Colloidal Sodium and Potassium Phosphates in Organic Medium: Synthesis and Analysis

JOURNAL OF COLLOID AND INTERFACE SCIENCE ARTICLE NO.

207, 218 –227 (1998)

CS985540

Colloidal Sodium and Potassium Phosphates in Organic Medium: Synthesis and Analysis Bruno Delfort,1 Laurent Normand, Philippe Dascotte, and Loı¨c Barre´ Institut Franc¸ais du Pe´trole, 1-4 av de Bois-Pre´au, Rueil-Malmaison, France Received September 8, 1997; accepted March 26, 1998

Colloidal dispersions of sodium and potassium phosphates in a hydrocarbon medium have been synthesized by reacting, under specific conditions, aqueous sodium or potassium hydroxyde with tetraphosphorous decasulfide in hydrocarbon solvent in the presence of a surfactant such as alkylarylsulfonic acid. The resulting products can be defined as a distribution of inorganic particles stabilized by sodium or potassium alkylaryl sulfonate in a reversemicelle type association. The chemical structures of the mineral cores are determined using 31P NMR, potentiometric titration and elemental analysis. The core particles were found to be respectively a mixture of sodium (poly)phosphates and a mixture of potassium (poly)phosphates. Inorganic core and organic surfactant balances were determined using thermogravimetric analysis. The weight ratio of mineral core part to organic surfactant part is found about to be 1.6 for Na derivative and 2.4 for K derivative. The morphological characterization of the colloidal dispersion has been investigated using Small Angle X-ray Scattering, Transmission Electron Microscopy and Wide Angle X-ray Scattering. The inorganic particles appear as cristallized particles with diameters ranging from 50 to 360 Å. © 1998 Academic Press Key Words: colloidal sodium phosphate particles; potassium phosphate; synthesis; small angle X-ray scattering; wide angle X-ray scattering; transmission electron microscopy.

INTRODUCTION

Colloidal dispersions of inorganic particles in hydrocarbon media have attracted large interest among others in the field of additives for lubricants (1, 2). These inorganic particles such as calcium or magnesium carbonate (3, 4), sodium or potassium carbonate (5), and calcium borate (6, 7) are surrounded by a surfactant in a reverse-micelle type association (8, 2). When they are used in lubricant compositions colloidal calcium or magnesium carbonate species act as detergents to keep the engine clean. Because of their alkaline reservoir they are able to neutralize the acidic by-products issuing from the oxidation of oil and from fuel combustion products. Moreover, it has been shown that some of these metal colloidal species could be used as antiwear additives in lubricant compositions. The mechanism of the antiwear action of 1

To whom correspondence should be addressed.

0021-9797/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.

colloidal particles of calcium carbonate micellized with calcium alkylarylsulfonate has been investigated. When the additive is subjected to friction constraints, its micellar structure is modified. Boundary film formation is related to the adsorption of calcium carbonate on the surface and to the expulsion of the surfactant sulfonate chains surrounding the micelle core. Surface film formation grows by agglomeration and crystallization of calcium carbonate cores. These colloidal species are usually prepared according to an in situ process performing the synthesis of the inorganic salt simultaneously to its micellization. For example, colloidal sodium carbonate is prepared by reacting carbon dioxide with sodium hydroxide under well-controlled solvent and temperature conditions in the presence of a sodium alkylarylsulfonate as surfactant (4, 9). The structure of these different compounds has been investigated by different techniques, such as small angle X-ray scattering (SAXS) or small angle neutron scattering (SANS), and transmission electron microscopy (TEM) (10 –15). A core shell model whereby salt spheres are surrounded by a stabilizing surfactant shell has been proposed. In previous studies, we reported the synthesis of new colloidal dispersions in oil, such as calcium thiophosphates (16) and calcium hydroxide (17). In this paper, we introduce two new colloidal dispersions in oil, in which the inorganic particles are made of sodium and potassium phosphates, respectively, and the surfactants are respectively sodium and potassium alkylarylsulfonates. Organic phosphates are well known to improve antiwear properties of lubricants, and the objective here was to prepare inorganic phosphates solubilized in an organic medium as colloidal particles. These two new colloidal species are obtained from the reaction of tetraphosphorus decasulfide and water with sodium or potassium hydroxide in the presence of alkylarylsulfonic acid. The synthesis process for both products is presented, as well as analysis from the chemical and physical standpoint. EXPERIMENTAL

Materials The surfactant used in this synthesis was an industrial C-24 monoalkylphenylsulphonic acid containing 30% by

218

COLLOIDAL SODIUM AND POTASSIUM PHOSPHATES

FIG. 1.

