Hydroalkoxylation of alkynes by a nitroxyl containing alcohol, 4-hydroxy-2,2,6,6-tetramethylpiperidin-1-oxyl: synthesis of spin-labeled enol ethers

A series of spin-labeled enol ethers has been synthesized in 53–67% yields by superbase-catalyzed (KOH/DMSO suspension as a catalyst) hydroalkoxylation of alkynes (acetylene, phenylacetylene, 4-tert -butylphenylacetylene and 3-ethynylpyridine) by 4-hydroxy-2,2,6,6-tetramethylpiperidin-1-oxyl (4-hydroxy-TEMPO) under mild conditions (70–80 °C, 1.5–2 h). With unsubstituted acetylene, the hydroalkoxylation readily occurs at atmospheric pressure, yield of the corresponding vinyl ether being 53%.


Introduction
2][3] Among such species, 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO) and its congeners are of special interest.They have found wide applications as reagents and catalysts in organic synthesis, 3 antioxidants, 4 materials for radical batteries, 5,6 magneto-active materials, 7 agents for DNP-NMR, 8,9 as well as radiation protective agents 10 and polymerization catalysts. 11urthermore, 4-hydroxy-TEMPO exhibits an antihypertensive effect. 12pecial attention has been devoted to the use of TEMPO derivatives as spin probes in various fields of biochemistry and biophysics. 13,14The incorporation of TEMPO labels into such vital molecules as sugars, lipids, proteins, DNA provides valuable information (by EPR spectroscopy) about key metabolic processes and drug docking and targeting.6][17] Clearly, for this purpose the TEMPO derivatives should contain reactive linkers, ensuring site-specific binding of the spin label with the target molecule under mild conditions.
For these considerations, the design of new TEMPO-based spin labels bearing appropriate chemically active linkers is a synthetic challenge.An attractive route to functionalization of the TEMPO skeleton might be the introduction into its molecule of the enol ether (vinyloxy) moiety that, thanks to its rich chemistry, [18][19][20][21] can be easily linked with the desired molecules such as sugars, amino acids, proteins, DNA, etc. Furthermore, the presence of easily polymerizable vinyloxy groups 22,23 in the molecules of paramagnetic compounds considerably extends their possible application in advanced technologies.
Therefore, this work focuses on the development of a general and efficient route to spinlabeled vinyl ethers by direct vinylation of the available 4-hydroxy-TEMPO with alkynes in the superbasic KOH/DMSO system under readily accessible conditions, viz. at atmospheric pressure and mild temperatures.
It is relevant to note that some KOH/DMSO-catalyzed vinylations, e.g.NH-and CHfunctionalization of pyrroles and indoles with acetylenes, [43][44][45][46] involve single electron transfer steps (the expected radicals were trapped as spin-adducts).Therefore, the KOH/DMSO-catalyzed vinylation of a free radical such as 4-hydroxy-TEMPO with acetylenes could be affected by involvement of this radical in the vinylation process.In view of these data, and the earlier result of predominant reduction of 4-hydroxy-TEMPO during its vinylation by acetylene in the KOH/DMSO system, it remains unclear whether efficient base-catalyzed hydroalkoxylation of nitroxyl-containing alcohols with alkynes is possible.This issue is of fundamental importance to the question of the mechanism of nucleophilic addition of alkoxide to the C≡C triple bond.

