Linear coordination polymers based on aluminum phosphates: Synthesis, crystal structure and morphology

Article abstract of DOI:10.1039/c6dt00153j

Several aluminum tris(diorganophosphates) have been synthesized and characterized via elemental analysis, NMR, FT-IR and Raman spectroscopy, as well as powder XRD, and SEM. Single-crystal X-ray diffraction studies revealed that the aluminum tris(diethylphosphate) crystal structure comprises two crystallographically nonequivalent catena-Al[O₂P(OEt)₂]₃ chains propagating along the c-axis. Their parallel orientation favors the formation of closely packed hexagonal domains. PXRD data suggest that other homologues have a similar structure, with the interchain distance closely corresponding to the dimensions of organic ligands. They are also susceptible to a reversible dissociation to ionic species under the effect of primary amines. This feature can be utilized for the synthesis of epoxy nanocomposites.The Royal Society of Chemistry.

Full text of DOI:10.1039/c6dt00153j

Dalton
Transactions
View Article Online
View Journal
PAPER
Linear coordination polymers based on
aluminum phosphates: synthesis, crystal structure
Cite this: DOI: 10.1039/c6dt00153j
and morphology†
M. Dębowski,* K. Łokaj, A. Wolak, K. Żurawski, A. Plichta, J. Zachara, A. Ostrowski
and Z. Florjańczyk*
Several aluminum tris(diorganophosphates) have been synthesized and characterized via elemental
analysis, NMR, FT-IR and Raman spectroscopy, as well as powder XRD, and SEM. Single-crystal X-ray
diffraction studies revealed that the aluminum tris(diethylphosphate) crystal structure comprises two
crystallographically nonequivalent catena-Al[O2P(OEt)2]3 chains propagating along the c-axis. Their paral-
lel orientation favors the formation of closely packed hexagonal domains. PXRD data suggest that other
homologues have a similar structure, with the interchain distance closely corresponding to the dimen-
sions of organic ligands. They are also susceptible to a reversible dissociation to ionic species under the
effect of primary amines. This feature can be utilized for the synthesis of epoxy nanocomposites.
Received 12th January 2016,
Accepted 27th March 2016
DOI: 10.1039/c6dt00153j
www.rsc.org/dalton
aluminum cations. The reaction of phosphoric acid mono-
esters with aluminum inorganic salts or trialkoxides has been
Introduction
The phosphate group is one of the essential components of a successfully employed in the synthesis of two-dimensional
variety of biological macromolecules and synthetic materials. plate-like particles consisting of aluminum phosphate phases
Considerable effort has been devoted to the development of separated by long organic chains,16,18,19 as well as nanosized
synthetic analogues of nucleic acids, lipids and enzymes, as octameric and decameric clusters which can serve as building
well as the mimicking of some processes that occur in living blocks for framework structures.36 It has been observed that
organisms,1–4 for example, selective cation transport in liquid dimers [R2AlO2P(OR′)2]2 or tetramers [(RO)2AlO2P(OR′)2]4 were
membrane systems1,5–7 and biomineralization.8–10 Esters of formed when phosphoric acid diesters were reacted with equi-
phosphoric acid are often used as starting materials in the syn- molar amounts of trialkylaluminum27,37 or trialkoxyalumi-
thesis of coordination polymers formed by metal cations num,27 respectively. Furthermore, chemical transformation of
linked by organic ligands, molecular clusters, or hybrid in- these cyclic oligomers allows for the fabrication of alumino-
organic–organic mesostructured lamellar systems.11–20 Some phosphate materials,27 silicoaluminum phosphates,38,39 and
of these hybrid materials, as well as their analogues with phos- 2D framework solids.40
phonate groups, are considered promising candidates for a
We have previously explored the use of boehmite and di-
variety of practical applications, e.g. organogelators, flame phenylphosphoric acid (DPhP)37 or triethyl phosphate
retardants, fillers in polymer composites, processable precur- (TEP)41–44 for the preparation of spherical nanoparticles,
sors of mesoporous zeolites, and thin films with high abrasion which consist of aluminum oxohydroxide cores covered by a
resistance and low refractive index.21–35
layer of organically modified aluminum phosphates. These
An important class of these materials consists of organi- products were shown to improve certain mechanical properties
cally modified aluminum phosphates synthesized from mono- of polymer composites based on polyurethane,44 or carboxyl-
or diesters of phosphoric acids and various sources of ated styrene–butadiene rubber.41,45 We have also observed
that, in such systems, increasing the reaction time favors the
formation of highly crystalline fibers composed of hexagonally
packed catena-Al[O2P(OR)2]3 chains (R = Et,44 Ph37). It has been
Faculty of Chemistry, Warsaw University of Technology, Noakowskiego 3, 00-664
Warsaw, Poland. E-mail: debowski@ch.pw.edu.pl, evala@ch.pw.edu.pl
suggested that these fibers are formed via the self-assembly of
aluminum cations and (RO)2PO2 anions in solution.44 This
−
†Electronic supplementary information (ESI) available: SEM images of DPhPAl,
FT-IR and Raman spectra of DOPAl samples, additional crystal structure details
including CIF, tabulated PXRD data for DOPAl samples, and NMR spectral data.
CCDC 1404032. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/c6dt00153j
type of hybrid material may be of practical interest because the
mechanical reinforcement gained from fibers is significantly
higher than that gained from spherical fillers.46 Some recent
This journal is © The Royal Society of Chemistry 2016
Dalton Trans.

View Article Online
Paper
Dalton Transactions
publications suggest also that the presence of Al[O2P(OR)2]3 meter at a spinning rate of 6–8 kHz with proton high-power
nanorods can affect other properties of the polymer matrix, for decoupling and a resonance frequency of 78.20 MHz. Solid-
example enhancing orientation in liquid-crystalline epoxy state 31P CP/MAS NMR measurements were conducted on a
resins cured in an external magnetic field.47,48
Bruker DSX 300 spectrometer at a spinning rate of 5–8 kHz,
To expand our knowledge of this class of coordination poly- resonance frequency of 121.50 MHz and contact time of
mers, we decided to investigate in detail their structural 1–1.5 ms. FT-IR spectra were recorded on a Thermo Scientific
and morphological properties, as well as the way they are influ- Nicolet 6700 spectrometer equipped with a Smart Orbit
enced by the synthetic procedures and the type of organic Diamond ATR accessory. Raman spectra were recorded on a
groups present in the phosphate moiety. In view of those Thermo Nicolet Almega XR spectrometer equipped with a
objectives we developed effective methods for the synthesis of 532 nm laser. The particle size distribution in the studied dis-
pure aluminum tris(diorganophosphate) (DOPAl) utilizing persions was determined using a dynamic light scattering
different sources of aluminum. We examined the physico- (DLS) method. The measurements were conducted on a
chemical and morphological properties of the resulting Malvern Zetasizer Nano ZS instrument equipped with a
materials using single-crystal or powder XRD studies, 27Al and 532 nm laser. SEM images were obtained on a Zeiss LEO 1530
31P NMR, Raman or FT-IR spectroscopy, as well as the SEM field emission scanning electron microscope equipped with a
technique. In this paper, we report the synthesis and charac- GEMINI column. Before the measurements, all samples were
terization of coordination polymers with the general formula coated with an electron conductive layer of carbon utilizing a
Al[O2P(OR)2]3 containing alkyl groups [R = Me (DMPAl)], Et high-vacuum carbon sputter coater.
(DEPAl), nPr (DnPPAl), Pr (DiPPAl), nBu (DBPAl), CH2CH(Et)
(CH2)3CH3 (BEHPAl) or phenyl moieties [R = Ph (DPhPAl)] in
i
Single-crystal structure analysis
aluminum phosphate monomeric units.
