), the average library yield for secondary carboxylic acids was 5% (versus 23% and 15% for the BF3K and CEC methods), and the median decarboxylative coupling yield was 0%. Furthermore, results were not consistent when comparing related alkyl acids (Figure 1B). Some α-oxy and α-amino acids worked reasonably well, like tetrahydro-2-furoic acid (2) and Boc-Pro-OH (3), while others did not (5 and 6). Similar observations were made comparing α-alkyl tetrahydropyran-4-carboxylic acid 4 (11% yield) versus cyclopentanecarboxylic acid 7 (3% yield). For a medicinal chemist, low yields and low predictability like this can result in product amounts insufficient for testing and consequently loss of SAR data.

Figure 1.

Figure 1

Open in a new tab

Summary of nickel/photoredox C(sp2)–C(sp3) cross-couplings from our method comparison study along with decarboxylative coupling examples.

The perceived limited substrate scopes of new methods, complexity of modern pharmaceutical intermediates, and technical difficulties users may experience when setting up photoredox reactions has suppressed adoption of transformations like the decarboxylative coupling. This disappoints us, as the structural diversity and availability of carboxylic acid building blocks makes this method attractive for analogue synthesis. Seeing this as an opportunity, we sought to use our expertise in HTE to develop tools to enable challenging decarboxylative couplings for discovery chemistry.

Chemical-coated glass beads, or ChemBeads, is a technology developed at AbbVie to enable high-throughput reaction screening and reliable dispensing of sub-milligram quantities of solid reagents.6 ChemBeads are prepared via resonant acoustic mixing (or other mixing techniques) to coat the glass bead hosts with solid reagent guests (Figure 2). Coating inert glass beads with solid reagents unifies the properties of the solids to the favorable properties of the glass beads and at the same time “bulks up” the solid reagents, which enables accurate dispensing of sub-milligram quantities into reaction vials and eliminates the need to prepare reagent stock solutions, saving time and material. This technology was first utilized to optimize Suzuki couplings, Buchwald–Hartwig aminations, and the nickel-catalyzed CEC.6,9,10 Concurrent with this work, we sought to extend this technology to nickel/photoredox dual catalysis optimization by preparing nickel catalyst- and photocatalyst-coated ChemBeads (Table 1). The nickel/photoredox ChemBeads shown in Table 1 were designed to deliver 5 mol% nickel catalyst and 2 mol% photocatalyst loadings for 4 μmol-scale reactions. We have observed that reaction profiles of these nickel/photoredox reactions can be sensitive to catalyst loadings, so it was a priority to stay close to the literature-reported loadings. As a result, dispensing accuracy was critical to realize the benefits of a photoredox ChemBeads HTE platform.

Figure 2.

Figure 2

Open in a new tab

Preparation of ChemBeads.

Table 1. Catalyst-Coated ChemBeads for Nickel/Photoredox Catalysis.

    Bead loading
    
ID Species (μmol/mg) (% w/w) Target bead dispense (mg) Amount of reagent delivered (mg)
PC1 [Ir(Fmppy)2(dtbbpy)]PF6 0.03 2.93 3 0.088
PC2 [Ir(dF(CH3)ppy)2(dtbbpy)]PF6 0.03 3.04 3 0.091
PC3 [Ir(dF(CF3)ppy)2(dtbbpy)]PF6 0.03 3.37 3 0.101
PC4 [Ir(dF(F)ppy)2(dtbbpy)]PF6 0.03 3.07 3 0.092
Ni1 NiCl2(dtbbpy) 0.075 2.98 3 0.09
Ni2 NiCl2(tmphen) 0.075 2.74 3 0.082
Ni3 NiCl2(dme) 0.075 1.65 3 0.049
(4-OMe)picolinimidamide·HCl 0.075 1.41 3 0.042
Ni4 Ni(TMHD)2 0.075 3.19 3 0.096
Open in a new tab

