Aluminum Phenylphosphonates: A Fertile Family of Compounds

Article abstract of DOI:10.1021/ic9802350

Six aluminum phenylphosphonates have been synthesized depending upon the synthetic conditions: Al₂(O₃PC₆H₅) ₃·2H₂O (I), Al₂(O₃PC₆H₅)₃ (II), α-Al(HO₃PC₆H₅)(O₃PC ₆H₅)·H₂O (III), β-Al(HO₃PC₆H₅)(O₃PC ₆H₅)·H₂O (IV), Al(HO₃PC₆H₅)₃·H₂O (V), and Al(OH)(O₃PC₆H₅) (VI). Thermal analysis, X-ray powder thermodiffractometry, IR spectroscopy, and ²⁷Al and ³¹P MAS NMR data have been obtained to study the structure and thermal stability of these materials. III crystallizes in the orthorhombic system, space group Pbca, with a = 9.7952(1) A?, b = 29.3878(4) A?, c = 9.3537(3) A?, and Z = 8. The structure was solved ab initio, from synchrotron data (λ ≈ 0.4 A?), using direct methods, and refined by Rietveld methods. The final agreement factors were R_wP = 6.73%, R_P = 5.24%, and R_F = 6.8%. The compound is layered with the aluminum atoms in an octahedral environment of oxygens and two crystallographically independent phosphonate groups, one being protonated. The powder patterns of V and VI have been indexed, and the experimental observations are consistent with layered structures. The unit cell of V contains one octahedral site for Al and three tetrahedral sites for P. Phosphonate I seems to have a three-dimensional tubular structure with aluminum atoms in both octahedral and tetrahedral environments and phosphorus atoms in three different types of sites.

Full text of DOI:10.1021/ic9802350

4168
Inorg. Chem. 1998, 37, 4168-4178
Ar ticles
Aluminum Phenylphosphonates: A Fertile Family of Compounds
Aurelio Cabeza,† Miguel A. G. Aranda,† Sebastian Bruque,*,† Damodara M. Poojary,‡,§
Abraham Clearfield,‡ and Jesus Sanz|
Departamento de Qu´ımica Inorga´nica, Universidad de Ma´laga, 29071 Ma´laga, Spain, Department of
Chemistry, Texas A&M University, College Station, Texas 77843, and Instituto de Ciencia de
Materiales, CSIC, Cantoblanco, 28049 Madrid, Spain
ReceiVed March 5, 1998
Six aluminum phenylphosphonates have been synthesized depending upon the synthetic conditions: Al2(O3-
PC6H5)3‚2H2O (I), Al2(O3PC6H5)3 (II), R-Al(HO3PC6H5)(O3PC6H5)‚H2O (III), â-Al(HO3PC6H5)(O3PC6H5)‚H2O
(IV), Al(HO3PC6H5)3‚H2O (V), and Al(OH)(O3PC6H5) (VI). Thermal analysis, X-ray powder thermodiffractometry,
IR spectroscopy, and 27Al and 31P MAS NMR data have been obtained to study the structure and thermal stability
of these materials. III crystallizes in the orthorhombic system, space group Pbca, with a ) 9.7952(1) Å, b )
29.3878(4) Å, c ) 9.3537(3) Å, and Z ) 8. The structure was solved ab initio, from synchrotron data (λ ≈ 0.4
Å), using direct methods, and refined by Rietveld methods. The final agreement factors were RwP ) 6.73%, RP
) 5.24%, and RF ) 6.8%. The compound is layered with the aluminum atoms in an octahedral environment of
oxygens and two crystallographically independent phosphonate groups, one being protonated. The powder patterns
of V and VI have been indexed, and the experimental observations are consistent with layered structures. The
unit cell of V contains one octahedral site for Al and three tetrahedral sites for P. Phosphonate I seems to have
a three-dimensional tubular structure with aluminum atoms in both octahedral and tetrahedral environments and
phosphorus atoms in three different types of sites.
Introduction
nates,4-8 and more recently about trivalent metal phospho-
nates.7b,9,10 The general interest in the chemistry of metal
organophosphonates is mainly due to the unusual compositional
and structural diversity varying from one-dimensional arrange-
ments to three-dimensional microporous frameworks, passing
by the most common layered networks. The importance of such
systems in several research areas such as electrochemistry,11,12
microelectronics,13 biological membranes,14,15 photochemical
mechanisms16 and catalysis17,18 has been widely recognized.
Several studies of trivalent metal phenylphosphonates have
been carried out. Two iron(III) derivatives, FeH(HO3PC6H5)4
and Fe(HO3PC6H5)(O3PC6H5)‚H2O, were synthesized, and the
structure of the first compound was solved from single-crystal
data. The main structural features are linear chains with iron
The metal phosphonates are a class of inorganic-organic
hybrid materials, characterized by the presence of covalent bonds
between the inorganic and organic moieties. In this family, the
chemistry of metal phosphonates has received increasing
attention for the last 15 years. The main research effort in the
metal phosphonates field was initially directed toward tetravalent
cations. A wide variety of tetravalent metal phosphonates (for
instance, Zr, Ti, Sn, and Ce) have been known since the early
1980s.1 The crystal structure of these materials was assumed
to be related to that of R-zirconium phosphate, Zr(O3POH)2‚
H2O.2 Later, the structure of zirconium phenylphosphonate,
Zr(O3PC6H5)2, was solved from powder diffraction data3 and
the close relationship between the structures became clear. Many
other investigations have been reported about the synthesis,
crystal structures, and properties of divalent metal phospho-
(4) Cunningham, D.; Hennelly, P. J.; Deeny, T. Inorg. Chim. Acta 1979,
37, 95-102.
(5) Cao, G.; Lee, H.; Lynch, V. M.; Mallouk, T. E. Inorg. Chem. 1988,
27, 2781-2785.
*
To whom correspondence should be addressed.
† Universidad de Ma´laga.
(6) Martin, K. J.; Squattrito, P. J.; Clearfield, A. Inorg. Chim. Acta 1988,
155, 7-9.
‡ Texas A&M University.
§ Current address: Symyx Technologies, 3100 Central Expressway, Santa
Clara, CA 95051.
(7) (a) Cao, G.; Lee, H.; Lynch, V. M.; Mallouk, T. E. Solid State Ionics
1988, 26, 63-69. (b) Cao, G.; Lee, H.; Lynch, V. M.; Swinnea, J. S.;
Mallouk, T. E. Inorg. Chem. 1990, 29, 2112-2117.
(8) (a) Zhang, Y.; Clearfield, A. Inorg. Chem. 1992, 31, 2821-2826. (b)
Bhardwaj, C.; Hu, H.-L.; Clearfield, A. Inorg. Chem. 1993, 32, 4294-
4299.
| CSIC.
(1) (a) Alberti, G.; Costantino, U.; Allulli, S.; Tomasini, N. J. Inorg. Nucl.
Chem. 1978, 40, 1113. (b) Dines, M. B.; Giacomo, P. D. Inorg. Chem.
1981, 20, 92-97. (c) Drumel, S.; Penicaud, V.; Deniaud, D.; Bujoli,
B. Trends Inorg. Chem. 1996, 4, 13-25. (d) Clearfield, A. In Progress
in Inorganic Chemistry; Karlin, K. D.; Ed.; John Wiley: New York,
1998, pp 371-510.
