Efficiency of the refocused 31P-29Si MAS-J-INEPT NMR experiment for the characterization of silicophosphate crystalline phases and amorphous gels

Article abstract of DOI:10.1021/ic061964f

One- and two-dimensional refocused MAS-J-INEPT NMR experiments in the solid state (through-bond polarization transfer) involving the highly abundant ³¹P spin and the rare ²⁹Si spin are described for the crystalline silicophosphate ph

Full text of DOI:10.1021/ic061964f

Inorg. Chem. 2007, 46, 1379−1387
Efficiency of the Refocused 31P 29Si MAS-J-INEPT NMR Experiment for
−
the Characterization of Silicophosphate Crystalline Phases and
Amorphous Gels
C. Coelho, T. Aza1ıs, L. Bonhomme-Coury, G. Laurent, and C. Bonhomme*
UniVersite´ Pierre et Marie Curie-Paris 6, UMR 7574, Laboratoire Chimie de la Matie`re
Condense´e de Paris, Paris, F-75005 France
Received October 13, 2006
One- and two-dimensional refocused MAS-J-INEPT NMR experiments in the solid state (through-bond polarization
transfer) involving the highly abundant 31P spin and the rare 29Si spin are described for the crystalline silicophosphate
phase Si5O(PO4)6 and complex mixtures of SiP2O7 polymorphs. The evaluation of the 2JP
coupling constants
-O-Si
for all 29Si sites is obtained by the careful analysis of the INEPT build-up curves under fast MAS. The results are
in agreement with the crystallographic data, taking into account the various J coupling paths. The efficiency of the
experiment is demonstrated by its application to more complex systems such as silicophosphate amorphous gels
(obtained by the sol−gel process).
Introduction
biomaterials, involving bioactive phosphosilicates and sub-
stituted hydroxyapatite (HAP) structures.9 This vivid field
of research includes the detailed study of silicon-substituted
HAP, in terms of bone interfaces and bioactivity.10 The
surface modification of HAP by silica or silicates has been
deeply investigated.11 The formation of calcium phosphate
in silica gels and the mechanisms of the bonelike formation
of bioactive implants have also been reviewed.12-14
Recently, silicophosphate gels and derivatives have gained
much attention in various fields. Among them, we can cite
(i) the study of gel-derived silicophosphate xerogels and
glasses,1-3 exhibiting interesting technological properties with
application as fast proton conductors,4 (ii) the study of surface
grafting reactions between silica gels or nanoparticles and
phosphate or phosphonate moieties,5-6 (iii) the study of gels,
precursors for the synthesis of microporous silicoalumino-
phosphates (SAPO),7-8 and (iV) the study of surfaces in
In the field of materials, the local structure of nuclei can
be probed efficiently by solid-state NMR spectroscopy. In
the case of silicophosphates, it is well-established that the
* To whom correspondence should be addressed. Phone: +33 1 44 27
41 35. Fax: +33 1 44 27 47 69. E-mail: bonhomme@ccr.jussieu.fr.
(1) (a) Clayden, N. J.; Esposito, S.; Pernice, P.; Aronne, A. J. Mater.
Chem. 2001, 11, 936-943. (b) Livage, J.; Barboux, P.; Vandenborre,
M. T.; Schmutz, C.; Taulelle, F. J. Non-Cryst. Solids 1992, 147-
148, 18-23. (c) Tian, F.; Pan, L.; Wu, X. J. Non-Cryst. Solids 1988,
104, 129-134. (d) Szu, S. P.; Klein, L. C.; Greenblatt, M. J. Non-
Cryst. Solids 1992, 143, 21-30. (e) Aronne, A.; Turco, M.; Bagnasco,
G.; Pernice, P.; Di Serio, M.; Clayden, N. J.; Marenna, E. Chem. Mater.
2005, 17, 2081-2090.
(8) (a) Huang, Y.; Machado, D. Microporous Mesoporous Mater. 2001,
47, 195-202. (b) Huang, Y.; Machado, D.; Kirby, C. W. J. Phys.
Chem. B 2004, 108, 1855-1865.
(9) (a) Hench, L. L. J. Am. Ceram. Soc. 1991, 74, 1487-1510. (b) Fujii,
E.; Ohkubo, M.; Tsuru, K.; Hayakawa, S.; Osaka, A.; Kawabata, K.;
Bonhomme, C.; Babonneau, F. Acta Biomater. 2006, 2, 69-74. (c)
Hayakawa, S.; Ohnishi, K.; Tsuru, K.; Osaka, A.; Fujii, E.; Kawabata,
K.; Babonneau, F.; Bonhomme, C. Key Eng. Mater. 2006, 309, 503-
506. (d) Hayakawa, S.; Ando, K.; Tsuru, K.; Osaka, A.; Fujii, E.;
Kawabata, K.; Bonhomme, C.; Babonneau, F. J. Am. Ceram. Soc.
2007, in press.
(2) Coelho, C.; Aza¨ıs, T.; Bonhomme-Coury, L.; Maquet, J.; Bonhomme,
C. C. R. Chim. 2006, 9, 472-477.
(10) (a) Porter, A. E.; Patel, N.; Skepper, J. N.; Best, S. M.; Bonfield, W.
Biomaterials 2004, 25, 3303-3314. (b) Porter, A. E.; Patel, N.;
Skepper, J. N.; Best, S. M.; Bonfield, W. Biomaterials 2003, 24, 4609-
4620.
(3) Coelho, C.; Babonneau, F.; Aza¨ıs, T.; Bonhomme-Coury, L.; Maquet,
J.; Laurent, G.; Bonhomme, C. J. Sol-Gel Sci. Technol. 2006, 40, 181-
189.
(4) Nogami, M.; Miyamura, K.; Abe, Y. J. Electrochem. Soc. 1997, 144,
2175-2178.
(11) (a) Borum, L.; Wilson, O. C., Jr. Biomaterials 2003, 24, 3681-3688.
(b) Liu, X.; Ding, C.; Chu, P. K. Biomaterials 2004, 25, 1755-1761.
(12) Korventausta, J.; Jokinen, M.; Rosling, A.; Peltola, T.; Yli-Urpo, A.
Biomaterials 2003, 24, 5173-5182.
(5) Lukes, I.; Borbaruah, M.; Quin, L. D. J. Am. Chem. Soc. 1994, 116,
1737-1741.
(6) Mutin, P. H.; Lafond, V.; Popa, A. F.; Granier, M.; Markey, L.; Dereux,
A. Chem. Mater. 2004, 16, 5670-5675.
(7) Fyfe, C. A.; Wong-Moon, K. C.; Huang, Y.; Grondey, H. Microporous
Mater. 1995, 5, 29-37.
(13) De Aza, P. N.; Luklinska, Z. B.; Santos, C.; Guitian, F.; De Aza, S.
Biomaterials 2003, 24, 1437-1445.
