Suzuki Coupling of Protected Aryl Diazonium Ions: Expanding the Knowledge of Triazabutadiene Compatible Reactions

Abstract

Aryl diazonium ions are important in synthesis and chemical biology, and the acid labile triazabutadiene can protect this handle for future use. We report a Suzuki coupling strategy that is compatible with the triazabutadiene scaffold, expanding the scope of synthetically available triazabutadienes. Shown herein, the triazabutadiene scaffold remains intact and reactive after coupling, as demonstrated by releasing the aryl diazonium ion to label a tyrosine-rich model protein.

Graphical Abstract

Aryl diazonium ions are coveted electrophiles in organic and bioconjugation chemistry due to their highly reactive nature (Scheme 1).¹ The inherent instability of diazonium ions makes them incompatible with many common reaction conditions, storage, and creates limitations when planning efficient syntheses. At the same time, diazotization of a primary arylamine occurs under acidic conditions, a method which can cause problems when acid sensitive groups have already been installed. The free diazonium ion can then be used to label proteins or for synthetic transformations onto an organic molecule.² The triazabutadiene scaffold has been used to protect the electrophilic warhead, and previous work in our lab has shown aryl diazonium ion release can be achieved under mildly acidic aqueous or organic conditions, which is compatible with a variety of functional groups compared to the harsh acidic conditions necessary for its formation (Scheme 1).^3,4 There are reports of triazabutadiene thermal decomposition occurring at high temperatures^5,6 and the scaffold acting as a reported metal chelator.^7-9 With these considerations in mind, initial ventures into triazabutadiene functionalization performed by our lab had avoided metal-based catalytic reactions that involved heating to maintain the protecting group’s integrity. Herein, we show the triazabutadiene scaffold to be tolerant of typical Suzuki coupling conditions, allowing for facile attachment of aryl cargo to the scaffold.

Scheme 1.

Routes of synthesis for aryl diazonium ions. Highly acidic conditions are used to transform primary amines to the reactive diazonium handle. The triazabutadiene protecting group can release an aryl diazonium ion under mildly acidic conditions (pH < 7).

Aryl diazonium ions are established substrates for Suzuki coupling, capable of reacting with aryl organoboranes.^10,11 Aryl triazenes have been reportedly used as Suzuki, Sonogashira, and Heck coupling substrates, in which the aryltriazene is believed to coordinate with boron trifluoride to allow for transmetalation with palladium and nickel metal centers. In this way, the triazene acts as a directing group for ligand free catalysis but unfortunately is not maintained at the end of the reaction pathway.¹² Similarly, alkenyl- and aryl-trifluorosilanes can be directly coupled to aryl triazenes in the presence of catalytic amounts of palladium (Figure 1a), with trifluorosilane facilitating proper orientation of the catalyst and triazene to allow for a concerted reaction pathway.¹³ Arylation of various azoles has been accomplished using catalytic amounts of palladium and an excess of copper at high temperatures (Figure 1b). It is theorized activation of the azole C-H bond occurs via copper insertion, while oxidative addition of palladium into the sacrificial triazene leads to the activated aryl-palladium complex to allow for Cu/Pd exchange and exclusion of the coupled product.¹⁴ Aryl triazenes have been implemented in numerous metal catalyzed cyclization reactions, but still never manage to survive catalysis.¹⁵ Adding to the archive of compatible attachment strategies for protecting groups like the triazabutadiene can save the electrophilic aryl diazonium handle for future utilization in a synthetic pathway.

Figure 1.

(a) Triazenes reacting with aryl silanes in the presence of palladium to form aryl-aryl bonds. (b) Palladium catalyzed arylation of ethynyl triazenes, using copper as a co-catalyst. (c) This work, wherein the diazonium protecting group is maintained allowing for the reactive handle to be utilized for future experiments.

Previous experiments performed in our lab led to the development of a copper-click method for the triazabutadiene scaffold, allowing for facile attachment of cargo to the protected aryl diazonium ion.¹⁶ Copper click chemistry performed between an azide and alkyne, forming C-N bonds, creates a limitation when it comes to scaffold design due to obligate triazole formation. The work herein expands the toolbox of triazabutadiene reactions, with an emphasis on directly altering the nature of the resulting benzene diazonium ion. By taking advantage of the palladium catalyzed Suzuki coupling strategy for installation of a variety of aryl rings to the triazabutadiene scaffold, this modular reaction readily affords biaryl diazonium products (Figure 1c). Suzuki coupling, a fundamentally different attachment strategy leading to the formation of C-C bonds, allows for the addition of different heterocycles and substituted benzene rings. These extended ring systems are versatile and pave the way for new synthetically achievable and functionalized triazabutadienes. To the best of our knowledge, this is the first report of a triazabutadiene or aryl triazene surviving metal catalysis at elevated temperatures and being utilized in late-stage synthesis.

