Yanina
Moglie‡
a,
María José
González-Soria
a,
Iris
Martín-García
a,
Gabriel
Radivoy
b and
Francisco
Alonso
*a
aInstituto de Síntesis Orgánica (ISO) and Departamento de Química Orgánica, Facultad de Ciencias, Universidad de Alicante, Apdo. 99, 03080 Alicante, Spain. E-mail: falonso@ua.es
bDepartamento de Química, Instituto de Química del Sur (INQUISUR-CONICET), Universidad Nacional del Sur, Avenida Alem 1253, 8000 Bahía Blanca, Argentina
First published on 24th May 2016
The hydrophosphination of carbon–carbon multiple bonds has been generally performed under acid, base or metal catalysis in different solvents. Herein, alkyl and alkenyl tertiary phosphines are obtained by the addition of diphenylphosphine to alkenes and alkynes, respectively, in the absence of a solvent and a catalyst. In the presence of elemental sulfur, the corresponding alkyl and alkenyl tertiary phosphine sulfides are synthesized in a three-component process. These simple methods, which meet most of the principles of Green Chemistry, are highly regioselective towards the anti-Markovnikov products and diastereoselective towards the Z alkenyl phosphines. The mechanistic aspects of the reactions are also tackled and the efficiency of the latter is compared with that of the catalytic methods.
![]() | ||
Scheme 1 Some general methods for tertiary phosphine synthesis; M = metal, X = leaving group; n = 2, m = 1 and n = 1, m = 2. |
Modern chemical research and production must advance on the basis of sustainable and environmentally benign practices.5 In this vein, the hydrophosphination of unsaturated compounds (i.e., alkenes and alkynes) appears as the most straightforward approach to form C–P bonds from readily accessible starting materials (Scheme 1, eqn (6));6 maximum atom economy is attained with no by-product formation.
Closely related to phosphines are the phosphine sulfides,7 a type of compound with multiple applications in different disciplines. Among others, phosphine sulfides have been utilized in anion-selective electrodes,8 lanthanide extraction from a nitrate medium,9 as sensor fluorescent materials for metal ions of environmental concern,10 as anchor units for single molecule junctions,11 in polymer chemistry,12 and as ligands for gold,13 catalysis14 and asymmetric synthesis.15 Base-promoted16 and free-radical17 initiated addition of secondary phosphine sulfides to alkenes (or alkynes)7 are the most practiced methods to prepare tertiary phosphine sulfides. The reaction of secondary phosphine sulfides with carbonyl compounds can provide tertiary α-hydroxy phosphine sulfides.7 The transformation of pre-formed tertiary phosphines into the corresponding sulfides can be readily accomplished by reaction with sulphur; in this case, however, hazardous solvents such as benzene, chloroform and dichloromethane are required.7 More recently, Trofimov et al. reported the one-pot synthesis of tertiary phosphine sulfides from styrenes, red phosphorus and elemental sulfur in a superbasic system containing hydroquinone under microwave irradiation.18 In addition to these, transition-metal catalyzed procedures have recently been published.19
On the other hand, both solvent-free reactions20 and catalyst-free organic synthesis21 notably simplify the reaction mixtures and experimentally, at the same time reduce the amount of waste which, in turn, depletes the environmental impact.
By virtue of our current interest in phosphorus chemistry,22 we found out that tertiary phosphines can be obtained in a very straightforward manner by addition of secondary phosphines to carbon–carbon double bonds under solvent- and catalyst-free conditions.22a Moreover, under these conditions but in the presence of sulfur, α,β-unsaturated carbonyl compounds have been converted into the corresponding β-substituted tertiary phosphine sulfide derivatives through a three-component approach (Scheme 2).22b Our intention is to present herein a comprehensive study on the substrate scope and mechanism of these environmentally friendly protocols, including the hydrophosphination and multicomponent hydrothiophosphination of alkenes and alkynes.
