CHAPTER 1: New Polymers From SuFEx Click Chemistry: Syntheses and Perspectives

Polymers are powerful tools for exploring nature. Condensation polymers, such as polyesters, polyurethanes, polycarbonates, and polyamides, appear in almost every corner of our daily life. From apparel to cell phones, from medical apparatus to aerospace crafts, they have evolved our approaches to discover the mystery of nature and are essential components of our modern society. Taking a close look at the polycondensation process to generate those polymers, functional groups (normally hydroxyl or amino groups) and other functional groups (normally carbonyl groups) are buckled together by reliable and robust reactions to form carbon-heteroatom linkages as shown in Scheme 1.1. Not surprisingly, Mother Nature is an expert synthesizer of polymers. She generates the backbones of proteins from the polycondensation of amino-carboxylic acids, and nucleic acids from the polycondensation of nucleotides.

Generally speaking, polycondensation may be regarded as the stepwise ligation of different functional groups with loss of water or an equivalent. Therefore, the reactivity of the functional groups in the building blocks is crucial for the construction of condensation polymers. Amidation and esterification are well adapted as polymer-forming chemical reactions, and the reactivity of carbonyl group towards O, N-nucleophiles has been well investigated and elucidated in the past century. For instance, the amidation process happens rapidly above 180 °C to 200 °C without catalysts and is usually free from side reactions to deliver the desired amides in very high yields. This is very important for the formation of high-molecular-weight polyamides as the principle of equal reactivity of all functional groups during polycondensations determines that the loss of one functional group through side reactions may significantly limit the molecular weight and yield that can be reached. From this perspective, the outcome of polycondensations highly depends on the reactivity of the functional groups of the monomers. Only those chemical reactions possessing an ideal efficiency (higher yield in shorter reaction time) and selectivity (fewer side reactions) are chosen by polymer scientists to create various useful materials.

Imagine a polycondensation between monomer A-A with B-B and define the yield of the chemical reaction A-A + B-B → A-A-B-B (assuming that the ratio of A-A : B-B is exactly 1 : 1) as Y. Figure 1.1 gives the cumulative yield Y_n of the reaction as a function of {2X_n − 1} in a linear stepwise sequence.

From Figure 1.1, we can tell that with Y = 90%, Y_n reached almost 0 when 2X_n − 1 = 40, which means that the degree of polymerization (X_n) is limited to around 20, thus constraining the size of the polymer. Pushing Y up to 97%, the yield of the polymers with X_n ∼ 50 (2X_n − 1 = 100) does not even exceed 10%. Elevating Y to 99% only improves the overall yield of condensation polymer with X_n ∼ 50 to 40%. One comment here is that for a chemical reaction yielding 99% desired product, it is already a near-perfect one for a synthetic chemist, but it is still not good enough for producing those big fellows. To achieve an overall yield of condensation polymer with X_n ∼ 50 above 90%, each step needs to proceed with a yield of 99.95% or even higher!

The calculations above illuminate the extraordinary dependence of those processes on yield per step. In short, the yields can never be too high! Carothers equation (where r is the stoichiometric ratio of reactants) determines that the ratio of the two functional groups involved in polycondensation will also influence the degree of polymerization (X_n), thus requiring the high purity and stability of monomers. Searching for reactions with strong thermodynamic driving forces and well-controlled, selective, efficient pathways is a permanent pursuit in polymer chemistry and its allied disciplines.

Discovering ideal reactivity is not enough for the polymer industry because producing polymers is unlike the syntheses of fine-scale chemicals in grams or kilograms, polymers require grand industrial plants to generate millions of tons of products per annum. Table 1.1 lists the world production of polyesters, polyurethanes polycarbonates, and polyamides in recent years.¹ 

In the polymer industry, the power of a company rests upon its access to monomers and their related polymers. After all, you can't change the world, or even the single market, with several kilograms of something in your hand. Most of the monomers for polycondensation are streamed from the portal of the petrochemical industry that provides the most basic raw materials for organic syntheses. For instance, MDI (methylene diphenyl diisocyanate) is an important monomer (world production in 2015: 7.5 million tons) for the production of polyurethanes (world production in 2015: 20 million tons).¹  The 4,4′ isomer (4,4′-diphenylmethane diisocyanate) is most widely used and is also known as Pure MDI. The synthetic route of MDI is depicted in Scheme 1.2.