(14) Handke, M.; Sitarz, M.; Rokita, M.; Galuskin, E. J. Mol. Struct. 2003,
651-653, 39-54.
10.1021/ic061964f CCC: $37.00
Published on Web 01/20/2007
© 2007 American Chemical Society
Inorganic Chemistry, Vol. 46, No. 4, 2007 1379

Coelho et al.
isotropic 31P and 29Si chemical shifts (δiso) can be assigned
to the QN and Q′N species, where QN is related to the
Si(OX)N(OX′)4-N moieties (with X ) Si, P; X′ ) H, Et)
and Q′N stands for P(O)(OY)N(OY′)3-N entities (with Y )
P, Si ; Y′ ) H).1d Nevertheless, we have shown recently
that these notations are rather loose, in the sense that various
combinations of P-O-P/P-O-Si/P-OH‚‚‚ bonds can lead
to the same δiso value.3 In other words, sole knowledge of
the chemical shift values is not sufficient for the fine
structural description of silicophosphate derivatives.
niques,23 as well as spin diffusion experiments,23a,24 were
successfully used in the frame of inorganic phosphates.
The second category includes J-mediated experiments,
where J stands for the heteronuclear or homonuclear isotropic
scalar coupling. In the early 1990’s, such NMR sequences
were implemented by Fyfe and co-workers25 and Eckert and
co-workers.26 In the late 1990’s, Emsley and co-workers
showed that the INADEQUATE (incredible natural abun-
dance double quantum transfer experiment)27 and the HMQC
(heteronuclear multiple quantum coherence)28 sequences
could be safely transposed in solid-state NMR for the study
of organic and bio-organic derivatives. The homonuclear
INADEQUATE and UC2QF COSY (uniform-sign cross-
peak double quantum filtered correlation spectroscopy)
sequences have been subsequently used for the following
spin pairs: 31P/31P, 29Si/29Si, 15N/15N, and 13C/13C.23c-e,29-32
The heteronuclear HMQC sequence was adapted for the
Various solid-state NMR techniques can be implemented
for establishing the connectiVities between nuclei (i.e., 29Si
and 31P). The experiments used so far can be divided roughly
in two main categories.
The first category includes D-mediated experiments, where
D stands for the heteronuclear or homonuclear dipolar
interactions. Such experiments establish spatial connectiVities
between X and Y nuclei. In the case of the heteronuclear
dipolar interaction (X * Y), REDOR (rotational-echo double
resonance),15 TEDOR (transferred-echo double resonance),
TRAPDOR (transfer of polarization in double resonance),16
and CP MAS17-18 (cross polarization magic angle spinning)
pulse sequences are usually implemented. In some cases, the
distances between nuclei were accurately measured by the
careful analysis of the dipolar oscillations.19-20 For X ) 31P
and Y ) 29Si, very few results have been reported so far in
the literature. These results are related to silicon phosphide
(involving direct 31P-29Si bonds)21 and silicophosphate
phases involving 31P-O-29Si groups.2,22 Triple resonance
experiments 1H f 31P f 29Si were used for editing purposes
in silicophosphate gels.2 The crystalline Si5O(PO4)6 phase
was used as a standard for the setup of the 31P f 29Si 2D
HETCOR (heteronuclar correlation) CP MAS experiment.22
It has to be noted that because of the small 31P-29Si dipolar
couplings (∼360 Hz), very long contact times (up to 40 ms)
were used under the Hartmann-Hahn condition. Such
experimental conditions are very demanding in terms of
probe coil and RF power levels. In the case of the
homonuclear dipolar interaction (X ) Y), recoupling tech-
1
following spin pairs: H/13C, 1H/15N, 31P/27Al, 27Al/17O, 31P/
29Si, and 31P/71Ga.26,28,33-36 More involved J-derived solid-
state NMR techniques were proposed recently in the
literature, including triple quantum correlation experiments,37
2D and 3D H-HSQC (homonuclear-heteronuclear single
quantum correlation) experiments,38 and frequency-selective
(24) (a) Hartmann, P.; Jana, C.; Vogel, J.; Ja¨ger, C. Chem. Phys. Lett. 1996,
258, 107-112. (b) Clayden, N. J.; Esposito, S.; Aronne, A. J. Chem.
Soc., Dalton Trans. 2001, 2003-2008.
(25) (a) Fyfe, C. A.; Feng, Y.; Gies, H.; Grondey, H.; Kokotailo, G. T. J.
Am. Chem. Soc. 1990, 112, 3264-3270. (b) Fyfe, C. A.; Wong-Moon,
K. C.; Huang, Y.; Grondey, H. J. Am. Chem. Soc. 1995, 117, 10397-
10398. (c) Bechmann, M.; Helluy, X.; Marichal, C.; Sebald, A. Solid
State NMR 2002, 21, 71-85.
(26) Franke, D.; Hudalla, C.; Eckert, H. Solid State NMR 1992, 1, 33-40.
(27) (a) Lesage, A.; Auger, C.; Caldarelli, S.; Emsley, L. J. Am. Chem.
Soc. 1997, 119, 7867-7868. (b) Lesage, A.; Bardet, M.; Emsley, L.
J. Am. Chem. Soc. 1999, 121, 10987-10993. (c) Sakellariou, D.;
Brown, S. P.; Lesage, A.; Hediger, S.; Bardet, M.; Meriles, C.; Pines,
A.; Emsley, L. J. Am. Chem. Soc. 2003, 125, 4376-4380.
(28) (a) Lesage, A.; Sakellariou, D.; Steuernagel, S.; Emsley, L. J. Am.
Chem. Soc. 1998, 120, 13194-13201. (b) Lesage, A.; Charmont, P.;
Steuernagel, S.; Emsley, L. J. Am. Chem. Soc. 2000, 122, 9739-9744.
(29) Fayon, F.; Massiot, D.; Levitt, M. H.; Titman, J. J.; Gregory, D. H.;
Duma, L.; Emsley, L.; Brown, S. P. J. Chem. Phys. 2005, 122, 194313.
(30) (a) Hedin, N.; Graf, R.; Christiansen, S. C.; Gervais, C.; Hayward, R.
C.; Eckert, J.; Chmelka, B. F. J. Am. Chem. Soc. 2004, 126, 9425-
9432. (b) Brouwer, D. H.; Kristiansen, P. E.; Fyfe, C. A.; Levitt, M.
H. J. Am. Chem. Soc. 2005, 127, 542-543.
(31) (a) Brown, S. P.; Pe´rez-Torralba, M.; Sanz, D.; Claramunt, R. M.;
Emsley, L. J. Am. Chem. Soc. 2002, 124, 1152-1153.(b) Brown, S.
P.; Pe´rez-Torralba, M.; Sanz, D.; Claramunt, R. M.; Emsley, L. Chem.
Commun. 2002, 17, 1852-1853.