Assembling HTE screening plates with ChemBeads has been demonstrated using fully automated robotic dispensers6 and manually using calibrated scoops.27 The automated approach enables accurate dispensing of hundreds of solids in small scales, carried out by solid dispensing robots. The clear advantage of automated robotic dispensing is the minimal human intervention, but this approach suffers from high instrument costs, lab space requirements, and potential downtime caused by instrument malfunctions (which can cripple operations for days). Manual assembling of screenings plates, on the other hand, requires minimal capital investment, but the low-throughput nature means this approach is best suited for screening sets of limited size or use. Automated solid dispensing was the first strategy tested to prepare photoredox screening plates, but problems were quickly observed. Utilizing automation meant preparing larger quantities of screening plates at once. In situations where demand was tough to forecast, these plates could sit unused for months, and if screening plate changes were required, then unused plates would need to be discarded. Most importantly, automated dispensing accuracy issues were observed with our robot when trying to dispense ChemBead amounts ≤3 mg. The dispensed ChemBead mass was sometimes 2–4 times the desired quantities, due to the high flowability of the beads, and was difficult for our robot to control, which resulted in real changes to catalyst loadings. Reaction scale could not simply be increased to accommodate larger-than-expected ChemBead dispenses, as this would present an undue material burden on chemists. Lowering ChemBead reagent loadings to accommodate increased ChemBead dispense weights was also not an option because excess beads would require additional solvent to ensure submersion of all beads, therefore affecting reaction concentration. It also was not clear whether excess glass beads would impact irradiation in any way. Considering these limitations, a manual parallel bead drop dispenser was developed, which is capable of dispensing ChemBeads into 8×12, 96-position arrays with only the push of a sliding plate (Figure 3). There are three major components of the bead dispenser: the top, middle, and bottom plates. The top plate has 96 individual compartments as repositories for ChemBeads, the middle plate has 96 holes aligned with the bottom of the top plate (load position), and each hole has dimensions calibrated to host 3 mg of ChemBeads. When the beads are ready for dispensing, the user slides the middle plate to align the holes with the openings of the bottom plate (drop position) and dispense the ChemBeads to the 96 receiving vials below. Once the parallel dispensing is done, the user can slide the middle plate back to the “load” position to allow ChemBeads to fill the void holes for the next dispense sequence. The parallel bead dispenser combines the throughput of a robotic dispenser with the low costs, low maintenance, and low barrier to use of manual dispensing. The consistency of the dispensed weights has been evaluated to be within ±10% error. In our experience, this small discrepancy in the dispensed quantity has minimal to no effect on the HTE experiment outcome. If more than one reagent is needed per vial for the experiment, additional parallel bead dispensers can be used to add additional reagents. Alternatively, users can premix reagents thoroughly prior to coating on beads to dispense multiple reagents at once (e.g., nickel/ligand combination Ni3 in Table 1).6 The parallel bead dispenser has enabled rapid on-demand assembly of comprehensive HTE screening plates, eliminating the need to stockpile screening plates anticipating future use.

Figure 3.

Figure 3

Open in a new tab

Picture and schematics of the parallel bead dispenser.

Having developed and validated the ChemBead drop dispenser, we next set out to design a 96-well decarboxylative coupling reaction screen and to validate the use of ChemBeads in photoredox HTE (Figure 4). Reaction conditions incorporated into this 96-well plate were sourced from nickel/photoredox decarboxylative coupling literature,18,28−30 other C(sp2)–C(sp3) coupling literature (specifically pyridine-2-carboxamidine ligands from nickel-catalyzed couplings),31 and our own internal expertise.32 Taken together, a 96-reaction array was constructed comprised of four photocatalysts (PC1–PC4, Table 1), four nickel catalysts (Ni1–Ni4, Table 1), three bases (Cs2CO3, 1,8-diazabicyclo[5.4.0]undec-7-ene [DBU], 2-tert-butyl-1,1,3,3-tetramethylguanidine [BTMG]), and two solvents (N,N-dimethylacetamide [DMA] and dimethyl sulfoxide [DMSO]), displayed in Figure 4A. Two parallel ChemBead dispensers were assembled to prepare this decarboxylative coupling screen: one dispenser containing the photocatalysts and one dispenser containing the nickel catalysts (see Supporting Information). The ChemBeads were dispensed sequentially into 8 × 30 mm glass vials followed by stock solutions of the remaining reagents. After all reagents were added, the plate was sealed and then shaken and irradiated overnight using an orbital mixer and an array of blue LEDs. Upon completion, the reactions were diluted, internal standard was added, and then the reactions were analyzed by UPLC. This array was designed to serve as a first-pass HTE tool to expedite decarboxylative coupling optimization. On an as-needed basis, more comprehensive ChemBeads-enabled HTE could then be performed to further evaluate variables such as ligands, catalyst loadings, and/or additives. Ease of setup, enabled by the drop dispensers, allows modification of the screen on demand, to adjust specific reaction conditions or update the screen by incorporating new literature conditions.

