An efficient method for the preparation of silyl esters of diphosphoric, phosphoric, and phosphorous acid

An efficient method for the preparation of silyl esters of diphosphoric, phosphoric, and phosphorous acid

Polyhedron 70 (2014) 133–137 Contents lists available at ScienceDirect Polyhedron journal homepage: www.elsevier.com/locate/poly An efficient method...

398KB Sizes 209 Downloads 242 Views

Polyhedron 70 (2014) 133–137

Contents lists available at ScienceDirect

Polyhedron journal homepage: www.elsevier.com/locate/poly

An efficient method for the preparation of silyl esters of diphosphoric, phosphoric, and phosphorous acid Ludger A. Wessjohann ⇑, Marco A. Dessoy Bioorganic Chemistry, Leibniz Institute of Plant Biochemistry, Weinberg 3, D-06120 Halle (Saale), Germany

a r t i c l e

i n f o

Article history: Received 30 September 2013 Accepted 22 December 2013 Available online 31 December 2013 Keywords: Silylations Diphosphates Protective groups Hydrolysis Phosphate esters Triphasic system

a b s t r a c t Tetrakis(trialkylsilyl) diphosphate (alkyl = Me, Et, iPr, tBu) can be obtained in quantitative yield by reacting commercial disodium dihydrogen diphosphate with the respective trialkyl chlorosilane in a triphasic system with formamide. The alkylsilane residues of the diphosphate silyl esters can be either partially or completely hydrolyzed without concurrent cleavage of the P–O–P bond of the diphosphate moiety. The method can be expanded to efficiently produce other persilylated or partially silylated phosphates and phosphites. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction The diphosphate moiety comprehends a critical structural element in a series of important biological metabolites and cofactors (i.e. diphosphorylated terpenols, sugars, nucleotides, and nicotineamide-based dinucleotides) which play central roles in both primary and secondary metabolisms of all living organisms. Silylated derivatives of diphosphoric acid are potential precursors for the chemical synthesis of these biologically relevant molecules. Tetraalkyl diphosphates (alkyl = Et, nPr, iPr, nBu) were shown to be toxic to some organisms [1]. E.g., tetraethyl diphosphate was used as a pesticide [2], as it expresses powerful anticholinesterase activity with actions similar to eserine and neostigmine [3]. Tetrabenzyl diphosphate has been used for the phosphorylation of inositol derivatives (e.g. D-myo-inositol-1,4,5-trisphosphate a cellular second messenger) [4–6] and in the synthesis of ( )-5-enolpyruvylshikimate-3-phosphate (a principle metabolite in the shikimic acid pathway) [7]. Trimethylsilyl (TMS) esters of a series of oxyanions of main group elements have been synthesized and spectroscopically studied in the past [8]. However, the effective silylation of diphosphoric acid or its salts on preparative scale to give pure product has not been reported until now. Formation has been observed in different reaction systems. Thus, tetrakis(trimethylsilyl) diphosphate (2a) is formed as a byproduct in the preparation of polyphosphoric acid trimethylsilyl ester (PPSE). PPSE is obtained from phosphorus pentoxide upon reaction with excess hexame⇑ Corresponding author. Tel.: +49 345 5582 1301; fax: +49 345 5582 1309. E-mail address: [email protected] (L.A. Wessjohann). 0277-5387/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.poly.2013.12.024