Aniline, formaldehyde, and phosgene are all bulk chemicals at the million-tonne level, allowing MDI production on a massive scale. Wanhua started their MDI project in 1983. Beginning with an obscure procedure, they thrived for 12 years increasing the capacity to 10 thousand tons per year. They spent another 20 years pushing up the annual capacity to 210 million tons, accounting for one-fourth of the total world production of MDI today.¹  Considering the rapidly growing demand for MDI (annual growth: ∼9%), all the oligarchs in this market have started an arms race that competes for the annual production of MDI, a number that determines who is the boss at the table.

The enormous demand of these polymers stems from a wide range of unique properties. There is huge demand from the military, clothing, electronics, automobiles, etc. Discovering a reaction combining ideal reactivity and availability of starting materials with unique properties is like fishing in the sea. It can't be designed, and usually just turns up in one of your experiments, even the failed ones. Someone might encounter or even record it decades before, but they did not see the beauty of it and let it go. In the journey of “fishing”, Sharpless and coworkers proposed the idea of “click chemistry” in 2001,¹  which focuses on the modular syntheses of molecules with desired functionality using highly efficient, selective, and benign reactions. From a synthetic perspective, polycondensation chemistry depends on the same basic principles as click chemistry: selectivity and efficiency are inherent in both polymer syntheses and modular click reactions. It is no wonder that the impact of click chemistry on materials science has grown rapidly over the past decade, especially in the field of polymer preparation and modification.

The first ideal click reaction to emerge in the Sharpless group is called CuAAC (copper-catalyzed azide–alkyne cycloaddition),²  a type of reaction that orthogonally ligates two organic synthons, azides and alkynes, to form triazoles catalyzed by Cu(i). Due to the highly energetic nature of these two functional groups, CuAAC reactions normally proceed very rapidly and show unprecedented selectivity even in complex environments, opening new horizons to modular syntheses in nearly any environment with great reliability. Since its discovery, CuAAC has also been beloved among biochemists, being extremely efficient for bioconjugations, such as in the scenario of installing probes to target proteins.³  In material science, CuAAC ligations have been applied in the postmodification of polymer surfaces and chains.⁴  However, one fact that limits the application of CuAAC in polymer syntheses is, as mentioned above, the availability of the azide and alkynes. The generation of organic azides involves the use of sodium azide, an explosive compound with 250 tons production in 2004, mostly consumed by the automobile industry as airbags. The industrial production of alkynes is also much smaller, compared with other hydrocarbons.

CuAAC is good for polymer functionalization and joining two blocks to form block polymers, but it is not good for polymerization. The journey continues then, until 2014, when researchers in the Sharpless group encountered another click reaction, Sulfur(vi) Fluoride Exchange (SuFEx), which is identified as a fringe acid–base reaction. S(vi)–F bonds are very stable in most conditions, but they show excellent reactivity towards O, N-nucleophiles under the right circumstances, e.g. when activated by protons or silicons, to form S(vi)–O and S(vi)–N bonds to make functional products.⁵  One application of the SuFEx process is the reaction between aromatic bis(silyl ethers) and bis(fluorosulfates) or bis(sulfonyl fluorides) in the presence of certain catalysts to generate polymers containing –SO₂- backbones.^5,6  The two monomers bis(silyl ethers) and bis(fluorosulfates) are easily synthesized from phenols, SO₂F₂ and a chlorotrialkylsilane, which are readily available as bulk industrial feedstocks. In this chapter, we will focus on the details of the processes for making new polysulfonate and polysulfate polymers, especially the evolution of the catalysts involved. Aside from leading to the discovery of new polymers, SuFEx click chemistry as a powerful tool in polymer modification will also be discussed herein.

Germans are the pioneers in the syntheses of compounds containing S(vi)–F bonds. Alkyl sulfonyl fluoride (R-SO₂F) and aryl sulfonyl fluoride (Ar-SO₂F) compounds appeared in the German literature in the early 20th century.⁷  They are mainly used in the dye industry due to the dye properties of sulfur(vi) derivatives of benzenes, naphthalenes, and anthracenes. They are identified as thermodynamically stable compounds with resistance to reduction. This chemically inert nature has kept them away from the core of synthetic chemistry and the Germans' work faded from the view of most organic chemists in the mid-20th century.