Figure 4.

Figure 4

Open in a new tab

Decarboxylative coupling HTE. Plate design, workflow validation, and use examples. Histograms and heatmap visualizations generated using UVPdt/UVIS data after min-max normalization. Additional head-to-head protocol comparisons for 6 and 7 are available in the Supporting Information.

This screening set and workflow was tested on the challenging couplings partners from Figure 1B. In the original comparative study with 1, these coupling partners resulted in low yields, unconsumed starting material, and/or byproducts such as debromination and regioisomeric Minisci addition.16 Starting with the coupling of methoxyacetic acid (5) to 8 (ent-1), the parallel bead dispenser protocol was tested against a traditional manual liquid handling protocol, looking for consistency between the two protocols and, importantly, improved coupling conditions (Figure 4B). These side-by-side experiments were analyzed by product-to-internal standard ratio (UVPdt/UVIS), percent conversion, and product/byproduct ratios. In both cases the same improved coupling conditions were found: 2 mol% [Ir(dF(CF3)ppy)2(dtbbpy)]PF6, 5 mol% NiCl2(dme), 5 mol% 4-methoxypicolinimidamide hydrochloride, 1.5 equiv of BTMG, and 0.05 M DMA (condition C11). Under the original coupling conditions, unconsumed core 8 remained and the major product was debromination. HTE condition C11 resulted in complete consumption of 8, elevated product formation (UVPdt/UVIS = 1.13), and suppressed debromination (UVPdt/UVdesBr = 16.77). The results generated from both protocols were consistent: the best conditions appeared in rows C and G, the top conditions were conserved between both, and the histogram UVPdt/UVIS distributions for each plate were very similar. Excited by this initial result, we then tested the ChemBeads reaction screen on the couplings of l-pyroglutamic acid (6) and cyclopentanecarboxylic acid (7) with 8 (Figure 4C). Applying the same array of conditions to these couplings resulted in improved conditions for both transformations, as shown by the UVPdt/UVIS heat maps. For 6, while many conditions delivered improved results, condition C11 (above) resulted in a clean reaction profile and the highest UVPdt/UVIS ratio (1.47). For 7, coupling under standard conditions resulted in a low yield (ca. 3%) and large quantities of debromination and Minisci byproducts (UVPdt/UVdesBr = 0.13, UVPdt/UVMinisci = 0.54). HTE condition G10 (2 mol% [Ir(dF(CH3)ppy)2(dtbbpy)]PF6, 5 mol% NiCl2(dme), 5 mol% 4-methoxypicolinimidamide hydrochloride, 1.5 equiv of BTMG, and 0.05 M DMSO) resulted in a cleaner reaction profile, the highest UVPdt/UVIS ratio (0.93), and minimized byproducts (UVPdt/UVdesBr = 2.33, UVPdt/UVMinisci = 16.63). The HTE conditions for all reactions scaled well, as shown in Table 2. The top condition (C11) from the reaction screen with 5 resulted in a 68% yield of 9, in stark contrast to the original result, where no product was isolated (from 1).16 Analogue 10 likewise was produced in a 69% yield from 6 using the HTE conditions, compared to 3% yield using standard conditions, and analogue 11 was produced in a 31% yield from 7 using HTE conditions, compared to 3% under standard conditions. Notably, the top conditions for all three couplings employ ligand 4-methoxypicolinimidamide·HCl in place of 4,4′-di-tert-butyl-2,2′-bipyridine. These examples represent the first uses of a pyridine-2-carboxamidine ligand to improve decarboxylative couplings outside of on-DNA decarboxylative arylation.33

Table 2. Summary of HTE Decarboxylative Coupling Optimization.