(2) Clearfield, A.; Troup, J. M. Inorg. Chem. 1977, 16, 3311-3314.
(3) Poojary, D. M.; Hu, H.-L.; Campbell, F. L., III; Clearfield, A. Acta
Crystallogr. 1993, B49, 996-1001.
(9) Bujoli, B.; Palvadeau, P.; Rouxel, J. Chem. Mater. 1990, 2, 582-
589.
(10) Wang, R.; Zhang, Y.; Hu, H.; Frausto, R. R.; Clearfield A. Chem.
Mater. 1992, 4, 864-871.
(11) Murray, R. W. Acc. Chem. Res. 1980, 13, 135-141.
(12) Facci, J. S. Langmuir 1987, 3, 525-530.
S0020-1669(98)00235-3 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/06/1998

Aluminum Phenylphosphonates
Inorganic Chemistry, Vol. 37, No. 17, 1998 4169
atoms in an octahedral coordination of oxygens.9 The structure
of the second compound could not be determined although it
seems to be related to that of R-ZrP. A series of phenyl- and
benzylphosphonates of the lanthanide elements were reported,
and the structures of La(HO3PC6H5)(O3PC6H5) and La(HO3-
PCH2C6H5)(O3PCH2C6H5)‚2H2O were solved from single-
crystal data.10 The synthesis of a second type of rare-earth
phenylphosphonates of general composition Ln2(O3PC6H5)3 was
also reported.10 Other reports describing trivalent metal phos-
phonates are those of V organodiphosphonates,19 the series of
compounds with general formula LnH(O3PR)2 (R ) alkyl,
phenyl; Ln ) La, Sm, Ce),8 and bismuth (carboxyethyl)-
phosphonate.20
Thermal analysis (TGA and DTA) was carried out in air on a Rigaku
Thermoflex apparatus at a heating rate of 10 K min-1 with calcined
Al2O3 as reference. IR spectra were recorded on a Perkin-Elmer 883
spectrometer in the spectral range 4000-400 cm-1, by using KBr
pellets.
Room-temperature powder diffraction patterns were collected with
a Siemens D-5000 automated diffractometer using graphite-monochro-
mated Cu KR radiation. The powder thermodiffractometric studies were
carried out with the same diffractometer but in a second goniometer
permanently equipped with an HTK10 heating chamber. The samples
were mounted directly on Pt foil, which is both the heating system
and the holder. The thermodiffractometric patterns were scanned over
the angular range 3-38° (2θ), with a step size of 0.05° and counting
1 s per step. The appropriate heating and cooling temperatures were
selected by using the Diffract AT software. Samples were held at
temperature 10 min, before recording any pattern, to ensure that any
transformations that may take place were allowed time to complete.
There are several reports dealing with aluminum phospho-
nates. The research effort has been focused on phenyl- and
21
methylphosphonates. Al2(O3PCH3)3 has at least two poly-
For Al(HO3PC6H5)(O3PC6H5)‚H2O, high-resolution synchrotron pow-
der data were collected on the diffractometer of the BM16 line of ESRF.
The sample was loaded into a borosilicate glass capillary (φ ) 0.5
mm) and rotated during data collection. The pattern was collected with
λ ) 0.39989(2) Å, in the angular range 1-30° in 2θ, for an overall
counting time of 10 h. Further experimental details regarding data
collection have already been reported.27
morphic forms. The phases show a tridimensional structure with
unidimensional channels, in which the methyl groups are situated
toward the interior of the pores, resulting in hydrophobic
microtubes.22,23 A layered aluminum methylphosphonate,
AlOH(O3PCH3)‚H2O, has also been reported,24 and two alu-
minum phenylphosphonates were very recently reported, Al(HO3-
PC6H5)(O3PC6H5)2‚H2O25 and Al2(O3PC6H5)3‚4H2O,26 although
their crystal structures were not determined.
In this paper, we report on a study of the Al(III) phenylphos-
phonate system. Six phosphonates have been isolated, depend-
ing upon the synthetic conditions. One structure has been solved
ab initio from powder diffraction data, and 27Al and 31P MAS
NMR and powder X-ray thermodiffractometry have been used
to gain some insights into the structures of these materials.
27Al and 31P MAS NMR spectra were recorded at 104.26 and 161.98
MHz, respectively, with a Bruker MSL-400 spectrometer. The external
magnetic field was 9.4 T. All measurements were carried out at 295
K, and the samples were spun around the magic angle (54°44′ with
respect to the magnetic field) at a spinning rate of 12 kHz. The NMR
spectra were obtained after π/8 (Al signal) and π/2 (P signal) excitations
(2 and 4 µs, respectively) and intervals between successive accumula-
tions of 5 and 30 s in both signals. The 27Al and 31P chemical shift
values are given relative to 1 M AlCl3 and 85% H3PO4 aqueous
solutions, respectively. In all cases, the mean error on chemical shift
values of components was lower than 0.3 ppm. The deconvolution of
the MAS NMR spectra was carried out by using the WINFIT program
(from Bruker) in order to determine the shape and intensities of the
components.
Experimental Section
Chemicals of reagent quality were obtained from commercial sources
and were used without further purification. The synthesized phospho-
nate samples were dissolved in HNO3 (1/1), and the aluminum content
was determined by atomic absorption spectrometry. Carbon and
hydrogen contents were determined by elemental chemical analysis in
a Perkin-Elmer 240 analyzer. The phosphorus content was deduced
from the carbon percentage found, assuming a C/P molar ratio of 6/1.
The water content was determined from the weight loss in the heating
process.
.
Synthesis of Al2(O3PC6H5)3 2H2O (I). Two solutions which
contained respectively 7.43 mmol of H2O3PC6H5 in absolute ethanol
and 1.48 mmol of anhydrous AlCl3 in absolute ethanol were mixed
(P/Al molar ratio: 5/1). The mixture had a pH of <3 to avoid the
hydrolysis of aluminum, and it was refluxed for a week (final volume
40 mL). The resulting white suspension was centrifuged, washed with
ethanol several times, dried in air, and stored at 58% relative humidity
(atmosphere from 40% w/w aqueous H2SO4). Anal. Found: Al, 9.43;
P, 16.66; C, 38.76; H, 3.49. Calcd for I: Al, 9.67; P, 16.65; C, 38.72;
H, 3.41. When I is heated above 80 °C, it is dehydrated to give Al2(O3-
PC6H5)3 (II).
(13) Roberts, G. G. AdV. Phys. 1985, 34, 475-512.
(14) Ringsdorf, G.; Schmidt, X.; Schneider, J. Thin Solid Films 1987, 152,
207-222.
(15) Fendler, J. H. Membrane Mimetic Chemistry; Wiley: New York, 1982.
(16) (a) Gratzel, M. Pure Appl. Chem. 1982, 54, 2369. (b) Thomas, J. K.
Acc. Chem. Res. 1988, 21, 275-280. (c) Colon, J. L.; Yang, C.-Y.;
Clearfield, A.; Martin, C. R. J. Phys. Chem. 1988, 92, 5777; 1990,
94, 874. (d) Vermeulen, L. A.; Thompson, M. E. Nature 1992, 358,
656. (e) Vermeulen, L. A.; Snover, J. L.; Sapochak, L. S.; Thompson,
M. E. J. Am. Chem. Soc. 1993, 115, 11767-11774. (f) Poojary D.
M.; Vermeulen, L. A.; Vicenzi, E.; Clearfield, A.; Thompson, M. E.