A single crystal of DEPAl was selected under a polarizing
microscope and mounted in an inert oil (Immersion Oil type
A, Cargill) and transferred to the cold gas stream of the
diffractometer. Diffraction data were recorded at −173(2) °C
using graphite-monochromated Mo-Kα radiation on an Oxford
Diffraction κ-CCD Gemini A Ultra diffractometer. Cell refine-
Experimental
Materials and characterization methods
All chemicals for syntheses were purchased from commercial ment and data collection, as well as data reduction and analy-
sources and were used without further purification: trimethyl sis, were performed with the Oxford Diffraction software
phosphate (TMP) (97%, Aldrich), TEP (99%, Aldrich), tri-n- program CrysAlisPRO.50 The structure was solved and refined
propyl phosphate (TnPP) (99%, Aldrich), triisopropyl phos- with SHELXL-97.51 Crystal data and the details of structure
phate (TiPP) (99%, Aldrich), tri-n-butyl phosphate (TBP) (97%, determinations are given in Table 1. A more detailed descrip-
Aldrich), triphenyl phosphate (TPhP) (99%, Merck Chemicals), tion of structure determination as well as some selected bond
di-n-butyl phosphate (DBP) (97%, Aldrich), bis(2-ethylhexyl) distances and bond angles (Tables S1–S5†) are given in the
phosphate (BEHP) (97%, Aldrich), DPhP (99%, Aldrich), alumi- ESI. Crystallographic data for the structure reported in this
num L-lactate (AlLact) (97%, Aldrich), triethylaluminum (TEA) paper have been deposited in the Cambridge Crystallographic
(1.9 M solution in toluene, Aldrich), and sodium bicarbonate Data Center with CCDC number 1404032.
(pure, POCh Gliwice). Toluene used in the reactions involving
TEA was distilled from sodium-benzophenone ketyl solution
Preparation of DOPAl samples
and stored over dry 4 Å molecular sieves under a nitrogen
atmosphere.
Reactions of AlLact with phosphoric acid triester (TOP).
Method M1. In the case of water-miscible TMP or TEP, DOPAl
The hydrogen and carbon contents were determined using was synthesized according to the following procedure: AlLact
a Perkin-Elmer CHNS/O II 2400 instrument. Powder X-ray diffr- (6.0 g, 20 mmol) was dissolved in redistilled water (20 mL) fol-
action (PXRD) patterns were recorded at room temperature on lowed by the addition of TOP (60 mmol). The mixture was
a Seifert HZG-4 automated diffractometer using Cu-Kα radi- heated under reflux (48 h), during which a white solid material
ation (λ = 0.15418 nm). The data were collected in the Bragg– precipitated. The crude product was isolated via filtration on a
Brentano (θ/2θ) horizontal geometry (flat reflection mode) Schott funnel and purified by washing with water and then
between 4° and 60° (2θ) in 0.04° steps at 10 s per step. In the methanol. The final product was obtained after drying at
case of DnPPAl, the TOPAS software49 (Bruker AXS) was uti- 70–80 °C under an air flow. In this manner, DMPAl (7.5 g,
lized for indexing the PXRD pattern using 22 reflections (2θ < 94%) and DEPAl (8.8 g, 92%) were synthesized.
38°), as well as Pawley refinement of the obtained cell para-
Method M2. All reactions of AlLact with TOP containing
meters. 1H, 27Al and 31P NMR spectra were recorded in solu- hydrophobic alkyl chains (C3, C4, and C8) or phenyl groups
tion on a Varian 400 MHz spectrometer operating at 400.11, required temperatures in the range of 130–150 °C and the use
104.25, and 161.96 MHz, respectively. Chemical shifts are of pressure vessels. A typical synthesis was performed as
reported relative to external standards: TMS (1H), [Al(H2O)6]3+ follows: AlLact (3.0 g, 10 mmol) was dissolved in redistilled
(27Al), and 85% H3PO4(aq) (31P). Solid-state 27Al MAS NMR water (20 mL) followed by the addition of TOP (30 mmol). The
measurements were performed on a Bruker DSX 300 spectro- mixture was placed in a stainless steel or thick-walled glass
Dalton Trans.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Dalton Transactions
Paper
Method M3. The procedure was identical to that described
previously (see the TOP synthetic route, Method M2) with the
exception that DOP was used as the phosphorus source. In this
manner, DPhPAl (6.9 g, 90% after 1 h of heating and 6.8 g,
89% after 24 h of heating) could be synthesized.
Table 1 Crystal and structure refinement data for DEPAl
Formula
C12H30AlO12P3
Molecular weight [g mol−1
Crystal system
Space group
]
486.26
Trigonal
ˉ
P3
Reactions of AlLact with phosphoric acid diester sodium salt
(DOPNa). A typical synthesis was performed in 2 steps. First,
DOP (30 mmol) was neutralized with a 10 wt% aqueous solu-
tion of sodium bicarbonate (2.5 g, 30 mmol). A solution of
AlLact (3.0 g, 10 mmol) in water (10 g) was added dropwise
under vigorous stirring to the resulting clear solution of
DOPNa. Even at room temperature, the ion-exchange reaction
proceeded very quickly. Depending on the type of organic
ligand used, an insoluble product (DOPAl) precipitated (in the
case of DBPAl and DPhPAl) or formed a stable dispersion (in
the case of BEHPAl); BEHPAl could be isolated by salting out
with NaCl. The crude DOPAl was isolated and purified accord-
ing to the previously described procedures (see the DOP
synthetic route). In this manner, DBPAl (5.8 g, 90%), BEHPAl
(9.1 g, 93%) or DPhPAl (3.5 g, 45%) was synthesized.
Reactions of TEA with DOP. All operations were performed
under a purified nitrogen atmosphere. A solution of DBP
(6.5 g, 30 mmol) in toluene (10 mL) was added dropwise to
a vigorously stirred mixture of TEA (5.3 mL, 10 mmol) and
toluene (10 mL) that was precooled to −70 °C. The reaction
mixture was allowed to warm up to room temperature for 1 h.
Stirring was continued at this temperature for an additional
12 h, resulting in the formation of an opalescent, viscous
liquid or gel. After completion of the reaction, all volatiles were
removed under reduced pressure, yielding DBPAl (5.6 g, 86%)
as a white solid. By using BEHP (10.0 g, 30 mmol) in place of
DBP and stirring the reaction mixture under reflux (12 h),
BEHPAl (9.7 g, 98%) was obtained.
a [Å]
b [Å]
c [Å]
α [°]
20.43184(12)
20.43184(12)
8.99609(6)
90
90
120
3252.36(3)
6
1.490
0.08 × 0.10 × 0.32
−173(2)
0.71073
0.370
3.2–33.0
101 704
7909
0.042
7212
−30 ≤ h ≤ 30
−30 ≤ k ≤ 31
−13 ≤ l ≤ 13
7909
272
0.0262
0.0717
1.031
β [°]
γ [°]
Cell volume [Å3]
Z
ρcalcd [g cm−3
]
Crystal dimensions [mm]
Measurement temperature [°C]
Radiation λ [Å]
μ [mm−1
]
θmin–θmax [°]
No. of measured reflections
No. of unique reflections
Rint
Observed data [I > 2σ(I)
Index ranges
Nref
Npar
R1 for observed data
wR2 for observed data
Goodness-of-fit
Residual density: min/max [e Å−3
]
−0.34/0.55
pressure reactor and heated in an oil bath (24 h). After cooling
to room temperature, the crude product was isolated via
filtration on a Schott funnel and purified by washing with
water and an organic solvent (methanol in the case of DnPPAl,
DiPPAl, and DBPAl or acetone in the case of DPhPAl). The
final product was obtained after drying at 70–80 °C under an
air flow. In this manner, DnPPAl (1.4 g, 25%), DiPPAl (3.9 g,
70%), DBPAl (3.9 g, 60%) or DPhPAl (1.4 g, 18%) was
synthesized.
Method M3. The procedure was identical to that described
in the case of Method M2, but the reaction was performed in a
mixture of water and 1,4-dioxane (1 : 1 v/v). In this manner,
DPhPAl (3.3 g, 43%) was obtained.
Reactions of AlLact with phosphoric acid diester (DOP).
Method M1. A typical synthesis proceeded according to the
following procedure: AlLact (3.0 g, 10 mmol) was dissolved in
redistilled water (20 mL) followed by the addition of DOP
(30 mmol). The mixture was vigorously stirred at room temp-
erature (15 min), during which a solid precipitated. The insolu-
ble product was isolated via filtration on a Schott funnel and
purified by washing with water and methanol. The final
product was obtained after drying at 70–80 °C under an air
flow. In this manner, DBPAl (6.4 g, 98%) was synthesized.
Method M2. The procedure was very similar to Method M1,
but the reaction mixture was heated under reflux (24 h). In
this manner, BEHPAl (7.4 g, 75%) was synthesized (the crude
insoluble product was washed with acetone).
Unless stated otherwise, the analytical, spectroscopic or
X-ray diffraction data presented in this paper for DBPAl refer
to the sample obtained from TEA and DBP (see the DOP-TEA
synthetic route), whereas in the case of BEHPAl or DPhPAl
such data relate to the products prepared from the respective
DOPNa (see the DOPNa synthetic route).