The inorganic molecules containing S(vi)–F bonds, such as SF₆, SO₂F₂, SOF₄, and NSF₃, were also isolated and investigated decades ago. Sulfur hexafluoride (SF₆) was first synthesized from the burning of sulfur in an atmosphere of fluorine in 1900 by Moissan and Lebeau.⁸  It is a colorless, odorless, non-toxic and chemically inert gas, widely used as an insulating atmosphere for high-voltage electrical equipment. Sulfuryl fluoride (SO₂F₂) is readily obtained by fluorination of sulfur dioxide with F₂⁹  or the replacement of chlorine in sulfuryl chloride by other reagents.¹⁰  Chemically, it is a much more inert gas than its sibling, sulfuryl chloride, which undergoes both chloride substitution at the S(vi) electrophilic center (alternatively, nucleophilic attack by a Grignard reagent at the chlorine takes place to afford the C–Cl bond) and reductive collapse yielding the S(iv) species and Cl⁻. SO₂F₂ hydrolyzes slowly in aqueous sodium hydroxide, and remains stable even in melted sodium metal.⁵  The Dow company developed the widespread use of SO₂F₂ as a fumigant insecticide. Thionyl tetrafluoride (SOF₄) is obtained by the reaction of fluorine with thionyl fluoride (SOF₂) or the reaction of oxygen with sulfur tetrafluoride (SF₄) in the presence of 20 mol% NO₂. This gas is prone to undergo hydrolysis generating SO₂F₂ and HF.¹¹  Thiazyl trifluoride (NSF₃) is both thermally and chemically stable. It reacts with metallic sodium only at 300 °C.¹²  Table 1.2 summarizes some properties of these gaseous molecules.

In spite of their importance in dye manufacture, S(vi) fluorides have wandered on the brink of chemistry for decades, not to mention their minimal influence on other sciences. Vikane has high-volume annual production (two production lines with 1500-ton annual capacity in China, owned by Yingpeng and Linhai, respectively), but mainly serves as a fumigant gas instead of a chemical reagent. William Firth synthesized fluorosulfates using bisphenol A and SO₂F₂ in the presence of pyridine, but the yield was low (34%).¹³  The reaction of SO₂F₂ with sodium¹⁴  and lithium phenolates¹⁵  had previously been shown to provide better yields of fluorosulfates, but all these procedures did not catch on. SO₂F₂, therefore, represents a curious combination of ton-scale application in industry and poor appreciation in the laboratory, as well as in the fine chemical industry.

The reviving moment of this chemistry started in 2011, with the introduction of SuFEx that demonstrated the stability–reactivity dual character of S^VI–F functional groups: they remained moribund under most conditions but turned reactive towards an appropriate nucleophile under conditions in which fluoride ejection is activated. More accurately speaking, the whole story starts from the serendipity in Scheme 1.3.⁵ 

The very stable gas SO₂F₂ reacts with phenol in the presence of triethylamine or DIPEA once, and it stops, delivering phenyl fluorosulfate almost quantitively. This protocol provided the first reliable route for the synthesis of fluorosulfates, offering the opportunity to explore their unique properties. SO₂F₂ was then found to derivatize phenols in all kinds of environments, even those in highly functional molecules such as vancomycin, leaving aliphatic alcohols, aliphatic and aromatic amines, and carboxylates untouched. This is of particular interest for biology and medicine, and pharmaceutical research followed up very soon. For instance, fluorosulfates have also been found to be unique warheads for covalent capture of protein side chains in a highly selective, context-dependent manner.

The use of aryl fluorosulfates finds another powerful set of applications when silicon is brought into play: for example, they react rapidly and quantitively with aryl silyl ethers to give diarylsulfates and a volatile F-SiR₃ as the only byproduct. Those SuFEx linkages can be constructed with complete selectivity in the presence of many other functional groups, thereby enabling a vast array of building blocks to be assembled into oligomers and polymers (e.g. the sulfate analogue of bisphenol A polycarbonate).

Polycarbonates are popular thermoplastic polymers that are easily worked, molded, and thermoformed. They can be found almost everywhere in our daily life: eyeglass, water bottles, computers, cell phones etc. However, their counterpart polysulfonates and polysulfates are not industrially produced due to a lack of reliable and scalable synthetic access.

Interestingly, in searching for new polymers that might also show evidence of high thermal stability, polysulfonates had been noticed and investigated in the Materials Laboratory of Wright-Patterson Air Force Base back in 1964. Thomas and coworkers synthesized a series of polysulfonates through polycondensations of aromatic sulfonyl chlorides with phenols according to the reaction shown in Scheme 1.4b.¹⁶  These polymers were found to have molecular weights up to 50 000 and thermal stability under nitrogen to above 300 °C. PDI of the polymers was shown to be 22, indicating a wide molecular weight distribution of mixed polymers. Nevertheless, tons of polysulfonates were produced using such a protocol. However, no updates of this new polymer family appeared in the literature. In 1972, Firth synthesized several aromatic polysulfates using the reaction of bisfluorosulfates with the disodium salt of bisphenol A.¹³  Preparation of these polymers required prolonged heating, and they were obtained in pure form in low yields only after repeated precipitation. Other attempts using sulfuryl chloride (SO₂Cl₂) with dihydroxy aromatic compounds or bischlorosulfate with bisphenol gave only ring chlorinated sulfate esters of low molecular weight, and the reaction of bisphenol A in pyridine solution with an equimolar amount of sulfuryl fluoride (SO₂F₂) stopped at the fluorosulfate stage. However, despite showing excellent mechanical properties and the high chemical stability needed for engineering polymers, polysulfates and polysulfonates lacking a “good reaction” for manufacture could not go further.