Chem. Mater. 1994, 6, 1848.
Synthesis of (a) r-Al(HO3PC6H5)(O3PC6H5)‚H2O (III). A 1.27
mmol sample of AlCl3‚6H2O and 12.66 mmol of H2O3PC6H5 (P/Al
molar ratio: 10/1) were dissolved in 30 mL of water, and then 2 mL
of 30% NaOH was added. The mixture was introduced into a 50 mL
Teflon-lined autoclave, which was sealed and placed in an oven at 200
°C for 9 days. A single-phase sample was isolated by filtration, washed
with water and acetone, and finally dried under vacuum. Anal.
Found: C, 40.35; P, 17.37; H, 3.65. Calcd for III: C, 40.22; P, 17.31;
H, 3.63.
(17) Soghomonian, V.; Haushalter C.; Zubieta, J. Chem. Mater. 1995, 7,
1648-1654.
(18) Janvier P.; Drumel, S.; Piffard, Y.; Bujoli, B. C. R. Acad. Sci. Ser. II
1995, 320, 29-35.
(b) â-Al(HO3PC6H5)(O3PC6H5)‚H2O (IV). To 25 mL of a 0.1 M
solution of Al2(SO4)3‚18H2O was added 3.9525 g of H2O3PC6H5
dissolved in 25 mL of water (P/Al molar ratio: 10/1). The pH was
increased from 1 to 1.5 through addition of 30% aqueous NaOH,
resulting in the precipitation of a white solid. The total mixture was
heated in a Teflon-lined autoclave at 200 °C for 3 days. A single
powdered phase was isolated by filtration. The powder diffraction
pattern of IV was different from that shown by III. Anal. Found: C,
41.28; P, 17.85; H, 3.44. Calcd for IV: C, 40.22; P, 17.31; H, 3.63.
(19) (a) Maeda, K.; Kiyozumi, Y.; Mizukami, F. Angew. Chem., Int. Ed.
Engl. 1994, 33, 2335-2337. (b) Maeda, K.; Kiyozumi, Y.; Mizukami,
F. J. Phys. Chem. B 1997, 101, 4402-4412. (c) Carter, V. J.; Wright,
P. A.; Gale, J. D.; Morris, R. E.; Sastre, E.; Perez-Pariente, J. J. Mater.
Chem. 1997, 7, 2287-2292.
(20) Maeda, K.; Akimoto, J.; Kiyozumi, Y.; Mizukami, F. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 1199-2001.
(21) Maeda, K.; Akimoto, J.; Kiyozumi, Y.; Mizukami, F. J. Chem. Soc.,
Chem. Commun. 1995, 1033-1034.
(22) Sawers L. J.; Carter, J. V.; Armstrong, A. R.; Bruce, P. G.; Wright,
P. A.; Gore, B. E. J. Chem. Soc., Dalton Trans. 1996, 3159-3161.
(23) Haky, J. E.; Brady, J. B.; Dando N.; Weaver, D. Mater. Res. Bull.
1997, 32, 297-303.
(25) Aranda, M. A. G.; Losilla, E. R.; Cabeza, A.; Bruque, S. J. Appl.
Crystallogr. 1998, 31, 16-21.
(24) Raki, L.; Detellier, C.; Chem. Commun. 1996, 21, 2475-2476.

4170 Inorganic Chemistry, Vol. 37, No. 17, 1998
Cabeza et al.
stable up to 40 °C, above which it begins to lose water. Between
45 and 55 °C, I and II coexist, and at 60 °C a single phase of
II is observed. Although II is present at room temperature on
cooling, we have observed by X-ray powder diffraction, that
the rehydration process does take place partially over several
days, by absorbing water from the atmosphere. It is important
to point out that the most intense peak for I, at 13.2 Å, is
displaced to 11.7 Å after the weight loss. Simultaneously, the
highest d-spacing peak moves from 26.1 to 23.2 Å and its
relative intensity increases notably.
(b) r-Al(HO3PC6H5)(O3PC6H5)‚H2O. The DTA-TGA curves
for III are shown in Figure 3. The first endotherm centered at
183 °C is due to the loss of hydration water (the observed and
calculated weight losses are 6.0 and 5.0%, respectively). There
are three small endotherms at 235, 275, and 310 °C, without
mass losses, which must be due to subtle phase transitions. The
second mass loss is due to the pyrolysis of the phenyl groups,
and it can be seen as a main endotherm at 537 °C with two
small associated endotherms at 400 and 460 °C. The overall
observed mass loss, 45.5%, agrees well with the theoretical loss
of 46.1% deduced from the overall thermal decomposition
reaction (2). There is no evidence of combustion of the organic
radicals in air as no exothermal effect is observed in the DTA
curve.9,38
Figure 1. TGA-DTA curves for Al2(O3PC6H5)3‚2H2O.
.
Synthesis of Al(HO3PC6H5)3 H2O (V). A mixture of 3.79 mmol
of anhydrous AlCl3 and 18.97 mmol of H2O3PC6H5 (P/Al molar ratio:
5/1) in 50 mL of toluene was refluxed in a round-bottom flask for 2
weeks. The product was isolated by filtration as a white powder. It
was washed with ethanol and acetone several times and dried in air.
Anal. Found: Al, 4.5; C, 43.37; P, 18.67; H, 3.87. Calcd for V: Al,
5.3; C, 41.86; P, 18.02; H, 3.64.
Synthesis of Al(OH)(O3PC6H5) (VI). To 25 mL of an aqueous
solution containing 18.96 mmol of AlCl3‚6H2O was added 25 mL of a
0.25 M solution of H2O3PC6H5 (P/Al molar ratio: 1/3). Initially no
precipitate was formed. The resulting solution was introduced into a
Teflon-lined autoclave, with stirring, and heated at 200 °C for 5 days.
A white powder was isolated by filtration. It was washed with water
and acetone several times and dried in air. This compound can also
be obtained from Al2(O3PC6H5)3‚2H2O (I) by heating 170 mg with 30
mL of water in an autoclave at 200 °C for 10 days. Anal. Found: C,
35.29; P, 15.19; H, 2.40. Calcd for VI: C, 35.98; P, 15.49; H, 2.99.
2R-Al(HO3PC6H5)(O3PC6H5)‚H2O + O2 f
Al2O3‚2P2O5 + 3H2O + [4C6H5] (2)
The X-ray powder diffraction pattern of the sample heated
at 800 °C showed no peaks, indicating that the thermal
decomposition of III leads to a glassy solid. The loss of the
hydration water was studied by powder thermodiffractometry
(Figure 4). Above 160 °C, the water is lost, and the resulting
powder patterns have much broader peaks and much lower
intensities. It is important to underline that the d-spacing value
for the first reflection (020), 14.6 Å, does not change during
this process. On heating, there are no further changes in the
powder pattern although several endotherms are evident in the
DTA curve of this material.
Results
.
Thermal Study. (a) Al2(O3PC6H5)3 2H2O. TGA-DTA
.