(32) (a) Grasso, G.; De Swiet, T. M.; Titman, J. J. J. Phys. Chem. B 2002,
106, 8676-8680. (b) Olsen, R. A.; Struppe, J.; Elliott, D. W.; Thomas,
R. J.; Mueller, L. J. J. Am. Chem. Soc. 2003, 125, 11784-11785. (c)
De Pae¨pe, G.; Lesage, A.; Steuernagel, S.; Emsley, L. Chem. Phys.
Chem. 2004, 5, 869-875. (d) Cadars, S.; Lesage, A.; Emsley, L. J.
Am. Chem. Soc. 2005, 127, 4466-4476. (e) Harris, R. K.; Joyce, S.
A.; Pickard, C. J.; Emsley, L. Phys. Chem. Chem. Phys. 2006, 8, 137-
143.
(15) Gullion, T.; Schaefer, J. J. Magn. Reson. 1989, 81, 196-200.
(16) (a) Hing, A. W.; Vega, S.; Schaefer, J. J. Magn. Reson. 1992, 96,
205-209. (b) Grey, C. P.; Vega, A. J. J. Am. Chem. Soc. 1995, 117,
8232-8242.
(17) Hartmann, S. R.; Hahn, E. L. Phys. ReV. 1962, 128, 2042-2053.
(18) Pines, A.; Gibby, G.; Waugh, J. S. J. Chem. Phys. 1973, 59, 569-
590.
(19) Aza¨ıs, T.; Bonhomme, C.; Bonhomme-Coury, L.; Vaissermann, J.;
Millot, Y.; Man, P. P.; Bertani, P.; Hirschinger, J.; Livage, J. J. Chem.
Soc., Dalton Trans. 2002, 609-618.
(20) Aza¨ıs, T.; Bonhomme-Coury, L.; Vaissermann, J.; Bertani, P.;
Hirschinger, J.; Maquet, J.; Bonhomme, C. Inorg. Chem. 2002, 41,
981-988.
(21) Franke, D.; Hudalla, R.; Maxwell, R.; Eckert, H. J. Phys. Chem. 1992,
96, 7506-7509.
(33) Massiot, D.; Fayon, F.; Alonso, B.; Trebosc, J.; Amoureux, J. P. J.
Magn. Reson. 2003, 164, 160-164.
(34) Iuga, D.; Morais, C.; Gan, Z.; Neuville, D. R.; Cormier, L.; Massiot,
D. J. Am. Chem. Soc. 2005, 127, 11540-11541.
(35) Coelho, C.; Aza¨ıs, T.; Bonhomme-Coury, L.; Maquet, J.; Massiot,
D.; Bonhomme, C. J. Magn. Reson. 2006, 179, 114-119.
(36) Montouillout, V.; Morais, C. M.; Douy, A.; Fayon, F.; Massiot, D.
Magn. Reson. Chem. 2006, 44, 770-775.
(37) Fayon, F.; Roiland, C.; Emsley, L.; Massiot, D. J. Magn. Reson. 2006,
179, 49-57.
(38) (a) Deschamps, M.; Fayon, F.; Montouillout, V.; Massiot, D. Chem.
Commun. 2006, 1924-1925. (b) Deschamps, M.; Massiot, D. J. Magn.
Reson. 2006, published online.
(22) Lejeune, C.; Coelho, C.; Bonhomme-Coury, L.; Aza¨ıs, T.; Maquet,
J.; Bonhomme, C. Solid State NMR 2005, 27, 242-246.
(23) (a) King, I. J.; Fayon, F.; Massiot, D.; Harris, R. K.; Evans, J. S. O.
Chem. Commun. 2001, 1766-1767. (b) Fayon, F.; Massiot, D.;
Suzuya, K.; Price, D. L. J. Non-Cryst. Solids 2001, 283, 88-94. (c)
Fayon, F.; Le Saout, G.; Emsley, L.; Massiot, D. Chem. Commun.
2002, 1702-1703. (d) Fayon, F.; King, I. J.; Harris, R. K.; Gover, R.
K. B.; Evans, J. S. O.; Massiot, D. Chem. Mater. 2003, 15, 2234-
2239. (e) Fayon, F.; King, I. J.; Harris, R. K.; Evans, J. S. O.; Massiot,
D. C. R. Chim. 2004, 7, 351-361. (f) Feike, M.; Ja¨ger, C.; Spiess, H.
W. J. Non-Cryst. Solids 1998, 223, 200-206.
1380 Inorganic Chemistry, Vol. 46, No. 4, 2007

Refocused 31P-29Si MAS-J-INEPT NMR Experiment
pulse schemes.39 Solid-state COSY (correlation spectros-
copy),25a,c TOBSY (total through-bond correlation spectros-
copy),40 and J-resolved experiments23d-e,33,41 were also used
for the characterization of organic and inorganic derivatives.
The INEPT (insensitive nuclei enhanced by polarization
transfer) sequence42 is a routine pulse block for a large variety
of solution-state multidimensional experiments.43 The main
advantage of this sequence is the efficient through-bond
transfer of polarization from an abundant spin system (such
1
as H or 31P) to a low-abundance spin system (such as 13C
or 29Si). In the frame of solid-state NMR, the INEPT
sequence was adapted for the study of mobile44 and rigid45
organic and hydrid systems. When dealing with inorganic
derivatives, very few examples were published in the
literature, involving mainly the 27Al/31P spin pair,25b,46-49 and
two specific minerals, namely, the microcline (27Al/29Si) and
the albite (23Na/29Si).25b The recent interest in J-MAS derived
NMR techniques is essentially due to the fact that the
involved coherence lifetimes are long enough for efficient
polarization transfer and that the refocused line widths are
much smaller than the apparent line widths, as discussed
recently in the literature.32c,50
In this paper, we show that the 31P-29Si MAS-J-INEPT
experiment can act as an invaluable tool of investigation for
silicophosphate derivatives. Efficient through-bond polariza-
tion transfer is first demonstrated for the crystalline Si5O-
(PO4)6 phase.51 The 1D MAS-J-INEPT build-up curves are
analyzed in terms of SIn (n e 6) or SInI′n (n ) 3) spin systems
(S ) 29Si, I ) 31P), leading to the determination of the
Figure 1. Description of the structure of Si5O(PO4)6 around the P and Si
atoms. The labelling scheme of atoms is given according to the literature.51
29Si MAS spectrum (single-pulse experiment) (Ø, 4 mm; rotation frequency
(RO), 14 kHz; number of scans (NS), 1760; recycle delay (RD), 10 s; 90°-
(29Si), 4.5 µs; LB ) 10 Hz).