thyldisiloxane. In an early account, Mileshkevich and Karlin reported the isolation of 2a in up to 18% from crude PPSE [9]. The authors, though, did not provide a definitive proof of structure and purity of the isolated product at that time. Later, Yamamoto and Watanabe disclosed the composition of crude PPSE by means of 31P NMR spectroscopy [10]. According to the latter authors, the share of 2a in the crude product mixture is somewhat influenced by the reaction conditions, but in no case it has been found to be higher than 7%. Similarly, the silyl ester 2a also forms in ca. 6% yield upon treatment of phosphorus pentoxide with hexamethyldisilazane [11]. The direct reaction of (nucleophilic) inorganic diphosphate and (electrocphilic) activated trialkylsilyl compounds, e.g. TMS-chloride, does not work without solvent mediation. All solvents commonly applied for such reactions either do not further the reaction, e.g. they are unable to dissolve the diphosphate, or they react with the activated TMS or with the product with its sensitive (activated) P–O–P or Si–O–P bonds. Furthermore, active strong acid or nucleophiles deteriorate the product, sometimes even catalytically. Isolation of the products thus was a major challenge not yet tackled. The stability of sterically hindered higher silyl esters is expected to be better, but they also promise less applicability at a much higher price. To the best of our knowledge, the only example of a purposeful preparation of a diphosphate silyl ester is given by Mawhinney [12]. Accordingly, both diphosphoric acid and its ammonium salt were converted into tetrakis(tert-butyldimethylsilyl) diphosphate (2d) with N-methyl-N-(tert-butyldimethylsilyl) trifluoroacetamide in N,N-dimethylformamide (DMF). The silyl es-

134

L.A. Wessjohann, M.A. Dessoy / Polyhedron 70 (2014) 133–137

ter 2d, however, was prepared for analytical purposes in no more than a few micro-grams and has not been isolated from the crude reaction mixture. Unfortunately, silyl protected diphosphates are also unavailable by an inverse approach starting from (electrophilic) diphosphoryl chloride and nucleophiles (like R3SiO ) because these cleave the P–O–P bond rather than the P–Cl bond [13,14]. The claim of Nifant’ev et al. to have synthesized 2a in 44% yield from bis(trimethylsilyl) phosphite (4a) via a Todd-Atherton-like reaction could not be verified (v.i.) [15]. In reality, the authors obtained tris(trimethylsilyl) phosphate (4b, d 31P NMR = 24.8 ppm) and have erroneously assigned the structure of this compound to 2a. Herein we wish to report the first efficient synthesis of a series of diphosphate trialkylsilyl esters (2a–2d) and improved access to silylphosphate and silylphosphite.

5 mmol), formamide (5 mL), and the corresponding trialkylsilyl chloride (20 mmol) is added. The ternary system is vigorously stirred at 55 °C during the time (time intervals for 2b–2d: 4, 10 and 5 h, respectively). Petrol ether (20 mL) is added and stirring is continued for ca. 5 min. The clear top layer is transferred into a second dry 50 mL Schlenk-tube and the solvent is evaporated under reduced pressure (ca. 12 mbar or lower, depending on the silylether formed) at 40 °C.

2.2.2.1. Tetrakis(triethylsilyl) diphosphate 2b. 3.05 g (96%) of a viscous, colorless liquid. 1H NMR (CDCl3, 400 MHz, d, ppm): 0.74 (q, 3 JHH = 7.8 Hz, 24H, CH2), 0.97 (t, 3JHH = 7.8 Hz, 36H, CH3). 13C NMR (CDCl3, 100 MHz, d, ppm): 5.14 (CH3), 6.31 (CH2). 31P NMR (CDCl3, 162 MHz, d, ppm): 30.59 (s). 29Si NMR (CDCl3, 99 MHz, d, ppm): 25.77 (bt, 2,4JSiP  3.6 Hz). HRMS: calculated for C24H61O7P2Si4 [M+H]+: 635.29641, found: 635.29729.