![]() | ||
Scheme 2 Solvent- and catalyst-free hydrophosphination22a and hydrothiophosphination22b of alkenes; EWG = electron-withdrawing group. |
In spite of the higher selectivity achieved by metal catalysis, most of its applications in alkene hydrophosphination does not meet some of the stringent criteria demanded for green and sustainable production because of the use of non-reusable precious metals or noxious solvents (e.g., benzene). In addition, transition metals can accelerate the undesired phosphine oxidation to the phosphine oxide. Ideally, this reaction should be conducted under metal-free and neutral conditions, with the latter also preventing side-reactions and use of aqueous work-up (which can favour oxidation). To the best of our knowledge, Gaumont's group was the first to report the uncatalysed hydrophosphination of alkenes: phosphine–borane complexes were added to inactivated alkenes under neutral conditions followed by either conventional or microwave heating.32
Following our recent discovery on the uncatalysed addition of secondary phosphines to carbon–carbon double bonds under solvent-free conditions,22a we expanded this practice to the addition of diphenylphosphine to a wide range of alkenes, including not only styrenes but also α,β-unsaturated carbonyl compounds and inactivated alkenes (Tables 1–3). All reactions were executed under an inert atmosphere of argon.
As regards styrenes, the simplest one (1a) reacted the fastest and resulted in good yield. p-Halostyrenes (1b and 1c), p-methoxy- and p-acetoxystyrene (1d and 1e, respectively) behaved similarly in terms of yield (around 85%), with a shorter reaction time for 2d. This hydrophosphination was also appropriate for vinyl pyridines (1f and 1g), permitting the synthesis of the P,N-bidentate ligand pyphos (2g), used in homogeneous transition-metal catalysis.33 Other vinyl aromatics, such as 2-vinylnaphthalene (1h) and 4-vinyl-1,1′-biphenyl (1i) were transformed into the tertiary phosphines in high yields. Furthermore, the standard conditions were also effectual for the less reactive 1,1-disubstituted alkene isopropenyl benzene (1j). We must underline that all products 2a–2j were produced with exclusively anti-Markovnikov regioselectivity.
We next explored the diphenylphosphine addition to α,β-unsaturated carbonyl and related compounds (Table 2). These substrates can be considered more activated alkenes than the aforesaid styrenes and, hence, more reactive under milder conditions. Indeed, most of the starting alkenes (except 1l and 1q) experienced hydrophosphination at room temperature (7–12 h); alternatively, the reaction times could be decreased (1–3 h) by warming at 70 °C with comparable yields (see 1k, 1n and 1p). Diverse functional groups were compatible with these conditions, whereby the synthesis of β-diphenylphosphino ketones (1k–1m), esters (1n and 1o), nitrile (1p), amides (1q and 1r) and phosphonate (1s) was carried out in good yields; products 2l and 2r were isolated as the phosphine oxides due to easy oxidation of the phosphine precursors when exposed to air. It is worth noting that the platinum-catalysed24 addition of diphenylphosphine to acrylonitrile (1p) can instead be effectuated in the absence of a catalyst at room temperature.
Heteroatom-bonded vinyl compounds, such as N-vinylphthalimide (1t), N-vinylpyrrolidin-2-one (1u) and phenyl vinyl sulfide (1v) gave the expected (diphenylphosphino)ethyl heteroatom products under the conventional conditions (70 °C) (Table 3). Compound 2v is a P,S-bidentate ligand also employed in catalysis.34 The usefulness of this protocol was validated by its exploitation in the hydrophosphination of inactivated alkenes such as allylbenzene (1x) and oct-1-ene (1y). The anti-Markovnivov regioselectivity was consistent with the trend displayed by all the substrates in Tables 1–3. Still, the tertiary phosphines originated from 1x and 1y were susceptible to oxidation and were isolated as the phosphine oxides 2x and 2y, respectively. It was gratifying to verify that the hydrophosphination of (–)-β-pinene (1z) followed an anti-Markovnivov addition and involved the opening of the cyclobutane ring to supply the enantiomerically pure p-menth-1-ene derived phosphine oxide 2z. An unmistakable assignment of the structure was done by X-ray crystallographic analysis (Fig. 1).35
In a previous study,22b we evidenced that alkene hydrothiophosphination could be achieved through a three-component approach involving the alkene, diphenylphosphine and sulfur, under solvent- and catalyst-free conditions in air. This is an advantageous strategy not only from the environmental point of view, but also because it circumvents the preparation of the secondary phosphine sulfide, which is generated in situ.36c Only α,β-unsaturated carbonyl compounds were covered in that preliminary study; the results obtained for other alkenes, either activated or not activated, are depicted in Table 4.