2
2JP-O-Si coupling constants. It is shown that the JP-O-Si
coupling constants are strongly dependent on the crystal-
lographic chemical paths. To the best of our knowledge, such
constants have not been reported so far in the literature even
in the frame of solution-state NMR. The extension of the
MAS-J-INEPT approach to 2D experiments leads to the
unambiguous description of crystalline silicophosphate mix-
tures, involving SiP2O7 polymorphs.24a,35,52 Finally, the MAS-
J-INEPT experiment is also adapted for the study of
amorphous silicophosphate gels,1-3 revealing interesting
structural features. The INEPT through-bond approach
appears very robust and less demanding in terms of RF power
levels, when compared to the 31P f 29Si CP MAS experiment
(indeed, short π/2 and π pulses are used in the INEPT
sequence instead of long contact time pulses). We strongly
believe that the MAS-J-INEPT technique is suitable for the
fine characterization of the derivatives cited above in points
i-iv.
(39) (a) Trebosc, J.; Amoureux, J. P.; Wiench, J. W.; Pruski, M. Chem.
Phys. Lett. 2003, 374, 432-438. (b) Trebosc, J.; Amoureux, J. P.;
Delevoye, L.; Wiench, J. W.; Pruski, M. Solid State Sci. 2004, 6,
1089-1095. (c) Amoureux, J. P; Trebosc, J.; Wiench, J.; Massiot,
D.; W.; Pruski, M. Solid State NMR 2005, 27, 228-232.
(40) (a) Baldus, M.; Meier, B. H. J. Magn. Reson. A 1996, 121, 65-69.
(b) Helluy, X.; Marichal, C.; Sebald, A. J. Phys. Chem. B 2000, 104,
2836-2845.
(41) Miura, H.; Terao, T.; Saika, A. J. Magn. Reson. 1986, 68, 593-596.
(42) (a) Morris, G. A.; Freeman, R. J. Am. Chem. Soc. 1979, 101, 760-
762. (b) Burum, D. P.; Ernst, R. R. J. Magn. Reson. 1980, 39, 163-
168. (c) Sorensen, O. W.; Ernst, R. R. J. Magn. Reson. 1983, 51,
477-489.
Experimental Section
Solid-state NMR experiments were performed on a Bruker
AVANCE 300 spectrometer at B0 ) 7 T with ν0(31P) ) 121.49
MHz and ν0(29Si) ) 59.63 MHz, using a 4 mm triple-resonance
Bruker MAS probe. The spinning rate was 14 kHz, and samples
were spun at the magic angle using ZrO2 rotors. The setting of the
magic angle was carefully checked. For the MAS-J-INEPT experi-
ments in one and two dimensions, the τ and τ′ delays were
(43) (a) Levitt, M. H. Spin Dynamics, Basics of Nuclear Magnetic
Resonance; Wiley: Chichester, U.K., 2002. (b) Keeler, J. Understand-
ing NMR Spectroscopy; Wiley: Chichester, U.K., 2005.
(44) (a) Soubias, O.; Reat, V.; Saurel, O.; Milon, A. J. Magn. Reson. 2002,
158, 143-148. (b) Alonso, B.; Massiot, D. J. Magn. Reson. 2003,
163, 347-352.
(45) Elena, B.; Lesage, A.; Steuernagel, S.; Bo¨ckmann, A.; Emsley, L. J.
Am. Chem. Soc. 2005, 127, 17296-17302.
(46) Kao, H. M.; Grey, C. P. J. Magn. Reson. 1998, 133, 313-323.
(47) Fyfe, C. A.; Meyer zu Altenschildesche, H.; Wong-Moon, K. C.;
Grondey, H.; Chezeau, J. M. Solid State NMR 1997, 9, 97-106.
(48) Wiench, J. W.; Pruski, M. Solid State NMR 2004, 26, 51-55.
(49) Kao, H. M.; Grey, C. P. J. Am. Chem. Soc. 1997, 119, 627-628.
(50) Lesage, A.; Duma, L.; Sakellariou, D.; Emsley, L. J. Am. Chem. Soc.
2001, 123, 5747-5752.
(52) (a) Krawietz, T. R.; Lin, P.; Lotterhos, K. E.; Torres, P. D.; Barich,
D. H.; Clearfield, A.; Haw, J. F. J. Am. Chem. Soc. 1998, 120, 8502-
8511. (b) Poojary, D. M.; Borade, R. B.; Campbell, F. L.; Clearfield,
A. J. Solid State Chem. 1994, 112, 106-112. (c) Mudrakovskii, I. L.;
Mastikhin, V. M.; Shmachkova, V. P.; Kotsarenko, N. S. Chem. Phys.
Lett. 1985, 120, 424-426.
(51) Mayer, H. Monatsh. Chem. 1974, 105, 46-53.
Inorganic Chemistry, Vol. 46, No. 4, 2007 1381

Coelho et al.
Table 1. 31P and 29Si Isotropic Chemical Shifts for Si5O(PO4)6 (Si5P6)
Table 2. Estimated T2′(29Si) and T2′(31P) Obtained using MAS
Spin-Echo Experiments and eq 1 for the Si5O(PO4)6 (Si5P6) Phase (Si1,
Si2, Si3, and P sites)
and the Various SiP2O7 Polymorphsa
phase
(JCPDS)
δ
iso (ppm)
δ
iso (ppm)
δiso (ppm)
31P MAS
31P INEPT
29Si INEPT
T2′(29Si)
(ms)
T2′(31P)
(ms) 35
Si5O(PO4)6
(70-2071)
-43.8
-43.9
-119.3
-213.5
-217.3
-213.0
Si1-O-P
Si2-O-P
Si3-O-P
41 ( 4
55 ( 4
63 ( 5
65 ( 3
SiP2O7 tetragonal
(22-1320)
SiP2O7 monoclinic 1
(39-0189)
SiP2O7 monoclinic 2
(25-0755)
SiP2O7 cubic
(22-1321)
-45.5
-52.8
-47.6
-55.3
-46.1
-49.4
∼ -50
∼ -58
∼ -70
-45.6
-52.9
-47.7
-55.4
-215.3
Si5O(PO4)6, TEOS (Si(OCH2CH3)4) (44.80 mmol, 5.16 g), ethanol,
and distilled water were used as precursors (TEOS/EtOH/H2O )
1:4:3). The phosphorus precursor H3PO4 (85%) was added (44.8
mmol, 5.16 g) at room temperature. The reaction was slightly
exothermic. The doping by a paramagnetic complex NiCl2·6H2O
(1% molar ratio) was performed for relaxation purposes (see below),
leading to a slightly green solution. After the mixture was stirred
at 25 °C, a transparent green wet gel was obtained (2 h). The final
powder was obtained after heat treatment at 800 °C for 2 h. For
the mixture of the SiP2O7 polymorphs and Si5O(PO4)6, after the
dissolution of 0.05 g (0.22 mmol) of NiCl2·6H2O in ethanol, 5.16
g (44.80 mmol) of H3PO4 were added, followed by 4.67 g (22.40
mmol) of TEOS. A gel was obtained after 2 h at room temperature,
then heated at 100 °C for 48 h, and subsequently, heated at 1000
°C for 2 h. The SiP-136 silicophosphate gel was prepared by the
sol-gel process using TEOS (Si(OCH2CH3)4, ethanol, and distilled
water as precursors (TEOS/EtOH/H2O ) 1:4:3). At room temper-
ature, the phosphorus precursor (H3PO4 85% phosphoric acid) was
added (Si/P ) 1:1).2 The doping by Ni2+ was performed as well,
leading to a slightly green solution. A transparent green wet gel
was obtained (2 h). A powder was obtained after 6 days at 136 °C.