SuFEx click chemistry now has the potential to accomplish this goal. In 2014, Dong and Sharpless et al. reported a simple and straightforward SuFEx-based method for the synthesis of high-molecular-weight polysulfate polymers from bis(aryl fluorosulfate) and bis(aryl silyl ether) under simple and mild reaction conditions: Bisphenol A (7.1 million tons annual production in 2017) was first treated with SO₂F₂ in the presence of triethylamine to generate monomer A-A, BPA-fluorosulfate. Bisphenol A then reacted with chlorosilanes, such as tert-butyldimethylsilyl chloride (TBSCl), to deliver another monomer B-B, bis(silyl ether). Combination of A-A and B-B in the presence of catalysts, such as DBU (20%) or BEMP (1%) almost quantitatively produced the desired polysulfate P-7 with high-molecular-weight. The reaction was compatible with many functional groups and easily scaled up due to its operational simplicity, as described in Scheme 1.5.⁵ 

Reactions of silylated and fluorinated compounds are, of course, well known in organic synthesis and in polymer chemistry. For example, the advantages of using silylated monomers have been insightfully summarized by Kricheldorf.¹⁷  He wrote: First, polycondensation of silylated monomers with halogen containing electrophiles avoids the liberation of free hydrochloride or hydrofluoric acid, and thus, acid catalyzed side reactions. Second, the liberated trimethylsilyl derivatives, such as fluoro or chlorotrimethylsilane, trimethylsilylacetate, or hexamethyldisiloxane are highly volatile and easy to remove from the reaction mixture. Third, silyation improves the solubility of monomers in a given reaction medium. Fourth, silylation enhances the volatility of monomers due to the elimination of H-bonds. One application of the “silyl method” was the synthesis of poly(aryl ethers) taking advantage of the strength of the Si–F bond and the innocuous nature of the silyl fluoride byproducts (Scheme 1.6).

The SuFEx-based polycondensation of bis(aryl fluorosulfate) and bis(aryl silyl ether) benefited from the advantages mentioned above. The byproduct, TBSF, was easy to distill off from the mixture. The polymerization process could be conducted neat by simply melting the two monomers together at elevated temperature. Table 1.3 showcases the performance of different catalysts under solvent-free polymerization conditions.⁶ 

With the polymers on hand, researchers compared the chemical stability of bisphenol A polysulfate P-7 (BPA-PS) with its polycarbonate analogue (BPA-PC). Treatment of P-7 with sodium hydroxide (1.3 M, 1 : 2 EtOH/H₂O) at ambient temperature or with a saturated solution of sodium carbonate at 80 °C for 16 hours caused no observable change in polymeric structure, whereas BPA-PC was not resistant to hydrolysis under the same conditions and decomposed to low-molecular-weight materials, as indicated by gel-permeation chromatography (Figure 1.2).⁶ 

The excellent thermal stability of the P-7 was shown by the TGA analysis. The P-7 started to decompose at approximately 350 °C or higher, which was in good agreement with the data obtained by Firth. Density, tensile properties, and oxygen permeability were also measured and compared to BPA-PC, as summarized in Table 1.4: P-7 was slightly denser, had a higher tensile modulus, and had a similar yield stress compared with BPA-PC. The oxygen permeability of P-7 was significantly lower than that of BPA-PC, indicating that P-7 had a smaller free volume at room temperature.⁶ 

A practical and facile preparation of polysulfates was established in the laboratory. However, several disadvantages of this base-catalyzed protocol limit its application at larger scale. First, the catalyst loading is quite high, for instance, no less than 10 mol% of DBU has to be used, making purification procedures tedious to remove catalyst residue before polymer processing. Second, though the loading of BEMP is lower (1.0 mol%), it is expensive and lacks bulk accessibility. Third, the strong basic nature of the catalysts restricts the substrate scope of the process. Better catalysts are needed to facilitate the scale-up of polysulfates.