A variety of electron-neutral, -rich and -deficient styrenes underwent the hydrothiophosphination reaction at 70 °C in air; the corresponding products (3a, 3b, 3aa, 3e and 3ab) were obtained in moderate-to-excellent isolated yields (Table 4). Likewise, styrenes bearing N and S atoms, as well as allylbenzene, were successfully converted into phosphine sulfides (3f, 3g, 3v and 3x). It is worth noting that the process was highly regioselective, giving rise in all cases to the anti-Markovnikov products. The same conditions were applicable to aliphatic substrates, such as oct-1-ene (1y) and 3,4-dihydro-2H-pyran (1ac). Despite the lack of regioselectivity observed for oct-1-ene, both regioisomers (3y and 3y′) could be separated by chromatography in reasonable yields, as a result of the quantitative reaction conversion. In contrast, the hydrothiophosphination of 3,4-dihydro-2H-pyran (1ac) led to a single regioisomer (3ac), derived from the addition of P to the α-O carbon atom.
In our preliminary communication,22a we observed that phenylacetylene could also undergo diphenylphosphine addition under solvent- and catalyst-free conditions. A variety of alkynes has been subjected to this simple procedure to produce vinylphosphines in a regio- and stereoselective manner (Table 5).
Starting alkyne | Product | Yieldb (%) |
---|---|---|
a Alkyne 4 (0.5 mmol) and Ph2PH (0.5 mmol) at 70 °C under argon, overnight.
b Isolated yield.
c As a 95![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
||
![]() |
![]() |
80c |
![]() |
![]() |
73 |
![]() |
![]() |
88 |
![]() |
![]() |
86 |
![]() |
![]() |
87 |
![]() |
![]() |
13d |
![]() |
![]() |
64e |
![]() |
![]() |
62f |
![]() |
![]() |
61g |
![]() |
![]() |
60 |
Electron-neutral and -rich arylacetylenes reacted nicely with diphenylphosphine at 70 °C yielding the expected anti-Markovnikov alkenylphosphines (5a–5e) as Z diastereoisomers.39 Converse behaviour was observed for arylacetylenes bearing electron-withdrawing substituents (e.g., 4f), which were reluctant to react under the same conditions. When methyl(phenyl)acetylene (4g) was subjected to the standard reaction conditions, the corresponding vinylphosphine was formed as a 90:
10 E/Z diastereomeric mixture, with opposite regioselectivity to that reported with a calcium complex.38d Besides arylacetylenes, aliphatic alkynes also experienced hydrophosphination. The alkyl-chain alkyne 4h was converted into the phosphine 5h with a high degree of stereocontrol (97
:
3 Z/E ratio), whereas the tertiary phosphines resulting from the cyclohexyl derivatives 4i and 4j were prone to rapid oxidation, being obtained as the phosphine oxides 5i and 5j, respectively. A 60
:
40 Z/E ratio was recorded for 5i, though the major Z isomer could be isolated in moderate yield. Different outcome arose for the hydroxyl derivative 5j, with absolute control of both the regio- and the stereochemistry.
It must be pointed out that all the alkenylphosphine sulfides were synthesised as single anti-Markovnikov regioisomers and Z stereoisomers. The stereochemistry of the alkyne hydrothiophosphination and, therefore, that of the alkyne hydrophosphination was unequivocally established by X-ray crystallographic analysis of alkenylphosphine sulfide 6a (Fig. 2).41
In order to gain an insight into the reaction mechanism, the hydrophosphination of styrene was performed in the presence of radical traps, such as cumene, TEMPO or 2,6-di-tert-butylphenol [Scheme 3, eqn (1)]. All reactions proceeded with >96% conversion and products derived from the radical traps and diphenylphosphinyl radicals were not detected. These results, together with the fact that the addition of diphenylphosphine to hepta-1,6-diene did not yield the corresponding cyclopentane derivative [Scheme 3, eqn (2)],43 point to a radical-free process. This hypothesis is consistent with the Z to E isomerisation observed for alkenylphosphine derivatives 5a and 5i under radical conditions (Scheme 4); that is, the thermodynamic E isomers, instead of the Z counterparts, should be largely formed if the hydrophosphination was driven by radicals. It must be highlighted that E/Z equilibration did not occur when the alkenyl phosphines 5a and 5i were subjected to prolonged heating either in the absence or presence of iodine.