The temperature has a major impact on the involved species. The
choice of 136 °C is related to the presence of both amorphous and
crystalline components. Moreover, the obtained gels are highly
sensitive to air moisture, leading to completely different spectro-
scopic characteristics upon aging. The gel was therefore carefully
kept under a dry atmosphere and subsequently analyzed.
a 31P MAS, single-pulse experiments; 31P/29Si INEPT, projections of the
2D 31P-29Si MAS-J-INEPT experiments.
The addition of NiCl2·6H2O led to a drastic reduction of T1(31P)
(∼1 s). Indeed, T1(31P) have been estimated to ∼40 s for gels and
∼480 s for crystalline silicophosphate phases.22 The percent molar
ratio of paramagnetic complex was adjusted to obtain shortened
T1(31P) and T1(29Si), without drastic modification of the line widths
and of the characteristic T2′ of the various resonances. The notion
of T2′ constant is essential for the evolution of the coherences
involved in the J-derived experiments and will be explained below.
Figure 2. (a) 1D refocused 31P-29Si MAS-J-INEPT pulse sequence. Phase
cycling: Φ1 ) +x, +x, +x, +x, +x, +x, +x, +x, -x, -x, -x, -x, -x,
-x, -x, -x; Φ2 ) +x, -x; Φ3 ) +y, +y, -y, -y; Φ4 ) +x, -x; Φ5 )
+x, +x, +x, +x, +y, +y, +y, +y, -x, -x, -x, -x, -y, -y, -y, -y; Φ6
) +x, -x, +x, -x, +y, -y, +y, -y; receiver ) +x, +x, -x, -x, +y, +y,
-y, -y. (b) Experimental evolution of 29Si MAS-J-INEPT intensities as a
function of τ (with τ′ ) 11 ms) for Si1, Si2, and Si3 (Si5O(PO4)6 phase).
Results and Discussion
Si5O(PO4)6 Crystalline Phase. The structure of
Si5O(PO4)6 described by Mayer51 (trigonal, R3h, a ) 7.869
Å, c ) 24.138 Å) involves one unique P site and three
inequivalent Si sites (two 6-fold coordinated SiVI atoms, Si1
and Si2, and one 4-fold coordinated, SiIV atom Si3). The Si1/
Si2/Si3 ratio is 1:2:2. The structure consists of [SiO6] and
[Si2O7] groups linked by [PO4] groups. Each [PO4] tetrahe-
dron is surrounded by three SiVI atoms (Si2 (×2) and Si1)
and one SiIV atom (Si3) (Figure 1). The Si1 atom is bonded
to six equivalent oxygen atoms O3 (Si1O3) 1.756 Å, Si1O3P
) 145.58°). Consequently, the six involved P-O3-Si1 bonds
are equivalent. The Si2 atom is bonded to six oxygen atoms,
but two inequivalent oxygen atoms (O2 and O5) are involved
(Si2O2 ) 1.791 Å, Si2O5 ) 1.756 Å, Si2O2P ) 131.15°,
Si2O5P) 151.38°). The three P-O4-Si3 bonds involved in
synchronized with the rotor period. The INEPT build-up curves
were established independently for the 4- and 6-fold coordinated
29Si sites, by varying the 29Si offset (to minimize off-resonance
effects) (see below). The same experimental approach was applied
for the measurement of T2′ constants by MAS spin-echo experi-
ments. 31P chemical shifts were referenced to 85% aqueous
H3PO4. 29Si chemical shifts were referenced to TMS. Full experi-
mental details are given in the figure captions.
The synthesis protocols of the Si5O(PO4)6 phase and of the
various polymorphs of SiP2O7 are the following. For crystalline
1382 Inorganic Chemistry, Vol. 46, No. 4, 2007

Refocused 31P-29Si MAS-J-INEPT NMR Experiment
Figure 3. (a-c). Experimental evolution of 29Si MAS-J-INEPT intensities as a function of τ′ (with τ ) 11.4 ms) for Si1, Si2, and Si3 (Si5O(PO4)6 phase),
respectively. The experimental data (O) were fitted by the eqs 2 for Si1 and Si3 and 4 for Si2. The fixed T2′ constants are extracted from spin echo experiments
(Table 2) (Ø, 4 mm; RO, 14 kHz; NS, 656; RD, 5 s; 90°(29Si), 5.3 µs; 90°(31P), 5 µs). (d) Theoretical curves (versus τ′, at fixed τ delay) for the SIn spin
systems (1 e n e 6), following eq 2 with one unique J coupling constant and neglecting all T2′ effects.
the Si3 tetrahedron are equivalent (Si3O4 ) 1.611 Å, Si3O4P
) 138.83°). The fourth Si3-O bond is connected to a SiIV
tetrahedron.
respect to silicon-29 is created after the first τ-π-τ period.
The two simultaneous 90° pulses lead to antiphase silicon-
29 coherence. The refocusing period τ′-π-τ′ is then applied
to obtain in-phase 29Si magnetization. Note that 31P decou-
pling is applied during the acquisition of the 29Si signal.
In the MAS-J-INEPT experiment, the two τ and τ′ delays
must be carefully optimized. Such parameters involve not
The 31P MAS spectrum reveals a unique resonance located
at δ(31P) ) -43.8 ppm, while the 29Si MAS spectrum
exhibits three resonances located at δ(29Si) ) -119.3,
-213.5, and -217.3 ppm (Figure 1 and Table 1). A broad
resonance centered at δ(29Si) ≈ -113.5 ppm associated with
amorphous silica SiO2 (side product) is also evidenced.