Results and Discussion
To develop an efficient synthesis of 4-vinyloxy-TEMPO, the reaction of 4-hydroxy-TEMPO (1)  with acetylene in the KОН/DMSO system was studied under atmospheric pressure (acetylene flow).The 1/KOH molar ratio, reaction temperature and time were optimization parameters.The GLC monitoring of the reaction revealed that, along with the hoped-for vinylation to afford vinyl ether 2, reduction leading to compounds 3 and 4 also took place (Table 1).As may be seen from Table 1, the selectivity of formation of the vinyl ether radical 2 significantly depends on the reaction temperature and time.Under the mildest conditions (60 °C), the rate of alcohol 1 vinylation is insignificant: after 1 h only traces of vinyl ether 2 are present in the reaction mixture (entry 1).At 80 °С, and with 75 mol% of КОН, increase of the reaction time (from 0.75 to 3 h) results in complete conversion of the starting alcohol 1, while the content of vinyl ether 2 in the crude product decreases from 62.2% to 51.6% due to its reduction to products 3 and 4 (entries 3-5, Table 1).The elevated temperature also facilitates the reduction processes.A longer reaction time (4 h, other conditions being the same) diminishes the content of 2 in the crude product from 46.2 to 2.8% (entries 6, 7, Table 1).In this case, the major product becomes the vinyl ether 3 (entry 7, Table 1), isolated in 59% yield.Eventually, the best result (65.4% content of 2 in the crude product) is achieved at 1 : KOH molar ratio of 2:1, 80 °С and the reaction time of 1.5 h (entry 2, Table 1).The vinyl ether 2 is isolated in 53% yield by column chromatography (basic Al 2 O 3 ) along with unreacted alcohol 1 (71% conversion).
The reduction of the nitroxyl moiety of 1 and/or 2 during the vinylation under the above conditions is nontrivial since the nitroxyl radicals are known to be quite stable in basic media.In this case the reducing agent may be either DMSO (being further oxidized to dimethylsulfone) or acetaldehyde (the product of side hydration of acetylene).
To suppress the reduction of alcohol 1 and vinyl ether 2, leading to the formation of side products 3, 4, the vinylation of 1 was conducted in an autoclave under elevated pressure of acetylene (initial loading pressure at room temperature 10-12 atm, maximum pressure at the reaction temperature 9 atm).At 70 °С with 100 mol% KОН the vinylation proceeds rapidly (about 20 min for 10 mmol loading of 1) so that the undesirable reduction of the nitroxyl moiety of 1 and 2 does not take place, and hence the target vinyl ether 2 is a major product; its isolated yield reaches 67%.
Further, aiming to develop an efficient synthesis of other spin-labeled enol ethers (hitherto unknown), we extended the reaction studied to aryl-and hetarylacetylenes.The experiments showed that, under the modified conditions, in a suspension of KОН (50 mol%) in DMSO, alcohol 1 regioselectively reacts with phenylacetylene, 4-tert-butylphenylacetylene and 3ethynylpyridine (70 °C, 2 h) to give vinyl ethers 5a-c as mixtures of E-and Z-isomers in 58-64% total yields (Table 2).Because 1 H NMR analysis of the products 5 is impossible because of their paramagnetic nature, to determine the Z/E-isomeric ratio of alkenes 5a-c their crude mixtures were reduced to the corresponding nonparamagnetic hydroxylamines with hydrazine hydrate under mild conditions (ethanol, 40-45 °C, 2 h), in which the Z/E-isomerization is assumed to be improbable.Thus, the ratio of the formed diamagnetic hydroxylamines, according to 1 H NMR analysis, closely corresponds to the ratio of parent vinyl ethers 5a-c.As can be seen from Table 2, Z-isomers always predominate over the E-adducts as is typical for nucleophilic addition to alkynes. 41c Isolated yields after chromatographic purification.
The fractional crystallization of crude products 5a,b from hexane allows their Z-isomers to be isolated in pure form.Interestingly, the heating of (Z)-5b in the KOH (50 mol%) / DMSO suspension (70 °C, 2 h) does not lead to its isomerization into the thermodynamically preferable E-isomer.This result clearly suggests that the E-isomers are formed during the nucleophilic addition of alcohol 1 to the alkynes, and their formation is not due to isomerization of the kinetically preferable Z-isomers.
The synthesized paramagnetic vinyl ethers 2 and 5a-c were characterized by X-ray crystallography (for 2, (Z)-5a and (Z)-5b), EPR, EI-MS and FT-IR techniques.In the EI-MS spectra of adducts 2 and 5a-c, the intense peaks of the molecular ions, [М] +• , are presented.In the FT-IR spectra of these compounds, the stretching vibration of the N-O bonds appears as a medium intensity band at 1364-1365 cm −1 .
Compounds 2, (Z)-5a and (Z)-5b crystallized from hexane in the monoclinic P21/c space group with four molecules per unit cell.Their molecular structures are shown in Figures 1-3.8][49] The N-O bonds are in an equatorial position of the piperidine ring.The N(1)-O(2) distances are nearly equal in length (1.286-1.2918][49]     The EPR spectra of all nitroxyl radicals in various solvents (CHCl 3 , benzene, DMSO) reveal a stable signal of three intense equidistant lines (see typical signal in Figure 4).The isotropic hyperfine coupling constants for the 14 N of all nitroxyl radicals are 15.4-15.8G and are almost independent of the functionalization of the vinyl fragment and solvent (Table 3).Clearly, the modification of the vinyl ethers 2 and 5a-b at hydroxyl oxygen atom does not change the geometry of the six-membered ring and the radical center on the oxygen atom at which the electron spin is localized.Stability of the electron-spin values, depending on structure of the substituents in the nitroxyl radicals, is one of the essential conditions for the successful application of the radicals as spin probes. 2,13,14 Expectedly, when dissolved in polar solvent like DMSO, the linewidth becomes significantly narrower (Table 3, Figure 5).In a DMSO solution (at 288 K), anisotropy of the EPR spectra is observed due to orientation of the radical relative to external magnetic field.Only in the polar solvent is resolution of each line of the triplet detected from coupling to 1 H nuclei.