2. Experimental 2.1. Materials and methods Formamide (cat. No. 47670), chlortrimethylsilane (cat. No. 92361), chlorotriethylsilane, chlortriisoproylsilane, chlordimethyl-tert-butylsilane, and pyridine (cat. No. 82704) were purchased from Fluka. P.a. grade H2Na2O7P2 was purchased either from Fluka (cat. No. 71501) or Aldrich (cat. No. 34,073-1) [the amount of phosphate impurity correlates directly with the later amount of impurity of silylated phosphate]. Petrol ether (b.p. 40–60 °C, Roth) was used as received. 1H, 13C and 31P NMR spectra were recorded on a Varian Mercury 400 spectrometer, operating at 400, 100, and 162 MHz, respectively. 29Si NMR spectra were recorded on a Varian Inova 500 spectrometer, at an operational frequency of 99 MHz. Unless otherwise stated, the spectra were recorded in CDCl3. 1H, 13 C and 29Si NMR spectra were referenced to tetramethylsilane (TMS, d = 0 ppm, s = singlet, d = doublet, t = triplet, bt = broad triplet, q = quartet, hep = heptet). 31P NMR spectra were externally referenced to an 85% H3PO4 capillary (d = 0 ppm). High-resolution mass spectra were recorded with a Bruker BioApex 70e FT-ICR (Bruker Daltonics, USA) and low-resolution mass spectra (MSESI) were recorded with an API 150Ex (Applied Biosystems) mass spectrometer, equipped with a turbo ion spray source. 2.2. Synthesis of tetrakis(trialkylsilyl) diphosphate esters 2a–2d 2.2.1. Tetrakis(trimethylsilyl) diphosphate 2a A 250 mL dry flask was loaded with finely powdered dihydrogen disodium diphosphate (22.2 g, 0.1 mol), formamide (50 mL), and under vivid stirring by trimethylsilyl chloride (45.5 g, 0.44 mol). Refluxing TMSCl must be cooled for larger scale synthesis (caution: exothermic reaction with hydrogenchloride gas evolving!). After 1 h of stirring, petrol ether (200 mL) was added and stirring was continued for ca. 5 min. The clear top layer was transferred into a dry 250 mL Schlenk-tube and the solvent as well as excess TMSCl were evaporated off under reduced pressure (ca. 12 mbar) at 40 °C to give 46 g (100%) of the title compound as an colorless liquid. 1H NMR (CDCl3, 400 MHz, d, ppm): 0.24 (s, CH3). 13 C NMR (CDCl3, 100 MHz, d, ppm): 0.60 (s, CH3). 31P NMR (CDCl3, 162 MHz, d, ppm): 30.76 (s). 29Si NMR (CDCl3, 99 MHz, d, ppm): 24.6 (t, 2,4JSiP  2.8 Hz); HRMS: calculated for C12H37O7P2Si4 [M+H]+: 467.10861, found: 467.10860. 2.2.2. General procedure for the preparation of tetrakis(trialkylsilyl) diphosphate esters 2b–2d A 50 mL dry Schlenk tube flushed with dry nitrogen is loaded with finely powdered dihydrogen disodium diphosphate (1.11 g,

2.2.2.2. Tetrakis(triisopropylsilyl) diphosphate 2c. 3.8 g (95%) of a highly viscous, colorless liquid. 1H NMR (CDCl3, 400 MHz, d, ppm): 1.11 (d, 3JHH = 7.4 Hz, 72H, CH3), 1.23 (hep, 3JHH = 7.4 Hz, 12H, CH). 13C NMR (CDCl3, 100 MHz, d, ppm): 12.65 (CH), 17.12 (CH3). 31P NMR (CDCl3, 162 MHz, d, ppm): 32.81. 29Si NMR (CDCl3, 99 MHz, d, ppm): 20.91 (t, 2,4JSiP  10.0 Hz). HRMS: calculated for C36H85O7P2Si4 [M+H]+: 803.48421, found: 803.48613.