219

Synthesis of colloidal alkaline (poly)phosphate, where M 5 Na (product A) and M 5 K (product B).

weight of a nonactive diluting oil, with an equivalent average molecular weight of 700. Tetraphosphorus decasulfide is a commercial crude product supplied by Atochem and it is mainly made of 56.1% by weight of P4S10 and 43.9% by weight of P4S9. Mineral diluting oil was a 130 neutral solvent. Synthesis Synthesis of colloidal sodium phosphate (product A). Tetraphosphorus decasulfide (78.8 mmol) was dispersed in a solution of crude alkylphenylsulphonic acid (20 g), mineral diluting oil (20 g), and xylene (300 cc). The mixture was heated to 70°C; then a solution of sodium hydroxide (2.36 mol) in water (99 g) was slowly added for 2 h to this dispersion. After an additional hour of reaction, distillation of water was performed by means of a Dean and Stark separator. After returning the mixture to room temperature, the solid residue was removed by filtration and the xylene was distilled off to yield 45 g of a clear brown liquid with the properties discussed below. Synthesis of colloidal potassium phosphate (product B). The same procedure was applied except potassium hydroxide was used instead of sodium hydroxide. The reaction yielded 58 g of clear brown viscous product. Analysis The infrared spectrum was recorded with a Nicolet Impact 410 spectrometer through two KBr plates. Elemental analysis of the calcium and sulfur was performed by X-ray fluorescence with a Philips PW 1480 spectrometer. Potentiometric analysis was performed using a Tacussel TT automatic titrator according to the ASTM 2896 method. Dialysis of colloidal dispersions was performed in n-heptane

through a rubber membrane according to a well-known procedure (18). Dialysis time ranged from 8 h to 16 h. Thermogravimetric analysis was performed with a Mettler TG50 apparatus at a heating rate of 5°C/min. Transmission electron microscopy (TEM) measurements were performed on a Jeol 100cx top entry type working at 100 kV. Preparation of specimens was made first by diluting dry powder of the product in xylene (90 ppm for K3PO4 and 10 ppm for Na3PO4) and then by placing a drop of the solution on a copper grid covered with carbon holly film. Electron micrographs were taken of the particles retained on the film after evaporation of the solvent. Small-angle X-ray scattering (SAXS) measurements were performed with a Cu rotating anode (Rigaku) providing an X-ray beam with a wavelength of 1.54 Å, a curved coated gold mirror combined with a Ni filter, and a one-dimensional position-sensitive proportional counter (Elphyse). The detector had a resolution of 150 mm (full width at half maximum). The X-ray generator was operated at 1 kW (40 kV p 25 mA). The range of scattering angles corresponded to wave vectors between 0.01 and 0.22 Å21 for a sampleto-detector distance of 80 cm. For the experiments, the colloidal suspension was diluted to approximately 1 wt% calcium hydroxide in xylene and was placed in a 1.85-mm diameter sealed glass capillary. After normalisation in terms of thickness, transmission, and measuring time, the solvent (xylene) signal was subtracted from the sample signal. Experimental data were converted to an absolute scale and desmeared following Lake’s method (19). Wide angle X-ray scattering measurements were performed on a u–2u Philips diffractometer. The incident wavelength was Co Ka 5 1.789 Å, and the detector moved step by step (D2u 5 0.05°) between 2 and 82°2u. The measuring time for each step was 5 s.

FIG. 2.

Composition of colloidal alkaline phosphates.

220 DELFORT ET AL.

221

COLLOIDAL SODIUM AND POTASSIUM PHOSPHATES

Chemical Structure Analytical data on the synthesized products are summarized in Fig. 2 and are discussed below. The 31P-NMR spectrum of each product exhibits different broad signals, as is usual for these colloidal species. Spectra are shown in Fig. 3. The signals peaked at 0 ppm, 4 ppm, and 10 ppm for sodium salt, and 21 ppm, 4 ppm, and 8 ppm for potassium. They are attributed to alkaline phosphate and polyphosphate according to the literature data (20, 21). The general formula of the synthesized phosphorized salts is assumed to be O O

MO©P©©O©P©©OM, n

OM

OM

where M is Na or K The general reaction of complete hydrolysis of tetraphosphorusdecasulfide in aqueous sodium or potassium hydroxide MOH, leading to a mixture of (poly)phosphates, could be illustrated by the following equation: P S 1 31nMOH 1 2n11H2O 3 M3PO4(MPO3)n 1 5/ 2~n11!H2S.

n11 4 10

FIG. 3. 31P NMR spectra of colloidal sodium and potassium phosphates.