(c) Al(HO3PC6H5)3 H2O. The DTA-TGA curves for V are
curves for I are shown in Figure 1. One endothermic event
centered at 80 °C and three exotherms centered at 600, 840,
and 950 °C are observed. The endothermic effect has an
associated weight loss of 6.4%, which corresponds to the release
of the two water molecules (calculated mass loss 6.45%). The
first strong exotherm is due to the combustion of the organic
groups. The other two exothermic effects correspond to
crystallizations and decompositions. The total mass loss at 1050
°C, 46.2%, is slightly larger than the theoretical value, 43.6%,
deduced from the overall thermal decomposition reaction (1).
given in Figure 5. The first endotherm centered at 240 °C, with
an associated mass loss of 4.0%, corresponds to the release of
the water molecule (calculated value 3.5%). There are two
endotherms at 470 and 512 °C, which are due to the pyrolysis
of the phenyl organic groups. As observed for R-Al(HO3-
(26) Garvey, R. LATTPARM autoindexing program. Department of
Chemistry, North Dakota State University, Fargo, ND 58105-5516.
Based on the Visser algorithm: Visser, L. W. J. Appl. Crystallogr.
1969, 2, 89.
(27) Werner, P. E.; Eriksson, L.; Westdahl, M. J. Appl. Crystallogr. 1985,
18, 367-370.
2Al2(O3PC6H5)3‚2H2O + 45O2 f
(28) Le Bail, A.; Duroy, H.; Fourquet, J. L. Mater. Res. Bull. 1988, 23,
447-452.
Al(PO3)3 + 3AlPO4 + 36CO2 + 19H2O (1)
(29) Larson, A. C.; von Dreele, R. B. (Program Version: PC, summer 96).
Los Alamos Natl. Lab. [Rep.] LA (U.S.) 1994, LA-UR-86-748.
(30) Thompson, P.; Cox, D. E.; Hasting, J. B. J. Appl. Crystallogr. 1987,
20, 79-83.
(31) Finger, L. W.; Cox, D. E.; Jephcoat, A. P. J. Appl. Crystallogr. 1994,
27, 892-900.
(32) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A.; Burla,
M.; Polidori, G.; Camalli, M. SIRPOW.92. J. Appl. Crystallogr. 1994,
27, 435-436.
(33) Dollase, W. A. J. Appl. Crystallogr. 1986, 19, 267.
(34) Rietveld, H. M. J. Appl. Crystallogr. 1969, 2, 65-71.
(35) Morris, M. C.; McMurdie, H. F.; Evans, E. H.; Paretzkin, B.; deGroot,
J. H.; Newberry, R.; Pyrros, N. P.; Hubbard, C. R.; Carmel, S. NBS
Monogr. (U.S.) 1977, No. 25, 1-5.
(36) Poojary, D. M.; Baolong, Z.; Cabeza, A.; Aranda, M. A. G.; Bruque,
S.; Clearfield, A. J. Mater. Chem. 1996, 6, 639-644.
The X-ray powder diffraction pattern of the sample heated
at 750 °C is amorphous with some low-intensity peaks.
However, the pattern of the sample heated at 1050 °C is
crystalline and shows diffraction peaks which correspond to Al-
(PO3)3 and AlPO4 (PDF database nos. 13-0430 and 11-0500,
respectively). There are also some very small peaks which
belong to a third unidentified phase, which may be responsible
for the difference between the observed and calculated overall
mass losses.
A thermodiffractometric study (Figure 2) was undertaken to
characterize the water loss of compound I. Compound I is only

Aluminum Phenylphosphonates
Inorganic Chemistry, Vol. 37, No. 17, 1998 4171
Figure 2. X-ray thermodiffractometric powder patterns for Al2(O3PC6H5)3‚2H2O.
PC6H5)(O3PC6H5)‚H2O, there is no evidence of combustion of
the organic radicals in air in the DTA curve. The overall
observed mass loss, 50.0%, is in good agreement with the
theoretical value, 48.8%, deduced from the overall thermal
decomposition reaction (3).
loss of 4.4%. This value is very closed to that calculated for
the condensation of two hydroxide groups, 4.5%, of Al(OH)-
(O3PC6H5) to release a water molecule. This transformation
causes an increment in the d spacing of the first diffraction peak
from 16.45 to 17.36 Å. The second endotherm, without mass
loss, must be due to subtle phase transitions. The exothermic
effects correspond to the combustion of the phenyl groups.
The powder pattern of a sample obtained by heating VI at
1000 °C is amorphous, with some low-intensity peaks. How-
ever, when the sample is heated to 1100 °C for 2 h, the resulting
compound is crystalline and it shows diffraction peaks which
mainly correspond to AlPO4 (PDF no. 20-0045). The overall
experimental weight loss is 40.8%, in good agreement with the
theoretical weight loss of 39.02% for the decomposition reaction
(4).
2Al(HO3PC6H5)3‚H2O + 3/2O2 f
Al2O3‚3P2O5 + 5H2O + [6C6H5] (3)
The powder pattern of the sample heated at 800 °C also
showed no peaks, indicating that the thermal decomposition of
V results in an amorphous solid. The loss of the hydration water
was also studied by powder thermodiffractometry (Figure 6).
Above 210 °C, the water is lost; the resulting powder patterns
have fewer peaks, and they are broader. The d-spacing value
for the first reflection (020), 14.5 Å, does not change during
the dehydration process (as for III), but a large new reflection
is present at 2θ ≈ 7°.
Al(OH)(O3PC6H5) + 15/2 O2 f AlPO4 + 3H2O + 6CO2
(4)
For both R-Al(HO3PC6H5)(O3PC6H5)‚H2O and Al(HO3-
PC6H5)3‚H2O samples, the thermal decomposition reactions
including oxygen and organic radicals are possible processes
compatible with the observed mass losses. However, further
studies are needed to confirm this suggested mechanism and to
rule out other decomposition reactions. For example, with the
experimental data it is not possible to discard a decompositional
reaction without oxygen participation. However, in this case,
the vitreous phosphates should contain phosphorus in V and
III oxidation states with overall stoichiometries Al2O3‚P4O8 and
Al2O3‚P6O12, respectively.
IR Spectroscopic Study. The IR spectra were recorded
between 4000 and 400 cm-1 and are shown in Figure 8 between
4000 and 1400 cm-1. This region was selected to study in detail
the hydration water and the H-O-P groups. In Figure 8a-d
are shown the spectra for I, III, V, and VI, respectively. One
important feature in the IR spectra is the band observed at 2340
cm-1 for V (curve c) that appears as a shoulder near 2400 cm-1
for III (curve b) and is absent for I and VI (curves, a and d,
respectively). This band is characteristic of the hydrogen
phenylphosphonate groups and it is more intense for compound
V, which has three such groups per formula. The bending
H-O-H vibration of the hydration water located at ∼1620
cm-1 is always partially overlapped with the sharp band due to
the phenyl group stretchings (ν C-C) at 1595 cm-1. The bands
(d) Al(OH)(O3PC6H5) (VI). The DTA-TGA curves for VI
are given in Figure 7. Two endothermic events, centered at
280 and 400 °C, and two exothermic events, at 660 and 730
°C, are observed. The first endotherm has an associated mass

4172 Inorganic Chemistry, Vol. 37, No. 17, 1998
Cabeza et al.
weakly through H-bonds. When I is heated above 100 °C to
give II, these bands and that due to the H-O-H bending
vibration disappear. The IR spectrum for III presents one band
at 3415 cm-1, showing that the water interacts notably by
H-bonding, and another at 3070 cm-1, due to the OH group.
