Study of the accessibility of phosphorus centres incorporated within ordered mesoporous organic–inorganic hybrid materials

Abstract

Phosphorus centres incorporated within surfactant-directed mesoporous hybrid materials are shown to be more easily accessible than those incorporated within the corresponding materials prepared in the absence of surfactant.

---

The one step preparation of organically functionalised ordered mesoporous silica by using non-ionic^1,3 and ionic surfactants^4–9 as structure-directing agents constitutes a new and very promising route to hybrid materials with organised functionalities. Our contribution in this field was the preparation of surfactant-directed mesoporous hybrid materials incorporating phosphino groups.³ We showed that such materials are easily accessible and undergo further chemical reactions at phosphorus (sulfuration and quaternization) without disrupting the ordered structure and without changing notably the textural characteristics, thus suggesting the location of the P centres within the mesopore channels.⁵ The potential applications of these materials containing P centres for catalysis¹⁰ and separation which are strongly connected with the chemical accessibility of the functionalities have led us to further studies. Here, we show that, in surfactant-directed mesoporous hybrid materials, phosphorus centres are more easily accessible than in the corresponding materials prepared in the absence of surfactants.

The hybrid materials X_n and X′_n were obtained by co-hydrolysis and polycondensation of mixtures Ph₂P(CH₂)₃Si(OMe)₃/nS i(OEt)₄ (n = 6, 9 and 19) in the presence of n-hexadecylamine as template for X_n and in the absence of template but in the presence of 1% TBAF as catalyst for X′_n (Scheme 1). The molar composition of each mixture was: 1 − x Si(OEt)₄:xPh₂P(CH₂)₃S i(OMe)₃: 0.27 n-hexadecylamine∶24.2 H₂O∶9.1 EtOH. The xerogels were prepared according to previously published procedures³ and some relevant physical properties are given in Table 1. The BET surface areas were determined by N₂ adsorption–desorption isotherm measurements. X_n (n = 6, 9, 19) exhibit type IV isotherms,¹¹ characteristic for mesoporous materials. In contrast, the isotherms of X′_n are indicative of macroporosity.¹¹ XRD patterns for X_n exhibit a single diffraction peak corresponding to d₁₀₀ spacing while no peak was observed for X′_n. Organic incorporation was calculated by thermogravimetric analysis. It is worth noting that the calculated values are slightly closer to theoretical values for the mesoporous xerogels than for the others and this, in particular for high ‘dilutions’ of the organic moiety in silica. Another interesting observation in relation to the effect of dilution was made from the ³¹P NMR data of the materials. As shown in Fig. 1, the ³¹P NMR spectra for X′_n displayed signals which are always broader than those for X_n. This is an indication of a greater mobility for the P centres in X_n than in X′_n. Furthermore, on going from the ‘dilution’ n = 9 to the higher ‘dilution’ n = 19, the Dn_1/2 values remain stable for the ordered hybrid materials while for the others X′_n, they regularly decrease as the ‘dilution’ increases. Thus it can be concluded that from n = 9 all the P centres within X_n have the same surroundings, the organic pendants being regularly dispersed at the surface of the mesopores while for the solids X′_n the distribution of the organic groups is random.

Scheme 1

Table 1 Physical characteristics for X_n (X′_n)

Xerogel	d 100 Lattice spacing/Å	BET surface area/m^2 g^−1	Total pore volume/cm^3 g^−1	Pore diameter^a/Å	^31P CP MAS NMR (δ)	Organic incorporation^b (%)
a Measured using the Horvath–Kawazoe pore size distribution model. b Percentage of Si atoms present as organosilane with respect to total Si content calculated by thermogravimetric analysis. c Theoretical values.
X 6 (X′6)	41.86 (—)	385 (149)	0.30 (—)	27 (≈300)	−15.84	14.2^c	11 (10)
X 9 (X′9)	34.41 (—)	1100 (350)	0.53 (—)	35 (≈300)	−15.71 (−15.61)	10^c	9.4 (8.8)
X 19 (X′19)	38.91 (—)	1380 (640)	0.61 (—)	34 (≈300)	−15.13 (−15.0)	5^c	5 (4.7)

---

Fig. 1 Half linewidth (Δν_1/2) of the solid state ³¹P NMR signals for X_n (———) and X′_n (-----) as a function of the ‘dilution’ n.

The accessibility of the P centres in these materials was investigated by using the reactions of quaternization and complexation of phosphorus centres. All the reactions at phosphorus were monitored by solid state ³¹P NMR spectroscopy which is a reliable and sensitive probe for immobilized phosphorus groups.¹²