Al[O2P(OMe)2]3 (DMPAl). Found: C, 18.18; H, 4.65%. Calc.
for C6H18AlO12P3: C, 17.92, H 4.51%. FT-IR (neat, cm−1):
2983(w), 2957(w), 2851(w), 1463(w), 1448(w), 1219(m),
1190(m), 1134(m), 1085(w), 1068(w), 1034(s). 837(s), 792(w),
778(w), 632(w), 544(m), 523(s), 494(s), 479(s). Raman (neat,
cm−1): 3018(m), 3000(m), 2968(s), 2863(w), 1472(w), 1198(w),
1059(w), 855(w), 789(m), 589(m). 27Al MAS NMR δ (78 MHz):
−27.3 ppm. 31P MAS NMR δ (121 MHz): −7.7 and −10.4 ppm,
integral ratio 16.59 : 10.00. 1H NMR δ (saturated solution,
400 MHz, D2O): 3.75 (d, J = 11.4 Hz), 3.67 ppm (dd, J = 11.1,
1.7 Hz), 3.58 ppm (dd, J = 10.7, 1.6 Hz), integral ratio
1.00 : 4.03 : 6.60. 1H NMR δ (dilute solution, 400 MHz, D2O):
3.77 (dd, J = 11.0, 2.4 Hz), 3.68 ppm (dd, J = 11.1, 1.4 Hz),
3.59 ppm (d, J = 10.7 Hz), integral ratio 1.00 : 11.25 : 26.82. 31P
NMR δ (saturated solution, 162 MHz, D2O): 5.82 (s), −3.47 (s),
and −8.79 ppm (s), integral ratio 10.17 : 6.09 : 1.00. 31P NMR
δ (dilute solution, 162 MHz, D2O): 5.87 (s) and −3.43 ppm (s),
This journal is © The Royal Society of Chemistry 2016
Dalton Trans.

View Article Online
Paper
Dalton Transactions
integral ratio 2.50 : 1.00. 27Al NMR
δ
(dilute solution, ing to the TOP synthetic route δ (104 MHz, 1 : 2 v/v mixture of
104.25 MHz, D2O): −1.16, −2.44, and −5.88 ppm.
CDCl3 and hexan-1-amine): 0–10 ppm (signal overlapping with
Al[O2P(OEt)2]3 (DEPAl). Found: C, 29.89; H 6.13%. Calc. an artificial probe signal). 31P NMR of sample prepared accord-
for C12H30AlO12P3: C, 29.64; H, 6.22%. FT-IR (neat, cm−1): ing to the TOP synthetic route δ (162 MHz, 1: 2 v/v mixture of
2981(w), 2933(w), 2905(w), 2868(w), 1483(w), 1438(w), 1397(w), CDCl3 and hexan-1-amine): −0.52 ppm.
1366(w), 1220(m), 1201(w), 1135(s), 1071(m), 1035(s), 972(s),
Al[O2P(OCH2CH(Et)(CH2)3CH3)2]3 (BEHPAl). The results of
962(s), 819(m), 782(m), 621(w), 557(m), 505(m), 474(m). elemental analysis are presented in Table 2. FT-IR (neat,
Raman (neat, cm−1): 2983(m), 2942(s), 2909(m), 2870(w), cm−1): 2934(s), 2928(m), 2872(m), 2861(m), 1462(m), 1379(w),
2787(w), 2730(w), 1454(w), 1294(w), 1106(w), 1047(w), 794(w), 1204(m), 1133(s), 1063(s), 1028(s), 884(m), 767(w), 726(w),
552(w). 27Al MAS NMR δ (78 MHz): −14.9 and −21.3 ppm. 622(m), 563(m), 479(s). Raman (neat, cm−1): 2934(s), 2896(s),
31P MAS NMR δ (121 MHz): −14.7 ppm.
2873(s), 2730(w), 1449(w), 1297(w), 1143(w), 1053(w). 27Al MAS
Al[O2P(OnPr)2]3 (DnPPAl). Found: C, 37.92; H, 7.48%. Calc. NMR δ (78 MHz) (overlapping signals): −18.6 and −30.7 ppm
for C18H42AlO12P3: C, 37.90; H 7.42%. FT-IR (neat, cm−1): (highest intensity). 31P MAS NMR δ (121 MHz) (overlapping
2965(w), 2937(w), 2877(w), 1464(w), 1378(w), 1214(m), 1136(s), signals): −0.1, −3.3, −4.4, and −13.4 ppm (highest intensity).
1062(s), 997(s), 906(w), 886(w), 868(m), 845(w), 749(m), 567(m), 27Al NMR of sample prepared according to the DOP synthetic
520(m), 473(m). Raman (neat, cm−1): 2983(w), 2943(w), route δ (104 MHz, CDCl3): −20 to 10 ppm (broad and overlap-
2889(w). 27Al MAS NMR δ (78 MHz): −26.0 ppm. 31P MAS NMR ping signals). 27Al NMR of sample prepared according to the
δ (121 MHz): −16.8 ppm.
DOP synthetic route δ (104 MHz, C6D6): −17.72 ppm (broad
Al[O2P(OiPr)2]3 (DiPPAl). Found: C, 37.95; H, 7.44%. Calc. peak with a downfield shoulder). 27Al NMR of sample prepared
for C18H42AlO12P3: C, 37.90; H, 7.42%. FT-IR (neat, cm−1): according to the DOP synthetic route δ (104 MHz, CD3OD):
2977(w), 2919(w), 2866(w), 1456(w), 1384(w), 1373(w), 1237(m), −4.60 and −15.32 ppm (broad and overlapping signals).
1154(m), 1096(m), 1018(m), 992(s), 898(w), 789(m), 624(w), 31P NMR of sample prepared according to the DOP synthetic
556(w), 528(m), 463(m). Raman (neat, cm−1): 2988(m), 2931(s), route δ (162 MHz, CDCl3) (broad and overlapping signals):
2878(m), 2741(w), 1461(w), 1371(w), 1226(w), 1123(w), 756(w). −2.14, −2.88, −6.41, −13.33, −15.00, and −17.80 ppm.
27Al MAS NMR δ (78 MHz): −28.2 ppm. 31P MAS NMR 31P NMR of sample prepared according to the DOP synthetic
δ (121 MHz): −9.3 and −21.5 ppm, integral ratio 1.00 : 96.57.
route δ (162 MHz, C6D6) (broad and overlapping signals):
Al[O2P(OnBu)2]3 (DBPAl). The results of elemental analysis −0.80, −6.14, −12.07, −12.92, −14.46, and −17.26 ppm.
are presented in Table 2. FT-IR (neat, cm−1): 2956(m), 2932(w), 31P NMR of sample prepared according to the DOP synthetic
2906(w), 2872(w), 1465(w), 1429(w), 1379(w), 1217(s), 1137(s), route δ (162 MHz, CD3OD): 3.15, −0.10, −1.63, −3.48, −11.89,
1064(s), 1025(s), 984(s), 911(m), 850(w), 811(w), 734(w), 624(w), and −15.22 ppm.
573(m), 523(w), 478(s). Raman (neat, cm−1): 2972(m), 2945(s),
Al[O2P(OPh)2]3 (DPhPAl). The results of elemental analysis
2920(s), 2883(s), 2739(w), 1462(w), 1310(w), 1134(w), 915(w), are presented in Table 2. FT-IR (neat, cm−1): 3063(vw),
844(w). 27Al MAS NMR δ (78 MHz): −24.0 ppm. 31P MAS NMR 1590(m), 1488(m), 1456(w), 1288(w), 1228(m), 1202(s), 1121(s),
δ (121 MHz): −16.9 ppm. 27Al NMR of sample prepared accord- 1071(m), 1026(m), 1006(m), 944(s), 901(w), 772(s), 749(s),
686(s), 628(w), 561(s), 499(s), 468(s). Raman (neat, cm−1):
3191(vw), 3069(s), 1601(w), 1239(w), 1167(w), 1037(w),
1015(m), 951(w), 742(w), 625(w). 27Al MAS NMR of sample
Table 2 Elemental analysis of DBPAl, BEHPAl and DPhPAl samples syn-
thesized by different methods
prepared according to the DOP synthetic route, Method M3
δ (78 MHz): −22.5 ppm. 31P MAS NMR of sample prepared
according to the DOP synthetic route, Method M3
δ (121 MHz): −26.1 ppm (unsymmetrical, with shoulder to
high-field region).