The SuFEx click reaction between aryl silyl ethers and aryl fluorosulfates has paved the way to the facile synthesis of polyfluorosulfate, in which the first generation of catalysts for this SuFEx reaction was identified as a variety of bases. The screening of more bases seems reasonable for the improvement of catalyst performance. However, what is fascinating in chemistry is that discoveries generally deny rationale and that is what happens here. While using NMR to monitor the base-catalyzed polycondensation process, Dong observed a unique proton peak from [HF₂]⁻. The “bifluoride phenomenon” also showed up previously: potassium bifluoride was much more efficient than potassium fluoride for the synthesis of sulfonyl fluorides from the corresponding chlorides (Scheme 1.7).⁵ 

Meanwhile, the scope of bases that were active in promoting the polymerization was limited after screening, whereas the use of most inorganic or organic bases either led to the decomposition of silyl ethers or failed to catalyze the reaction. The researchers were then suspicious of whether the base-catalyzed SuFEx polycondensation was the mechanism behind the reaction. The suspicion reached its peak when Dong read an interesting story about the discovery of the catalyst for Group Transfer Polymerization (GTP), a silicon-mediated process, at DuPont over 20 years ago:¹⁸  Sogah stored his TASF in an old dry box (he was a new employee and had not yet gotten one of the new highly efficient boxes like the one Farnham had). This lucky break led us to the first “leave-in” catalyst for GTP. Farnham's good sample of TASF did not give a living GTP at room temperature, while the sample Sogah used worked perfectly. Very small amounts gave living polymer at room temperature… Water was reacting with the TASF to produce bifluoride.

The surprising emergence of bifluoride in the GTP story along with the “bifluoride phenomenon” observed in SuFEx reactions aroused the interest and courage of Dong to try [TAS]⁺[FHF]⁻ as the catalyst in the SuFEx polycondensation. It worked, with only 0.1 mol% catalyst loading! Therefore, a totally different type of catalyst, acidic bifluoride salt (M⁺[FHF]⁻, where M⁺ refers to a wide range of organic and inorganic cations), emerged as the second generation of catalysts for the SuFEx polycondensation, exhibiting a superior catalyzing ability than the basic DBU-like catalysts.¹⁹ 

Besides [TAS]⁺[FHF]⁻, a number of bifluorides (M-4 to M-13) were synthesized by ion-exchange of their chloride/bromide precursors (M⁺X⁻, X = Cl, Br) with silver bifluoride (Ag⁺[FHF]⁻). The performance of these as-synthesized catalysts was tested in the model reaction between monomer A-A and B-B to produce polymer P-7, as shown in Scheme 1.8. It was found that the use of 2 mol% potassium bifluoride (K⁺[FHF]⁻) as catalyst for the polycondensation of A-A and B-B in NMP at 130 °C for 17 hours was able to promote the formation of polysulfates, but it was much slower compared to the case catalyzed by DBU (10 mol%) and BEMP (1 mol%) (17 h versus 1 h). A much higher molecular weight of P-7 (as high as 100 kDa) could be achieved by combining 2.0 mol% potassium bifluoride with 1.0 mol% tetrabutylammonium chloride (M-2) or 18-crown-6 ether (M-3). The activities of other biflouride catalysts (M-4 to M-13) were subsequently investigated, most of which were highly active at catalyzing the reaction. 0.1 mol% catalyst loading was sufficient to accomplish the polycondensation within one hour, generating P-7 with M_n^ps ranging from 30 kDa to 100 kDa (PDI = 1.4–1.7). Among all the candidates screened, tris(dimethylamino)sulfonium bifluoride (M-8) was chosen as the catalyst as it was able to generate P-7 with a molecular weight up to 100 kDa at only 0.05 mol% loading (Scheme 1.8).¹⁹  The scale-up of polysulfate P-7 to 100 gram scale in the laboratory was performed by combining 0.2 mol of A-A and 0.2 mol of B-B at 120 °C (internal temperature) in NMP (50 ml) in the presence of only 0.05 mol% catalyst M-8. Polysulfate P-7 was obtained as white fibrous materials in near quantitative yield (112.7 g) and GPC analysis showed a higher M_n^ps of 110 kDa (PDI = 1.7), compared with the P-7 product obtained at the 2.0 mmol scale (M_n^ps = 92 kDa, PDI = 1.7).¹⁹ 

With M-8 on hand, a variety of monomers were synthesized and used as building blocks to produce polymers with different backbones. The silyl ether and fluorosulfate monomers from the same phenol polymerized efficiently under the catalyst condition (Scheme 1.9).¹⁹  Compared with the results from DBU-catalyzed polycondensations, polysulfates with comparable molecular weights and polydispersities were obtained. For instance, the M_n^ps/PDI of P-9 was 46 kDa/1.5 in the case of the DBU catalyst compared with 84 kDa/1.6 in the case of the M-8 catalyst.