The pulse sequence for the solid-state refocused 31P-29Si
MAS-J-INEPT experiment is shown in Figure 2a. The
original solution-state technique42 was adapted here for
rotating solids and for a new pair of nuclei 31P-29Si. Fast
magic angle spinning (usually 14 kHz) averages the chemical
shift anisotropy and the heteronuclear dipolar couplings to
zero, leaving only the scalar couplings and the isotropic
chemical shifts.45 The refocused INEPT pulse sequence
consists of the following steps: a 90° pulse is applied to the
31P channel, followed by an evolution delay optimized to
achieve 31P antiphase magnetization. The refocusing of the
isotropic chemical shifts is obtained by the simultaneous
application of 180° pulses on both phosphorus-31 and silicon-
2
only the value of the JP-O-Si coupling constants but also
the relaxation time T2′ associated to each site. The T2′ time
constant corresponds to the non-refocusable line width
(namely ∆′ ) 1/πT2′) which is usually significantly narrower
than the “apparent” line width. Following the discussion of
Emsley and co-workers,27-28,32c-e,50 we use the notation T2′
which clearly states the experimental nature of this relaxation
time and makes no hypothesis on its underlying mechanism.33
It is possible to estimate the T2′ parameter by using a simple
MAS spin-echo experiment 90°-τ-180°-τ. The corre-
sponding decay curves versus 2τ (not presented here) were
fitted in the time domain by a single-exponential time
constant, according to
Iecho ) exp (-2τ/T2′)
(1)
2
29 channels (only the JP-O-Si couplings have to be taken
into account). For a pair of 31P and 29Si nuclei involved in a
leading therefore to an estimate of T2′ for the 29Si and
31P resonances (see Table 2). For the refocused INEPT
31P-O-29Si group, antiphase phosphorus-31 coherence with
Inorganic Chemistry, Vol. 46, No. 4, 2007 1383

Coelho et al.
τ′max is given by53
τ′max ) 1/2πJ arcsin[n-1/2
]
(3)
The intensity evolution is governed by the product of two
terms, namely, cos(n-1)(2π 2JP-O-Siτ′) (0 for τ′ ) 1/4J) and
sin(2π 2JP-O-Siτ′) (0 for τ′ ) 1/2J). For n ) 1, a simple sine
evolution is observed. An important parameter is the parity
of the multiplicity n. Indeed, for odd n values, the signal
intensity is always positive for τ′ varying between 0 and 1/2J.
For eVen n values, the signal intensity is positive for τ′ ) 0
f 1/4J and then negative for τ′ ) 1/4J f 1/2J. For
increasing eVen n values, a flat evolution of the intensity is
observed around τ′ ≈ 1/4J. The eq 2, as well as the T2′
constants presented in Table 2, were used for the fitting of
the experimental curves for Si1 and Si3. For both 29Si sites,
Figure 4. Theoretical curves (versus τ′, at fixed τ delay) for SI3, SI6, and
SI3I′3 spin systems following eq 4 with various (J1, J2) coupling constants.
T2′(29Si) for Si2 is extracted from spin echo experiments (Table 2).
2
a unique JP-O-Si coupling constant is involved (see Figure
1). The Si1 build-up curve (Figure 3a) is characteristic for
an even n value (positive and negative intensity) with a flat
evolution around τ′ ≈16 ms (i.e., n ) 6). The extracted J
2
experiment and considering a unique JP-O-Si coupling
constant, the signal intensity is proportional to the
2
value corresponds to JP-O-Si ≈ 15 Hz. The Si3 build-up
product43,53-54
curve (Figure 3c) is characteristic of an odd n value (i.e., n
2
) 3). The extracted J value corresponds to JP-O-Si ≈ 12
IINEPT(τ ,τ′) ) I0 sin(2π 2JP-O-Siτ)
Hz. Obviously, the Si2 build-up curve differs drastically from
the one obtained for Si1, even though Si1 and Si2 correspond
both to 6-fold coordinated silicon atoms. In the case of Si2,
two different 2JP-O-Si coupling constants have to be a priori
considered corresponding to Si2-O2-P and Si2-O5-P
bonding paths (see Figure 1). With two distinct J coupling
constants (J1 and J2, SI3I′3 spin system), the product operator
formalism,43 leads to the MAS-J-INEPT intensity given by
sin(2π 2JP-O-Siτ′) cos(n-1)(2π 2JP-O-Siτ′)
× exp(-2τ/T2′(31P)) exp(-2τ′/T2′(29Si)) (2)
where n is the bond multiplicity (i.e., SIn spin system with
S ≡ 29Si and I ≡ 31P). Neglecting all T2′ effects, the 29Si
31P-O-29
signal is maximized by choosing τ ) 1/4J
Si. Figure
2b presents the experimental 29Si signal intensities obtained
by varying τ (τ′ being fixed to 11 ms) for Si1, Si2, and Si3.
An optimum located at τ ) 11.4 ms is observed and will be
used as a constant thereafter. As in solution state NMR, the
MAS-J-INEPT experiment can be safely implemented for
editing purposes. The build-up curves obtained versus τ′ (at
fixed τ ) 11.4 ms) for Si1, Si2, and Si3 are presented in Figure
3a-c. The variation of the intensities is clearly distinct,
depending on both the number of (-OP) bonds around the
silicon atoms (n ) 6 for Si1, Si2; n ) 3 for Si3) and the
crystallographic characteristics of the various sites (for Si1
and Si2). To the best of our knowledge, we present here for
the first time the build-up curves related to SIn spin systems
with n > 3 (in the solid state). Examples dealing with n <
4 are related mainly to CHn groups in solution-state NMR.
Editing is achieved. (i) For τ′ ) 4.3 ms, all resonances are
positive. (ii) For τ′ ) 32.1 ms, the (Si1, Si2) resonances are
almost zero, and the Si3 resonance is positive. (iii) For τ′ )
42.8 ms, all resonances are negative or almost zero. The
theoretical curves corresponding to n ) 1, 2, ..., 6 (SIn spin
systems) are presented in Figure 3d (neglecting all T′2
relaxation effects). A unique J coupling constant is consid-
ered. The maximum transfer ranges from 1 (in arbitrary units)
for a SI spin system to 1.55 for a SI6 spin system.53 Moreover,
IINEPT(τ ,τ′) ) I0 cos2(2πJ1τ′) cos2(2πJ2τ′) [sin(2πJ1τ)
sin(2πJ1τ′) cos(2πJ2τ′) + sin(2πJ2τ) sin(2πJ2τ′)
cos(2πJ1τ′)] × exp(-2τ/T2′(31P)) exp(-2τ′/T2′(29Si)) (4)
For J1 ) J2, eq 4 is comparable to eq 2 with n ) 6. For J2
) 0 Hz, eq 4 is comparable to eq 2 with n ) 3. By
considering the Figures 3a and 3b, we can exclude J1 ≈ J2
because no flat evolution is observed for Si2 as it would be
for a SI6 case. Various theoretical curves using eq 4 are
presented in Figure 4 for various (J1, J2) values. The curves
corresponding to J1 ) J2 and n ) 3, 6 (SI3 and SI6 spin
systems) are also presented. At a fixed J1 value (here J1 )
14 Hz), strong differences in the build-up curves are observed
for J2 varying from 14 to 0 Hz. In particular, the MAS-J-
INEPT intensity remains strictly positive for τ′ e 30 ms and
for J2 < 8 Hz. The fitting of the Si2 curve (Figure 3b) leads
to J1 ≈ 14 Hz and J2 ≈ 4 Hz. The obtained J2 value appears
small, but several authors previously mentioned differences
in J coupling constants, corresponding to different crystal-
lographic paths.33 We note that two distinct bond angles are
involved in the case of Si2 (Si2-O2-P ≈ 131° and Si2-
O5-P ≈ 151°) and could explain the strong difference
between J1 and J2. The errors on the extracted J-coupling
constants can be estimated to (2 Hz.