2.2.2.3. Tetrakis(tert-butyldimethylsilyl) diphosphate 2d. 3.01 g (94%) as a white, amorphous solid. 1H NMR (CDCl3, 400 MHz, d, ppm): 0.25 (s, 3H, CH3), 0.26 (s, 3H, CH3), 0.91 (s, 9H, C(CH3)3). 13C NMR (CDCl3, 100 MHz, d, ppm): 4.01 (d, 3JCP = 3.8 Hz, CH3), 17.98 (C(CH3)3), 25.27 (C(CH3)3). 31P NMR (CDCl3, 162 MHz, d, ppm): 30.02. 29Si NMR (CDCl3, 99 MHz, d, ppm): 25.98 (bt, 2,4JSiP  3.6 Hz). HRMS: calculated for C24H61O7P2Si4 [M+H]+: 635.29641, found: 635.29802.

2.3. General procedure for the preparation of 4a and 4b To a dry, nitrogen flushed 100 mL Schlenk-tube containing a solution of phosphorous acid (1.64 g, 20 mmol for the synthesis of 4a) [or potassium dihydrogen phosphate (2.72 g, 20 mmol for the synthesis of 4b)] in formamide (10 mL) trimethylsilyl chloride (5.4 g, 50 mmol) is added. The resulted binary (4a) or ternary system (4b) was vigorously stirred (caution: exothermic reaction, HCl gas formation). After 1 h stirring, petrol ether (40 mL) was added and stirring was continued for ca. 5 min. The top, clear layer of the new binary or ternary system was transferred to a dry 50 mL Schlenk-tube, and the solvent as well as excess TMSCl and TMS–O–TMS were evaporated off under reduced pressure (ca. 12 mbar) at 40 °C.

2.3.1. Bis(trimethylsilyl) phosphite 4a [18] Yield 4.41 g (97%), colorless liquid. 1H NMR (CDCl3, 400 MHz, d, ppm): 0.22 (s, 9H, CH3), 6.75 (d, 1JHP = 700 Hz, 1H, PH). 13C NMR (CDCl3, 100 MHz, d, ppm): 0.72 (d, 3JCP = 1.6 Hz, CH3). 31P NMR (CDCl3, 162 MHz, d, ppm): 13.07 (d, 1JPH = 700 Hz, CH3). MS(ESI): 227 [M+H]+.

2.3.2. Tris(trimethylsilyl) phosphate 4b [9,17,19] Yield 6.20 g (98%), colorless liquid. 1H NMR (CDCl3, 400 MHz, d, ppm): 0.19 (s, 9H, CH3). 13C NMR (CDCl3, 100 MHz, d, ppm): 0.44 (d, 3 JCP = 1.5 Hz, CH3). 31P NMR (CDCl3, 162 MHz, d, ppm): 25.11.