Conclusions
In summary, a general and efficient methodology for the synthesis of enol ethers bearing paramagnetic TEMPO moieties has been developed using the KOH/DMSO-catalyzed regioselective hydroalkoxylation of diverse alkynes by available 4-hydroxy-TEMPO.This reaction readily proceeds under environmentally benign transition metal free conditions at atmospheric pressure and mild heating (up to 80 o C).The methodology represents a facile and concise route to prospective spin labels containing reactive enol ether groups.

Experimental Section
General.The FT-IR spectra were recorded on a Bruker Vertex 70 spectrometer.NMR spectra were recorded on Bruker DPX-400 spectrometer (400.1 MHz for 1 H, 100.6 MHz for 13 C) at ambient temperature for CDCl 3 solutions.Chemical shifts were reported in  (ppm) relative to CDCl 3 ( 1 H, 13 C) as internal standard.Prior to recording of NMR spectra, the paramagnetic compounds were reduced with ~5 equivalets of hydrazine hydrate (EtOH, 40-45 o C, 2 h).The EI-MS spectra were on an Agilent 5975C MSD instrument.Samples were introduced into the source by means of the gas chromatograph GC-6890N (Agilent Technologies) through capillary column HP-5MS (30 m × 0.25 mm × 0.25 mm), the helium being the gas-carrier.The source temperature was approximately 150 °C.The elemental analyses were carried out on a Flash EA 1112 analyzer.Melting points were determined on a Kofler micro-hot stage.The high-pressure experiments were carried out in a stainless steel, high-pressure batch reactor (Parr 4572, Parr Instrument Co.) equipped with an electrical heating jacket, a gas inlet, a mechanical stirrer and 4848A temperature controller.Basic Al 2 O 3 was used for column chromatography and SiO 2 plates for TLC (50% Et 2 O/hexane).Visualization was performed with iodine vapor.Alcohol 1, KOH•0.5H 2 O, DMSO (up to 0.5% of water) and all other reagents are commercially available and were used without further purification.

Crystallography
The single crystals of 2, 5a and 5b were obtained by slow evaporation of solutions in hexane.The data were collected on a Bruker D8 Venture diffractometer by using graphite monochromatic MoKα radiation (λ = 0.71073 Å).Structures were solved by direct method and were refined against the least-squares methods on F 2 with the SHELXL-97 package, 50 incorporated in SHELXTL-PC V6.14.8.All non-hydrogen atoms were refined anisotropically.CCDC 990980 (2), 990592 (5a) and 1054445 (5b) contain the supplementary crystallographic data for this paper.These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif.  3 .24498 reflections measured (2.32° < 2θ < 30.04°), 3440 of which were independent (R int = 4.69%) and 2760 (80.23%) were greater than 2σ(F 2 ).The final anisotropic full-matrix least-squares refinement on F 2 with 131 variables converged at R1 = 4.28%, for the observed data and wR2 = 11.49% for all data.The goodness-of-fit was 1.088.

Table 1 .
Vinylation of 4-hydroxy-TEMPO (1) with acetylene: effect of the conditions on the product yields and ratio a b Relative to 1.

Table 3 .
EPR characteristics of 2 and 5a-c in different solvents (288 K) Compound Solvent g-factor Constant A N , G Width ΔH, G