Alkyl acid/product Original conditions Original result (screen) Top HTE conditions Top HTE result (screen) Original scale-up (from 1) HTE result scale-up (from 8)
5 / 9 Well A3: PC3, Ni1, Cs2CO3, DMA UVPdt/UVIS 0.24 Well C11: PC3, Ni3, BTMG, DMA UVPdt/UVIS 1.13 0 mg 23 mg
19% conv. 94% conv. 0% yield 68% yield
6 / 10 Well A3: PC3, Ni1, Cs2CO3, DMA UVPdt/UVIS 0.34 Well C11: PC3, Ni3, BTMG, DMA UVPdt/UVIS 1.47 1 mg 26 mg
24% conv. 88% conv. 3% yield 69% yield
7 / 11 Well E10: PC2, Ni1, BTMG, DMSO UVPdt/UVIS 0.29 Well G10: PC2, Ni3, BTMG, DMSO UVPdt/UVIS 0.93 1 mg 11 mg
18% conv. 67% conv. 3% yield 31% yield
Open in a new tab

Library synthesis facilitates thorough SAR exploration. In the case of α-oxy acid 5 and α-amino acid 6, the top HTE condition was the same (C11, Table 2), while the top HTE condition for α-alkyl acid 7 was unique (G10, Table 2). Encouraged by these results, we shifted our focus to testing how library-amenable these novel conditions were (compared to the standard conditions used during our comparative study, Figure 1) and to exploring the substrate scope (using structurally and electronically diverse carboxylic acids). A standard library synthesis workflow comprised of parallel synthesis, reverse-phase mass-directed preparative HPLC purification, and automated liquid handling (for transfers) was used to determine if these conditions were library-amenable. To assess the scope and generality of these conditions, a series of 74 alkyl carboxylic acids was chosen using the same logic which was applied to alkyl group selection in our original comparative study.16 Primary (22), secondary (43), and tertiary (9) coupling partners were selected which scan a wide range of alkyl space, and alkyl groups of interest to internal projects were prioritized for inclusion, such as all-carbon alkyl groups (cyclic and acyclic), alkyl ethers, and Boc-protected alkyl amines. Furthermore, this coupling library contained 26 α-alkyl, 20 α-amino, 10 α-aryl, and 18 α-oxy carboxylic acids.

Combining the HTE conditions with a parallel synthesis workflow enabled the synthesis of a greater range of decarboxylative coupling products, at higher yields, compared to the initial standard reaction conditions. A summary of this comparative study is shown in Figure 5. For the coupling of α-alkyl acids to 8, product was detected for 19/26 reactions using HTE conditions, and the average yield was 20%, compared to just 3% under the original conditions (Figure 5A). Primary α-alkyl acids were competent coupling partners under the HTE conditions, albeit in modest yields, as shown by 12 (13% under HTE conditions versus 1% under original conditions) and 13 (17% versus 0%). The HTE conditions resulted in significant yield improvements for secondary α-alkyl acids, as shown by examples 14 (50% versus 5%), 15 (55% versus 8%), 16 (23% versus 4%), and 17 (54% versus 8%). Notable exceptions to this trend were cyclopropyl and 3-oxetane groups, which failed to couple under both original and HTE conditions. All-carbon cycloalkyl groups such as cyclobutyl, cyclopentyl, and cyclohexyl were installed in high library yields. When incorporating an inductive withdrawing group into the ring, such as an ether or N-Boc group, it was observed that yields increased as ring size increased (3-oxetane to 3-tetrahydrofuran to 4-tetrahydropyran: 0% to 23% to 54% yield) and as proximity of the withdrawing group to the carboxylic acid decreased (3-tetrahydropyran, 33%, to 4-tetrahydropyran, 54%). For some reactions, the presence of a withdrawing group appeared to have a deleterious effect. Comparing primary acids, yields for butyronitrile and methyl 3-propionate were low relative to n-butyl (12). Comparing secondary acids, yields for N-Boc-3-azetidine and 3-oxetane were very low relative to cyclobutyl. Finally, tertiary alkyl groups failed to couple under the original conditions and the HTE conditions. This study showed that these HTE conditions were library-amenable for 8 and increased yields overall, despite being limited to primary and secondary alkyl acids. Acids possessing α-amino/oxy/aryl functionality are common substrates for this transformation.18,28 These acids generate stabilized radicals upon decarboxylation, but despite this results were initially underwhelming using 8 and standard literature conditions. The top condition from the HTE screens to prepare 9 and 10 resulted in significant improvements for the couplings of 8 to α-amino, α-aryl, and α-oxy acids, as shown in the Figure 5B library comparison. Focusing on the 20-member α-amino acid series, the HTE conditions resulted in formation of 18/20 products compared to just 10/20 products under the original conditions. These conditions therefore enabled synthesis of analogues which were previously missed or extremely low-yielding. Primary N-Boc and N-acyl α-amino acids were competent coupling partners, and the HTE conditions resulted in improved yields (46% versus 10% on average). Secondary N-Boc/N-acyl α-amino acids performed well, too, while tertiary α-amino acids failed. Examples include N-alkyl indazole coupling 18 (56% versus 1%), Boc-Val-OH coupling 19 (31% versus 0%), and 2-azetidine coupling 20 (66% versus 4%). Alpha-aryl carboxylic acids coupled poorly under the original conditions, where only 1/10 reactions produced isolable product (4-methoxyphenylacetic acid, 3 mg, 8%, Figure 5B). Under the HTE conditions this rose to 6/10, and electron-deficient (4-trifluoromethylphenyl, 5%), electron-neutral (phenyl, 10%), and electron-rich (21, 4-methoxyphenyl, 34%) arenes were tolerated. Some α-heteroaryl acids coupled under the HTE conditions as well, such as 5-pyrimidineacetic acid (24% yield). Ultimately only primary α-aryl acids coupled under these HTE conditions; secondary and tertiary α-aryl acids remained limitations. Finally, for α-oxy acids, 10/18 analogues were isolable under the HTE conditions, compared to 8/18 under the original conditions. Alpha-hydroxy carboxylic acids were unsuccessful under either set of conditions, as were oxetane-2-carboxylic acid and tertiary α-oxy acids. However, in general, yields for primary and secondary α-alkoxy couplings were improved under the HTE conditions. Two examples include acyclic secondary coupling with 22 (67% versus 6%) and cyclic secondary coupling with 23 (18% versus 0%). A surprising exception to this was 2,3-dimethoxypropanoic acid (Figure 5B), which coupled under the original conditions (5%) but was unsuccessful under the HTE conditions (0%). It is unclear why this is the case, as other α-branched acids worked well under HTE conditions (22), as did acids possessing remote withdrawing groups (23). This and other low-yielding examples (e.g., tetrahydropyran-2-carboxylic acid) represent areas of ongoing investigation.