The IR spectrum for V has a main band at 3250 cm-1 with a
shoulder at 3500 cm-1, showing that this water interacts quite
strongly by H-bonding. Compound VI shows a strong sharp
band at 3660 cm-1, characteristic of the stretching vibration of
the hydroxide groups (νO-H), and no band is observed in the
H-O-H bending region, close to 1640 cm-1
.
.
Crystal Structure Study. (a) Al2(O3PC6H5)3 2H2O. This
phosphonate shows a high d-spacing value for the first reflection,
26.3 Å. This peak is very sensitive to the humidity, and its d
spacing decreases quickly to give II when it is placed in a dry
atmosphere or is heated above 50 °C. The compound has to
be stored in a controlled humidity environment (i.e. H2SO4,
40%) to avoid the loss of water.
The X-ray powder patterns for both phases (Figure 2) do not
show sharp diffraction peaks mainly due to the small size of
the crystallites. This problem and the complexity of the
structure (with a very large unit cell) have made it difficult to
index the patterns. LATTPARM28 and TREOR9029 autoindex-
ing programs suggested several possible solutions. We have
selected those shown below because they result in good values
of volume per atom (Vat) and they show a clear relation between
the hydrated and anhydrous phases. However, higher crystalline
samples or single-crystal data are necessary to verify these initial
indexings. The powder pattern of I was indexed in a hexagonal
unit cell, with parameters a ) 29.17(1) Å, c ) 26.38(1) Å, and
V ) 19440 Å3. The powder pattern of II was indexed in a
hexagonal unit cell, with parameters a ) 29.94(1) Å, c ) 23.44-
(1) Å, and V ) 18195 Å3. The observed decrease in the unit
Figure 3. TGA-DTA curves for R-Al(HO3PC6H5)(O3PC6H5)‚H2O.
at 1490 and 1440 cm-1 are due to the skeletal vibrations of the
rings, and the bands at 3000 cm-1 are due to the C-H stretching
vibrations.
All spectra, except that of VI, show an intense and broad
band in the O-H stretching vibration region between 3600 and
3200 cm-1, which it is consistent with the presence of hydration
water. The IR spectrum for I has two bands in this region, one
very sharp located at 3640 cm-1 and another broader near 3510
cm-1. The sharp band is probably due to a zeolitic water that
does not interact by H-bonding, and the position of the second
broader band clearly indicates that this water molecule interacts
Figure 4. X-ray thermodiffractometric powder patterns for R-Al(HO3PC6H5)(O3PC6H5)‚H2O.

Aluminum Phenylphosphonates
Inorganic Chemistry, Vol. 37, No. 17, 1998 4173
The powder pattern of â-Al(HO3PC6H5)(O3PC6H5)‚H2O (IV)
is similar to that already published,25 and therefore, it is not
reported here. The powder patterns of these two phases are
clearly different, indicating two distinct crystal structures.
Although the basal spacings are quite similar for both poly-
morphs, the (hkl) reflections are different, indicating different
intralayer atomic arrangements. It is worthwhile to point out
that diffraction peaks are sharper in the R polymorph than in
the â-phase which reflects a better crystallized R sample.
The crystal structure of compound III was solved from high-
resolution powder synchrotron diffraction data. The structure
factor amplitudes were extracted from a limited region of the
diffractogram (2 < 2θ < 21.2°) using the Le Bail method30
with the pattern decomposition option of the GSAS package.31
The pattern was fitted without any structural model by refining
the following parameters: zero-point error, unit cell, and peak
shape. In this case, the pseudo-voight peak shape function32
was used together with the asymmetry correction of Finger, Cox,
and Jephcoat.33 The background was fitted manually because
of the presence of a small amount of amorphous matter. A
total of 1140 reflections were used as input to the direct methods
of the SIRPOW.92 program.34 An E map calculated for a set
with the best figure of merit (CFOM ) 0.78) yielded the position
of the two phosphorus atoms and the aluminum atom. The
positions of these atoms were used as a starting model in the
Rietveld refinement using the overall parameters obtained in
the last cycle of the Le Bail extraction. We then came back to
the Rietveld method. With the inclusion of these three atoms
and refining only the scale factor, RwP dropped to 33%. At
this stage, a difference Fourier map revealed all the oxygen
atoms as well as the two carbon atoms bonded to the two
crystallographically independent phosphorus atoms. No other
atoms were found from a difference Fourier map. So, the rest
Figure 5. TGA-DTA curves for Al(HO3PC6H5)3‚H2O.
cell volume after the dehydration process (17.4 Å3 per H2O
molecule) is consistent with the water loss.
(b) r-Al(HO3PC6H5)(O3PC6H5)‚H2O. The X-ray laboratory
powder pattern of III was autoindexed using the LATTPARM
autoindexing program, giving an orthorhombic unit cell with a
) 9.798 Å, b ) 29.387 Å, c ) 9.372 Å, V ) 2698 Å3, Z ) 8,
Vat ) 15.3 Å3/atom, and M20 ) 14. The X-ray powder
diffraction data exhibit strong preferred orientation. Hence,
intensities are not reliable and it is better to calculate the powder
diffraction intensities from the structural model given below.
Systematic absences indicated the space group Pbca (No. 61).
Figure 6. X-ray thermodiffractometric powder patterns for Al(HO3PC6H5)3‚H2O.

4174 Inorganic Chemistry, Vol. 37, No. 17, 1998
Cabeza et al.
Table 1. Crystallographic Data for
R-Al(HO3PC6H5)(O3PC6H5)‚H2O
fw
359.17
Dc/g cm-3
1.702
0.39989(2)
cryst system orthorhombic λ/Å
a/Å
b/Å
9.7952(1)
29.3878(4)
9.3537(3)
2692.5(1)
8
no. of overall params 104
RwP
RP
RF
ø2
0.098
0.078
0.068
1.65
c/Å
V/Å3
Z
space group
Pbca (No.61)
Table 2. Positional Parameters and Uiso Values for
R-Al(HO3PC6H5)(O3PC6H5)‚H2O
atom
x
y
z
Uiso/Å2
Al
0.73354(30)
0.68796(28)
0.43969(31)
0.5356(4)
0.7021(6)
0.7663(6)
0.0796(4)
0.3375(5)
0.3842(7)
0.6111(7)
0.7595(7)
0.8507(9)
0.9028(11)
0.8419(11)
0.7465(10)
0.6977(9)
-0.0320(6)
0.5782(8)
0.5978(9)
0.5002(10)
0.3928(11)
0.3667(7)
0.25023(14)
0.32379(9)
0.21105(9)
0.33016(17)
0.29836(19)
0.30042(19)
0.23214(16)
0.2158(20)
0.23097(18)
0.28812(21)
0.38058(11)
0.38883(18)
0.43268(23)
0.46916(16)
0.46090(15)
0.41682(16)
0.15042(11)
0.12977(17)
0.08299(18)
0.05664(13)
0.07789(20)
0.12412(17)
0.4060(7)
0.1780(5)
0.3315(5)
0.2214(9)
0.0367(6)
0.3006(7)
0.2093(10)
0.2073(7)
0.4710(6)
0.5263(14)
0.1566(8)
0.0455(10)
0.0270(13)
0.0980(13)
0.2047(13)
0.2269(12)
0.1391(9)
0.2935(14)
0.3073(15)
0.3774(15)
0.4488(15)
0.4227(13)
0.003(1)
0.011(1)
0.006(1)
0.015(3)
0.015(3)
0.012(3)
0.003(2)
0.013(3)
0.012(3)
0.013(2)
0.016(4)
0.012(4)
0.016(4)
0.030(5)
0.017(4)
0.027(4)
0.008(4)
0.017(4)
0.013(4)
0.022(4)
0.035(5)
0.005(4)
P(1)
P(2)
O(1)
O(2)
O(3)
O(4)
O(5)
O(6)
O(7)
C(1)
C(2)
C(3)
C(4)
C(5)
C(6)
C(7)
C(8)
C(9)
C(10)
C(11)
C(12)
Figure 7. TGA-DTA curves for Al(OH)(O3PC6H5).