Results of elemental analysis
DOPAl formula
Method of
synthesisb
DOPAl
DBPAl
(theoretical values)a
C [%]
H [%]
C24H54AlO12P3
TOP (M2)
DOP (M1)
DOPNa
DOP-TEA
DOP (M2)
DOPNa
DOP-TEA
TOP (M2)
TOP (M3)
DOP (M3)
43.72
43.82
43.51
43.87
56.53
57.14
57.41
55.67
54.14
55.61c
55.42d
54.79
8.08
8.29
8.21
(C, 44.04; H, 8.32%)
Results and discussion
Synthesis and morphology of DOPAl
8.49
BEHPAl
DPhPAl
C48H102AlO12P3
(C, 58.16; H, 10.37%)
10.11
10.25
10.59
4.03
TOP synthetic route. The major route for synthesizing
DOPAl applied in the present study involved reactions of phos-
phoric acid triester (TOP) with AlLact (Scheme 1a). These reac-
tions were performed in water at temperatures between 100
and 150 °C, depending on the type of TOP used. In the case of
water-miscible TMP or TEP, a simple heating of the reaction
mixture under reflux resulted in the formation of DMPAl or
DEPAl in yields exceeding 90%. Under these conditions, TOP
underwent gradual hydrolysis, yielding the respective diester of
C36H30AlO12P3
(C, 55.83; H, 3.90%)
4.12
3.93c
4.07d
4.07
DOPNa
a Values of carbon and hydrogen contents calculated for the assumed
DOPAl formula. b TOP, DOP, DOPNa, DOP-TEA – synthetic routes; M1,
M2 and M3 – methods of synthesis within the respective synthetic
route. c Reaction time of 1 h. d Reaction time of 24 h.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Dalton Transactions
Paper
i
70%. The morphology of DOPAl containing nPr or Pr groups
was very similar to that of DMPAl or DEPAl, as indicated by
SEM images (an example in Fig. 1B for DnPPAl). Particles with
much more regular shapes were easily obtained when TBP or
TPhP was utilized. The compounds DBPAl and DPhPAl formed
rod-like structures with cross-sections similar to a regular
hexagon and diameters of 0.2–2.0 μm (Fig. 1C for DBPAl). In
contrast, BEHPAl exhibited a wax-like consistency.
DOP, DOPNa, and DOP-TEA synthetic routes. DOPAl can be
synthesized directly from an appropriate amount of DOP or its
sodium salt (DOPNa). The aluminum source can be AlLact
(Scheme 1b) or TEA (Scheme 1c). In our studies, these
methods were tested using n-butyl, 2-ethylhexyl, or phenyl
derivatives. The reactions of AlLact with DBP or its sodium salt
were performed at room temperature, using water as the
solvent. After mixing the solutions of both reagents, DBPAl
almost immediately precipitated. DBPAl synthesized in this
manner formed large, irregular agglomerates composed of
spherical particles with diameters of 0.5–1.0 μm (Fig. 1D). Due
to the higher hydrophobicity of BEHP, its reaction with AlLact
required prolonged heating under reflux. Under these con-
ditions, a wax-like solid of BEHPAl was separated from the
mixture, and its amount increased over time; after 24 h, the
overall yield was 75%. The morphology of this product
(Fig. 1E) was very similar to that of DBPAl synthesized via an
analogous process. BEHP neutralization with an aqueous solu-
tion of sodium bicarbonate followed by the slow addition of an
AlLact solution led to the formation of a stable dispersion con-
taining particles with an average size of approximately 200 nm.
BEHPAl could be precipitated from this system via salting out
with a saturated NaCl solution. SEM analysis of the resulting
product (Fig. 1F) indicated that it consisted mainly of irregular
agglomerates. However, there was also a small fraction of
rod-like domains.
Scheme 1 General synthetic routes to DOPAl applied in this paper:
(a) TOP route, (b) DOP or DOPNa route, and (c) DOP-TEA route.
Despite its lack of solubility in water, DPhP easily yielded
the desired product when heated to 130 °C in an aqueous solu-
tion of AlLact. Even after 1 h, rod-like particles of DPhPAl
(Fig. S1A and S1B, ESI†) could be obtained almost quantitat-
ively. However, increasing the reaction time resulted in the for-
mation of crystallites with more regular and uniform shapes.
These crystallites measured a few micrometers long and
exhibited hexagonal cross-sections with diameters less than
1 μm (Fig. S1C and S1D, ESI†). Samples of DPhPAl with very
similar morphologies could be obtained at lower temperatures
and over shorter reaction times when the sodium salt of DPhP
was used (Fig. S1E and S1F, ESI†).
The reactions of DOP with TEA were performed according
to a modified procedure described in our previous paper.37
They resulted in the formation of the appropriate DOPAl in
nearly quantitative yield. In SEM images of the as-synthesized
compounds BEHPAl (Fig. 2A) or DBPAl (Fig. 2B), one cannot
distinguish any submicron rod-like particles. The large struc-
tures present in these samples possessed an irregular (DBPAl)
or continuous surface (BEHPAl). Heating DBPAl in toluene led
to very small changes in the sample morphology, e.g., a slight
increase in the porosity of the particle surface (Fig. 2C).
Fig. 1 SEM images of DMPAl (A), DnPPAl (B), DBPAl (C, D), and BEHPAl
(E, F). The products were synthesized from AlLact and the appropriate
amount of TOP (images A–C), DOP (images D–E), or DOPNa (F).
phosphoric acid (DOP), which subsequently reacted with the
aluminum substrate.44
Initially, these reactions were performed in solution, and
after approximately 4 h, a white solid consisting of DOPAl
began to precipitate; the amount of solid increased with time.
SEM images revealed that the products formed fibrous par-
ticles with particle diameters in the range of 0.1–2.0 μm
(Fig. 1A). The particles were placed irregularly and formed ir-
regular unwoven fabrics. We also observed that, under suitable
conditions (T
= 130 °C and pressure of approximately
1.5 MPa), DEPAl could be synthesized in the form of large
(approximately 100 μm in diameter) crystals that are suitable
for X-ray structure determination.
All syntheses involving TOP containing hydrophobic alkyl
chains or phenyl groups required temperatures in the range of
130–150 °C and the use of pressure vessels. Under such con-
ditions pure DOPAl were obtained in yields between 20 and
This journal is © The Royal Society of Chemistry 2016
Dalton Trans.

View Article Online
Paper
Dalton Transactions
The collected crystallographic data indicate that both types
of DEPAl chains show considerable similarities in their struc-
ture when considering the coordination spheres of aluminum
and phosphorus. The chains are formed by the linking of the
adjacent aluminum centers by three O–P–O bridges, each of
which belongs to a different phosphate group. The plane of
the O–P–O bridge is inclined at an angle of 22.7° with respect
to the c-axis, which creates the potential for a change in the
Al⋯Al distance due to temperature or interactions between
chains.
The Al–O distances vary between 1.8803(7) and 1.8917(12)
Å. The cis O–Al–O angles are in the range of 88.65(4)–91.35(4)°,
whereas angles between O ligands in trans positions differ
from 180° by less than 0.2°. The lengths of the P–O bonds vary
between 1.4944(8) and 1.5843(11) Å, whereas the O–P–O angles
range from 101.37(5) to 118.82(4)° (Tables S2 and S3, ESI†).
The coordination numbers of aluminum and phosphorus
in each type of DEPAl chain are equal to 6 and 4, respectively.
All P and Al centers are surrounded by oxygen ligands located
at the vertices of the corresponding polyhedra. Each PO4 tetra-
hedron shares two of its O atoms with two AlO6 octahedra.
The calculated distortion of the AlO6 octahedron (ΔAl),
O–Al–O angle variances (σ2), aluminum bond valence sums
(sAl),52 and the length of the resultant bond-valence vector
(|vAl|) for each crystallographically independent aluminum
atom are summarized in Table 3. The ΔAl values can be esti-
mated by using the modified general equation (eqn (1)) that
was successfully applied in the case of TiO6 octahedra in
titanates.53
Fig. 2 SEM images of BEHPAl (A) and DBPAl (B–D) synthesized from
TEA and DOP. For DBPAl: the as-synthesized product (B), the product
heated in toluene for 12 h (C), and the product heated with 1-butanol/
water mixture (40 : 60 v/v) (D).
However, replacing toluene with a mixture of 1-butanol and
water (40 : 60 v/v) and heating to 130 °C induced the formation
of 2D nanoparticles (Fig. 2D) similar in shape to those
obtained from AlLact and TBP.