Figure 5.

Figure 5

Open in a new tab

Comparative study of initial coupling conditions versus HTE coupling conditions using parallel library synthesis. Boc-Pro-OH was present in duplicate. Acids are presented in the following order: primary then secondary then tertiary acids.

Overall, HTE rendered the decarboxylative coupling library-amenable for pyrimidyl bromide 8, as shown in Figure 6. A histogram comparison (Figure 6A) shows that most reactions run under the original conditions resulted in 0–10% yields (65/75). The yield distribution for the libraries using HTE conditions was significantly improved, spanning 0%–84%, and importantly, the number of reactions occupying the 0–10% yield bin was reduced to 32/75. In addition to improving yields, the HTE conditions enabled failed reactions. The failed reaction count decreased from 30 to 21, and very low-yielding reactions (<5% yield, approximately 2 mg) decreased from 20 to 4. Breaking down the results by carboxylic acid functionality highlights where improvements were realized and where opportunities still exist (Figure 6B). Looking at carboxylic acid steric substitution shows that average yields for primary and secondary carboxylic acids improved significantly by leveraging HTE, to 24% and 29% respectively from 4% and 5% under the original coupling conditions. Tertiary carboxylic acids remained a limitation, with an average yield <1%, and remain an opportunity for innovation (along with secondary α-aryl acids). Similar improvements were observed when comparing the results focusing on α-position functionality. The average yield for an α-alkyl acid rose from 3% to 20% under HTE conditions. Under HTE conditions, the average yield with an α-amino acid increased to 37% (from 7%), the average α-aryl acid yield increased to 9% (from 1%), and the average α-oxy acid yield increased to 24% (from 4%). Taken together, this data indicates that HTE identified library-amenable coupling conditions for 8 and has elevated the decarboxylative coupling to parity with the CEC and BF3K couplings, when considering library success and average yields (cf. Figure 1A). We are working to further define the substrate scope of these coupling conditions and to better understand the differentiation/benefit that 4-methoxypicolinimidamide provides. This ChemBeads-enabled photoredox HTE platform has since been extended internally to other nickel/photoredox couplings and to tailored screens with more focused goals, such as comprehensive scans of catalyst/ligand space via larger nickel catalyst × photocatalyst arrays.

Figure 6.

Figure 6

Open in a new tab

Comparative study breakdown and summary.