Initial concerns when considering the Suzuki reaction for facile cargo attachment to the triazabutadiene centered around the high temperature and metal catalyst used for coupling. Work performed by Bielawski and Khramov showed at elevated temperatures (170 °C) the triazabutadiene scaffold will undergo an intramolecular rearrangement, in which collapse of a tetrahedral carbon intermediate led to the formation of a substituted guanidine species through the exclusion of nitrogen gas. They found bulky substitutions about the benzimidazole-2-ylidene ring heightened the thermal stability of these compounds due to steric hindrance of the benzylidene carbon atom center.⁶ Bugarin et al. were able to tell a similar story through consideration of the electronics surrounding the triazabutadiene motif. When heated to 120 °C, triazabutadienes with electron donating groups were found to undergo significant rearrangement after 2 hours, while electron withdrawing groups were a stabilizing factor, with no reported rearrangement even after 24 hours.⁵ Our group has found that bismesityl triazabutadienes can with-stand refluxing in methanolic potassium hydroxide (about 65 °C) for upwards of five hours with an electron withdrawing ester substituent, which agrees with other reports of triazabutadiene thermal stability.¹⁷ With Suzuki chemistry in mind, we had reservations about the stability of our target products after removal of the electron donating halogen once coupling was achieved.

We began our work with a stable bismesityl triazabutadiene with a bromine handle for the initial investigation of Suzuki coupling compatibility. This scaffold was selected due to the larger σ_p value of bromine compared to boronic acid (σ_p = 0.23 versus σ_p = 0.12, respectively) which would produce a more thermally stable triazabutadiene.¹⁸ Depending on the type of aromatic ring substituted about the para position, such ring systems can act in an electron donating or withdrawing fashion.^18,19 The initial screen included a range of aromatic rings to determine if electronic substitutions changed the synthetic feasibility of target compounds. Triazabutadiene 1 (Figure 2) was synthesized by reacting 1,3-bismesitylimidazolium chloride (2) with 1-azido-4-bromobenzene (3) in the presence of potassium tert-butoxide.

Figure 2.

Synthesis of model bismesityl triazabutadiene (1) and Suzuki Products 4a-i.

Planning an efficient synthetic route towards 4a involved consideration of the catalyst of choice. Palladium-tri-phenylphosphine complexes generated in situ are commonly employed for this coupling strategy and was tried first.²⁰ The oxidized phosphine side product proved difficult to separate away from target product 4a and as a result a different palladium-ligand complex was selected. Purification difficulties were circumvented using the commercially available palladium-ferrocene catalyst PdCl₂(dppf). In addition to being a good solution to purification challenges, the heightened electron donating capability of the diphenylphosphonium ferrocene complex has been found to increase the catalytic activity of palladium for Suzuki coupling compared to triphenyl phosphine alone.²¹ With the proper catalyst in hand, we sought to explore the limitations of catalyst loading and how it affected product formation. We found this coupling method provided robust yields for as low as 1 mol% catalyst loading (Figure 3, 4a). To ensure consistent catalyst loading, 11 mol% was used moving forward.

Figure 3.

(a-e) Crystal structures of triazabutadienes 1, 5, 6, 4a and 7a, respectively, highlighting the N1, N2 and N3 of the triazabutadiene motif (C in gray, N in blue, Br in red).

Figure 4.

(a) Workflow for modification of model protein MSP (PDB: 2LEM)²⁴ using 8a. (b) Deconvolved mass distribution of the native mass spectrum of protein MSP1D1T2(−) showing the monomeric protein (purple circle) with no modifications. (c) Following chemical modification, new peaks are seen at with 1×181 Da (blue down triangle), 2×181 Da (light blue up triangle), 3×181 Da (green right triangle), and 4×181 Da modifications (dark blue square).

We hypothesized the thermal stability of Suzuki coupled triazabutadiene products 4a-i would follow a similar trend to the observations made by Bugarin’s group, in which electronics would be the dominating factor in predicting the thermal stability of the resulting triazabutadiene.⁵ We were pleased to find this coupling method provided robust results with yields of 75-90% for different meta- and para- substituted aryl rings, affording 4a-f. This method was also compatible with some heterocycles that were explored, affording yields of 63-83% for triazabutadienes 4g-i. Starting triazabutadiene 1 was stable to flash chromatography on silica gel, but 4a-i experienced extensive degradation when subjected to the same conditions.