L.A. Wessjohann, M.A. Dessoy / Polyhedron 70 (2014) 133–137

3. Results and discussion 3.1. Synthesis of tetrakis(trimethylsilyl) diphosphate ester 2a Disodium dihydrogen diphosphate (1) is quantitatively converted into tetrakis(trimethylsilyl) diphosphate (2a, Scheme 1) when treated with trimethylsilyl chloride (TMSCl) in formamide. The use of the highly polar formamide as solvent is crucial for the accomplishment of the reaction and subsequent product isolation. While the very high polarity of formamide ensures the partial solubilization of inorganic diphosphate 1, this property renders the rather apolar TMSCl immiscible in this solvent. Consequently, the resulting reaction mixture constitutes a ternary system of one solid and two liquid phases. Nevertheless, the heterogeneous mixture displays high reactivity if inorganic diphosphate 1 is used in a finely powdered form. If these criteria are met, an exothermic reaction with HCl formation quickly sets in with the temperature rising to 60 °C, forcing the TMSCl to gently reflux on the walls of the reaction vessel (or condenser in larger setups) for a few minutes. The reaction is completed within 1 h. A small excess (ca. 10%) of TMSCl is routinely used in order to ensure the complete conversion of 1 into 2a. Taking this measure, initially no special caution concerning moisture exclusion has to be taken, i.e. drying of glassware is not required for the reaction, but eventually the product formed is highly sensitive to the presence of moisture. Given the high polarity of formamide, which renders products 2a insoluble in this solvent, product isolation/purification is an exceptionally easy task. Straightforward extraction with ordinary petrol ether (equally immiscible with formamide) gives an alkane layer that is separated and the solvent is evaporated under reduced pressure. In larger batches even direct separation of the pure product is possible, without the help of alkane extraction. Excess TMSCl and hexamethyl disiloxane, which is eventually formed in consequence of the partial hydrolysis of TMSCl, are equally eliminated during solvent evaporation. Following this simple procedure, the silylated diphosphate is obtained in a highly pure form. The only measurable impurity of 2a is tris(trimethylsilyl) phosphate (1% to up to 5%), which is mainly formed by the silylation of the phosphate-contamination already present in the diphosphate salt 1, i.e. the purity of the starting material is directly reflected in the product. Nevertheless, the reaction should not be extended over 1 h in order to avoid product decomposition. Upon extended reaction, the free HCl probably induces the decomposition of 2a. When the reaction is performed in the presence of a stoichiometric amount of pyridine (2 mmol pyridine per 1 mmol 1) no decomposition of 2a is observed. One should be aware of the fact that in the presence of pyridine, the reaction time increases to ca. 2.5 h. Crude tetrakis(trimethylsilyl) diphosphate (2a) is a colorless liquid with a shelf life of only a few days at room temperature. Shelf life increases with better removal of HCl traces. 2a dissolved in petrol ether is stable for several weeks at room temperature, if the HCl liberated during the reaction is removed under reduced pressure prior to storage, e.g. by distillation of the petrol ether obtained after the extraction. HCl-free, neat 2a is stable for months when stored at 30 °C. Disodium dihydrogen diphosphate (1) can also be persilylated with TMSCl in polar solvents like pyridine or N,N-dimethyl form-

Scheme 1. Synthesis of tetrakis(trimethylsilyl) diphosphates.

135

amide. The use of these solvents, though, is quite disadvantageous because the reaction is rather sluggish and product isolation proved to be extremely difficult. 3.2. Synthesis of tetrakis(trialkylsilyl) diphosphate esters 2b–2d The protocol that has been developed for the synthesis of 2a also can be applied for the synthesis of the bulkier trialkylsilyl diphosphate esters 2b–2d (Scheme 1). The preparation of these compounds, though, requires somewhat more elaborate reaction conditions. The difficulties arise from the fact that the trialkylsilyl chlorides required as starting materials as well as the corresponding hexaalkyl disiloxanes and/or trialkyl silanols eventually formed through concurrent hydrolysis have relatively high boiling points. Thus, they cannot be separated from the respective diphosphate esters as easily as in the case of 2a. Therefore, the reactions should be conducted under strict moisture exclusion and the amounts of the trialkylsilyl chloride used should be as close as possible to stoichiometry. The bulkier trialkylsilyl chlorides are expectedly less reactive than TMSCl, thus longer reaction times and heating is required. The shielded diphosphates 2b–2d are much less susceptible to the cleavage of the P–O–P bond, and the HCl-catalyzed decomposition to the corresponding phosphate silylester under reaction conditions is less pronounced than for 2a. Therefore these silylations also can be conducted in the absence of pyridine. The isolation of the esters 2b–2d by alkane extraction proceeds analogously to the isolation of 2a. The former esters, though, are not as pure as the latter. Besides the discussed tris(trialkylsilyl) phosphate, the bulkier esters 2b–2d can be contaminated (<5%) with a diphosphate species that is silylated only threefold, i.e. possessing one free OH-group, and with the corresponding silanols or bissilylethers. 3.3. NMR spectroscopy The 1H, 13C, 31P and 29Si NMR spectroscopic data sets recorded for the esters 2a–2d are summarized in the Experimental section. All synthesized diphosphate esters (2a–2d) delivered satisfactory high-resolution mass spectrometric data. The trialkylsilyl groups of the trialkylsilyl diphosphate esters 2a, 2b, and 2d can be fully hydrolyzed in aqueous base, leading to inorganic diphosphate and the corresponding hexaalkyl disiloxanes or trialkyl silanols. The rather labile P–O–P anhydride bond is not markedly affected during this process [13,14]. As expected, the hydrolysis rate declines with the bulkiness of the trialkylsilyl group. The hydrolysis of the trimethylsilyl groups of 2a is almost instantaneous at 4 °C and pH 10, whereas the full hydrolysis of the triethylsilyl groups of 2b takes ca. 30 min and that of the tert-butyl-dimethylsilyl groups of 2d takes a few hours (at room temperature). At pH 4–5 and room temperature, the tert-butyldimethylsilyl groups of 2d are not completely cleaved even after 18 h. Hydrolysis of the even bulkier triisopropylsilyl groups of 2c is not complete even after 30 h at room temperature and pH 10. The partial hydrolytic cleavage of the trialkylsilyl groups of trialkylsilyl diphosphate esters is also possible. In order to achieve this, they have to be treated with a defined amount of water in a watermiscible solvent. Fig. 1b shows the 31P NMR spectrum of a mixture containing 0.74 mmol of 2a and 0.36 mmol of water in DMSO/CH3CN (4 mL) at 23 °C. In contrast to the single, sharp line displayed by 2a (Fig. 1a), this spectrum presents very broad bands, indicating that various chemical identities are in a very fast dynamic equilibrium showing 31P signals of –OPO(OTMS)2, –OPO(OH)(OTMS), and –OPO(OH)2 from high to low field in the statistical ratio 1:2:1 expected from the