It should be noted that the carbon or hydrogen contents in
DOPAl samples prepared by different chemical routes but con-
taining the same organic ligands are in good agreement with
those calculated for the respective aluminium tris(diorgano-
phosphates) (Table 2). This strongly indicates that the choice
of the method of preparation only affects the DOPAl
morphology, with no changes in the chemical composition of
the resulting product.
X
2
ΔAl ¼ 6ꢀ1
ꢁ
½ðdi ꢀ daveÞ=dave
ꢂ
ð1Þ
i
where di and dave represent the interatomic distances of Al–O
and the average value, respectively.
Crystal structure of DEPAl at −173 °C
To estimate the strains at phosphorus centers, one can
One of the main objectives of our studies was to obtain DOPAl apply the bond-valence vector (BVV) model proposed by
in a form suitable for single-crystal X-ray analysis. We antici- Zachara.54 Table 4 contains data regarding the length of the
pated that solving its structure could provide us the reference resultant bond-valence vector (|vP|) for each crystallographi-
data required to properly characterize other organically modi- cally independent phosphorus center in the DEPAl structure.
fied aluminum phosphates using powder XRD. We managed
As is shown, the ΔAl, σ2 and |vAl| values are very low, which
to synthesize such a material via the TEP and AlLact reaction indicates that the octahedral environments of the oxygen
performed in a dilute system under hydrothermal conditions ligands around all of the aluminum metal centers remain
(130 °C, 1.5 MPa). The obtained DEPAl crystals were subjected nearly undistorted. Furthermore, the bond valence sums calcu-
to a reliable single-crystal X-ray analysis at −173 °C.
lated for each type of Al center are close to 3, the theoretically
DEPAl crystallizes into a trigonal crystallographic system, expected value for trivalent elements. In contrast, in the case
ˉ
space group P3 and forms a linear coordination polymer of the PO4 group, the deviation of its geometry from that of a
propagating along the c-axis of its unit cell (Fig. 3). In its struc- regular tetrahedron is significant. Generally, the P–O bonds
ture, one can distinguish two crystallographically nonequiva- involved in the bridging of aluminum atoms are on average
lent chains that are characterized by a 3-fold symmetry axis 5% shorter than those involved in connecting alkyl groups. In
parallel to the [001] direction. The symmetries of these struc- contrast, the angles between them, which approach 120°, are
ˉ
tures can be described by p3 and p3 rod groups, respectively. 12–15% greater than those between ester linkages. The largest
The chains propagating along the edges of the DEPAl unit cell, constraints on the geometry of the coordination sphere were
where aluminum atoms are also located at inversion centers, estimated for the phosphorus centers in the B-type DEPAl
belong to the former group and are hereafter referred to as chains, for which the resultant bond-valence vectors were the
A-type chains. The chains exhibiting lower symmetry (B-type longest (Table 4). These vectors are directed outwards from the
chains) are situated within the unit cell (Fig. S2, ESI†).
chain and obliquely to the symmetry axis (Fig. S3, ESI†), which
Dalton Trans.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Dalton Transactions
Paper
Fig. 3 ORTEP diagram of the DEPAl crystal structure with an A-type chain on the left side and a B-type chain on the right side. a–p are the symbols
of translation symmetry for equivalent positions defined in Table S1.†
Table 3 Octahedral distortions (ΔAl), O–Al–O angle variances (σ2),
bond valence sums (sAl), and the length of the resultant bond-valence
vector (|vAl|) for each crystallographically independent aluminum coordi-
nation center in DEPAl
indicates that the largest strains are caused by the arrange-
ment of the organic ligands. It should also be noted that,
according to the BVV model, the smallest strains occur in the
phosphate tetrahedra forming A-type chains for which the
value of |vP| is almost equal to 0.
Atom
ΔAl × 105
σ2 [(°)2]
sAl
|vAl| [v.u]b
a
The main structural difference between A-type and B-type
DEPAl chains is the conformation of the ethyl groups around
the phosphorus centers. In A-type chains, all P atoms are crys-
tallographically equivalent, and the O–P–O–C torsion angles
characterizing each of the organic substituents connected to
them are equal to 65.57(8) and 69.26(8)° (Table S4, ESI†). In
agreement with IUPAC recommendations,55 these ethyl groups
have a gauche–gauche (G+G+) conformation. In contrast, the
chains with lower symmetry (B-type chains) contain two crys-
tallographically unequal phosphorus centers, each of which is
characterized by a gauche–trans arrangement of alkyl groups
(G−T and G+T for the P11 and P12 centers, respectively). Based
on the theoretical calculations performed for dimethyl-
phosphate anions, the gauche–gauche conformation is thermo-
dynamically more stable than the gauche–trans one.56
Al1
Al2
Al11
Al12
0
0
0.11
0.03
1.5
2.0
1.0
1.1
3.05
3.02
3.06
2.97
0.000
0.000
0.003
0.003
a Bond valences were calculated according to the method proposed in
the literature41 (for details, see the ESI). b Valence units.
Table 4 Bond valence sums (sP) and the length of the resultant bond-
valence vector (|vP|) for each crystallographically independent phos-
phorus coordination center in DEPAl
Atom
sP
|vP| [v.u.]b
a
P1
P11
P12
4.98
4.97
4.98
0.003
0.074
0.077
A detailed analysis revealed that the periodicity of each
DEPAl chain is equal to the c-axis vector, and there are no
directional interactions, e.g., hydrogen bonds, beyond van der
a Bond valences were calculated according to the method proposed in
the literature41 (for details, see the ESI). b Valence units.
This journal is © The Royal Society of Chemistry 2016
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Waals packing contacts. Each polymer chain is surrounded by
six others, where the aluminum centers are shifted alternately
in the c-axis direction by +1/6 and −1/6 of its vector value
(Fig. S4, ESI†). As a result, the DEPAl structure is characterized
by hexagonal close-packing.
Powder XRD patterns of DOPAl samples
Fig. 4 shows a simulation of the DEPAl powder XRD pattern
calculated based on the data collected by single-crystal X-ray
analysis and the results of PXRD measurements performed at
room temperature. As shown, at −173 °C, the largest diffrac-
tion peak occurs at 2θ = 8.65°. This peak is attributed to the
diffraction of X-rays on the (2–10) lattice plane, which is per-
pendicular to the a-axis and passes through the aluminum
metal centers of one A-type and one B-type DEPAl chain. The
interplanar distance characterizing this plane (d2–10 spacing) is
equal to 10.22 Å. It is worth noting that, with increasing temp-
erature, the main XRD reflection did not disappear but shifted
to a lower value of 2θ, suggesting that DEPAl retained its hexa-
gonally packed structure and the d2–10 spacing increased.
However, one cannot exclude a certain degree of translational
movement of the DEPAl chains in the c-axis direction.
The analysis of PXRD diffractograms obtained for other
DOPAl samples (Fig. 5 and Tables S6–S13, ESI†) allows for the
hypothesis that they may form crystal structures analogous to
the structure of DEPAl with polymeric chains arranged in a
hexagonal close-packing pattern. This behavior is suggested by
the similarities in the shapes of the respective PXRD patterns,
particularly the presence of a single, high-intensity diffraction
peak at low values of 2θ. These reflections most likely arise
from the diffraction phenomena occurring on the lattice
planes parallel to the axis of the DOPAl chain and passing
through its aluminum centers – the exact analogues of the
DEPAl (2–10) plane. We tried to carry out a full-profile refine-
ment of the cell parameters of different DOPAl complexes,
for example by applying Pawley fitting. Unfortunately, for a
majority of DOPAl samples such calculations gave unreliable
results due to an insufficient number of reflexes observed,
especially in the area of low values of 2θ. The only exception
occurred in the case of DnPPAl (Fig. S5, ESI†), for which we
Fig. 5 PXRD patterns for different DOPAl samples measured at room
temperature.
obtained the following cell parameters: a = 23.6800(11) Å, c =
3
ˉ
9.2316(10) Å, and cell volume of 4475.4(7) Å (space group P3),
consistent with the model of the DOPAl structure proposed by
us. In order to verify our hypothesis about the similarities in
crystal structures formed by different aluminium tris(diorgano-
phosphates) we decided to try another approach. By assuming
that the average Al⋯Al distance in each DOPAl chain is con-
stant and equal to approximately 4.6 Å, one can evaluate the
volume attributable to
a
single monomeric unit in
Al[O2P(OR)2]3 crystals (VXRD) based on the position of the
highest, low-angle reflex observed on the DOPAl PXRD pat-
terns. These values can also be calculated using empirical
relationships between the volume of each structural unit and
the unit’s arrangement in space (Vcalcd).57 The results obtained
using both methods are summarized in Table 5.