The cross-condensation, such as A-A-8 and B-B, gave polymer P-16 with high molecular weight and good polydispersity (M_n^ps = 58 kDa, PDI = 1.7). The self-condensation of A-B-1 was also successfully catalyzed by M-8 to produce P-7 with a M_n^ps of 65 kDa and PDI of 1.5. For the first time, the direct polycondensation of A-A with SO₂F₂ (vikane) was realized by the aid of M-8, affording P-7 with moderate molecular weight (M_n^ps = 22 kDa, PDI = 1.4). This protocol was also applicable to other monomers like B-B-7 to prepare polysulfate P-14 in near quantitative conversion (M_n^ps = 28 kDa, PDI = 1.5). The orthogonality of the SuFEx reaction to many other chemical reactions allowed the build-up of polymer backbones bearing functional groups latent for further modification. For example, the propargyl groups in P-19 represent a tail handle for ejecting off CuAAC-annulated triazole decorations (Scheme 1.10).¹⁹ 

While DBU or BEMP failed to catalyze the reactions of aryl silyl ether with alkyl sulfonyl fluoride to prepare polyfluorosulfonates, only affording a mixture of oligomers, bifluoride catalysts proved successful with those alkyl and aryl sulfonyl fluorides and the resulting polysulfonates were obtained (Scheme 1.11). For instance, the aliphatic bis(sulfonyl fluorides) A-A-12, A-A-13 and A-A-14 reacted with B-B respectively in NMP in the presence of 5 mol% catalyst M-8 at 130 °C, giving P-20, P-21 and P-22 in quantitative yields and moderate molecular weights (13–23 kDa). The polycondensation of the aromatic bis(sulfonyl fluoride) A-A-15 and B-B required lower catalyst-loading (0.5 mol%) and gave the polysulfonate with a higher molecular weight (65 kDa).¹⁹ 

The bifluoride-catalyzed SuFEx polycondensation overcame several limitations of the DBU-catalyzed protocol. First, this new class of catalysts enabled a broader substrate scope, such as aliphatic sulfonyl fluorides and SO₂F_2. More importantly, the bifluoride catalysts are significantly more active than DBU or BEMP, therefore much lower catalyst-loading was required, which is essential for cost-effective industrial production. Undoubtedly, this work sets a foundation for future translation of the synthesis of polysulfates or polysulfonates from laboratory to industry.

Mechanism investigation seems to fall behind the improvement of the catalysts in the SuFEx-based polycondensation of fluorosulfates and silyl ethers. Once triggered, the polymerization took place rapidly, leaving tiny windows for monitoring the active species and the pathways during the reaction. The switch from basic catalysts to acidic bifluorides overrode all the assumptions based on the rationale with respect to “base-catalyzed”. The bifluoride-catalyzed hypothesis might borrow some evidence from the GTP process ([TAS]⁺[FHF]⁻ as the catalyst), but considering the arguments in the mechanism study of GTP,¹⁸  it is too early to propose one mechanism and ignore all the other possibilities (what if a different and better catalyst is discovered?).

At this moment, we are still not clear whether the reaction proceeds on the sulfur center or the silicon center in a disassociation fashion, or whether it undergoes an association pathway to form a concerted intermediacy in the middle of the process since both silicon and sulfur can have hypervalent coordination (Scheme 1.12). One interesting observation by Dong is that the use of an excess amount of fluorosulfates affords polysulfate with high molecular weight while an excess amount of silyl ethers is harmful to the polycondensation, leading to the formation of oligomers and polymers with low molecular weight. Locklin et al. discovered that the surface modification is successful only when fluorosulfates are immobilized on the surface followed by the reaction with silyl ether, not the other way around, which gives a mess.²⁰  More efficient catalysts are under investigation. Hopefully, we will have more clues to solve this puzzle.

Since 2014, investigations on SuFEx click chemistry have been sparked, leading to a variety of intriguing discoveries that showcase the distinctive reactivity of SuFEx-able compounds. In material science, SuFEx reactions have been applied in the modification of polymer surfaces and syntheses of functional polymers. Below are some highlights on how SuFEx click chemistry reactivity has been explored and advanced by research groups dedicated to finding even more useful properties in the domain of polymers.

In polymer science, post-polymerization modification (PPM) involves the polymerization of monomers carrying reactive functional groups that are completely inert towards the initial polymerization conditions but can be quantitatively transformed in a later step. Therefore, S(vi)–F functionalities are beloved in this realm due to their unique stability-reactivity dual character that allows them to be tolerated under various polymerization environments and then modified using appropriate protocols.

