Metal-free synthesis of phosphinoylchroman-4-ones via phosphinoylation/cyclization cascade mediated by K2S2O8

A variety of chroman-4-ones bearing phosphine oxide motifs were conveniently synthesized from readily available diphenylphosphine oxide and alkenyl aldehydes via a metal-free tandem phosphinoylation/cyclization protocol. The reaction utilizes K2S2O8 as oxidant and proceeds in DMSO-H2O at environmentally benign conditions with broad substrate scope and afforded the title compounds in moderate yields.

It is well known that organophosphorus compounds, with their medicinal, biological or specific material-relating properties, have found wide applications in pharmaceutical chemistry, biochemistry and material science [21][22][23]. There are also excellent ligands for many metals and have been used in catalyzing a huge of significant organic reactions [21][22][23]. Due to the importance of the chroman-4-one scaffolds and the organophosphorus compounds, the development of concise and 3 efficient approaches for the synthesis of the chroman-4-one derivatives containing phosphorus functionality frameworks [24,25] that combine both characteristics together may find useful application. By far, there are only few ways to prepare such compounds. For example, in 2008 Rovis group [24] reported an intramolecular Stetter reaction of alkenyl aldehydes to synthesize a series of phosphine oxide and phosphonate functionalized chroman-4-ones, but preparation of the substrates involved Rh-catalyzed hydrophosphinylation of a protected functional alkyne and subsequent deprotection with Hg(O2CCF3)2, which is not environmentally benign (Scheme 1a). Besides, in 2016 Li's group [25] reported a silver catalyzed straightforward approach to synthesize phosphonate chroman-4-ones via a phosphoryl-radical-initiated cascade cyclization of 2-(allyloxy)arylaldehydes using K2S2O8 as an oxidant, however, diphenylphosphine oxide (DPPO) was not suitable for the transformation (Scheme 1b). So the development of metal-free and greener methods to approach chroman-4-ones bearing phosphine oxide moiety is still highly desirable. Herein, we present a transition-metal-free radical cascade cyclization to access the above chroman-4-ones in one pot under environmentally benign conditions (Scheme 1c).

Results and Discussion
Motivated by the desire to develop a metal-free and environmentally benign protocol for the construction of phosphine oxide functionalized chroman-4-ones, we focused on the cascade cyclization employing 2-(allyloxy)benzaldehyde 1a and diphenylphosphine oxide (DPPO) 2a as model substrates with K2S2O8 as an oxidant, which is cheap, readily available, and versatile oxidant. On the basis of the literature reports [26,27] and our continuing interest in green chemistry [28,29], we set the temperature as 70 °C based on the fact K2S2O8 thermally decomposes to form sulfate radical (SO4 •-) [26,27], which may react with the substrates to finish such a cascade cyclization. To our delight, the anticipated product 3aa was obtained in 42% yield in DMSO/H2O (4:1) system in one pot (Table 1, entry 1). The structure of 3aa 5 was unambiguously confirmed by X-ray diffraction analysis of a single crystal ( Figure   2) and NMR spectroscopy (see ESI) [30]. Then the increase of the amount of K2S2O8 to 3 equiv. resulted in the improvement of the yield of 3aa to 52% (    With the optimal reaction conditions in hand (Table 1, entry 16), we next sought to explore the scope and generality of this protocol using various 2-(allyloxy)arylaldehydes 1 with diphenylphosphine oxide (DPPO) 2a. As shown in Table 2, substrates 1 with a range of functional groups, such as electron-donating 7 groups Me- (1b and 1c), t-Bu- (1d and 1e), electron-withdrawing groups Cl-(1f), Br-(1g), F-(1h) were well tolerated in this transformation, providing the desired products (3aa-3ha) in 48-62% yields. Furthermore, the transformation can also proceed when there is a naphthyl ring in the substrate (1i) giving the desired product 3ia in 45% yields. Notably, when the substrate is 2-allylbenzaldehyde 1j, this protocol is also compatible affording the product 3ja as an indanone derivative. The indanone derivatives are also privileged structural motifs found in numerous natural products and pharmaceuticals with extraordinary biological and pharmaceutical activities [31,32]. However, no desired product (3ka) was obtained when there was an NO2-group in the substrate (1k).  (2e and 2f) were also well tolerated under these reaction conditions. Gratifyingly, diphenylphosphine oxides bearing F-groups (2g) reacted smoothly to furnish the anticipated products 3ag in 58% yield. Furthermore, 1-Naphthyl-DPPO (2h) was suitable for this transformation, to give the product 3ah in 50% yield. The reaction between diethyl phosphonate 2j and 1a proceeded less efficiently under the conditions and low yield of 3aj was obtained. Dimethylphosphine oxide 2k did not participate in the reaction, perhaps because of its high oxidation potential and poor ability to undergo tautomerization [33]. To demonstrate the practicability of this methodology, a gram-scale experiment was then performed, employing 1b and 2a as substrates under optimized conditions (Scheme 4). The reaction afforded the desired product 3ba in a good yield of 56%, and the structure was also confirmed by X-ray diffraction (see ESI) [30].
To gain an insight into the reaction mechanism, some control experiments were carried out (Scheme 5). When the reaction was conducted in the presence of radical scavengers such as 2,2,6,6-tetramethyl-1-piperidinyloxyl (TEMPO) and butylated hydroxytoluene (BHT), the reactions were completely shut down. Also, we successfully separated a little by-product 4 which was identified by NMR spectroscopy. These experiments clearly support the phosphorus centered radical reaction pathway. It has been reported that phosphorus centered radicals can be generated from phosphine oxides in the presence of potassium persulfate [34][35][36].
Based on literature precedent [26,27,[34][35][36][37][38] and preliminary mechanistic experiments, a plausible mechanism was proposed in Scheme 5, which should be different from the predominant mechanism with Ag-catalyzed radical cascade for the 10 preparation of phosphonate functionalized chroman-4-ones [25]. Initially, K2S2O8 thermally decomposes to form sulfate radical anion (SO4 •-) [26,27], which reacts with the diphenylphosphine oxide (DPPO) 2 to give a phosphorus centered radical I [34][35][36]. Then I undergoes an intermolecular addition to the C-C double bond of 1, leading to the formation of a new radical intermediate II. Sequential radical II attack on the aldehyde moiety, and an oxygen radical III is generated and then a formal 1,2-H shift occurred to deliver the benzyl radical IV [37,38]. Finally, the sulfate radical anion (SO4 •-) abstract a hydrogen from the benzyl radical IV to give the final products 3 [37,38].

Conclusion
In conclusion, we have developed an environmentally benign and practical protocol for the synthesis of phosphonate chroman-4-ones from 2-(allyloxy)benzaldehydes and diphenylphosphine oxide under metal-free conditions. The efficiency of this protocol was further enhanced by using K2S2O8 as oxidant. This transformation proceeded through a tandem C-P and C-C bond formation with easy-handing and good functional-group compatibility.

Supporting Information
Supporting Information File 1: Experimental procedures, spectroscopic and X-ray data and copies of NMR spectra.

Funding
We are grateful for the financial supports from Science Foundation for Distinguished Scholars of Dongguan University of Technology (No. GC200906-10) and Postdoctoral Foundation of Dongguan University of Technology (No. 196100040006).