The solids X_n (n = 6, 9, 19) were treated with 1 equiv. of benzyl bromide 1 or 0.5 equiv. of α,α′-dibromoparaxylene 2 per P centre in CH₂Cl₂ at 20 °C. It is to be noted that after treatment, the ³¹P NMR spectra of materials exhibited no signal corresponding to the oxidation of the P centres. The results given in Table 2 show that the percentage of quaternization of P centres increase with the dilution of the organic moieties in silica with both reagents. The reaction is quantitative with 1 after 72 h from X₉ and X₁₉. It is also quantitative with the bifunctional compound 2 from X₁₉ after 72 h and almost quantitative from X₉ which is noteworthy as in that case all the P centres are bridged. The same reactions were carried out on X′_n and the percentage of quaternization was calculated after 24 h of reaction. These results are compared to those obtained from X_n after the same reaction time in Table 3. It appears that the percentage of reaction with 1 is always lower from X′_n than from X_n, whatever the dilution. Thus the P centres are more easily accessible within ordered mesoporous materials X_n than within the amorphous X′_n. Interestingly, with the bifunctional reagent 2, the gap between the percentage of quaternization of P centres obtained from X_n and X′_n increases with the dilution in favour of the ordered materials X_n. This result is due to the regular distribution of organic moieties in X_n. Indeed, while the substitution of the second bromide of 2 is promoted by the dilution within X_n, it is rendered difficult within X′_n because of the random distribution of the organic groups in silica. The distribution of the organic moieties within the solids was further studied by treating the materials with 0.5 equiv. cis-(PPh₃)₂PdCl₂ per P centre. We observed that the solid state ³¹P NMR spectra of the materials after 24 h of reaction at 20 °C in CH₂Cl₂ displayed a resonance at δ −16 corresponding to the remaining starting phosphine and a resonance at δ +16 which was assigned to the trans anchored phosphine Pd^II complex.¹² A further resonance at δ +36 was attributed to starting phosphine oxidation, which never exceeded 15%. Treatment of the materials with a large excess of the nucleophile PBuⁿ₃ allowed the complete elimination of the palladium liberating the starting phosphine (δ −16). The presence of the unchanged signal at δ +36 after this treatment confirmed the oxidation of the starting phosphine. The extent of the anchored phosphine Pd^II complex obtained from X_n and X′_n determined by ³¹P NMR spectroscopy is shown in Fig. 2. The difference between the reactivity of materials X_n and X′_n (n = 9, 19) is important. Thus the diffusion of a bulky molecule like (PPh₃)₂PdCl₂ through materials X′_n is much more difficult than through the ordered mesoporous materials X_n. The very low extent of reaction for X₆ compared to X₉ and X₁₉ should be due to steric hindrance, which implies also that P centres are mostly located within the mesopores of X_n.

Table 2 Percentage of quaternization of P centres within X_n treated with 1 or (2) as reagent

Reaction time/h	X 6	X 9	X 19
2	48 (35)	56 (58)	78 (74)
16	63 (68)	72 (82)	95 (90)
72	83 (83)	100 (93)	100 (100)

---

Table 3 Percentage of quaternization of P centres within X_n and X′_n after treatment for 24 h at 20 °C with 1 or (2) as reagent

n	6	9	19
X n	71 (73)	87 (92)	94 (95)
X′ n	60 (72)	64 (82)	79 (70)

---

Fig. 2 Percentage of anchored phosphine Pd^II complex within X_n (———) and X′_n (-----).

In conclusion, we have shown that in surfactant-directed mesoporous hybrid materials, phosphorus centres are more easily accessible than in the corresponding materials prepared in the absence of surfactants. Furthermore the accessibility of phosphorus centres within ordered mesoporous hybrid materials depends on the ‘dilution’ of the organic part in silica. The higher the ‘dilution’, the greater the accessibility of the organic part in particular towards bulky reagents for which a minimum ‘dilution’ (n > 6) seems to be required. The study of hybrid materials with very low concentration of organo groups is in progress.

References

1. D. J. Macquarrie, Chem. Commun., 1996, 1961.
2. D. J. Macquarrie, D. B. Jackson, J. E. G. Mdoe and J. H. Clark, New J. Chem., 1999, 23, 539.
3. R. J. P. Corriu, A. Mehdi and C. Reyé, C.R. Acad. Sci. Paris, Sér. IIc, 1999, 35.
4. S. L. Burkett, S. D. Sims and S. Mann, Chem. Commun., 1996, 1367.
5. M. H. Lim, C. F. Blanford and A. Stein, J. Am. Chem. Soc., 1997, 119, 4090.
6. C. E. Fowler, S. L. Burkett and S. Mann, Chem. Commun., 1997, 1769.
7. M. H. Lim, C. F. Blanford and A. Stein, Chem. Mater., 1998, 10, 467.
8. F. Babonneau, L. Leite and S. Fontlupt, J. Mater. Chem., 1999, 9, 175.
9. S. R. Hall, C. E. Fowler, B. Lebeau and S. Mann, Chem. Commun., 1999, 201.
10. B. F. G. Johnson, S. A. Raynor, D. S. Shephard, T. Mashmeyer, J. Meurig Thomas, G. Sankar, S. Bromley, R. Oldroyd, L. Gladden and M. D. Mantle, Chem. Commun., 1999, 1167.
11. S. J. Gregg and K. S. W. Sing, Adsorption, Surface Area and Porosity, Academic Press,New York, 2nd edn., 1982..
12. R. A. Komoroski, A. J. Magistro and P. P. Nicholas, Inorg. Chem., 1986, 25, 3917.

---