RESULTS AND DISCUSSION

In this system, the synthesis and the micellization of sodium or potassium salts occur simultaneously in a one-step process. The reactions take place in a multiphasic system where dispersed solid tetraphosphorusdecasulfide in hydrocarbon medium reacts with aqueous sodium or potassium hydroxide in the presence of a surfactant. This is illustrated in Fig. 1. The reactions are exothermic and are controlled by the rate of introduction of aqueous base to the system. In these experiments, sodium or potassium alkylarylsulfonates, which are the effective surfactants, were prepared in situ by neutralization of the starting sulfonic acid. In this paper, products are characterized first from a chemical standpoint, including the compositions and the chemical structures of the micellized species, and then from a more physical standpoint for size, shape, and distribution of the colloidal particles.

These alkaline (poly)phosphates are inorganic species intrinsically insoluble in hydrocarbon. It is therefore suggested they are incorporated in oil as colloidal inorganic particles in a reverse-micelle type association. The ratio of colloidal and noncolloidal fractions in the crude product were determined by the dialysis technique in n-heptane through a rubber membrane following a well-known procedure (18). In this operation the diluting oil dialyses through the membrane while the colloidal species, because of their larger size, are retained by the membrane. The colloidal fraction consists of the inorganic particles plus the surrounding surfactant chains. The ratio of these two fractions depends on the total amount of diluting oil, comprising the oil from the crude surfactant plus the oil added in the synthesis in order to obtain a fluid-resulting product. From these determinations, the colloidal fractions of sodium and potassium phosphate respectively are found to be 58.5% by weight and 65.1% by weight of the crude products. Elemental analysis was performed in the

TABLE 1 Composition of Crude Colloidal Products

Inorganic alcaline phosphate core Surfactant Diluting oil

A

B

Crude colloidal Na phosphate

Crude colloidal K phosphate

36.2 wt% 22.3 wt% 41.5 wt%

46.3 wt% 18.8 wt% 34.9 wt%

222

DELFORT ET AL.

FIG. 4.

Thermogravimetric behavior of alkaline (poly)phosphates. Surfactant and mineral particles content determination.

colloidal and noncolloidal fractions. For both alkaline species, sodium or potassium, sulfur and phosphorus were exclusively located in colloidal fraction. The presence of whole phosphorus in the colloidal fraction confirms that the previously reported 31-P NMR data are specifically representative of the composition of the colloidal particles. Assuming that the known balance between surfactant and diluting oil in the initial step of each synthesis is unchanged in the final product and taking account of the dialysis data about the balance between colloidal and noncolloidal fractions, it is now possible to estimate the composition of each crude colloidal product (Table 1). The potentiometric titration of alkaline phosphates indicates 5.35 basic meq per gram in crude colloidal sodium (poly)phosphate (A), and 5.17 basic meq per gram in potassium (B). From these results and considering the compositions of the two considered crude products, it is possible to determine the average structure of each sodium or potassium (poly)phosphate inorganic colloidal core. The general formula of the sodium (poly)phosphate core is assumed to be Na3PO4(NaPO3)n with n 5 1.1 and that of potassium (poly)phosphate to be K3PO4(KPO3)n with n 5 2. An additional estimation of the composition of the colloidal fraction allowed us to differentiate the respective shares of

inorganic particles and surfactant. This was performed by thermogravimetric analysis (TGA), assuming that the organic surfactant molecules are thermally decomposed in the 300°C to 400°C range, while the inorganic species remain stable. Hence, the measurement of the percentage of residual weight after thermal degradation of organic surfactant molecules could

FIG. 5. SAXS results. Comparison between the experimental data (F) and the theoretical (———) for potassium phosphate.