A series of deuterium-labelling experiments have provided additional evidence on the reaction course. The addition of diphenylphosphine to deuterated cyclohexylacetylene D1-4i gave the Z α-deuteriovinylphosphine oxide D1-5i with 80% D incorporation. This essay ratifies the anti-addition of the P–H bond across the carbon–carbon triple bond which, apparently, can involve some H/D scrambling (Scheme 5).
We also compared the rate of addition of Ph2PH to styrene at 70 °C with that of Ph2PD (Fig. 3). A kinetic isotopic effect was manifested, which was especially dramatic in the range 0–50 min (induction period for Ph2PD) and points to the cleavage of the P–H bond as being the rate-determining step of the reaction.
![]() | ||
Fig. 3 Kinetic profiles for the reaction of styrene (1a) with protio- and deuteriodiphenylphosphine, DPP{H} and DPP{D}, respectively, at 70 °C under Ar. |
The effect of the stoichiometry of the reactants was also investigated for the addition of diphenylphosphine to styrene, in this case, in the range 10–60 min (Fig. 4). It is clear that, with respect to a 1:
1 stoichiometric ratio, the excess of styrene has a negligible effect on the formation of 2a, whereas an excess of diphenylphosphine speeds up its formation (e.g., >5-fold at 10 min by doubling the amount of diphenylphosphine). This effect was expected to be more prominent at shorter reaction times.44 Larger amounts of diphenylphosphine (e.g., 1
:
3 ratio) were shown to be unproductive with respect to the 1
:
2 ratio, displaying a very similar trend.
![]() | ||
Fig. 4 Effect of the styrene (1a)/diphenylphosphine (DPP) ratio on the formation of 2a in the 10–60 min range, at 70 °C under Ar. The amount of 2a is referenced to the limiting substrate. |
Considering the close electronegativities of P (χP = 2.19) and H (χH = 2.20), in principle it might be presumed that the anti-Markovnikov regioselectivity for the hydrophosphination (particularly, that of inactivated substrates) is primarily governed by steric factors rather than by electronic factors. In this context, substrates leading to more sterically hindered products, such as stilbene of diphenylacetylene, reacted sluggishly with diphenylphosphine either in the presence or absence of sulfur. A hydroboration-type model, in which the P–H bond is added through a four-membered ring transition state must be disregarded because (a) the B–H bond (χB = 2.04) is more polarized than the P–H bond and this circumstance works in the same direction as the steric factor in the former, and (b) the hydroboration is syn whereas the hydrophosphination is anti [Scheme 6, eqn (a)].
![]() | ||
Scheme 6 Mechanistic proposals for the hydrophosphination of carbon–carbon double bonds following (a) a syn addition (hydroboration-type model) and (b) an anti addition. |
The above experiments and considerations lend weight to the argument that the hydrophosphination reaction follows an ionic pathway with anti-addition, where each of the P and H atoms emerges from two different phosphine molecules. The ability of tertiary phosphines to act as nucleophilic catalysts in the addition to alkenes (especially, α,β-unsaturated carbonyl compounds) is well known.45 Therefore, it seems reasonable to propose that one phosphine molecule has a nucleophilic role, by addition to the terminal carbon atom of the alkene, while a second molecule behaves as the electrophilic partner through its hydrogen atom [Scheme 6, eqn (b)].
With the proposed transition state model, electronic factors also come into play as the negative charge density developed in the transition state is especially stabilised at the benzylic position (e.g., in the styrenes) as well as at the α-position with respect to heteroatoms (in Het-vinyl substrates) and carbonyl groups (in α,β-unsaturated carbonyl compounds). In the case of 3,4-dihydro-2H-pyran (1ac), with a near symmetric carbon–carbon double bond, electronic factors seem to prevail with the addition of the P atom to the most electrophilic α-C atom (Table 4). Further support for the interpretation made of the hydrophosphination comes from the addition of Ph2PH to β-pinene (1z): the formation of the ring-opened product 2z (Table 3) seems more feasible if two molecules of diphenylphosphine are implicated in the reaction.