As shown, both methods yield very similar values for the
monomeric unit volume with differences of only a few percent.
It should be noted that in the case of DnPPAl the same situ-
ation occurs, if we compare its VXRD value with that evaluated
based on the cell volume obtained from Pawley fitting of the
PXRD pattern and the assumption that the Z parameter is
equal to 6 (VPawley = 746 Å3). This good agreement between the
Fig. 4 PXRD patterns of DEPAl measured at room temperature and cal-
culated from the data obtained in single-crystal X-ray analysis carried
out at −173 °C.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Dalton Transactions
Paper
relaxation time. However, one cannot exclude that the presence
of aluminum centers in non-equivalent crystallographic
environments may also contribute to the line shape.
Table 5 The main peak position in PXRD diffractograms measured at
room temperature (2θ), the corresponding interplanar distance (d2θ), and
monomeric unit volumes evaluated for different DOPAl samples
Similarly, the 31P MAS NMR spectra of the majority of the
DOPAl samples exhibit only one symmetric signal at δ between
−10 and −26 ppm, depending on the type of organic ligand
used (Fig. 6). The spectral feature suggests that there is only
one type of phosphorus chemical environment, which can be
attributed to the tetrahedral phosphorus nuclei in aluminum
phosphates.18,64,65 In contrast, the 31P MAS NMR spectra of
DMPAl or BEHPAl show several signals from non-equivalent
phosphorus nuclei, whose precise origin remains unclear.
DOPAl
2θ [°]
d2θ [Å]
Vcalcd [Å3]
VXRD [Å3]
ΔVa [%]
DMPAl
DEPAl
DnPPAl
DiPPAl
DBPAl
BEHPAl
DPhPAl
9.68
8.47
7.52
7.76
6.64
5.35
6.73
9.13
10.44
11.75
11.38
13.31
16.50
13.12
429
575
722
738
868
1471
898
443
578
734
688
941
1445
915
3.3
0.5
1.7
−6.8
8.4
−1.8
1.9
a ΔV = 100 × (VXRD − Vcalcd)/Vcalcd
.
FT-IR and Raman spectra of DOPAl
values of VXRD and Vcalcd (or VPawley) supports the initial
hypothesis that DEPAl and other DOPAl compounds possess
analogous hexagonal close-packed polymeric structures, where
the distance between chains depends on the dimensions of
the organic ligands.
Based on the literature data concerning the FT-IR and Raman
spectra of ammonium,66 sodium56 or barium67 dialkyl phos-
phates, as well as those obtained from theoretical studies on
the vibrational spectra of dimethyl56,66,68 or diethyl67 phos-
phate anions, one can expect four diagnostic bands associated
with the stretching modes of phosphate ligands to appear over
the wavenumber range of 750–1250 cm−1. Analogously to pre-
vious assignments, we suggest that the two strong bands
observed in the FT-IR spectra of DOPAl samples between
1115–1140 and 1190–1230 cm−1 (Fig. S6–S9, ESI†) can be
attributed to the symmetric and asymmetric stretching
vibrations of P–O bonds in Al–O–P–O–Al bridges, respec-
tively.56,67 The asymmetric stretching mode is prominent only
in the FT-IR spectra, whereas the corresponding symmetric
mode is also active in the Raman spectra (Fig. S6, ESI†).
However, the position of the latter is not exactly the same as
that in the FT-IR spectra, most likely due to contributions
from the symmetric C–O stretching mode. The absorption
bands attributable to the symmetric O–P–O stretching mode of
the phosphodiester group in DOPAl occur in the range of
750–790 cm−1 in both the infrared and Raman spectra.56,67
The Raman intensities of these signals are significantly greater
than those of the symmetric P–O stretching bands in the brid-
ging units. The corresponding asymmetric mode gives rise to
Solid-state MAS NMR spectra of DOPAl
The 27Al MAS NMR spectra of the majority of the DOPAl
samples recorded at room temperature show a single, broad
resonance peak with a maximum located between −21 and
−31 ppm (Fig. 6). The observed peak positions are consistent
with the literature data previously reported for aluminum
centers in AlO6 octahedra forming crystalline structures of
aluminum methylphosphonates58–61 and natural or synthetic
hydrated aluminum phosphates.62,63 The broadening of the
signal most likely arises from the high quadrupole moments
and electric field gradients of aluminum nuclei, suggesting
that the lower symmetry effectively reduces the spin–lattice
IR bands in the range of 820–850 cm−1 56,67
which can also be
,
observed as weak signals or shoulders in the Raman spectra of
DOPAl.
The characteristic absorption bands for O–Al–O vibrations
in the bridging units are most likely located between 480 and
620 cm−1. In the FT-IR spectra of the investigated DOPAl, one
can distinguish 5 highly intense signals in this region.
However, these signals may overlap with weak bands that are
attributable to the deformation modes of the phosphate
moiety.
The other stretching and bending modes of C–H, C–C, or
C–O bonds give rise to several additional bands with wave-
numbers between 900 and 3000 cm−1. However, these bands
are not of immediate interest in this study.
It is worth noting that the FTIR spectra of BEHPAl, DPhPAl
or DBPAl samples obtained in different chemical reactions are
almost identical (Fig. S7–S9 in the ESI,† respectively). This
observation, in combination with the results of elemental ana-
Fig. 6 27Al (left) and 31P (right) MAS NMR spectra of DOPAl recorded at
room temperature.
This journal is © The Royal Society of Chemistry 2016
Dalton Trans.

View Article Online
Paper
Dalton Transactions
lysis (see Table 2), suggests that for a given organic group in
the phosphate moiety the use of different synthetic routes
results in the formation of the same DOPAl complex.
DOPAl in solution
DOPAl compounds are insoluble in aprotic organic solvents
and even the presence of large hydrocarbon ligands merely
results in the formation of a stable colloid dispersion, as in
the case of BEHPAl mixed with benzene, hexane, chloroform,
diethyl ether or THF. Based on the NMR studies of those
systems, one can assume that the particles are stabilized by a
small fraction of soluble aluminum phosphate species. The
signals observed in the 27Al NMR spectra of BEHPAl (Fig. S10,
ESI†) are broad and often overlap with the artificial probe
signal located at δ ≈ 50 ppm. Nevertheless, in a spectrum
recorded in C6D6, one can distinguish a distinct asymmetric
resonance peak with a chemical shift (−17.7 ppm) close to
that detected in the MAS NMR measurements, which suggests
the presence of polymeric species structurally similar to the
particles in the solid sample. In contrast, the corresponding
31P NMR spectra of BEHPAl (Fig. S10, ESI†) show at least four
different signals at δ values between −20 and 10 ppm, depend-
ing on the choice of solvent. This indicates that, in solution,
diorganophosphate ligands can coordinate aluminum centers
in various ways.
Fig. 7 SEM image of an epoxy resin cured with a solution of DPhPAl in
triethylenetetramine.
a solid matrix (Fig. 7). The resulting polymer composites
exhibit reduced flammability and some improvements in their
mechanical properties.41,69
Among all of the DOPAl species we synthesized, only those
containing methyl or ethyl groups exhibited very limited solu-
bility in water, approximately 0.4–0.5 g per 100 g of H2O at
room temperature. The 27Al and 31P NMR studies of DMPAl
solutions in D2O (Fig. S12 and S13 in the ESI,† respectively)
revealed the presence of several populations of phosphorus or
aluminum nuclei, each of which existed in a different chemi-
cal environment. The signal positions for the aluminum
species are reasonable for aluminum cations complexed by
water (δ value of approximately 1.2 ppm) or aluminum cations
exhibiting a mixed surrounding of water and dimethyl-
phosphate ligands (broad and overlapping signals with
δ values of approximately −2.3 and −5.9 ppm).70,71 The sharp
peak observed in the 31P NMR spectrum at a δ value of
5.8 ppm is most likely related to the free [(CH3O)2PO2]− ions,
One of the most important features of DOPAl is its ability to
form true solutions in primary amines. We hypothesize that,
in these systems, DOPAl undergoes dissociation to diorgano-
phosphate anions and complex aluminum cations, as indi-
cated in eqn (2). This conclusion is supported by the results of
the NMR studies of the DBPAl/hexan-1-amine model system
(Fig. S11, ESI†).