COLLOIDAL SODIUM AND POTASSIUM PHOSPHATES

223

considering the accuracy of the dialysis and the thermogravimetric measurements. Results with potassium polyphosphate indicate a general composition of 25% of organic surfactant corrected to 27% with counterions, and 75% of inorganic species, corrected to 73% without counterions. Dialysis and potentiometric data give 28.9% and 71.1% respectively. The agreement is also acceptable. From these determinations, the average weight ratio of inorganic core to surfactant can be expressed at about 1.6 for the sodium derivative and at about 2.4 for the potassium. The average molar ratio of inorganic polyphosphate to surfactant chain, considering the two assumed general formulas is found to be around 2.9 for sodium species and 2.8 for potassium. FIG. 6. Volume distribution deduced from SAXS measurements for potassium phosphates (--) and sodium phosphates (———).

offer a convenient approach to the inorganic particle and surfactant contents. Figure 4 shows the typical behavior of such colloidal products. As expected, a weight loss occurs in the 350°C range, after which the residue remains stable up to 800°C. For sodium polyphosphate, the measurements indicate a general composition of 34% organic surfactant and 66% inorganic species, including counterions, and after correction, 35.6% of surfactant including counterions and 64.4% of sodium polyphosphate. Data from dialysis and potentiometric titration indicate 38.1% of surfactant and 61.9% of sodium polyphosphate. This is assumed to be in acceptable agreement,

FIG. 7.

Physical Structure Small angle X-ray scattering. The relevant scattering theory is outlined as follows (22): the micelles were assumed to be spherical particles. The scattering intensity per unit volume, I(q), from particles in a solvent can be expressed in the form I~q! 5 A 2e nv~1 2 nv!~ r 1 2 r 2! 2F~q!S~q!,

[1]

where q is the scattering vector which has a magnitude of 4p sin u/l, A e2 is the intensity scattered by one electron, i.e., 7.9 3 10226 cm2, and Dr2 5 (r1 2 r2)2 is the contrast term expressing the electronic density difference between particles and solvent. It has been demonstrated (23) that the alkylaryl

SAXS spectrum of sodium phosphate: (a) “Guinier” representation; (b) “Porod” representation.

224

DELFORT ET AL.

where R g is the z average radius of giration of the particles. In the “Guinier” representation of ln(I(q)) versus q 2 , the slope gives the radius of giration of the particles. For large q values and a well-defined interface between particles and solvant, the “Porod” representation is used, I~q! 5 A 2e D r 2

FIG. 8. SAXS results. Comparison between the experimental data (F) and the theoretical (———) for sodium phosphate.

chain has nearly the same electronic density as the xylene used as solvent. In this case, particles are considered to be only cores of the micelles, i.e. sodium or potassium (poly)phosphates; n is the number of particles per unit volume; and v is the volume of one particle. The product nv is the volume fraction of scattering matter. F(q) is the form factor and can be expressed for spheres of radius R by F~q! 5 v

S

D

3 sin~qR! 2 qR cos~qR! 2 . ~qR! 3

[2]

S(q) is the structure factor which takes into account the correlation between particles. In the diluted regime, the structure factor is 1. In this case, Eq. [1] can be extended to polydispersed systems, I~q! 5 A

2 e

O n v ~r 2 r ! F ~q!, 2

k k

1

2

[3]

k

2 p A 1,2 , q4 V

[6]

where A 1,2 is the particle surface in the scattering volume V. For monodispersed spheres of radius R, the product q 4 I(q) goes through a maximum for q ' 2.7/R (23). Potassium phosphate. The experimental and calculated curves are shown in Fig. 5. The reasonable agreement between these two curves on the whole q range confirms the choice of the model, i.e. identical and uncorrelated spherical particles. The results of the fitting procedure reveal particles with an average radius of 80 Å and a standard deviation of 0.26. The resultant volume distribution is plotted on Fig. 6. Sodium phosphate. The SAXS diagram of sodium phosphate is shown on Fig. 7. In the “Guinier” representation, two linear regions are seen, indicating two different sizes of particles, namely R g1 5 85 Å and R g2 5 28 Å. In the “Porod” representation, one maximum is seen at q 1 5 0.0316 Å21, followed by a plateau for q . 0.1 Å21. The first maximum is linked for monodispersed spheres to a radius of 85 Å. The beginning of the plateau is linked to a second distribution of smaller particles. These results could indicate a bimodal distribution of particles. The fitting procedure has to be modified to take into account the two distributions referred to in subscripts 1 and 2: I~q! 5 A 2e D r 2@

O n v F ~q! 1 O n v F ~q!#. k1 k1

k1

k1

k2 k2

k2

k2

k

The fitted parameters are R1, s1, R2, s2, and f1/f 5 n1v1/(n1v1 1 where n k and vk are the number and the volume of particles of radius R k . Log-normal distributions were used to represent the particle size distribution F~R! 5

1

S

D

1 exp 2 2 @ln~R! 2 ln~R 0!# 2 , 2 s s R Î2 p

[4]

where R is the particle radius, R 0 is the geometric mean, and s is the standard deviation. SAXS spectra were analysed using a calculated intensity based on Eqs. [2], [3], and [4] through fitting programs which provide R 0 and s, i.e. the size distribution of particles. For small q values (qR , 1), the scattering intensity is reduced to the Guinier approximation, I~q! 5 I~0!exp~2q 2R 2g/3!,

[5]

FIG. 9. Comparison of the X-ray diffractometry diagram of dialyzed sodium (poly)phosphate with sodium phosphate Na4P2O7 (sticks).