converged to RwP ≈ 18%. The refinement of the preferred
orientation along the [010] and [012] directions dropped the
RwP factor to 10%. This compound is layered (along the b axis),
and it crystallizes as fine needles. The high compaction level
in the capillary induced preferred orientation with the needles
mainly oriented along the rotation axis. The GSAS program
allowed the use of the March-Dollase35 correction, and the
coefficients converged to 2.181(15) (fraction ) 0.5) for [010]
and 0.841(4) (fraction ) 0.5) for [012]. The refined peak shape
parameters were: LX ) 0.23(3), LY ) 8.86(32), S/L ) 0.0099-
(3), H/L ) 0.0080(3), stec ) 12(1), and ptec ) 15.9(7). These
parameters indicate that the peak shape is anisotropic with a
significant broading along the [010] direction. Finally, an
isotropic temperature factor was refined for each atom. The
last refinement converged to RwP ) 9.76%, RP ) 7.82% and
RF ) 6.83%; R factors are as defined by Rietveld36 and by
Larson and Von Dreele,31 with a final weighting factor for the
soft constraints of -40.
Crystallographic data for the refinement of R-Al(HO3-
PC6H5)(O3PC6H5)‚H2O are given in Table 1. Final atomic and
isotropic thermal parameters for III are given in Table 2, and
bond lengths and angles, in Table 3. The final observed,
calculated, and difference profiles are given in Figure 9. The
coordination around the Al and P atoms is given in Figure 10,
where all atoms are labeled. The arrangement of atoms in a
layer and the packing of layers down the a axis are shown in
Figures 11 and 12, respectively. The aluminum atoms have an
octahedral environment of oxygens, five belonging to five
different phosphonate groups and the sixth being the oxygen
of the water molecule, O(7). There are two crystallographically
independent phosphonate groups, one being protonated, P(1).
The O(1) oxygen atom is the hydrogen phosphonate oxygen
Figure 8. IR spectra for (a) Al2(O3PC6H5)3‚2H2O, (b) R-Al(HO3-
PC6H5)(O3PC6H5)‚H2O, (c) Al(HO3PC6H5)3‚H2O, and (d) Al(OH)(O3-
PC6H5).
of the carbon atoms were modeled, as they belong to two phenyl
groups, and they were included in their calculated positions.
At this stage, the refinement of this model was carried out using
soft constraints to keep a reasonable geometry for the CPO3
tetrahedron and the phenyl rings. The following soft constraints
were used. Bonding distances: P-O, 1.53(1) Å; P-C, 1.80-
(1) Å; C-C, 1.40(1) Å. Nonbonding distances: C‚‚‚C, 2.40-
(1) Å; C‚‚‚C, 2.78(1) Å. Initially, the weight for the soft
constraints was high, -1500, but it was reduced as the
refinement progressed. Under these conditions, the refinement

Aluminum Phenylphosphonates
Inorganic Chemistry, Vol. 37, No. 17, 1998 4175
Figure 9. Observed, calculated, and difference synchrotron X-ray (λ ≈ 0.4 Å) powder diffraction profiles for R-Al(HO3PC6H5)(O3PC6H5)‚H2O
between 2 and 30° (2θ). The tic marks are calculated 2θ angles for Bragg peaks.
Table 3. Bond Lengths (Å) and Angles (deg) for R-Al(HO3PC6H5)(O3PC6H5)‚H2O
Al-O(2)
Al-O(3)
Al-O(4)
Al-O(5)
Al-O(6)
Al-O(7)
1.905(7)
1.803(7)
1.929(9)
1.786(6)
1.951(6)
1.986(10)
O(2)-Al-O(3)
O(2)-Al-O(4)
O(2)-Al-O(5)
O(2)-Al-O(6)
O(2)-Al-O(7)
O(3)-Al-O(4)
O(3)-Al-O(5)
O(3)-Al-O(6)
173.2(5)
91.48(28)
92.69(29)
87.5(4)
87.6(4)
93.4(4)
O(3)-Al-O(7)
O(4)-Al-O(5)
O(4)-Al-O(6)
O(4)-Al-O(7)
O(5)-Al-O(6)
O(5)-Al-O(7)
O(6)-Al-O(7)
87.6(4)
87.5(4)
177.7(5)
89.96(31)
94.57(26)
177.5(5)
87.5(4)
92.2(4)
87.49(28)
P(1)-O(1)
P(1)-O(2)
P(1)-O(3)
P(1)-C(1)
1.558(3)
1.525(4)
1.541(4)
1.821(3)
O(1)-P(1)-O(2)
O(1)-P(1)-O(3)
O(2)-P(1)-O(3)
111.81(29)
109.70(30)
112.40(26)
O(1)-P(1)-C(1)
O(2)-P(1)-C(1)
O(3)-P(1)-C(1)
106.70(23)
108.59(26)
107.37(27)
P(2)-O(4)
P(2)-O(5)
P(2)-O(6)
P(2)-C(7)
1.552(4)
1.540(4)
1.529(4)
1.824(3)
O(4)-P(2)-O(5)
O(4)-P(2)-O(6)
O(5)-P(2)-O(6)
110.7(4)
111.77(30)
112.25(25)
O(4)-P(2)-C(7)
O(5)-P(2)-C(7)
O(6)-P(2)-C(7)
107.02(22)
107.38(26)
107.43(27)
P(1)-O(2)-Al
P(2)-O(5)-Al
159.8(6)
150.3(4)
P(1)-O(3)-Al
P(2)-O(6)-Al
133.1(4)
151.6(5)
P(2)-O(4)-Al
131.5(6)
by the P(2) atoms and then interconnected through the O(4)
oxygen to give rise to the inorganic layer. This phosphonate
group has the phenyl ring approximately perpendicular to the
layers. The stacking of layers along the b axis is evident in
Figure 12.
.
