Thread by @Daniel_J_Blair

We are excited to report in @Nature a new class of modular molecular building blocks, TIDA boronates, which pull automated lego-like chemical synthesis into the third dimension 1/ #ChemTwitter #LEGO
nature.com/articles/s4158… @ChemistryUIUC @LASillinois @NewsAtIllinois @UofIllinois

Previous work showed MIDA as a powerful ligand for reversibly attenuating boron reactivity within Suzuki-Miyaura cross-coupling reactions - allowing iterative assembly of B-protected haloboronic acid building blocks. 2/

A 1st gen automated synthesis platform based on MIDA boronates reported in @ScienceMagazine in 2015 enabled small organic molecules to be prepared using a single automated platform. But it came with a key limitation... 3/
science.org/doi/10.1126/sc…

..don't get them wet! Prior studies with @LloydJonesLab revealed that frustrated Lewis pair-like activation of water and hydroxide mediated ester hydrolysis as two potential pathways for MIDA boronate cleavage. 4/
nature.com/articles/nchem…

Sensitivity to water as well as an incompatibility with nucleophilic/basic reagents limited extension of this automated synthesis platform into saturated and potentially stereogenic assembly chemistries (i.e. Csp3-C). So we needed a better boronate. 5/

We began with a hunt for a hydrolysis-resistant boronate, large nitrogen substituents (blue) were destabilizing... and modifications to the backbone (green) were more promising... but not sufficiently stable to enable iterative synthesis. 6/

Substitution of the last remaining spot on the backbone provided tetramethyl N-methyliminodiacetic acid (TIDA) which was stable under our hydrolysis conditions. 7/

Head to head with its MIDA boronate counterpart, TIDA was completely retained during a Csp3-bond forming aqueous basic Suzuki-Miyaura cross-coupling, whereas MIDA was completely cleaved! 8/

Building on these findings @SriyaChitti found that TIDA boronates remarkably tolerate RMgX and RLi. This allowed us to develop bifunctional sulfoxide-TIDA boronates to access iterable 1,2-metallate reactions. MIDA again was incompatible. 9/

Despite these exciting results TIDA just didn't make sense, because from what we read in the literature methylation of iminodiacetic acid boronates would be expected to activate frustrated Lewis-pair behavior rather than deactivate it?! 10/

Crystallographically determined electron density allowed us to "see" what was going on. The methyl groups on TIDA caused twist around the water-activating N-B bond and introduced stabilizing hyperconjugation along the N-B bond, deactivating water mediated cleavage. 11/

What about RLi and RMgX? Well, the backbone methyls played a second important role, shielding the carbonyl carbons from attack. The observed arrangement of methyl groups on TIDA closely mirror the steric shielding within butyl-lithium resistant Beak-type benzoates. 12/

Although more stable, we found that with a little push (read: heat) we could convert TIDA boronates into cross-coupling and 1,2-metallate active boronic acids, trilfluoroborate salts, and boronic esters. 13/

A challenge in automating chemical synthesis is generalizing purification, however TIDA (like MIDA) can serve as a selective tag for generalized purification in catch-and-release chromatography. This broadly enabled automated purification of TIDA boronates. 14/

Bringing all these features together, with REVOLUTION Medicines, we were able to automate the assembly of TIDA boronate building blocks with new synthesis machines. Captured in all their glorious greenness by @FredZwicky 15/

Using these machines we automated the assembly of Csp3 boronate building blocks and halo-aryl TIDA boronates using aqueous/protic Pd-catalyzed Suzuki-Miyaura coupling reactions. Such products were completely inaccessible using MIDA boronates which were universally cleaved. 16/

In a similar fashion we were able to execute a series of 1,2-metallate reactions in a manual and automated fashion to form Csp3-Csp3 linkages. 17/

Leveraging TIDAs elevated stability and orthogonality vs diaminonaphthalene (dan) boronates we were able to complete a Lego-like synthesis of Ieodomycin C. 18/

In a final example we employed our sulfoxide-TIDA boronates to perform iterative 1,2-metallate reactions in an uninterrupted fashion using automated synthesis. Here we could now form two Csp3-Csp3 linkages with complete stereocontrol to complete a short synthesis of sch725674 19/

In closing, we are excited for the new molecular functions TIDA boronates will enable us to explore through automated lego-like synthesis. 20/

A special mention to @SriyaChitti who spearheaded development of sulfoxide-TIDA boronates and our adventures into 1,2-metallate chemistry. A poke to twitterless Marty Burke for transforming the way I think about science. 21/

HUGE thanks to @DamonRunyon for supporting this work, and @illini4000 for support/visits over the years, with other support for the team from @HLuceFdn @ACSorganic @FWF_at 22/

Final thanks to @VarinderAggar11 the creator of the coolest homologation in town, and who has taught me so much! 23/