COLLOIDAL SODIUM AND POTASSIUM PHOSPHATES

225

and f1/f 5 0.74. The complete volume distribution is presented in Fig. 6. Nevertheless, several models can lead to the same intensity distribution, and models with two typical lengths like the core-shell one can account for the SAXS data.

FIG. 10.

X-ray diffractogramm of potassium (poly)phosphate.

n2v2). In this model, the small and large particles are reckoned to have the same composition and the same contrast. The fitted curve, shown in Fig. 8, is in very good agreement with the experimental one. The extracted parameters are R1 5 67 Å, s1 5 0.31 for distribution 1, R2 5 17 Å, s2 5 0.36 for distribution 2,

FIG. 11.

X-ray diffractometry. The X-ray diffractometry spectra of the dialysed products (Figs. 9 and 10) exhibits Bragg peaks. A broad reflexion is detected at 2u ' 21° (d 5 4.9 Å) and can be attributed to distances between the alkyl parts of the surfactant. In the case of sodium (poly)phosphates, the other peaks can be attributed partly to sodium phosphate Na4P2O7 (JCPDS:10-187) but another phase is needed to explain extra peaks (2u ' 26, 37.5, 48°). Because of phase heterogeneity, the classical size determination from peak broadening is not possible. For potassium (poly)phosphate, no references from the JCPDS database match the observed reflexions. X-ray diffractometry shows that the products are crystallized but no structural analysis can be done yet. Transmission electron microscopy. The transmission electron micrographs of the direct deposit diluted micelle

Transmission electron micrograph of the direct deposit diluted micelle dispersion for (a) potassium phosphate and (b) sodium phosphate.

226

DELFORT ET AL.

TABLE 2 Comparison of Particle Dimensions Measured by SAXS and TEM

Comments SAXS •K3PO4 Spheres •Na3PO4 Spheres, bimodal distribution TEM

•K3PO4 •Na3PO4

Spheres, fast evolution under electron beam Spheres, merging under electron beam

Diameter n°1 (Å)

Diameter n°2 (Å)

100–360 50–320

— ,90

60–300

—

50–500

—

dispersions are shown on Figs. 11a and 11b for potassium phosphate and sodium phosphate, respectively. Micelles are seen as circular objects, either with an undercontrasted core (for both materials) or as a full contrasted disc (for K3PO4). In both cases, the materials are very sensitive to the electron beam. The statistical visualisation of only a circular type of shape in each material can be interpreted as spherical objects. This is in close agreement with SAXS modelisation hypothesis. Potassium phosphate. Very short incident electron exposed particles are usually present as full contrasted discs (presenting facets in most of the cases). Micelles with a low contrast core have also been observed, but this specific morphology is certainly the result of a fast evolution of particles under the incident beam. Particle diameters range from 60 Å to 300 Å. This seems to correspond fairly well to the dimensions calculated with SAXS (Table 2). Sodium phosphate. Micelles are present as circular objects with low contrast core. Since the material is very sensitive to the incident beam, this is probably the result of electron damage. In addition, nonisolated micelles seem to merge with adjacent particles. Once more, this points out the problem of evolution of the preparation under the high energy incident beam. Internal diameters of observed discs range from 50 Å to 500 Å. In this case, it cannot be confirmed that the biggest particles are truly representative of the dispersion or that they are the merging consequence of electron beam damage during the TEM characterisation. Core dimensions correspond reasonably well to the characteristic length of one of the population calculated with SAXS (Table 2). CONCLUSION

The synthesis of new colloidal inorganic species in an organic medium involve the reaction of sodium or potassium hydroxide, tetraphosphorus decasulfide, and water in the presence of a surfactant. The reaction yields respectively

a mixture of sodium or potassium (poly)phosphates stabilized in a colloidal state in a reverse micelle-type association. Since sodium and potassium (poly)phosphates are the major products, it appears that the reaction involves the complete hydrolysis of tetraphosphorus decasulfide followed by neutralisation of the phosphoric acid generated. The inorganic particles appear as polydispersed crystallized spheres. NMR and X-ray diffraction measurements indicate mixtures, rather than pure phases. The diameter of these particles, as measured by SAXS and TEM, are in the range 50 –360 Å. ACKNOWLEDGMENTS The authors thank I. Cle´menc¸on and B. Rebours for the X-ray diffraction experiments.