Excited by the success of the bismesityl series, we were interested to see if our coupling strategy was compatible with more unstable triazabutadienes that are known to release aryl diazonium ions under milder protic conditions. As mentioned above, work performed by Bielawski and Khramov highlighted the importance of R₁ and R₂ in contributing to triazabutadiene thermal stability.⁶ Work in our lab has also shown the contributions of R₁ and R₂ to influence pH sensitivity of this scaffold and how quickly the protecting group releases the cargo.²² Based on previous findings, we were inspired to synthesize a less stable methyl mesityl triazabutadiene, 5, and the highly unstable tert-butyl methyl triazabutadiene, 6 (Table 1). Both triazabutadienes 5 and 6 were synthesized the same way as 1, starting with 3-mesityl-1-methyl-1H-imidazol-3-ium iodide for 5 and 1-(tert-butyl)-3-methyl-1H-imidazol-3-ium iodide for 6.

Table 1.

Synthesis of the less stable methyl mesityl (5) and highly unstable tert-butyl methyl (6) triazabutadiene (TBD) Suzuki products 7a-b and 8a-b.

TBD	R1	R2	X	K3PO4 equiv.	# (yield)
5	Me	Mes	H	2.5	7a (67%)
5	Me	Mes	NH2	4.5	7b (85%)
6	t-Bu	Me	H	2.5	8a (60%)
6	t-Bu	Me	NH2	4.5	8b (56%)

Using the Suzuki coupling procedure developed on model triazabutadiene 1, we were able to synthesize triazabutadienes 7a-b and 8a-b with moderate yields. As expected, we found yields of each triazabutadiene type to be highly dependent upon R1 and R2, wherein the more stable bismesityl series 4a-i provided high, robust yields and the more unstable tert-butyl methyl substituted 8a-b gave moderate yields. Similarly, we found products 7a-b and 8a-b to be unforgiving of purification conditions, with flash being an unsuitable tool to aid in the chromatography process.

We were able to crystallize triazabutadienes 1, 5, 6, along with Suzuki products 4a and 7a and successfully solved the crystal structures (Figure 3a-e). We were pleased to see the crystal structures of 5 and 6 were consistent with previous NOESY data published by our lab.²² As predicted, triazabutadiene 5 (Figure 3b) sits such that a stabilizing T-shaped π-stack occurs between the mesityl and phenyl ring of the triazabutadiene scaffold. Triazabutadiene 6 (Figure 3c) orients to minimize steric repulsions between the tert-butyl substitution, which acts as the dominating factor for predicting triazabutadiene conformations when the stabilizing π interaction is removed. When comparing the Suzuki products 4a and 7a (Figure 3d and 3e) to starting triazabutadienes 1 (Figure 3a) and 5 (Figure 3b), respectively, we find the largest differences in structure arise from the tilt of the triazabutadiene motif, where N3-N1 become coplanar with the imidazole ring for the Suzuki coupled products. Starting materials 1 and 5 skew the triazabutadiene and extended phenyl ring such that they tilt out of plane with the imidazole ring. Comparatively, triazabutadienes 4a and 7a twist the phenyl ring and N3-N1 to sit in a planar fashion with the substituted imidazole, resulting in an overall extension of the π-conjugated system.

The true utility of this coupling strategy is the ability to maintain the aryl diazonium protecting group, the triazabutadiene, and use the reactive handle in late-stage synthesis or for bio-chemical applications. To confirm reactivity of the triazabutadiene post coupling, we decided to exploit the reactive nature of the aryl diazonium ion by labeling the tyrosine mimic p-cresol under neutral conditions using triazabutadiene 8a to afford conjugated azo adduct 9 (see Supplemental Information (SI)). Extending this reactivity past the small molecule level, we sought to label a membrane scaffold protein (MSP) as a model protein, which contains multiple tyrosine residues and confirmed protein modification via mass spectrometry (Figure 4a).²³ Labeling experiments were performed at pH 9 to increase the nucleophilicity of tyrosine. It is important to note that even at pH 9, the reactive triazabutadiene could release the aryl diazonium. Under these conditions, a distribution of labeling occurred with upwards of four modifications per protein clearly visible (Figure 4c) when compared to the control (Figure 4b), confirming the reactive nature of the biaryl diazonium ions.

In conclusion, we have demonstrated a method for Suzuki coupling which is compatible with an acid labile aryl diazonium ion protecting group, the triazabutadiene. Unlike established protocols for metal catalyzed coupling of aryl triazenes where the protecting group is sacrificed, we find the triazabutadiene to remain intact and reactive, providing biaryl diazonium ions under mildly acidic conditions for future utilization. With this chemistry in hand, we believe the protecting group could be embedded into drug like molecules, opening avenues for the integration of triazabutadienes into pharmaceutical applications. The diverse reactivity of aryl diazonium ions allows for their transformation into a variety of functional groups, creating a diversification point for synthetic strategies. Future efforts will be put towards investigating closely related metal catalyzed reactions and their compatibility with the triazabutadiene scaffold, such as Sonogashira and Heck coupling. It is our hope that this work helps highlight the usefulness of the triazabutadiene scaffold, inspiring others to integrate it into their own synthetic systems.