Second Generation Catalyst System for Ag-mediated Liebeskind-Srogl Coupling of Tetrazine Thioethers with Arylboronic Acids

Abstract

An improved catalyst system, PdI₂/DiPPF, is described for the silver-mediated Liebeskind-Srogl coupling of thioalkyltetrazines with arylboronic and heteroarylboronic acids. The methodology advances the synthesis of functionalized tetrazines, offering greater efficiency, broader compatibility, and practical utility for accessing tool molecules for bioorthogonal chemistry applications.

Graphical Abstract

INTRODUCTION

The rapid kinetics of bioorthogonal reactions between tetrazines and strained dienophiles have enabled numerous applications in biology, medicine, and materials science. There is also a growing interest in developing improved methods for synthesizing tetrazine derivatives.^1–4 Traditionally, tetrazines have been synthesized by condensing hydrazine with nitriles or alkyl imidates, followed by oxidation.^{5, 6} Catalytic versions of these methods, using Zn(OTf)₂ or Ni(OTf)₂ with anhydrous hydrazine, have expanded applicability to the synthesis of unsymmetrical tetrazines.⁷ Thiol catalysis has enabled the synthesis of tetrazines from nitriles using hydrazine-hydrate,⁸ while the sulfur-catalyzed reaction of nitriles with hydrazine-hydrate and dichloromethane has been used to produce 3-aryltetrazines.⁹ Although these methods for synthesizing tetrazines have significantly advanced the field of bioorthogonal chemistry, there remains an ongoing demand for safe alternatives that offer better compatibility with a wider range of functional groups.

Metal-catalyzed cross-coupling reactions have emerged as valuable alternatives for synthesizing tetrazine derivatives. These transformations can be challenging due to the sensitivity of tetrazine substrates to basic conditions, which has caused their cross-coupling chemistry to lag behind other aromatic heterocycles. Early work in this area involved using tetrazine substrates bearing reactive groups tethered to the tetrazine core, including Stille and Sonogashira reactions of haloaryltetrazine analogs.^10–13 In 2014, Devaraj employed a sequential elimination-Heck reaction starting from 3-methoxyethyl-6-methyltetrazine to construct tetrazine conjugates.¹⁴ In 2020, Audebert reported the first optimized Buchwald-Hartwig amination using 3-(4-bromophenyl)-1,2,4,5-tetrazine.¹⁵ Hierso developed a Pd-catalyzed halogenation method for diphenyl tetrazines, allowing for the stepwise installation of up to four halogens at the ortho positions of the pendant rings.^16–18 Subsequent Suzuki couplings enabled the incorporation of ortho-substituted aromatic groups.¹⁷ In 2020, Xu introduced an iridium-catalyzed C–H amidation of methylphenyl tetrazine, enabling selective installation of sulfonamide groups at the ortho position of the phenyl ring.¹⁹

The earliest cross-coupling to the tetrazine core was reported in 2003 by Kotschy, who described Sonogashira and Negishi couplings between alkynes and 3-amino-6-chlorotetrazines.²⁰ In 2007, Guillaumet demonstrated a Liebeskind-Srogl coupling using 3-morpholine-6-thiomethyltetrazines.²¹ Avarvari later reported the first Stille coupling with dichlorotetrazine in 2013.²² In 2017, Lindsley introduced the first Suzuki coupling using 3-amino-6-chlorotetrazines.²³ Many of these initial studies used 3-aminotetrazine derivatives due to their relative stability. However, the electron-donating nature of the 3-amino group also reduces the activity of the tetrazine ring toward cross-coupling, prompting efforts to develop methodologies compatible with 3-carbo-substituted tetrazines. A significant advancement came in 2017 when Wombacher and colleagues demonstrated that 3-bromo-6-methyl-s-tetrazine serves as an effective precursor for Stille coupling with tributylstannyl-substituted fluorophores, providing access to 3-carbo-substituted tetrazines.²⁴ Using this strategy, tetrazine-fluorophore conjugates such as fluorescein and Oregon Green derivatives were prepared in moderate yields.

In 2019, our group introduced the first silver-mediated Liebeskind–Srogl coupling using 3-(([1,1'-biphenyl]-4-ylmethyl)thio)-6-methyl-1,2,4,5-tetrazine (b-Tz) as the tetrazine donor. This reagent is commercially available (Sigma-Aldrich 919950) or can be synthesized on large scale in just two steps.²⁵ Instead of the typical use of CuTc as an additive in Liebeskind–Srogl couplings, this method employed Ag (I) to activate the reaction. The reaction conditions enabled the efficient installation of methyl tetrazine on various complex molecules, including amino acids, estrone, BODIPY dyes, and a Tz-probe for monoacylglycerol lipase (MAGL) inhibition. Liebeskind-Srogl couplings followed by Pd-catalyzed reduction were also used to produce unsymmetrical tetrazines from thiomethyltetrazine precursors derived from carboxylic ester precursors.²⁶ In 2021, Gademann reported a Pd-catalyzed cross–coupling using 3-bromotetrazine to prepare 3-aryltetrazines.²⁷ More recently, Liang and Wu described cross-coupling reactions of thiomethyltetrazines with 1-trimethylsilyl-2-tributylstannylacetylene, which could be used in subsequent CuAAC chemistry with alkylazides.²⁸