(53) Pegg, D. T.; Doddrell, D. M.; Brooks, W. M.; Bendall, M. R. J. Magn.
Reson. 1981, 44, 32-40.
(54) Lux, P.; Brunet, F.; Desvaux, H.; Virlet, J. Magn. Reson. Chem. 1993,
31, 623-631.
We have shown that the MAS-J-INEPT experiment was
suitable for the fine description of Si-O-P bonding in terms
1384 Inorganic Chemistry, Vol. 46, No. 4, 2007

Refocused 31P-29Si MAS-J-INEPT NMR Experiment
Figure 5. (a) 1D refocused 31P-29Si MAS-J-INEPT spectrum of
Si5O(PO4)6 (Ø, 4 mm; RO, 14 kHz; NS, 1440; RD, 5 s; 90°(29Si), 5.3 µs;
90°(31P), 5 µs; τ ) 16.6 ms; τ′) 11.1 ms; LB ) 10 Hz). The red lines
correspond to the simulation of the broad components. (b) 29Si MAS
spectrum (single-pulse experiment) of Si5O(PO4)6 (Ø, 4 mm; RO, 14 kHz;
NS, 1760; RD, 10 s; 90°(29Si), 4.5 µs; LB ) 10 Hz).
of J coupling constants. The versatility of the experiment is
further demonstrated by the fine-tuning of the (τ, τ′)
parameters, to reveal interesting chemical details. In that
perspective, the Figure 5 shows the MAS-J-INEPT spectrum
of Si5O(PO4)6 for τ ) 16.6 ms and τ′ ) 11.1 ms, as well as
the corresponding 29Si MAS spectrum (single pulse experi-
ment). In the 29Si MAS spectrum, a broad component located
at δ ≈ -113.5 ppm is observed and primarily assigned to
amorphous SiO2. However, the 31P-29Si MAS-J-INEPT
spectrum exhibits at least two broad slightly shielded
components (centered at δ ≈ -114.5 ppm and δ ≈ -125.5
ppm, respectively, the latter being hardly discernible). These
shielded components could be assigned to Q4(29Si) units
involving at least one 29Si-O-31P group.1d,3
Figure 6. (a) 2D refocused 31P-29Si MAS-J-INEPT pulse sequence. (b)
Structural scheme for SiP2O7 pyrophosphate groups: a given P atom is
bonded to one P atom and three SiVI atoms. (c) 31P and 29Si MAS spectra
(single-pulse experiments) of the mixture of Si5O(PO4)6 and SiP2O7
polymorphs (tetragonal, monoclinic 1, monoclinic 2, and cubic, see Table
1) (31P Ø, 4 mm; RO, 14 kHz; NS, 8; recycle delay (RD), 5 s; 90°(31P), 6
µs; LB ) 0 Hz; 29Si Ø, 4 mm; RO, 14 kHz; NS, 800; RD, 5 s; 90°(29Si),
5.3 µs; LB ) 20 Hz). The 2D refocused 31P-29Si MAS-J-INEPT spectrum
of the mixture of Si5O(PO4)6 and SiP2O7 polymorphs is also shown (Ø, 4
mm; RO, 14 kHz; NS, 496 for each t1 increment; RD, 5 s; 90°(29Si), 5.7
µs; 90°(31P), 4.3 µs; τ ) 11.4 ms; τ′) 4.6 ms; States mode with 128 t1
increments, 88 h, LB ) 20 Hz in F(29Si), LB ) 20 Hz in F(31P). The
expansion of the boxed region is presented in the bottom of the figure. The
projections of the 2D spectrum are also given.
SiP2O7 Polymorphs and 2D MAS-J-INEPT Experi-
ment. The extension of the MAS-J-INEPT experiment to
the 2D spectra is now presented in the frame of the SiP2O7
polymorphs. The synthesis of SiP2O7 phases (see the
Experimental Section) leads generally to mixtures of poly-
morphs and to complex 31P MAS NMR spectra.24a,35 52 We
’
show that the 2D 31P-29Si MAS-J-INEPT experiment
(Figure 6a) is suitable for the characterization of the various
phases involved. The X-ray diffraction (XRD) powder pattern
of the sample (not shown here) indicates that, in addition to
the Si5O(PO4)6 crystalline phase presented above, three
polymorphs of SiP2O7 were synthesized as major constitu-
ents, namely, a tetragonal (JCPDS 22-1320) and two mono-
clinic forms (JCPDS 39-0189 monoclinic 1 and 25-0755
monoclinic 2). The cubic form of SiP2O7 (JCPDS 22-1321)
is also present, as a very minor component. The various
SiP2O7 phases involve pyrophosphate groups (Figure 6b). It
is known from XRD data that the pyrophosphate groups
(involving generally two nonequivalent P sites) are linked
exclusively to SiVI atoms. Figure 6c shows the 31P MAS
spectrum, exhibiting the tetragonal, monoclinic 1, monoclinic
2, and cubic forms of SiP2O7, as well as Si5O(PO4)6. The
assignments of the chemical shifts of the various phases were
previously reported in the literature.35 It shows also the
corresponding 29Si MAS NMR spectrum. Resonances cor-
responding to 4- and 6-fold coordinated Si atoms are
observed, with resolved components in the -210/-220 ppm
region. The 31P and 29Si isotropic chemical shifts of the
various SiP2O7 polymorphs are reported in Table 1.
Figure 6c shows also the 2D 31P-29Si MAS-J-INEPT
spectrum of the Si5O(PO4)6/SiP2O7 mixture (τ ) 11.4 ms
and τ′ ) 4.6 ms). Three cross-peaks associated to Si5O(PO4)6
are observed (the 31P resonance at -43.8 ppm correlates with
three 29Si resonances located at -119.3, -213.5, and -217.3
ppm). Moreover, the 2D spectrum reveals the presence of
four other cross-peaks (shown in the expansion of the 2D
spectrum). The 29Si resonance at δ(29Si) ) -213.0 ppm
correlates with two 31P resonances located at -45.6 and
-52.9 ppm, which are assigned to the tetragonal form of
SiP2O7. The 29Si resonance at δ(29Si) ) -215.3 ppm
correlates with two 31P resonances located at -47.7 and
-55.4 ppm, which are assigned to the monoclinic 1 form of
Inorganic Chemistry, Vol. 46, No. 4, 2007 1385

Coelho et al.