In summary, we developed an on-demand platform to rapidly assemble and conduct photoredox high-throughput experimentation using ChemBeads and a parallel ChemBead dispenser as enabling tools. These tools have made possible the rapid testing of multiple variables in parallel, such as catalysts, solvents, and bases, to identify improved (previously unknown) reaction conditions for decarboxylative couplings. The coupling conditions identified in this study were more general than the original conditions and resulted in improved library success, both measured by average yields and number of analogues prepared. We nominate the reaction conditions disclosed here to the chemistry community for challenging photoredox decarboxylative couplings, as they have been effective internally for other projects, as well. Through building internal expertise into photoredox C(sp2)–C(sp3) coupling methods and offering reliable optimization strategies for challenging reactions, we have observed an increase in adoption of photoredox C(sp2)–C(sp3) couplings for internal analogue and library synthesis.

Acknowledgments

All authors are employees of AbbVie. The design, study conduct, and financial support for this research were provided by AbbVie. AbbVie participated in the interpretation of data, review, and approval of the publication. We thank the following AbbVie scientists for their technical support of this project: Rick Yarbrough, Leena Bhatt, Anita McGreal, Frank Wagenaar, and Jay Bhanot. We thank Andrew Radosevich (AbbVie) and James Sawicki (AbbVie) for helpful discussions.

Glossary

Abbreviations

UPLC

ultraperformance liquid chromatography

HPLC

high-performance liquid chromatography

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsmedchemlett.2c00538.

  • General information, experimental procedures, and characterization data (PDF)

  • Weight data for accuracy determination of parallel ChemBead dispenser (XLSX)

  • Full HTE data table used to prepare Figure 4 (XLSX)

  • Full parallel library synthesis data table used to prepare Figure 5 (XLSX)

The authors declare no competing financial interest.

Supplementary Material

ml2c00538_si_001.pdf (10MB, pdf)
ml2c00538_si_002.xlsx (26KB, xlsx)
ml2c00538_si_003.xlsx (356.3KB, xlsx)
ml2c00538_si_004.xlsx (217.4KB, xlsx)