While the scope of our previously described silver-mediated Liebeskind-Srogl coupling was applicable across a range of electronically rich and neutral arylboronic and heteroarylboronic acids, a limitation was that electron-deficient substrates coupled in relatively low yields. Here, we describe an improved catalyst system using PdI₂(DiPPF) that functions for a broader range of substrates, including electronically deficient arylboronic and heteroarylboronic acids. The optimized conditions also require lower palladium and silver loadings than our original conditions, further improving the cost-effectiveness of the reaction conditions.

RESULTS AND DISCUSSION

Previously described conditions for thioalkyltetrazine cross-coupling were optimized for b-Tz with phenyboronic acid, and it was found that the bidentate ligand 1,1'-bis(diphenylphosphino) ferrocene (dppf, J) gave the best results. In those studies, we also observed hits with the Xantphos (G) ligand system. To develop a catalyst system with better compatibility toward electron-deficient arylboronic acids, we screened a library of ferrocene and Xantphos ligands with palladium halides (PdCl₂, PdBr₂, PdI₂) (Scheme 1). 4-(Trifluoromethyl) phenylboronic acid (1.9 equiv) was chosen as an electron-deficient substrate for screening, as it gave only a moderate yield (61%) under 1^st-generation conditions using PdCl₂(dppf) and excess boronic acid. All reactions were carried out with 15 mol% palladium halide, 16 mol% ligand, and 2.5 eq Ag₂O in dimethylformamide. Of the 17 ligands tested, 5 of them performed better than dppf when PdCl₂ was used. If PdI₂ was used instead, the yields of tetrazine (2a) increased dramatically, with 7 ligands performing better than dppf. The best combination was PdI₂ and 1,1′-bis(diisopropylphosphino)ferrocene (DiPPF), ligand Q, which gave a 93% yield of 1a.

Scheme 1.

Optimization of ligand and Pd-source for the reaction of b-Tz with an electron deficient boronic acid.

As a starting point for further optimization, b-Tz was coupled with 4-(trifluoromethyl)phenylboronic acid (1.9 equiv) using Ag₂O (2.5 equiv) in DMF. Prestirring PdI₂ (10 mol%) with DiPPF (11 mol%) for 1 h gave 1a in 76% yield (Scheme 2, entry 1), an equivalent result to using a PdI₂(DiPPF) precatalyst (10 mol%) (Scheme 2, entry 2); a slight yield decrease was observed when the prestirring step was skipped. As increasing the PdI₂ loading to 15 mol% had a deleterious impact (Scheme 2, entry 3), screening efforts were carried out with 10 mol% Pd. A solvent screen identified DMF as the best solvent (Scheme 2, Entries 4–7). Ag₂O and Ag₂CO₃ were the only silver(I) sources found to mediate the previously described PdCl₂(dppf)- catalyzed Liebeskind-Srogl coupling of b-Tz, with Ag₂O yielding the highest yields for most substrates. Decreasing the silver loading to 1.5 or 1.1 equiv did not affect the yield (Scheme 2, entries 8–10). A PdI₂/DiPPF catalyst system with Ag₂CO₃ (1.1 equiv) gave the product in 90% yield—the highest yield observed (Scheme 2, entry 10). Reducing the silver loading to substoichiometric levels had a deleterious impact (Scheme 2, entries 11–12). Using Pd(OAc)₂ or Pd₂dba₃ in place of PdI₂ was also detrimental, giving product in 4% and 27% yield, respectively (Scheme 2, entries 13–14).

Scheme 2.

Optimization studies. Changes from Entry 1 are indicated in red.

Next, we screened various catalyst loadings (0.5 – 15 mol%) using Ag₂CO₃-mediated conditions. The 4-(trifluoromethyl)phenylboronic acid loading was lowered to 1.1 equiv for these experiments. A Pd-loading of 10 mol% was optimal, providing a 69% yield. Lower yields were observed when more or less catalyst was used (Scheme 3, entries 1–7). The lower yield at 15 mol% PdI₂ loading may be due to the sacrifice of 2 equivalents of ArB(OH)₂ for every equiv. of PdI₂ in the catalyst activation step. The ratio of product to b-Tz conversion drops significantly at catalyst loadings below 5 mol%, likely due to thermal decomposition of b-Tz in reaction solutions where overall catalytic turnover was slower. Finally, screening temperatures confirmed that 60 °C is the optimal temperature. The reaction proceeded slowly at room temperature, resulting in only 8% yield. Lower yields were also recorded for reactions at 45 °C or 75 °C (Scheme 3, entries 8–10).