at δ ) -30.8 ppm (hardly discernible because of the
superimposition with a very broad component) can be
assigned to Si(HPO4)2·H2O.52a Both phases were clearly
identified by powder XRD (not shown here) (Si5O(PO4)6,
JCPDS 70-2071; Si(HPO4)2·H2O, JCPDS 18-1168). The
protonation of the phosphate groups in the Si(HPO4)2·H2O
phase has been unambiguously proved by 1H f 31P CP MAS
experiments reported in recent papers.2-3 The 29Si MAS
spectrum of the SiP-136 gel is presented in Figure 7b. A
rather broad peak at δ ≈ -212 ppm is observed and
corresponds to 6-fold coordinated Si atoms. This peak
includes the resonances corresponding to the Si5O(PO4)6 and
Si(HPO4)2·H2O phases. Moreover, a rather sharp line centered
at δ ≈ -119 ppm is superimposed to a much broader line
(centered at δ ≈ -114 ppm). This sharp line is safely
assigned to the pyrosilicate species involved in the
Si5O(PO4)6 structure (Figure 1). We assume that the broad
component corresponds to highly condensed Q4 species with
a certain amount of 29Si-O-31P groups. Figure 7c shows
the 1D 31P-29Si MAS-J-INEPT spectrum of the SiP-136 gel
(τ ) 11.4 ms and τ′ ) 4.6 ms). The obtained spectrum
exhibits two major resonances as expected, which are
assigned to 4- and 6-fold coordinated Si atoms (correspond-
ing to Si5O(PO4)6 and Si(HPO4)2·H2O). The broad component
centered at δ ≈ -114 ppm is efficiently suppressed,
demonstrating again the efficiency of the INEPT sequence
in terms of editing. Si-O-P linkages are not clearly seen
(though the signal-to-noise ratio may not be sufficient). For
this particular spectrum, the experimental time was 4 h,
precluding therefore the use of the 2D MAS-J-INEPT
correlation sequence. In a near future, efficient 1H decoupling
under fast MAS will be implemented to drastically increase
the lifetime of the coherences involved in the MAS-J-INEPT
sequence.27-28,32c-e,50 2D spectra should then be obtained in
a reasonable time.
Figure 7. (a) 31P MAS spectrum of the SiP-136 gel (Ø, 4 mm; RO, 14
kHz; NS, 16; RD, 5 s; 90°(31P), 4.5 µs; LB ) 20 Hz). (b) 29Si MAS spectrum
of the SiP-136 gel (Ø, 4 mm; RO, 14 kHz; NS, 1332; RD, 5 s; 90°(29Si),
1
5.7 µs; LB ) 20 Hz; H decoupling). (c) 1D refocused 31P-29Si MAS-J-
INEPT spectrum of the SiP-136 gel (Ø, 4 mm; RO, 14 kHz; NS, 2640;
RD, 5 s; 90°(29Si), 5.7 µs; 90°(31P), 4.3 µs; τ ) 11.4 ms; τ′ ) 4.6 ms; LB
) 100 Hz).
SiP2O7 (Table 2). The 29Si isotropic shift corresponding to
each SiP2O7 polymorph is determined with great accuracy.
The absence of observable cross-peaks for the monoclinic 2
and the cubic forms is surely related to the low amount of
these particular phases or to the much lower 2JP-O-Si coupling
constants and shorter T2′ constants. Such conclusions were
also derived in the case of the MAS-J-HMQC experiment.35
The HMQC and INEPT sequences are comparable when the
experimental time is considered. However, the INEPT
approach allows the clear distinction of the various Si
resonances in terms of J coupling constants.
SiP-136 Silicophosphate Gel. Finally, the efficiency of
the 31P-29Si MAS-J-INEPT technique is demonstrated for
the study of complex systems, such as silicophosphate gels.
The 31P MAS spectrum of the SiP-136 gel is shown in Figure
7a. Broad resonances are observed and correspond to Q′N
units (Q′N stands for P(O)(OY)N(OY)3-N entities, with Y )
P, Si ; Y′ ) H).1d,3 In addition to a resonance at ∼0 ppm
corresponding to H3PO4 (Q′0), three complex resonances are
located at δ ≈ -11, -35, and -44 ppm. Shoulders and
minor components are also found. All these resonances can
be assigned to Q′N species with N ranging from 0 to 3. The
resonance at -11 ppm can be assigned to the Q′1 units, but
the chemical nature of the involved species is unfortunately
not clearly defined (the notation Q′1 implies one P-O-P
bond and several P-(OX) bonds). The resonance at -44
ppm can be safely assigned to the crystalline Si5O(PO4)6
phase described above. Moreover, the resonance centered
Conclusion
We have shown that the extension of the original refocused
INEPT pulse scheme to the 31P-29Si MAS-J-INEPT experi-
ment is suitable for the description of silicophosphate
derivatives (crystalline phases and amorphous gels). Complex
spin systems, involving up to seVen spins, were taken into
account for the detailed analysis of the INEPT build-up
2
curves. JP-O-Si coupling constants were shown to depend
on both the number of (-OP) bonds around the Si atoms
and the various crystallographic paths involved. At this stage,
first principles calculations of such coupling constants would
be of crucial interest for the fine description of the P-O-
Si bonds.55 Moreover, such calculations could be validated
by J-derived techniques, such as the MAS-J-INEPT experi-
ment described above. From a fundamental point of view,
the Si5P6 phase and the SiP2O7 polymorphs are characterized
by very interesting SIn spin systems with n e 6. This
particular situation is rarely encountered, and such spin
(55) (a) Pickard, C. J.; Mauri, F. Phys. ReV. B 2001, 63, 245101-1-13. (b)
Gervais, C.; Dupree, R.; Pike, K.; Bonhomme, C.; Profeta, M.; Pickard,
C. J.; Mauri, F. J. Phys. Chem. A 2005, 109, 6960-6969.
1386 Inorganic Chemistry, Vol. 46, No. 4, 2007

Refocused 31P-29Si MAS-J-INEPT NMR Experiment
systems could be used as experimental standards for the set
up of “optimized sequences”.56 Indeed, such sequence are
designed for achieving the maximum transfer of polarization
between the S spin (here, 29Si) and the n spins I (here, 31P).
The versatility of the MAS-J-INEPT approach in 1D and
2D versions has been demonstrated for the study of complex
mixtures of crystalline phases and ill-crystallized gels obtai-
ned by sol-gel routes. Work is in progress in the laboratory
for the extension of the 31P-29Si INEPT approach to the
study of biomaterials involving silica and phosphate groups.
(56) Glaser, S. J.; Schulte-Herbru¨ggen, T.; Sieveking, M.; Schedletzky, O.;
Nielsen, N. C.; Sorensen, O. W.; Griesinger, C. Science 1998, 208,
421-424.
IC061964F
Inorganic Chemistry, Vol. 46, No. 4, 2007 1387