References

  1. Shevlin M. Practical High-Throughput Experimentation for Chemists. ACS Med. Chem. Lett. 2017, 8 (6), 601–607. 10.1021/acsmedchemlett.7b00165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Krska 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 (12), 2976–2985. 10.1021/acs.accounts.7b00428. [DOI] [PubMed] [Google Scholar]
  3. Cernak T.; Gesmundo N. J.; Dykstra K.; Yu Y.; Wu Z.; Shi Z.-C.; Vachal P.; Sperbeck D.; He S.; Murphy B. A.; Sonatore L.; Williams S.; Madeira M.; Verras A.; Reiter M.; Lee C. H.; Cuff J.; Sherer E. C.; Kuethe J.; Goble S.; Perrotto N.; Pinto S.; Shen D.-M.; Nargund R.; Balkovec J.; DeVita R. J.; Dreher S. D. 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 (9), 3594–3605. 10.1021/acs.jmedchem.6b01543. [DOI] [PubMed] [Google Scholar]
  4. Santanilla A. B.; Regalado E. L.; Pereira T.; Shevlin M.; Bateman K.; Campeau L.-C.; Schneeweis J.; Berritt S.; Shi Z.-C.; Nantermet P.; Liu Y.; Helmy R.; Welch C. J.; Vachal P.; Davies I. W.; Cernak T.; Dreher S. D. Nanomole-Scale High-Throughput Chemistry for the Synthesis of Complex Molecules. Science 2015, 347 (6217), 49–53. 10.1126/science.1259203. [DOI] [PubMed] [Google Scholar]
  5. Perera 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 (6374), 429–434. 10.1126/science.aap9112. [DOI] [PubMed] [Google Scholar]
  6. Tu 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 (24), 7987–7991. 10.1002/anie.201900536. [DOI] [PubMed] [Google Scholar]
  7. DiRocco 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 (19), 4802–4806. 10.1002/anie.201402023. [DOI] [PubMed] [Google Scholar]
  8. Corcoran E. B.; Pirnot M. T.; Lin S.; Dreher S. D.; DiRocco D. A.; Davies I. W.; Buchwald S. L.; MacMillan D. W. C. Aryl Amination Using Ligand-Free Ni(II) Salts and Photoredox Catalysis. Science 2016, 353 (6296), 279–283. 10.1126/science.aag0209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Aguirre A. L.; Loud N. L.; Johnson K. A.; Weix D. J.; Wang Y. ChemBead Enabled High-Throughput Cross-Electrophile Coupling Reveals a New Complementary Ligand. Chem. - Eur. J. 2021, 27 (51), 12981–12986. 10.1002/chem.202102347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Rago A. J.; Vasilopoulos A.; Dombrowski A. W.; Wang Y. Di(2-Picolyl)Amines as Modular and Robust Ligands for Nickel-Catalyzed C(Sp2)-C(Sp3) Cross-Electrophile Coupling. Org. Lett. 2022, 24 (46), 8487–8492. 10.1021/acs.orglett.2c03346. [DOI] [PubMed] [Google Scholar]
  11. Li P.; Terrett J. A.; Zbieg J. R. Visible-Light Photocatalysis as an Enabling Technology for Drug Discovery: A Paradigm Shift for Chemical Reactivity. ACS Med. Chem. Lett. 2020, 11 (11), 2120–2130. 10.1021/acsmedchemlett.0c00436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Candish L.; Collins K. D.; Cook G. C.; Douglas J. J.; Gómez-Suárez A.; Jolit A.; Keess S. Photocatalysis in the Life Science Industry. Chem. Rev. 2022, 122 (2), 2907–2980. 10.1021/acs.chemrev.1c00416. [DOI] [PubMed] [Google Scholar]
  13. Palaychuk N.; DeLano T. J.; Boyd M. J.; Green J.; Bandarage U. K. Synthesis of Cycloalkyl Substituted 7-Azaindoles via Photoredox Nickel Dual Catalytic Cross-Coupling in Batch and Continuous Flow. Org. Lett. 2016, 18 (23), 6180–6183. 10.1021/acs.orglett.6b03223. [DOI] [PubMed] [Google Scholar]
  14. Raynor K. D.; May G. D.; Bandarage U. K.; Boyd M. J. Generation of Diversity Sets with High Sp3 Fraction Using the Photoredox Coupling of Organotrifluoroborates and Organosilicates with Heteroaryl/Aryl Bromides in Continuous Flow. J. Org. Chem. 2018, 83 (3), 1551–1557. 10.1021/acs.joc.7b02680. [DOI] [PubMed] [Google Scholar]
  15. Zhang R.; Li G.; Wismer M.; Vachal P.; Colletti S. L.; Shi Z.-C. Profiling and Application of Photoredox C(Sp3)-C(Sp2) Cross-Coupling in Medicinal Chemistry. ACS Med. Chem. Lett. 2018, 9 (7), 773–777. 10.1021/acsmedchemlett.8b00183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Dombrowski A. W.; Gesmundo N. J.; Aguirre A. L.; Sarris K. A.; Young J. M.; Bogdan A. R.; Martin M. C.; Gedeon S.; Wang Y. Expanding the Medicinal Chemist Toolbox: Comparing Seven C(Sp2)-C(Sp3) Cross-Coupling Methods by Library Synthesis. ACS Med. Chem. Lett. 2020, 11 (4), 597–604. 10.1021/acsmedchemlett.0c00093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Grimm I.; Hauer S. T.; Schulte T.; Wycich G.; Collins K. D.; Lovis K.; Candish L. Upscaling Photoredox Cross-Coupling Reactions in Batch Using Immersion-Well Reactors. Org. Process Res. Dev. 2020, 24 (6), 1185–1193. 10.1021/acs.oprd.0c00070. [DOI] [Google Scholar]