(c) Al(HO3PC6H5)3 H2O. The X-ray powder diffraction
pattern of V was indexed, using the program TREOR90, in a
monoclinic unit cell with the parameters a ) 8.402(3) Å, b )
29.082(9) Å, c ) 9.176(4) Å, â ) 110.19(3)°, V ) 2104.3 Å3,
Z ) 4, and Vat ) 16.4 Å3/atom (Table 4). The X-ray powder
pattern was recorded with a sample diluted and blended with
spherical particles of Cab-O-Sil M-5 (from Fluka), to reduce
the preferred orientation.37
Figure 10. The atoms in one asymmetric unit of R-Al(HO3PC6H5)(O3-
PC6H5)‚H2O showing the octahedral environment around the Al.
and does not bond to the metal atoms. As observed in metal
hydrogen phosphates2 and hydrogen phosphonate compounds,
the P(1)-O(1)H bond is longer than the other P-OM bonds
(Table 3). Both phosphonate groups show the regular tetrahe-
dral geometry typical of such groups, O3PC. Inside an inorganic
layer, two oxygen atoms, O(2) and O(3) that belong to P(1),
link two consecutive aluminum atoms in the c-axis direction,
giving rise to infinite Al‚‚‚P(1)‚‚‚Al chains. The phosphonate
P(1) groups alternatively point the phenyl and the hydroxyl
groups up and down the chains. These chains are also bridged
(d) Al(OH)(O3PC6H5). The X-ray powder diffraction pattern
of VI was indexed in a monoclinic unit cell with the parameters
a ) 16.400(3) Å, b ) 5.530(1) Å, c ) 8.270(3) Å, â ) 90.51-
(2)°, V ) 750 Å3, Z ) 4, and Vat ) 15.6 Å3/atom (Table 5).
Both the TREOR90 and Lattparm programs converged to the
same solution. The X-ray powder pattern was recorded as
described above to reduce the preferred orientation.
27Al and 31P MAS NMR Study. 27Al MAS NMR and 31P

4176 Inorganic Chemistry, Vol. 37, No. 17, 1998
Cabeza et al.
Table 4. X-ray Powder Diffraction Data for Al(HO3PC6H5)3‚H2Oa
dobs/Å
dcal/Å
hkl
I/Im × 100
14.542
7.617
7.279
6.969
6.109
5.766
5.560
4.945
4.843
4.685
4.435
4.225
4.070
4.014
3.941
3.803
3.669
3.620
3.484
3.348
3.305
14.526
7.603
7.263
6.967
6.111
5.766
5.555
4.945
4.842
4.676
4.436
4.222
4.073
4.014
3.938
3.801
3.672
3.632
3.488
3.347
3.3076
020
100.0
11.89
1.51
0.93
1.91
1.12
2.85
7.89
5.65
2.04
9.98
2.28
3.44
8.99
2.27
2.53
2.67
6.43
4.43
3.40
2.40
110
040
1h11
130
1h31
041
111
060
150
1h12
061
132
1h61, 2h21
200, 032
220
170
2h41, 2h21
161
081
112
a a ) 8.402(3) Å, b ) 29.082(9) Å, c ) 9.176(4) Å, â ) 110.19(3)°,
V ) 2104 Å3, Vat ) 16.4 Å3/atom Z ) 4.
Figure 11. View of a layer of R-Al(HO3PC6H5)(O3PC6H5)‚H2O (ac
plane) with atoms labeled.
Table 5. X-ray Powder Diffraction Data for Al(OH)(O3PC6H5)a
dobs/Å
dcal/Å
hkl
100
I/Im × 100
16.443
5.481
4.600
4.426
4.104
4.015
3.708
3.512
3.313
3.283
3.055
2.927
2.900
2.766
2.730
2.621
2.558
2.444
2.345
2.291
2.278
16.434
5.474
4.597
4.423
4.102
4.016
3.707
3.509
3.317
3.285
3.0670
2.926
2.899
2.764
2.726
2.620
2.558
2.442
2.342
2.289
2.277
100.0
0.52
3.16
8.46
10.13
9.04
2.27
6.02
2.15
2.79
1.66
0.81
0.47
0.95
1.86
0.83
0.51
1.07
1.26
0.53
0.55
300
011
111
400
2h11, 1h02
2h02
311
3h02
302
4h11
4h02
402
020
120
220
502
1h13
700, 611
6h02
1h22
a a ) 16.400(3) Å, b ) 5.530(1) Å, c ) 8.270(3) Å, â ) 90.51(2)°,
V ) 750 Å3, Vat ) 15.6 Å3/atom, Z ) 4.
MAS NMR spectra for I are shown in Figure 13 (a) and Figure
14 a, respectively. The 27Al spectrum for I shows two main
broad peaks at 44 and -21 ppm with differently shaped
shoulders, which correspond to aluminums in tetrahedral and
octahedral coordinations, respectively (Figure 13a). The 31P
spectrum for I is formed by three components at -3, -6, and
-8 ppm, due to at least three inequivalent O3PC tetrahedral
groups (Figure 14a).
27Al MAS NMR spectra for III and V are shown in Figure
13b and Figure 13c, respectively. The 27Al spectrum for III
shows a sharp peak at -13.1 ppm, typical of aluminum with
an octahedral oxygen environment. The 27Al spectrum for V
shows a broad asymmetric band at -20 ppm, due again to
octahedral aluminum. In this case, three maxima are detected
at -16.4, -19.9, and -23.9 ppm, which are produced by
second-order quadrupolar interactions of a single aluminum with
the electric field gradient at the nuclear metal site.
Figure 12. Three-dimensional packing of layers for R-Al(HO3-
PC6H5)(O3PC6H5)‚H2O down the a axis (b axis vertical).
31P MAS NMR spectra for III and V are displayed in Figure
14b and Figure 14c, respectively. The 31P spectrum for III
shows a broad band at 1 ppm, which is the result of two partially
overlapped signals at 1.5 and 0.3 ppm. The 31P spectrum for
compound V shows two broad bands at 5.1 and 1.7 ppm. The
first band has double the intensity of the second one. A third
small peak is observed at 19 ppm, which is probably due to an
impurity.
Discussion
The crystal structures of several metal phenylphosphonates
are known from single-crystal or powder diffraction studies.
Layered materials, independently of the metal cation size, the

Aluminum Phenylphosphonates
Inorganic Chemistry, Vol. 37, No. 17, 1998 4177
Figure 13. 27Al MAS NMR spectra for (a) Al2(O3PC6H5)3‚2H2O, (b)
R-Al(HO3PC6H5)(O3PC6H5)‚H2O, and (c) Al(HO3PC6H5)3‚H2O.
Figure 14. 31P MAS NMR spectra for (a) Al2(O3PC6H5)3‚2H2O, (b)
R-Al(HO3PC6H5)(O3PC6H5)‚H2O, and (c) Al(HO3PC6H5)3‚H2O.
metal coordination polyhedra, and the metal/phosphonate molar
ratio, always display a strong diffraction peak at ∼15 Å
characteristic of an organic bilayer of phenyl groups between
an inorganic layer formed by the metal-phosphonate bonding.
For example peaks at 15.7 Å for LaIII(HO3PC6H5)(O3PC6H5),10
15.6 Å for PbII(HO3PC6H5)2,38 14.8 Å for ZrIV(O3PC6H5)2,3 and
14.33 Å for MnII(O3PC6H5)‚H2O5 are known for layered metal
phenylphosphonates. Although the crystal structures of three
polymorphs of FeIII(HO3PC6H5)(O3PC6H5)‚H2O are not known,
the main d spacings range from 14.66 to 15.11 Å, indicating
that these are also probably layered materials.9 The most intense
peaks of AlIII(HO3PC6H5)(O3PC6H5)‚H2O and Al2III(O3PC6H5)3‚
4H2O were reported at 14.8 Å25 and 14.35 Å,26 respectively,
and the authors suggested layered structures.