Al½O2PðORÞ ꢂ þ 6R′NH2 ! ½AlðR′NH2Þ ꢂ3þ þ 3½ðROÞ PO2ꢂꢀ
2 3
6
2
ð2Þ
The 27Al (Fig. S11a, ESI†) and 31P (Fig. S11b, ESI†) NMR whereas the additional low-intensity signals located between 0
spectra of DBPAl in a CDCl3/hexan-1-amine mixture (1 : 2 v/v) and −10 ppm arise from the same ligands complexing alumi-
do not exhibit any signals at chemical shifts typical of poly- num centers.70,71 The relative intensity of the latter decreases
meric DOPAl species. Instead, the phosphorus resonance spec- with increasing dilution of the DMPAl solution. The presence
trum shows a single peak at a δ value of approximately 0 ppm, of several different dimethylphosphate species is also con-
1
which is also present in an analogous spectrum recorded for firmed by the shape of the H NMR spectrum (Fig. S14, ESI†),
the hexan-1-amine salt of DBP (Fig. S11c, ESI†). 27Al NMR in which one can distinguish three groups of signals related to
measurements revealed that the resonance signals of the protons of methoxy groups. These species are characterized by
aluminum cations are not clearly visible and only account for chemical shifts of approximately 3.58, 3.67 and 3.75 ppm.
1
a broadening of the probe signal at approximately 0 ppm.
A careful analysis of the relative integral intensities of H and
It is worth noting that amine-induced dissociation of 31P NMR signals and their dependence on the DMPAl solution
DOPAl is a reversible process. For example, a polymeric form concentration suggests that the farthest upfield signal in the
of DBPAl can again be obtained as a precipitate if the saturated 1H NMR spectra is related to the free dimethylphosphate
solution prepared at room temperature from 16 g of DBPAl anions. The two other resonance peaks that decreased with
and 100 g of hexan-1-amine is cooled to 4 °C. The same dilution most likely arose from dimethylphosphate species
applies to the case in which amine is consumed in a chemical coordinating aluminum centers. It is worth noting that, even
reaction and can be utilized for the preparation of polymer in a concentrated DMPAl aqueous solution, approximately
nanocomposites. We examined this approach by curing an 57% of the dimethylphosphate groups existed as free ions
epoxy resin with a DPhPAl solution in triethylenetetramine, (Table 6). It is also interesting that, after the evaporation
which led to the formation of small nanoparticles dispersed in of water from such systems, one can recover the DMPAl
Dalton Trans.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Dalton Transactions
Paper
Table 6 Relative integral intensities of signals observed in the 1H or 31P
NMR spectra of aqueous DMPAl solutions and the content of the corres-
ponding dimethylphosphate groups
favors the formation of closely packed hexagonal crystalline
domains. Both the type of organic group R and the method of
preparation strongly affect the morphology of the product.
Therefore one can obtain DOPAl samples containing highly
crystalline fibers with hexagonal cross-sections, irregular
agglomerates or amorphous wax-like phases.
DOPAl species are insoluble in aprotic organic solvents or
strong inorganic acids. Only BEHPAl forms solutions or stable
colloidal dispersions in selected organic media. DOPAl con-
taining methyl or ethyl groups are sparingly soluble in water,
in which they act as dynamers susceptible to changes in the
solution concentration. Similarly, under the effects of primary
amines, samples of DOPAl undergo a reversible dissociation to
ionic species and form true solutions that can be successfully
applied as curing systems for epoxy resins.
δP
[ppm]
Relative
integral
δH
[ppm]
Relative
integral
Solution
mol%
mol%
Saturated
5.82
−3.47
−8.79
5.87
−3.43
—
10.17
6.09
1.00
2.50
1.00
—
59
35
6
71
29
—
3.58
3.65
3.75
3.59
3.68
3.77
6.60
4.03
1.00
26.82
11.25
1.00
57
35
8
69
29
2
Dilute
coordination polymer. Therefore, it can be assumed that,
during synthesis, the morphology of each DMPAl particle con-
stantly changes and depends on the balance between dis-
solution and precipitation. This behavior is supported by the
fact that the solubility of DMPAl in water at the boiling point
of the reaction mixture is approximately 22 times higher than
that estimated at room temperature.
It is worth noting that DOPAl samples exhibit very interest-
ing thermal properties that are the subject of the ongoing
studies and will be discussed in detail in a separate paper.
Our studies indicate that, over a given period (24 h), diorgano-
phosphate moieties in DMPAl or DEPAl are hydrolytically
stable at temperatures of up to approximately 100 °C. However,
in the case of the methyl derivative, an increase in temperature
results in gradual hydrolysis of P–OR linkages. We observed
that an insoluble product precipitated from aqueous solutions
of DMPAl heated for 1 h at 150 °C. The FT-IR spectrum of this
solid (Fig. S15, ESI†) does not contain absorption bands
characteristic of the C–H stretching modes that, in an ana-
Acknowledgements
This research was partially funded through the Structural
Funds in the Operational Programme – Innovative Economy
(IE OP) financed from the European Regional Development
Fund – Project “Modern material technologies in aerospace
industry”, no. POIG.01.01.02-00-015/08-00.
logous DMPAl spectrum, show maxima at 2957 and 2851 cm−1
.
Instead, one can distinguish two very strong bands at wave-
numbers of 3359 and 3177 cm−1 attributable to the O–H
stretching modes of hydroxyl groups formed during the hydro-
lysis of phosphoester linkages. Substantial differences also
occur with respect to the MAS NMR spectra (Fig. S16, ESI†).
The insoluble product of DMPAl hydrolysis is characterized by
two resonance peaks of 31P nuclei located at −23.80 and
−29.23 ppm that can be attributed to PO43− ligands in partially
hydrated aluminum phosphate. A corresponding 27Al MAS
NMR spectrum exhibits several signals within the ranges of
chemical shifts typical of tetrahedral and octahedral alumi-
num centers in AlPO4.65
Based on the results of studies performed on DEPAl,
DBPAl, and BEHPAl, one can conclude that these compounds
have high chemical resistance to concentrated solutions of
strong mineral (HCl, HNO3, H3PO4, H2SO4) or organic (acetic,
trichloroacetic, p-toluenesulfonic) acids, even when heated to
50 °C for 24 h. However, they easily transform into aluminum
hydroxide when treated with 10 or 20 wt% solutions of NaOH
in water or ethanol.
Notes and references
1 S. Penczek and P. Klosinski, in Biomimetic Polymers, ed.
C. G. Gebelein, Plenum Press, New York, 1990, ch. 12, pp.
223–241.
2 S. Penczek and P. Klosinski, in Models of Biopolymers By
Ring-Opening Polymerization, ed. S. Penczek, CRC Press,
Inc., Boca Raton, FL, 1989, ch. 4, pp. 291–378.
3 A. Eschenmoser, Chimia, 2005, 59, 836.
4 S. Penczek, J. Pretula and K. Kaluzynski, Biomacromolecules,
2005, 6, 547.
5 R. Wódzki and P. Kłosiński, Macromol. Chem., 1990, 191,
921.
6 R. Wódzki, M. Świątkowski and G. Łapienis, React. Funct.
Polym., 2005, 62, 195.
7 R. Wódzki, M. Świątkowski and G. Łapienis, React. Funct.
Polym., 2010, 70, 463.
8 K. Kaluzynski, J. Pretula, G. Lapienis, M. Basko,
Z. Bartczak, A. Dworak and S. Penczek, J. Polym. Sci., Part A:
Polym. Chem., 2001, 39, 955.
9 J. Rudloff, M. Antonietti, H. Cölfen, J. Pretula,
K. Kaluzynski and S. Penczek, Macromol. Chem. Phys., 2002,
203, 627.
Conclusions
We have shown that aluminum tris(diorganophosphates) form 10 S. Penczek, J. Pretula and K. Kaluzynski, J. Polym. Sci., Part
catena-Al[O2P(OR)2]3 polymeric chains, whose parallel location
A: Polym. Chem., 2005, 43, 650.
This journal is © The Royal Society of Chemistry 2016
Dalton Trans.

View Article Online
Paper
Dalton Transactions
11 V. G. Lebedev, K. K. Palkina, S. I. Maksimova, 41 M. Debowski, Z. Florjanczyk, K. Łokaj, A. Wolak, J. Zachara,
E. N. Lebedeva and O. V. Galaktionova, Zh. Neorg. Khim.,
1982, 27, 2980.