Scheme 3.

Optimization of Pd-loading, ligand loading, and reaction temperature.

^a Conversion not determined

We investigated the substrate scope with these new conditions for cross-coupling thioalkyltetrazines and arylboronic acids. Arylboronic acids were tested and compared to yields obtained using previously described conditions. With PdI₂/DiPPF, a range of electron-deficient arylboronic acids were coupled in improved yields compared to the 1^st generation PdCl₂(dppf) catalyst. Phenylboronic acids (1.6 equiv) functionalized by trifluoromethyl (1a), nitro (1b), cyano (1c), 4-methylsulfonyl (1d), or ester (1e) functional groups coupled with b-Tz in poor yields under 1^st generation conditions (29–65%). These yields improved significantly with the PdI₂/DiPPF catalyst system, giving products 1a-e in 73–90% yields. With the 2^nd generation catalyst system, good yields of 1a-e were also obtained when the ArB(OH)₂ loading was reduced to 1.3 equiv. 2^nd Generation conditions also improved results, albeit less strikingly, for the synthesis of product 1f, which was produced in 63% and 48% yields when b-Tz was coupled with 1.6 and 1.3 equiv of 4-formylboronic acid, respectively.

The 2^nd generation conditions for coupling b-Tz and electron-neutral arylboronic acids also proceed with superior (1 g-h) or similar (1i) yields. A caveat, however, is that 15 mol% PdI₂(DiPPF) and 2.1 equiv Ag₂CO₃ were necessary to achieve the best yields. Boc-protected aniline 1m was prepared in 82%. Heterocyclic arylboronic acids were also tolerated under the 2^nd generation coupling conditions. 3-(4-Fluoropyridin-3-yl)-tetrazine (1j) could be synthesized in 80% and 65% yield from 1.6 and 1.3 equiv of arylboronic acid, respectively, representing a slight improvement over 1^st generation conditions. Using 1.3 equiv of ArB(OH)_2, additional heterocyclic substrates that were synthesized in moderate yields include quinoline 1k (59%) and indole 1l (58%). N-methylpyrazole 1n, furan 1o, and styryl product 1 could also be prepared using only 1.3 equiv of boronic acid precursor, albeit in lower yields (29–33%). o-Methoxyphenylboronic acid was coupled with b-Tz to provide 1q in 39%, which is higher than the 19% obtained with our old catalyst system. We also note that o-tolylboronic acid successfully couples with b-Tz, yielding the product in 50% yield by ¹H NMR; however, the product was inseparable from the unreacted b-Tz and could not be obtained in pure form. Substrate limitations include 2-nitrophenylboronic acid and 2-pyridylboronic acid, which were unsuccessful and did not give the desired products.

Additional substrate scope was demonstrated by coupling 3-phenyl-6-thiomethyltetrazine (2) with electron-deficient aryl boronic acids—a low-yield coupling using our first-generation conditions. As shown in Scheme 4, phenylboronic acids (1.6 equiv) functionalized by trifluoromethyl (3a), nitro (3b), or ester (3c) groups, as well as electron-deficient heterocyclic aryl boronic acids (3d-e), coupled in moderate yields with the new catalyst system.

Scheme 4.

Comparison of 2^nd and 1^st generation conditions for Ag-mediated Liebeskind-Srogl coupling of b-Tz with arylboronic acids.

^a 10 mol% of PdI₂ and 11 mol% DiPPF were used with 1.1 equiv Ag₂CO₃. ^b 15 mol% of PdI₂ and 16 mol% DiPPF were used with 2.1 equiv Ag₂CO₃. ^c ‘equiv’ refers to number of equivalents of boronic acid.

Additionally, we demonstrated the ability of the new catalyst system to engage commercially available bis-thiomethyltetrazine 4 in reactions to produce 3-aryl-6-thiomethyltetrazines. Recently, our group demonstrated that thiomethyltetrazines can act as reversible covalent warheads whose reactivity can be “switched off” in live cells through bioorthogonal chemistry using trans-cyclooctene, enabling improved stability to downstream proteomic workflows (Scheme 5A).²⁹ Previously, we prepared tetrazine thioethers from (3-methyloxetan-3-yl)methyl carboxylic esters via sequential OBO-orthoester formation/condensation with methyl thiocarbohydrazidium iodide. As a complementary approach, coupling 4 with 1.3 equiv arylboronic acid gave tetrazines 5a-d in 22–27% isolated yields, along with recovered 4 and symmetrical diaryl-tetrazine products. Attempts to invert the reagent stoichiometry by using excess 4 (2.0 equiv) relative to ArB(OH)₂ did not lead to an improvement, yielding product 5c in only 20% yield. For this substrate class, the modest yields are offset by the simplicity of this method for accessing this emerging class of reversible covalent chemical probes.