On the other hand, several attempts to solve the structures of
V and VI from laboratory X-ray powder diffraction data have
been unsuccessful. However, the patterns present strong
preferred orientation effects and main peaks at 14.6 and 16.4
Å, respectively, which suggests layered structures.
The 27Al MAS NMR spectrum for III shows a single peak
due to a distorted octahedral environment of oxygens at -13.1
ppm, in agreement with the single-crystal structural analysis.
Quadrupolar constants deduced from the experimental profile
of this peak were CQ ) 0.23 MHz and η ) 0. The 31P MAS
NMR spectrum for III shows a broad peak due to two signals
at 1.5 and 0.3 ppm, which correspond to the nonresolved HO3P-
(1)C and O3P(2)C groups, respectively. In a first approximation,
the hydrogen phosphonate groups are expected to be at higher
chemical shifts than the corresponding phosphonate groups.
Typical variations of 3-6 ppm are observed between mono-
hydrogen phenylphosphonate and phenylphosphonate groups.39
However, some other factors can affect the 31P signal position
such as the P-C and P-O bond distances, the POAl angles,
and even the van der Waals interactions between phenyl rings.
The structure of III has been solved ab initio from powder
diffraction data; it is layered with the main peak at 14.7 Å,
representing the basal spacing. Compound IV, with the main
peak at 14.8 Å, is a polymorph of the R-phase with a very related
layered structure. The existence of several polymorphs depend-
ing on the synthetic conditions is quite common in the chemistry
of metal phosphonates. Even the nature of the salt that is the
source of the metal cation may slightly affect the structure of
the final product.24
(37) Aranda, M. A. G.; Cabeza, A.; Bruque, S.; Poojary, D. M.; Clearfield,
A. Inorg. Chem. 1998, 37, 1827-1832.

4178 Inorganic Chemistry, Vol. 37, No. 17, 1998
Cabeza et al.
The result of these factors appears to prevent the resolution of
the two overlapped 31P signals.
probably indicating the existence of more than three components.
A determination of quadrupolar constants is difficult due to the
overlapping of signals. These results are compatible with the
powder diffraction study, as the very large unit cell (V ) 19440
Å3) may contain several nonequivalent P and Al sites. However,
the indexation of the powder pattern of this material is only a
preliminary result, and further study with a better crystallized
sample is necessary.
The 27Al MAS NMR spectrum for V has a band composed
of three maxima that correspond to the same distorted octahe-
dron. From the analysis of the experimental profile, the values
of the quadrupolar constants were CQ ) 0.37 MHz and η ) 1.
The 31P MAS NMR spectrum for V shows two broad bands at
5.1 and 1.7 ppm, which must be due to at least three signals, as
the peak at 5.1 ppm has double the intensity of that at 1.7 ppm.
The peak at lower chemical shift must correspond to one
crystallographically independent phosphorus atom in a general
position. Then, the peak at 5.1 ppm must be the result of two
overlapped signals of the same intensity. These spectra are
again compatible with a layered framework of strongly distorted
octahedral aluminums sanwiched between hydrogen phenylphos-
phonate tetrahedral groups.
The 31P MAS NMR study of three aluminum phenylphos-
phonates with different stoichiometries has revealed a clear
trend. I has no hydrogen phosphonate groups and the 31P signals
range between -2 and -9 ppm. III has a hydrogen phospho-
nate and a phosphonate group with a broad peak centered at 1
ppm. V has three hydrogen phosphonate groups with peaks
ranging between 5 and 2 ppm. The influence of the HO-P
groups in the positions of the 31P signals is, hence, clearly
observed. As should be expected, a shift in the 31P signal
position to higher positive chemical shift values occurs as the
phosphonate groups become more protonated.
It is important to underline the thermal stabilities of these
types of phosphonates. I dehydrates quickly above 60 °C but
II remains stable up to 550 °C. Above this temperature, it
undergoes combustion in air. III and V lose crystallinity after
the dehydration processes at 180 and 240 °C, respectively. The
water is bonded to the aluminum atom in III, and the high
temperature at which water is lost in V supports that water is
bonded to Al, in the same way that is observed in III. Both
compounds undergo pyrolysis above 400 °C in the atmosphere.
The temperature of the water loss processes and the νO-H
stretching vibration frequencies indicate that water is weakly
retained in I, it is strongly bonded in III, and it presents the
strongest interaction in V. Finally, compound VI undergoes
condensation of the hydroxide groups at 280 °C. Then it
remains stable up to 600 °C, whereas the combustion of the
organic matter in air takes place.
Compound I has the main diffraction peak at 13.2 Å, being
the second-order reflection of the first peak observed at 26.4
Å. In contrast to what happens for compounds III and V, where
the main peak does not change on heating, the main peak of I
decreases from 13.2 to 11.7 Å suggesting a different structure
from that proposed for Al2(O3PC6H5)3‚4H2O.26 The relationship
between the powder patterns of I and II (Figure 2) indicates a
topotactic dehydration that we have found to be reversible. This
dehydration process takes place at a very low temperature ∼60
°C which indicates a labile water that is weakly retained. The
powder patterns have broad peaks probably due to a small
microcrystal average size. However, some insights about the
structure may be derived by comparing the X-ray and MAS
NMR results with those published for related materials.
The 27Al MAS NMR spectrum for I has two bands due to
octahedral (-21 ppm) and tetrahedral (46 ppm) oxygen
environments. A similar spectrum is observed for â-Al2(O3-
PCH3)3‚H2O23 with two signals, at -17.6 and 41.2 ppm. This
compound crystallizes in a trigonal cell of dimensions a ) 24.65
Å, c ) 25.30 Å, γ ) 120 °, and Z ) 36. The structure is tubular
with small channels running parallel to the c axis and the methyl
groups projected toward the interior of the channels resulting
in a microporous material with tunnels of approximately 5.8 Å
of cross-section. The relationship between the unit cells and
mainly the very similar 27Al MAS NMR spectra suggest that I
has a related tubular structure. However, N2 adsorption-
desorption isotherms showed that no micropores are present in
this material. The analysis of these data yielded a low specific
BET surface area (25 m2/g). The existence of different maxima
detected in the corresponding 31P and 27Al signals suggests that
the numbers of crystallographic sites are higher than the
chemical formula (three P and two Al). From the analysis of
the 27Al envelope, the presence of two tetrahedral and two
octahedral environments for Al was deduced. For the 31P signal,
the three detected components exhibit different line widths,
In the course of this investigation, other slightly different
aluminum phenylphosphonates have been obtained. A full
characterization of these compounds is in progress and will be
reported elsewhere. It is noteworthy that minor changes in the
syntheses lead to different stoichiometries and/or structures.
Acknowledgment. We thank NATO for funding through
NATO CRG Program 951242. The work in Ma´laga was also
supported by Research Grants FQM-113 of Junta de Andaluc´ıa
(Spain) and MAT-97-326-C4-4 of CICYT, Ministerio de
Educacio´n y Ciencia, Spain. We also thank Drs. E. Dooryhee,
G. Vaugahn, and A. Fitch for assistance during data collection
on BM16 and the ESRF for the provision of synchrotron
facilities. A.C. wishes to thank the Robert A. Welch Foundation
for partial support of this effort under Grant No. A-673.
IC9802350