In 2015, Locklin et al. reported the synthesis of brushes with integrated sulfonyl fluoride. They fabricated the brushes using UV-photopolymerization on surfaces and further reacted them with silyl ethers containing other click chemistry “handles” under ambient conditions in the presence of catalysts (Scheme 1.13). This process displayed that the SuFEx reaction between sulfonyl fluorides and silyl ethers was orthogonal to all other known click reactions. They also noted that the SuFEx linkages were even faster than the CuAAC linkages.²¹ 

Multiple orthogonal modification of the surface of a copolymer containing fluorosulfate, aromatic silyl ether and azide functionalities was achieved in Fokin's lab in 2016. Sulfur(vi) fluoride exchange (SuFEx) and CuAAC chemistries were both applied to attach three different fluorescent dyes bearing the corresponding reactive sites to the backbone of the copolymers in a successive fashion. A rare example of a triply functionalized copolystyrene was obtained via this approach under ambient conditions without protecting groups (Scheme 1.14).²² 

Locklin's group further evaluated the advantages and limitations of SuFEx reaction in the post-polymerization modification process. Three different polymer brushes containing alkyl sulfonyl fluorides (FSPMA), aromatic sulfonyl fluorides (PEFA), and aromatic fluorosulfonates (VPSF) were synthesized and then reacted with three different tert-butyldimethysilyl (TBDMS) ether derivatives respectively (Scheme 1.15). Surprisingly, the protocol that silyl-decorated brushes were reacted with fluorosulfonate derivatives resulted in no surface reaction. It was also found that product stability was dependent on the nature of the TBDMS ether derivatives while reaction rate was dependent on base selection, with stronger guanidine bases yielding faster reaction rates.²⁰ 

Sequence-regulated polymers were synthesized in Niu's lab recently by combining two orthogonal click reactions: CuAAC and SuFEx, to enable the one-pot synthesis of polydispersed sequence-controlled polymers through step-growth copolymerization. Furthermore, repeatedly conducting SuFEx and CuAAC Click reactions on a solid support yielded monodispersed sequence-defined oligomers. Both the one-pot fashion and the successive protocol represented the orthogonality of SuFEx and CuAAC click reactions, not only to each other, but also to a broad scope of other functionalities (Scheme 1.16). This protocol provided new opportunities to facilely access sequence-regulated synthetic polymers with a high degree of structural diversity.²³ 

Zuilhof et al. demonstrated the quantitative and orthogonal surface functionalization utilizing “SuFEx-able” monolayers. The surfaces bearing primary and secondary amines were reacted with ethenesulfonyl fluoride (ESF) to afford (di-SuFEx/CuAAC–SuFEx/SPOCQ–SuFEx) click platforms (Scheme 1.17). The Michael addition of the SuFEx linker, ethenesulfonyl fluoride, with widely available amines was proved to be an easy and efficient preparation of the sulfonyl fluoride motif with quantitative yields and high modularity. Different click reactions could be conducted on a single platform in high yields to form diverse surfaces.²⁴ 

Averick's group showcased a SuFEx-aided functionalization of chain-ends of four Aryl-TBDMS terminated polymers. Polymers were prepared via activators generated by electron transfer atom transfer radical (AGET ATRP) polymerization and then reacted with aryl flourosulfates in nearly quantitative conversion (Scheme 1.18).²⁵ 

Perrier et al. found that the monomers bearing sulfonyl fluorides were compatible with the conditions in reversible-addition fragmentation chain transfer (RAFT) polymerization (Scheme 1.19). Sulfonyl fluoride terminated polymers were synthesized and coupled together by an efficient SuFEx reaction with bis(tbutyldimethylsilyl) ether (Monomer B-B).²⁶ 

The combination of benzophenone photochemistry with SuFEx chemistry was introduced by Chen, in which the surface of an important biomedical material, PVC (polyvinyl chloride), was decorated by the UV-irradiated immobilization of 4-((3-benzoylphenyl) carbamoyl) benzenesulfonyl fluoride (BPSF) and the SuFEx reaction with silyl ethers catalyzed by TBD in a rapid and efficient one-pot fashion (Scheme 1.20).²⁷ 

A SuFEx click reaction was also applied in the modification of a carbon fibre surface. Moses et al. reported the direct installation of the sulfur(vi) fluoride moiety onto the surface via electrochemical deposition of the fluorosulfate phenyldiazonium tetrafluoroborate salt, or by exposing a phenol on the carbon surface to gaseous SO₂F₂ (Scheme 1.21). These surfaces decorated with SuFEx-able functionalities were reacted with aryl silyl ethers to generate diarylsulfate linkages that were stable under electrochemical redox conditions. They showed that the amino group installed onto the fiber surface improved interfacial shear strength up to 130% in epoxy resin.²⁸ 