136

L.A. Wessjohann, M.A. Dessoy / Polyhedron 70 (2014) 133–137

Fig. 1. 31P NMR spectrum of: (a) Compound 2a at 100 MHz in petroleum ether; (b) 2a (0.74 mmol) and water (0.36 mmol, one water molecule cleaves two TMS groups [10]) in DMSO/CH3CN 3:1 (4 mL) at 160 MHz and 23 °C. (P = silylated phosphate species, PP = silylated diphosphate species).

maximum equilibration for an average of semihydrolized 2a, i.e. P2O(OTMS)2(OH)2 (3a). This equilibrium is a consequence of the pronounced migratory aptitude exhibited by the TMS group [10], which has been compared to the mobility of a proton [16]. The fast dynamic equilibrium displayed between tetrakis(trimethylsilylated) diphosphate 2a and the (partially) hydrolyzed tris and eventually bis(trimethylsilylated) diphosphate 3a (Scheme 2 and Fig. 1b) can be ‘‘freeze-framed’’ if the analogous but less reactive tetrakis(triethylsilyl)diphosphate (2b) is submitted to the same hydrolytic treatment yielding bis(triethylsilylated) diphosphate 3b (Fig. 2). This reaction system evolves to the equilibrium as depicted in Fig. 2 in two phases. First, each water molecule quickly (ca. 15 min) cleaves one triethylsilyl-group from 2b to predominantly give rise to a mono desilylated derivative of 2b. Then, the just formed triethylsilanol slowly (over a period of ca. 67 h) cleaves off a second triethylsilyl-group from 2b and/or from the mono desilylated derivative thereof, leading to the formation of hexaethyl disiloxane, along with mono-, symmetrically bi-, and asymmetrically bi- and/or threefold desilylated derivatives of 2b. Due to the fact that the migratory aptitude of the triethylsilylgroup is much less pronounced than that of the TMS group, the half-life of each partially desilylated species in the equilibration mixture is long enough to give rise to distinct resonance lines (3JP,P) in the 31P NMR spectrum, as depicted in Fig. 2. The cleavage of each trialkylsilyl group causes a downfield shift of ca. 10 ppm on the affected phosphate moiety whereas the chemical shift of the distant phosphate moiety remains almost unchanged. The partially silylated diphosphoric acid species originating from partial desilylation of 2a and 2b can be neutralized with amines. Tertiary alkyl amines like triethyl amine or diethyl isopropyl amine form salts, which are highly soluble in polar solvents like acetone, acetonitrile, N,N-dimethyl formamide, dimethyl sulfoxide, etc. Salts carrying ammonium ions derived from aromatic or aliphatic primary and secondary amines or from pyridine are only soluble in highly polar

Scheme 2. Partial hydrolysis of silyldiphosphate esters 2a (R = Me) and 2b (R = Et).