Concerning the three-component syntheses of alkyl and alkenyl phosphine sulfides, we earlier confirmed that in situ formation of diphenylphosphine sulfide comes off prior to the addition (Scheme 7).22b
![]() | ||
Scheme 7 Diphenylphosphine sulfide as an intermediate in the three-component synthesis of alkylphosphine sulfides. |
Entry | Catalyst (mol%) | Solvent | T (°C) | Time (h) | Yielda (%) |
---|---|---|---|---|---|
a Isolated yield unless otherwise stated; conversion in parentheses. b 2.0 equiv. of styrene; 1 equiv. of Et3N. c Isolated yield of the phosphine oxide. d Conversion determined by NMR. e Isolated as the borane complex. f 1.82 equiv. of alkene. g This work. | |||||
146 | t-BuOK (20) | DMSO | rt | 1 | 83 |
227b | Ni[P(OEt)3]4 (5) | C6H6![]() |
130 | 20 | 92 |
326a | Ca-amide complex (10) | C6H6 | 75 | 20 | 64c (95)d |
430 | Cu(OTf)2·PhMe (10) | Dioxane-d8 | 100 | 18–24 | 83 |
531 | FeCl2 (30) | MeCN | 60 | 12 | 87e |
625e | Ca-amide complex (2) | C6D6 | 25 | 3 | (100)d |
725f | Ba-amide complex (2) | C6D6 | 60 | 0.4 | (>96)d |
825g | Yb(II) complex (1) | C6D6 | 60 | 4 | (92)d |
928 | Fe(III) complex (0.5) | MeCNf | rt | 24 | 89 |
10 | None | None | 70 | 4 | 82g |
Similar comments can be extended to the hydrophosphination of alkynes, in this case exemplified by the addition of diphenylphosphine to phenylacetylene (Table 8). The vinyl phosphine 5a can be obtained without a catalyst and a solvent, not only in good isolated yield but also in the highest E/Z diastereoselectivity (entry 6). Taking into account different parameters (catalyst, solvent, temperature, time, yield and selectivity), together with the simplicity of the procedure, we can state that this approach towards the addition of the P–H bond to alkenes and alkynes distinctly outperforms others based on catalytic approaches.
Entry | Catalyst (mol%) | Solvent | T (°C) | Time | Yielda (%) |
---|---|---|---|---|---|
a
Z/E ratio in parenthesis.
b ![]() |
|||||
137 | Pd(PPh3)4 (1.2) | MeCN | 130 | 18 h | 95b (86/14) |
238a | Yb-imine complex (5) | THF | rt | 5 min | >99c (24/76) |
325e | Ca-amide complex (5) | C6D6 | 75 | 38 h | 78d (76/24) |
425e | Yb-amide complex (5) | C6D6 | 75 | 38 h | 91d (10/90) |
538f | Y complex (2) | C6D6 | 70 | 72 h | 100d (42/58) |
6 | None | None | 70 | 10 | 80e (95/5) |
Furthermore, by introducing elemental sulfur into the reaction mixture in air, alkyl and alkenyl phosphine sulfides can be readily synthesised through a one-pot three-component approach in a regio- and stereoselective manner. Compelling experimental evidence suggests an ionic-type anti addition reaction mechanism governed by steric and electronic effects different from that of the alkene/alkyne hydroboration.
These methods are in agreement with some of the twelve principles introduced by Anastas et al.,5a namely: (a) waste is prevented because of no by-product formation, (b) high atom economy, as all the starting materials are incorporated into the final products, (c) unnecessary use of solvents, (d) some reactions proceed at room temperature and most of them at 70 °C, (e) unnecessary derivatisation (the secondary phosphine sulfides are generated in situ) and (f) neither stoichiometric nor catalytic reagents are employed because the processes are catalyst free. In addition to this, similar or better results are obtained with respect to the catalytic methods.
In short, this study supports the statements by Sheldon that “the best catalyst is no catalyst” and “the best solvent is no solvent”.47
Footnotes |
† Electronic supplementary information (ESI) available: Experimental procedures, compound characterisation, NMR spectra, kinetic graphics and X-ray data. CCDC 1468810 and 1437629. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6gc00903d |
‡ Present address: Departamento de Química, Instituto de Química del Sur (INQUISUR-CONICET), Universidad Nacional del Sur, Avenida Alem 1253, 8000 Bahía Blanca, Argentina. |
This journal is © The Royal Society of Chemistry 2016 |