William D. G.
Brittain
* and
Steven L.
Cobb
*
Department of Chemistry, Durham University, South Road, Durham, DH1 3LE, UK. E-mail: william.d.brittain@durham.ac.uk; s.l.cobb@durham.ac.uk
First published on 9th January 2019
Phenols are extremely valuable building blocks in the areas of pharmaceuticals, natural products, materials and catalysts. In order to carry out modifications on phenols, the phenolic oxygen is routinely protected to prevent unwanted side reactions. Presently many of the protecting groups available can require harsh conditions, specialist equipment, expensive or air/moisture-sensitive reagents to install and remove. Here we introduce the use of the tetrafluoropyridyl (TFP) group as a general protecting group for phenols. TFP can be installed in one step with no sensitivity to water or air, and it is stable under a range of commonly employed reaction conditions including acid and base. The TFP protecting group is readily cleaved under mild conditions with quantitative conversion to the parent phenol, observed in many cases in less than 1 hour.
As highlighted, current phenol protecting groups are often less than ideal in certain circumstances, and thus there is a need to develop new strategies to meet these challenges.
To address this need, we envisaged that one reaction class which could be exploited to develop a general phenol protecting group was SNAr (nucleophilic aromatic substitution).12 In the literature, phenols have been shown to react in SNAr reactions with a range of electron deficient aryl halides.13 Within these reported reactions, SNAr processes between phenols and perfluoroaromatics/perfluoroheteroaromatics have been observed to be facile and rapid. For example, pentafluoropyridine (PFP) has been observed to undergo rapid SNAr with phenolic compounds.14 In addition, our own group has reported the use of PFP as a tagging reagent in peptide chemistry. We observed that PFP could undergo SNAr reactions with peptides containing nucleophilic side-chains, including tyrosine.15 We have also developed a range of unnatural amino acids using the same chemistry.16
The reaction between PFP and phenols is highly regioselective, with the 4-position of the pyridine ring being the most susceptible to substitution.17 Simply mixing a 1:1 ratio of phenol and PFP in the presence of base will give the tetrafluoropyridyl (TFP) bi-aryl ether at room temperature with no sensitivity to water or air.18
Previous studies by Vlasov et al.19 and Aksenov et al.20 have shown that the addition of potassium fluoride (KF) to a reaction between a phenol and PFP leads to a complex mixture of products. They rationalised this observation by proposing that the fluoride ion can cleave the pyridine ether bond, thus regenerating a perfluoroheteroaromatic species which can undergo further SNAr reactions to give a complex mixture of products (Scheme 2).
Scheme 2 Previous study by Vlasov et al.; proposed mechanism for the reaction of pentafluoropyridine (PFP) with pentafluorophenol in the presence of potassium fluoride. |
This led us to postulate that if it were possible to intercept the regenerated PFP after exposure to KF, then the tetrafluoropyridyl (TFP) group could act as a readily-cleavable protecting group for phenols. Herein, we explore this idea in detail and demonstrate that TFP ethers can be installed across a wide range of phenolic compounds and cleaved under mild conditions utilising a combination of fluoride salts and methyl thioglycolate.
Entry | Conditions | Time/h | Conversiona/% |
---|---|---|---|
a Conversion was determined by 1H NMR analysis of the reaction mixture (see ESI). All reactions were carried out in a water bath set to 50 °C. | |||
1 | KI (2 equiv.) | 24 | — |
2 | KBr (2 equiv.) | 24 | — |
3 | KCl (2 equiv.) | 24 | — |
4 | KF (2 equiv.) | 24 | <1 |
5 | KI (2 equiv.), 18-C-6 (3 equiv.) | 24 | <1 |
6 | KF (2 equiv.), 18-C-6 (3 equiv.) | 24 | 1 |
7 | NaI (2 equiv.), 18-C-6 (3 equiv.) | 24 | <1 |
8 | KF (2 equiv.), 18-C-6 (3 equiv.), methyl thioglycolate (10 equiv.) | 1 | 78 |
9 | KF (2 equiv.), methyl thioglycolate (10 equiv.) | 1 | 61 |
10 | KF (1 equiv.), 18-C-6 (1 equiv.), methyl thioglycolate (2 equiv.) | 1 | 57 |
11 | Methyl thioglycolate (10 equiv.) | 24 | <1 |
12 | Methyl thioglycolate (10 equiv.). K2CO3 (2 equiv.) | 24 | 90 |
We also tested the thiol on its own, but only a trace conversion to the phenol 1a was observed after 24 h (Table 1, entry 11). Finally, we decided to try adding base and thiol to the TFP ether; this also led to deprotection with the reaction reaching 90% conversion after 24 h (Table 1, entry 12). Due to the rapid nature of deprotection, we decided to continue the study using the conditions from Table 1, entry 8, but it should be noted that deprotection will occur (but at a slower rate) with fewer equivalents of KF, 18-C-6 and methyl thioglycolate or by the application of methyl thioglycolate in the presence of potassium carbonate.