Fig. 2. (a) 31P NMR spectrum of (TES)4P2O5 2b in petroleum ether at 160 MHz; and 31 P NMR spectra of a mixture of 2b (0.78 mmol) and water (0.39 mmol, one water molecule cleaves two TES groups) in DMSO/CH3CN 3:1 (3 mL) at 23 °C after, (b) 15 min; (c) 10 h; (d) 36 h; (e) 67 h. Spectra from (a) to (e) are each offset by ca. 0.9 ppm in relation to the preceding one. [P = silylated phosphate species, PP = (partially) silylated diphosphate species].

solvents like N,N-dimethyl formamide or dimethyl sulfoxide and tend to precipitate in less polar solvents like acetone or acetonitrile. The salts, of which some in principle have potential as ionic liquids, unfortunately are rather unstable and often decompose within a few days at room temperature. 3.4. Synthesis and reactivity of silylesters of phosphite 4a and phosphate 4b As shown in Scheme 3, the new method for the synthesis of silyl diphosphate esters 2a–2d described above is also splendidly suited to accomplish improved and straightforward synthesis of known bis(trimethylsilyl) phosphite (4a) [18] and tris(trimethylsilyl) phosphate (4b) [9,17,19]. The advantages of this new reaction system over the reported methods for the preparation of both 4a [18] and 4b [9,17,19] are: (a) excellent yields combined with short reaction times; (b) simple apparatus set up; and (c) an uncomplicated product isolation and purification protocol allowing to obtain analytically pure products in only one reaction step without the need

Scheme 3. Synthesis of silylesters of phosphite 4a and phosphate 4b.

L.A. Wessjohann, M.A. Dessoy / Polyhedron 70 (2014) 133–137

to purify intermediates or the final products (4a and 4b) by troublesome distillation. It is worth to note that the homogeneous mixture of phosphorous acid and trimethylsilyl chloride does not undergo any reaction at room temperature. Upon the addition of formamide an exothermic reaction sets in leading to the formation of 3a. Thus, formamide must have a catalytic effect on the silylation reactions described herein, probably comparable to that displayed by N,N-dimethylformamide in the conversion of carboxylic acids into the corresponding acylchlorides upon reaction with thionyl chloride [20]. The Todd-Atherton reaction has been reported as a suitable method for the conversion of the phosphite 4a into the silyl diphosphate ester 2a in moderate yields (44%) [12,13]. Accordingly, a compound showing a 31P NMR resonance at 24.8 ppm has been isolated and its structure was assigned to 2a. As we and others have found [8], a chemical shift in the indicated range is rather due to the phosphate ester 4b than to diphosphate 2a (dP 30.76 ppm, in CDCl3). In view of this divergence, we reacted equimolar amounts of the phosphite, tetrachloromethane, and triethylamine as reported [12,13]. In sequence, the crude product mixture was diluted with petrol ether. An aliquot of the clear supernatant obtained after short centrifugation was then combined with CDCl3 (1:1 v/v) and the sample subjected to 31P NMR measurement. The 31 P NMR spectrum shows three major signals, with the following relative intensities: dP 30.53 ppm (ca. 8%); dP 24.76 ppm (ca. 47%); dP 16.05 ppm (ca. 40%). The addition of authentic 2a caused an increase on the intensity of the high-field signal ( 30.53 ppm), whereas the addition of authentic 4b amplified the signal at 24.76 ppm. In addition, the hydrolyzate of the crude reaction mixture gave inorganic phosphate as the main component, and only a very minor amount of inorganic diphosphate and other unidentified phosphorous compounds. Based on these observations, it can be concluded that, opposing earlier claims [15], the Todd-Atherton reaction is unsuitable for the effective conversion of phosphite 4a into diphosphate 2a. 4. Conclusions In conclusion, we have developed the first efficient method for the facile preparation of bulk amounts of tetrakis(trialkylsilyl) diphosphate esters in large-scale and excellent yields. The silyl diphosphate esters are amenable to both partial and total desilyla-