  18. Zuo Z.; Ahneman D. T.; Chu L.; Terrett J. A.; Doyle A. G.; MacMillan D. W. C. Merging Photoredox with Nickel Catalysis: Coupling of α-Carboxyl Sp3-Carbons with Aryl Halides. Science 2014, 345 (6195), 437–440. 10.1126/science.1255525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Tellis J. C.; Primer D. N.; Molander G. A. Single-Electron Transmetalation in Organoboron Cross-Coupling by Photoredox/Nickel Dual Catalysis. Science 2014, 345 (6195), 433–436. 10.1126/science.1253647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Zhang P.; Le C.; MacMillan D. W. C. Silyl Radical Activation of Alkyl Halides in Metallaphotoredox Catalysis: A Unique Pathway for Cross-Electrophile Coupling. J. Am. Chem. Soc. 2016, 138 (26), 8084–8087. 10.1021/jacs.6b04818. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gutiérrez-Bonet A.; Tellis J. C.; Matsui J. K.; Vara B. A.; Molander G. A. 1,4-Dihydropyridines as Alkyl Radical Precursors: Introducing the Aldehyde Feedstock to Nickel/Photoredox Dual Catalysis. ACS Catal. 2016, 6 (12), 8004–8008. 10.1021/acscatal.6b02786. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Jouffroy M.; Primer D. N.; Molander G. A. Base-Free Photoredox/Nickel Dual-Catalytic Cross-Coupling of Ammonium Alkylsilicates. J. Am. Chem. Soc. 2016, 138 (2), 475–478. 10.1021/jacs.5b10963. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Knauber T.; Chandrasekaran R.; Tucker J. W.; Chen J. M.; Reese M.; Rankic D. A.; Sach N.; Helal C. Ru/Ni Dual Catalytic Desulfinative Photoredox Csp2-Csp3 Cross-Coupling of Alkyl Sulfinate Salts and Aryl Halides. Org. Lett. 2017, 19 (24), 6566–6569. 10.1021/acs.orglett.7b03280. [DOI] [PubMed] [Google Scholar]
  24. Dong Z.; MacMillan D. W. C. Metallaphotoredox-Enabled Deoxygenative Arylation of Alcohols. Nature 2021, 598 (7881), 451–456. 10.1038/s41586-021-03920-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lovering F.; Bikker J.; Humblet C. Escape from Flatland: Increasing Saturation as an Approach to Improving Clinical Success. J. Med. Chem. 2009, 52 (21), 6752–6756. 10.1021/jm901241e. [DOI] [PubMed] [Google Scholar]
  26. Ishikawa M.; Hashimoto Y. Improvement in Aqueous Solubility in Small Molecule Drug Discovery Programs by Disruption of Molecular Planarity and Symmetry. J. Med. Chem. 2011, 54 (6), 1539–1554. 10.1021/jm101356p. [DOI] [PubMed] [Google Scholar]
  27. Martin M. 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 (9), 1900–1907. 10.1021/acs.oprd.9b00213. [DOI] [Google Scholar]
  28. Noble A.; McCarver S. J.; MacMillan D. W. C. Merging Photoredox and Nickel Catalysis: Decarboxylative Cross-Coupling of Carboxylic Acids with Vinyl Halides. J. Am. Chem. Soc. 2015, 137 (2), 624–627. 10.1021/ja511913h. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Hsieh 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 (4), 542–550. 10.1021/acs.oprd.8b00018. [DOI] [Google Scholar]
  30. Le C.; Wismer M. K.; Shi Z.-C.; Zhang R.; Conway D. V.; Li G.; Vachal P.; Davies I. W.; MacMillan D. W. C. A General Small-Scale Reactor To Enable Standardization and Acceleration of Photocatalytic Reactions. ACS Cent. Sci. 2017, 3 (6), 647–653. 10.1021/acscentsci.7b00159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Hansen E. C.; Pedro D. J.; Wotal A. C.; Gower N. J.; Nelson J. D.; Caron S.; Weix D. J. New Ligands for Nickel Catalysis from Diverse Pharmaceutical Heterocycle Libraries. Nat. Chem. 2016, 8 (12), 1126–1130. 10.1038/nchem.2587. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Phthalimide was recently described as a beneficial additive for photoredox decarboxylative couplings:; Prieto Kullmer C. N.; Kautzky J. A.; Krska S. W.; Nowak T.; Dreher S. D.; MacMillan D. W. C. Accelerating reaction generality and mechanistic insight through additive mapping. Science 2022, 376, 532–539. 10.1126/science.abn1885. [DOI] [PMC free article] [PubMed] [Google Scholar]; Photoredox ChemBeads and the screening plate described here were designed and validated before this disclosure. Conditions containing phthalimide as an additive have since been validated and used internally.
  33. Kölmel D. K.; Meng J.; Tsai M.-H.; Que J.; Loach R. P.; Knauber T.; Wan J.; Flanagan M. E. On-DNA Decarboxylative Arylation: Merging Photoredox with Nickel Catalysis in Water. ACS Comb. Sci. 2019, 21, 588. 10.1021/acscombsci.9b00076. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ml2c00538_si_001.pdf (10MB, pdf)
ml2c00538_si_002.xlsx (26KB, xlsx)
ml2c00538_si_003.xlsx (356.3KB, xlsx)
ml2c00538_si_004.xlsx (217.4KB, xlsx)

Articles from ACS Medicinal Chemistry Letters are provided here courtesy of American Chemical Society

RESOURCES