With the selected deprotection conditions in hand, we then progressed to study the stability of the bi-aryl tetrafluoropyridine moiety. 2a was dissolved in chloroform-d (non-polar), methanol-d4 (polar, protic) or acetonitrile-d3 (polar, aprotic) and exposed to a range of common synthetic reagents at room temperature for 24 h (see ESI† for details). Under all of the conditions tested, compound 2a was seen to be stable in chloroform-d (Table 2, entries 1–11). Notably, both acid and base were tolerated (Table 2, entries 1 and 5). In methanol-d4 the TFP ether was seen to be unstable in the presence of base (Table 2, entries 3, 4 and 12). It has previously been reported that perfluoroaromatic ethers are unstable in the presence of methoxide and thus it was not surprising that the TFP-ether was also unstable under such conditions.21 In acetonitrile-d3 the TFP ether was unstable in the presence of strong base (Table 2, entries 4 and 5). Across all the other conditions tested, the TFP ether was observed to be stable.
Entry | Conditionsa | Stability (CDCl3) | Stability (MeOD) | Stability (CD3CN) |
---|---|---|---|---|
a All reactions were carried out in 0.7 mL of solvent with 10 mg of the tetrafluoropyridyl ether 2a. b 1H NMR resonances were seen to shift in the presence of TEA. Upon concentration under reduced pressure and resuspension in chloroform-d, the resonances were observed to shift back to their original positions. | ||||
1 | 0.1 mL TFA | Stable | Stable | Stable |
2 | 0.1 mL 12 M HCl | Stable | Stable | — |
3 | 20 mg K2CO3 | Stable | Unstable | Stable |
4 | 20 mg tBuOK | Stable | Unstable | Unstable |
5 | 20 mg NaH | Stable | — | Unstable |
6 | 0.1 mL TEAb | Stable | Stable | Stable |
7 | 0.1 mL DIPEA | Stable | Stable | Stable |
8 | 0.1 mL SOCl2 | Stable | Stable | Stable |
9 | 20 mg NaBH4 | Stable | Stable | Stable |
10 | 20 mg I2 | Stable | Stable | Stable |
11 | 20 mg DCC | Stable | Stable | Stable |
12 | 20 mg NaOH | — | Unstable | — |
Next we tested the TFP protecting group in the presence of other nucleophiles in a basic environment. It has been previously reported17 that PFP can react sequentially with multiple nucleophiles and we wished to confirm that the fluoropyridine group could still be cleaved even if additional fluorine substitutions did occur. Exposing 2a to forcing reaction conditions of 2.1 equivalents of piperidine, and 2.1 equivalents of potassium carbonate in refluxing acetonitrile for 24 h resulted in recovery of compound 3 (Scheme 4).
Scheme 4 Exposure of TFP aryl ether 2a to nucleophiles under forcing conditions followed by deprotection to the parent phenol. |
Exposure of 3 to the developed deprotection conditions showed conversion to the free phenol, but the rate of conversion was slower than that for 2a (Scheme 4). This reaction demonstrated that it was still possible to cleave the fluoropyridine group even if it has undergone unwanted SNAr reactions.
In order to probe the utility of the protecting group, we synthesised a range of TFP bi-aryl ethers in good to excellent yields (70–99%) (Fig. 1, 2a–2n). The only substrate tested which was not compatible with TFP-ether formation was di-tert-butyl 2f presumably due to its steric bulk. All of the TFP bi-aryl ethers prepared were susceptible to the selected deprotection conditions with complete quantitative conversion to the phenol, observed in some cases in less than 1 h (2g, 2l, 2n see ESI, pages S-107/S-112/S-114†).
Fig. 1 Scope of installation (yields shown in red) and subsequent cleavage of TFP aryl ethers (conversion to the phenol shown in blue). (a) Isolated yield following purification. (b) Conversion to the phenol following 1 h under deprotection conditions, measured by integration of 1H NMR spectra (see ESI†). (c) TFP ether formation carried out in DMF due to solubility. (d) Reaction time = 96 h. (e) TFP ether formation carried out in a mixture of MeCN and DMF. |
It should also be noted that di-ortho-substituted TFP bi-aryl ethers took longer to deprotect than mono substituted varients. For example, di-methyl 2d had only reached 18% conversion to its parent phenol after 1 h. Presumably, this is due to the increased steric bulk surrounding the aryl-ether bond.