137

tion. The neutralization of the acidic hydroxyls formed as a result of the partial desilylation with tertiary alkyl amines leads to the formation of the corresponding organic solvent soluble trialkylammonium salts. These facts make the silylated diphosphates potential reagents for the preparation of both further organic and inorganic diphosphates as well as for the silylation of further oxyanions derived either from main group elements, transition metals, or organic oxo compounds. Acknowledgements We thank Prof. Dr. Stefan Berger (Institute of Analytical Chemistry, University of Leipzig) and Dr. A. Porzel (Leibniz Institute of Plant Biochemistry) for the recording of the 31P decoupled and 29 Si NMR spectra of compound 2d. M.A.D. thanks the German Academic Exchange Service (DAAD) for a PhD grant. Part of this work was conducted at the Vrije Universiteit Amsterdam. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.poly.2013.12.024. References [1] A.D.F. Toy, J. Am. Chem. Soc. 70 (1948) 3882. [2] C.R. Assis, A.G. Linhares, V.M. Oliveira, R.C. França, E.V. Carvalho, R.S. Bezerra, L.B. De Carvalho Jr, Sci. Total Environ. 441 (2012) 141. [3] A.S.V. Burgen, C.A. Keele, D. Slome, J. Pharmacol. Exp. Ther. 96 (1949) 396. [4] Y. Watanabe, H. Nakahira, M. Bunya, S. Ozaki, Tetrahedron Lett. 28 (1987) 4179. [5] J.F. Marecek, G.D. Prestwich, Tetrahedron Lett. 30 (1989) 5401. [6] S.J. deSolms, J.P. Vacca, J.R. Huff, Tetrahedron Lett. 28 (1987) 4503. [7] P.M. Chouinard, P.A. Bartlett, J. Org. Chem. 51 (1986) 75. [8] H.Z. Schmidbaur, Z. Anorg. Allg. Chem. 326 (1964) 272. [9] V.P. Mileshkevich, A.V. Karlin, Zh. Obshch. Khim. 40 (1970) 2573. [10] K. Yamamoto, H. Watanabe, Chem. Lett. (1982) 1225. [11] M.D. Mizhiritskii, V.O. Reikhsfel’d, Zh. Obshch. Khim. 55 (1985) 1883. [12] T.P. Mawhinney, J. Chromatogr. 257 (1983) 37. [13] H. Grunze, E. Thilo, Angew. Chem. 70 (1958) 73. [14] H. Grunze, Chem. Ber. 92 (1959) 850. [15] É.E. Nifant’ev, M.A. Kharshan, S.A. Lysenko, Zh. Obshch. Khim. 63 (1993) 776. [16] T. Hiyama, M. Obayashi, Tetrahedron Lett. 24 (1983) 395. [17] R.O. Sauer, J. Am. Chem. Soc. 66 (1944) 1707. [18] K. Issleib, A. Balszuweit, A.-J. Richter, W. Tonk, Z. Chem. 23 (1983) 434. [19] M. Schmidt, H. Schmidbaur, A. Binger, Chem. Ber. 93 (1960) 872. [20] Z. Chen, L. Jiang, W. Su, Z. Xu, J. Chem. Soc. Pakistan 34 (2012) 1003.