12 Y. Han, Z. Pan, N. Shi, L. Liao, C. Liu, G. Wu, Z. Tang and
Y. Xiao, Wuji Huaxue Xuebao, 1990, 6, 17.
13 C. G. Lugmair, T. D. Tilley and A. L. Rheingold, Chem.
Mater., 1997, 9, 339.
14 M. Fröba and M. Tiemann, Chem. Mater., 1998, 10, 3475.
15 M. Fröba and M. Tiemann, Chem. Mater., 2001, 13, 3211.
16 H. Tanaka and M. Chikazawa, J. Mater. Chem., 1999, 9,
2923.
17 H. Tanaka and M. Chikazawa, Mater. Res. Bull., 2000,
35, 75.
M. Heneczkowski, M. Oleksy and H. Galina, presented
in part at the 8th IUPAC International Conference on Novel
Materials and their Synthesis (NMS-VIII) & 22nd International
Symposium on Fine Chemistry and Functional Polymers
(FCFP-XXII), Xi’An, China, 2012.
42 Z. Florjańczyk, A. Lasota and M. Leśniewska-Mizak,
Polimery, 2004, 49, 389.
43 E. Zygadło-Monikowska, Z. Florjańczyk, A. Wolak, A. Lasota
and A. Plichta, PL Pat., 197233, 2008.
44 Z. Florjańczyk, A. Wolak, M. D¸ebowski, A. Plichta,
J. Ryszkowska, J. Zachara, A. Ostrowski, E. Zawadzak and
M. Jurczyk-Kowalska, Chem. Mater., 2007, 19, 5584.
18 M. Schulz, M. Tiemann, M. Fröba and C. Jäger, J. Phys. 45 Z. Florjańczyk, M. Dębowski, A. Wolak, M. Malesa and
Chem. B, 2000, 104, 10473. J. Płecha, J. Appl. Polym. Sci., 2007, 105, 80.
19 Y. Huang and Y. Zhimin, J. Am. Chem. Soc., 2005, 127, 46 H. Fischer, Mater. Sci. Eng., C, 2003, 23, 763.
2731.
47 B. Mossety-Leszczak, M. Włodarska, M. Kowalik and
K. Łokaj, Macromol. Symp., 2013, 329, 193.
20 T. Kimura, Microporous Mesoporous Mater., 2005, 77, 97.
21 A. Vioux, J. Le Bideau, P. H. Mutin and D. Leclercq, Top. 48 B. Mossety-Leszczak, B. Strachota, A. Strachota,
Curr. Chem., 2004, 232, 145. M. Steinhart and M. Slouf, Eur. Polym. J., 2015, 72, 238.
22 Metal Phosphonate Chemistry: From Synthesis to Applications, 49 TOPAS V.5, Bruker AXS, Karlsruhe, Germany, 2014.
ed. A. Clearfield and K. Demadis, Royal Society of 50 Oxford Diffraction CRYSALISPRO, Oxford Diffraction Ltd.,
Chemistry, Cambridge, 2011.
23 R. Murugavel, A. Choudhury, M. G. Walawalkar, 51 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystal-
R. Pothiraja and C. N. R. Rao, Chem. Rev., 2008, 108, 3549. logr., 2008, 64, 112.
24 G. P. Funkhouser, N. Tonmukayakul and F. Liang, Lang- 52 I. D. Brown and D. Altermatt, Acta Crystallogr., Sect. B:
Abingdon, England, 2009.
muir, 2009, 25, 8672.
Struct. Sci., 1985, 41, 244.
25 M. R. Mason, J. Cluster Sci., 1998, 9, 1.
53 A. Aimi, D. Mori, K. Hiraki, T. Takahashi, Y. J. Shan,
Y. Shirako, J. Zhou and Y. Inaguma, Chem. Mater., 2014, 26,
2601.
54 J. Zachara, Inorg. Chem., 2007, 46, 9760.
55 IUPAC Gold Book Home Page, http://goldbook.iupac.org/
T06406.html (accessed December 21, 2015).
56 J. Florián, V. Baumruk, M. Štrajbl, L. Bednárová and
J. Štěpánek, J. Phys. Chem., 1996, 100, 1559.
26 K. Maeda and F. Mizukami, Catal. Surv. Jpn., 1999, 3, 119.
27 C. G. Lugmair, T. D. Tilley and A. L. Rheingold, Chem.
Mater., 1999, 11, 1615.
28 E. S. Martin and L. F. Wieserman, US Pat., 4929589, 1990.
29 S. T. Meyers, J. T. Anderson, D. Hong, C. M. Hung,
J. F. Wager and D. A. Keszler, Chem. Mater., 2007, 19, 4023.
30 R. N. Rothon, Thin Solid Films, 1981, 77, 149.
31 Y. Huang, J. Ma, J. Yang, K. Cao and Z. Lu, J. Phys. Chem. C, 57 S. Beaucamp, D. Mathieu and V. Agafonov, Acta Crystallogr.,
2012, 116, 22518. Sect. B: Struct. Sci., 2007, 63, 277.
32 Y. Cao, Y. Ju, F. Liao, X. Jin, X. Dai, J. Li and X. Wang, RSC 58 K. Maeda, Y. Kiyozumi and F. Mizukami, Angew. Chem., Int.
Adv., 2016, 6, 14852. Ed. Engl., 1994, 33, 2335.
33 J. Ma, J. Yang, Y. Huang and K. Cao, J. Mater. Chem., 2012, 59 K. Maeda, J. Akimoto, Y. Kiyozumi and F. Mizukami,
22, 2007. Angew. Chem., Int. Ed. Engl., 1995, 34, 1199.
34 X. Lin, S. Du, J. Long, L. Chen and Y. Wang, ACS Appl. 60 K. Maeda, J. Akimoto, Y. Kiyozumi and F. Mizukami,
Mater. Interfaces, 2016, 8, 881. J. Chem. Soc., Chem. Commun., 1995, 1033.
35 Z. Ding, B. Chen, J. Ding, L. Wang and Y. Han, J. Colloid 61 K. Maeda, Y. Hashiguchi, Y. Kiyozumi and F. Mizukami,
Interface Sci., 2014, 425, 102. Bull. Chem. Soc. Jpn., 1997, 70, 345.
36 R. Murugavel and S. Kuppuswamy, Angew. Chem., Int. Ed., 62 D. Zhou, J. Xu, J. Yu, L. Chen, F. Deng and R. Xu, J. Phys.
2006, 45, 1. Chem. B, 2006, 110, 2131.
37 Z. Florjańczyk, A. Lasota, A. Wolak and J. Zachara, Chem. 63 Y. Huang and Z. Yan, J. Am. Chem. Soc., 2005, 127,
Mater., 2006, 18, 1995. 2731.
38 K. L. Fujdala and T. D. Tilley, J. Am. Chem. Soc., 2001, 123, 64 M. R. Mason, R. M. Matthews, M. S. Mashuta and
10133. J. F. Richardson, Inorg. Chem., 1996, 35, 5756.
39 J. Pinkas, D. Chakraborty, Y. Yang, R. Murugavel, 65 M. P. J. Peeters, J. W. de Haan, L. J. M. van de Ven and
M. Noltemeyer and H. W. Roesky, Organometallics, 1999,
18, 523.
40 R. Murugavel, M. G. Walawalkar, M. Dan, H. W. Roesky and
C. N. R. Rao, Acc. Chem. Res., 2004, 37, 763.
J. H. C. van Hooff, J. Phys. Chem., 1993, 97, 5363.
66 Y. Guan, G. S. C. R. Glaser and G. J. Thomas Jr., J. Phys.
Chem., 1995, 99, 12054.
67 Y. Guan and G. J. Thomas Jr., Biopolymers, 1996, 39, 813.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Dalton Transactions
Paper
68 B. Jayaram, M. Mezei and D. L. Beveridge, J. Am. Chem. 70 R. F. Mortlock, A. T. Bell and C. J. Radke, J. Phys. Chem.,
Soc., 1988, 110, 1691. 1993, 97, 767.
69 H. Galina, M. Oleksy, M. Heneczkowski, Z. Florjanczyk, 71 R. F. Mortlock, A. T. Bell and C. J. Radke, J. Phys. Chem.,
M. Debowski and K. Lokaj, EP Pat., 2628766, 2014.
1993, 97, 775.
This journal is © The Royal Society of Chemistry 2016
Dalton Trans.