We were able to demonstrate the flexibility of the TFP group in protecting a range of phenols (Fig. 1). The TFP moiety was also installed and cleanly removed from hydroxypyridine, showing quantitative conversion after 1 h (Fig. 1, 2q). Tyrosine was another phenolic species we wished to test, due to its utility in peptide chemistry, and as part of our ongoing research program into the synthesis of new amino acids.22 The TFP group was installed to give ether 2o in 75% yield, and after 1 h under deprotection conditions the reaction had reached 75% conversion.
Modification of 1,1′-bi-2-naphthol (BINOL) through SNAr with PFP has been previously disclosed by Koltunov et al.23 BINOL and substituted BINOLs have been utilised extensively in asymmetric catalysis.24 Exposing BINOL to 2 equivalents of PFP in the presence of K2CO3 gave the bis-TFP aryl ether 2r in 69% yield. This product was susceptible to the cleavage conditions, but the rate of deprotection was observed to be slow, with 96 h required to reach quantitative conversion back to BINOL.
To demonstrate the TFP group's applicability to steroid chemistry, we synthesised the TFP protected estrone derivative 2s in 87% yield. Exposure of the steroid derivative 2s to the deprotection conditions cleanly removed the TFP ether, with 83% conversion to estrone observed after 1 h. It should be noted that other perfluoroaromatics have been previously employed as protecting groups in steroid synthesis, but these required the use of sodium methoxide in DMF to recover the parent steroids.21
To confirm that the phenol was being successfully liberaterated using our developed conditions NMR spiking experiments were undertaken on a selection of the synthesised substrates. In all cases, spiking the reaction sample with reference compounds clearly demonstrated that the phenols were being successfully generated (see ESI pages S-123 to S-129†).
We also probed the fate of the methyl thioglycolate using mass spectrometry. Analysing the deprotection reaction mixture of 2a with LCMS, we observed a species with a mass equating to a pentafluoropyridine (PFP) with 2 positions substituted by methyl thioglycolate (see ESI†). This supports our hypothesis in terms of a mechanism for the deprotection in that after the fluoride regenerates the free PFP it is intercepted in situ by the thiol (e.g. 2× methyl thioglycolate) to form a species which is unable to undergo an SNAr reaction with the released phenol.
To further validate the developed conditions we undertook the deprotection of 2i using non-deuterated solvents on a larger scale. After 2 h, we were able to successfully isolate 4-nitrophenol from the reaction mixture in 98% isolated yield (see ESI page S-15†).
During the synthesis of the TFP aryl ethers (Fig. 1), we observed that may of the examples were highly stable crystalline solids. Therefore, we were able to obtain crystal structures for compounds 2r and 2i (Fig. 1) which confirmed the regiochemistry of the products.25
Finally, to further demonstrate the synthetic scope of the methodology developed we carried out a range of chemical transformations on our synthesised TFP ethers (Scheme 5). The free amine 4 was prepared by exposing the tyrosine TFP ether 2o to standard Boc-deprotection conditions. 4 was allowed to react with Boc-Ala-OH using common amide bond-forming reagents (e.g. PyBOP) to afford the dipeptide 5 in 94% yield (Scheme 5a). No cleavage of the TFP moiety was observed during the process.
Using Suzuki–Miyaura cross-coupling conditions, the iodo-TFP ether 2k reacted effectively with phenylboronic acid to give compound 2n in 86% yield (Scheme 5b). 2n had previously been found to be readily susceptible to the developed deprotection conditions (Fig. 1). We found that the alkyne-appended TFP ether 2m could undergo copper-catalysed azide–alkyne cycloaddition (CuAAC) (click chemistry) to give triazole 6 in excellent 93% yield (Scheme 5c).
Exposure of 6 to the developed deprotection conditions afforded the phenolic triazole 1b with quantitative conversion observed by 1H NMR spectroscopy (see ESI page S-120†). We carried out a borohydride-mediated reduction of the TFP-estrone derivative 2s (Scheme 5d). The reaction between 2s and sodium borohydride proceeded smoothly to give the secondary alcohol 7 in 90% yield. We deliberately allowed the reaction to stir for an extended period of 24 h to make sure that the TFP group was stable; no cleavage of the ether was observed under the reaction conditions.
Footnote |
† Electronic supplementary information (ESI) available. CCDC 1856218–1856219. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c8ob02899k |
This journal is © The Royal Society of Chemistry 2019 |