Cellulose acetate (CA) nanofibers were constructed as a versatile platform for biofunctionalization by introducing sulfonyl fluoride moieties via electrospinning fabrication of CA/poly(3-(fluorosulfonyl)-propyl methacrylate) (PFPM) mixed solutions of varying composition (CA-0, CA-10, and CA-20 corresponding to 0, 10, and 20 wt% PFPM, respectively, in the solution). Immersing the three nanofibers in a silyl ether protected poly(ethylene glycol)methyl ether (TBDMS-PEG) solution led to the installment of PEG chains onto the surface of CA-10 and CA-20. Due to the anti-biofouling nature of PEG, the adhesion of E. coli bacteria on the CA-10 and CA-20 nanofibers was very low, whereas that on CA-0 was high (Scheme 1.22).²⁹ 

SuFEx has been utilized for the modification of the UiO-67 series of metal−organic frameworks (MOFs) containing the sulfonyl fluoride group, which was later treated with different silyl ethers to respectively yield five modified MOFs without degrading the crystalline structure. Introduction of an imidazolium group into the MOF turned it into an efficient heterogeneous catalyst for the benzoin condensation reaction (Scheme 1.23).³⁰ 

In 2017, Wu and Sharpless et al. reported a SuFEx-based polycondensation between bisalkylsulfonyl fluorides and bisphenol bis(tbutyldimethylsilyl) ethers catalyzed by [Ph₃P=N-PPh₃]⁺[HF₂]⁻. The bisalkylsulfonyl fluorides were prepared via the Michael addition of ethenesulfonyl fluoride (ESF) and amines/anilines while these silyl ethers were synthesized from the silylation of bisphenols by TBSCl (Scheme 1.24).³¹  The protocol utilized readily available amines, anilines, and bisphenols as starting materials to generate a variety of polysulfonates with a variety of side chain functionalities in >99% conversion within 10 min to 1h, exhibiting excellent efficiency and functional group tolerance.

Lu's group reported the synthesis of two functional polysulfates (PolyTPP-NI and CPTPP-NI) from fluorosulfate and silyl ether monomers bearing large conjugated chains via SuFEx reactions in high yields under mild reaction conditions. The as-synthesized polymers were used to prepare sandwiched memory devices with stable ‘flash’ electron storage behaviour (Scheme 1.25).³²  Later, four more functional polymers with large conjugate moieties were prepared using the same protocol. These polymers displayed satisfactory thermal stability and solution processability and were used as the active layer for memory devices.³³ 

In this chapter, we have reviewed the discovery and development of SuFEx click chemistry in polymer formation. The polycarbonate and polyester analogues, polysulfates and polysulfonates, have been reliably produced in SuFEx-based polycondensation at mole-scale. Bifluorides are identified as the best catalysts at this moment, replacing the previous basic catalysts, such as DBU and BEMP. The search for cheaper, more stable and more efficient catalysts is ongoing, to realize the translation of these new polysulfate and polysulfonate polymers into industrial production. SuFEx click chemistry is emerging as a powerful tool for the postmodification of polymers and the discovery of other new polymers. Enrichment of the SuFEx universe's access to new substances and materials also relies on the discovery of more SuFEx-able connectors. For instance, multidimensional S(vi)–F functionalization has been achieved using SOF₄ as the connector.¹¹  Despite considerable efforts to understand this remarkable SuFEx catalysis, so fast and so universal, we think it best for now to leave this puzzle open. The reactivity is, in its own way, as extraordinary as that of CuAAC. SuFEx ligations, like CuAAC ligations, have arrived at the scene walking almost perfectly from the start. So well in fact that they walk everywhere and are ready to use as is. The understanding of both CuAAC and SuFEx catalysis will probably be sought for many thoughtful years to come, but meanwhile, it seems urgent to get on with what they can already do that is unique.

J.D. is grateful to the One Hundred Talents Program supported by SIOC, CAS. J.D. is financially supported by the National Natural Science Foundation of China (NSFC 21672240), Strategic Priority Research Program of the Chinese Academy of Sciences (XDB200203), Science and Technology Commission of Shanghai Municipality (18JC1415500), Shanghai Pujiang Program (16PJ1410800), and Key Research Program of Frontier Sciences, CAS, Grant No. QYZDB-SSWSLH-028.