Scheme 5.

(A) Thioalkyltetrazines are reversible covalent electrophiles toward cysteines in live cells. (B) PdI₂/DiPPF catalyzed, Ag₂CO₃-mediated coupling of 3,6-bis-thiomethyltetrazine with arylboronic acids.

CONCLUSIONS

Improved conditions are described for the silver-mediated Liebeskind–Srogl coupling of tetrazines with aryl and heteroaryl boronic acids using PdI₂ and the bidentate phosphine ligand DiPPF. Optimization studies showed that 10 mol% PdI₂, 11 mol% DiPPF, and 1.1 equiv Ag₂CO₃ in DMF at 60°C gave improved performance. This lower silver loading was tolerated without yield loss, improving cost efficiency. Electron-deficient boronic acids (e.g., trifluoromethyl-, nitro-, cyano-substituted) were coupled in significantly better yields (73–90%) than under prior conditions. Electron-neutral and heteroaryl boronic acids (e.g., pyridines, quinolines, indoles) also gave good yields, especially with higher catalyst loadings. The new method enabled access to 3-aryl-6-thiomethyltetrazines, which we have recently shown to be functional as reversible covalent warheads in chemical biology. We note that a recent publication by Mehl and co-workers, citing our ChemRxiv preprint,³⁰ uses the method described herein to prepare tetrazine unnatural amino acids required for genetic code expansion.³¹ By providing more efficient access to functionalized tetrazines, we anticipate that our second-generation PdI₂/DiPPF catalyst system will provide practical access to tool molecules for various chemical biology applications.

EXPERIMENTAL SECTION

Materials and Methods.

Reactions were conducted under nitrogen in flame-dried 4 mL sealed vials or in flame-dried round-bottom flasks. Silica gel chromatography was performed on Silicycle Siliaflash P60 silica gel (40–63 μm, 60Å) or on Yamazen reverse-phase prepacked Universal Column C18-silica gel (40–60 μm, 120Å). Automated column chromatography was performed on a Teledyne Isco Combiflash Rf. A Bruker AV400 was used to record NMR spectra (¹H: 400 MHz, ¹³C: 101 MHz, ¹⁹F: 376 MHz). Chemical shifts are reported in ppm and all spectra are referenced to their residual non-deuterated solvent peaks as follows: CDCl₃ (¹H: 7.26 ppm, ¹³C: 77.16 ppm), Methanol-d4 (¹H: 3.31 ppm, 13C: 49.00 ppm), DMSO-d₆ (¹H: 2.50 ppm, ¹³C: 39.52 ppm). ¹³C NMR resonances are proton decoupled and an APT pulse sequence was used to determine type of carbon as follows: quaternary and methylene (C or CH₂) carbons appear ‘up’ and methine and methyl (CH or CH₃) carbons appear ‘down’. Low resolution mass spectra were taken on a Water SQD2 detector which was attached to a Waters Acquity H-Class UPLC. High resolution mass spectra were obtained using a Waters GCT Premier.

Caution! Like all compounds containing nitrogen-rich functional groups, tetrazine products should be handled on a small scale until energetic properties are evaluated.

General Procedure for Silver-mediated Liebeskind-Srogl Cross-Coupling with b-Tz.

Palladium (II) iodide (13.5 mg, 38 μmol, 0.10 equiv or 20.3 mg, 56 μmol, 0.15 equiv) and 1,1'-bis(diphenylphosphino)ferrocene (DiPPF: 17.2 mg, 41 μmol, 0.11 equiv or 25.1 mg, 60 μmol, 0.16 equiv) were added to a flame dried 4 mL vial equipped with a stir bar. Dimethylformamide (3.75 mL, 0.1 M) was added and the vial was sealed under a blanket of N₂ and stirred for 1 h at 60 °C. In one portion, b-Tz (110 mg, 375 μmol, 1 equiv), arylboronic acid (488–713 μmol, 1.3–1.9 equiv), and silver (I) carbonate (114 mg, 413 μmol, 1.1 equiv or 217 mg, 788 μmol, 2.1equiv) were added and the vial was resealed under a blanket of N2 and stirred for 18–20 h in an oil bath at 60 °C. The vial was removed from the bath, allowed to cool, and solvent was removed by rotary evaporation. The crude solids were chromatographed directly on silica gel.

Supplementary Material

The Supporting Information is available free of charge at Full experimental details, complete characterization data, copies of NMR spectra (PDF)

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.