REFERENCES 1. Marsh, J. F., Colloidal lubricant additives, Chem. Indus. 20, 470 (1987). 2. Bray, U. B., Dickey, C. R., and Voorhees, V., Dispersions of insoluble carbonate in oils, Ind. Eng. Chem., Prod. Res. Dev. 14(4), 295 (1975). 3. Brooks, B. W., and Shhlimkan, R. V., Manufacture of calcium carbonate dispersions in lubricating oil, Colloid Polym. Sci. 257, 981 (1979). 4. Whittle, J., Method of preparing overbased calcium sulfonates, U.S. Patent 4,427,559 (1984). 5. Hunt, M. W., Overbased alkali metal sulfonates, U.S. Patent 4,867,891 (1989). 6. Inoue, K., Calcium borate overbased metallic detergent, J. Soc. Trib. Lub. Eng. 49, 263 (1993). 7. Inoue, K., and Nose, Y., Method for producing alkaline earth metal borate dispersions, U.S. Patent 4,683,126 (1987). 8. Marsh, J. F., Colloidal lubricant additives, Chem. Indus. 20, 470 (1987). 9. Belle, C., Beraud, C., Faure, D., Gallo, R., Hornaert, P., Martin, J. M., and Rey, C., Polyphasic reaction mechanism: Kinetic of overbased calcium sulfonate synthesis in non-polar medium, J. Chim. Phys. 87, 93 (1990). 10. O’Sullivan, T. P., Vickers, M. E., and Heenan, R. K., The characterization of oil-soluble calcium carbonate dispersions using small-angle X-ray scattering (SAXS) and small-angle neutron scattering (SANS), J. Appl. Cryst. 24, 732 (1991). 11. Markovic, I., Ottewill, R. H., Cebula, D. J., Field, I., and Marsh, J. F., Small-angle neutron scattering studies of non-aqueous dispersions of calcium carbonate, Colloid Polym. Sci. 262(8), 648 (1984). 12. Markovic, I., and Ottewill, R. H., Small-angle neutron scattering studies of non-aqueous dispersions of calcium carbonate. Part 2. Determination of the form factor for concentric sphere, Colloid Polym. Sci. 264(1), 65 (1986). 13. Markovic, I., and Ottewill, R. H., Small-angle neutron scattering studies of non-aqueous dispersions of calcium carbonate. Part 3. Concentrated dispersions, Colloid Polym. Sci. 264(5), 454 (1986). 14. Ottewill, R. H., Sinagra, E., McDonald, I. P., Marsh, J. F., and Heenan, R. K., Small-angle neutron scattering studies of non-aqueous dispersions of calcium carbonate. Part 5. Magnesium carbonate dispersions in hydrocarbon media, Colloid Polym. Sci. 270(6), 602 (1992). 15. Giasson, S., Espinat, D., Palermo, T., Ober, R., Pessah, M., and Morizur, M. F., Small-angle x-ray scattering (SAXS) on calcium sulfonate disper-

COLLOIDAL SODIUM AND POTASSIUM PHOSPHATES sions: Effect of friction on microstructure, J. Colloid Interface Sci. 153(2), 335 (1992). 16. Delfort, B., Chive´, A., and Barre´, L., Colloidal calcium thiophosphates in organic medium. Synthesis and analysis, J. Colloid Interface Sci. 186, 300 (1997). 17. Delfort, B., Born, M., Chive´, A., and Barre´, L., Colloidal calcium hydroxide in organic medium. Synthesis and analysis, J. Colloid Interface Sci. 189, 151 (1997). 18. Altgrelt, K. H., and Gouv, T. H., in “Chromatography in Petroleum Analysis,” Vol. 11, p. 417. Dekker, New York, 1979.

227

19. Lake, J. A., An iterative method of slit-correcting small angle X-ray data, Acta Cryst. 23, 191 (1967). 20. Grayson, M., and Griffith, E. J., “Topics in Phosphorus Chemistry,” p. 5. Wiley Interscience, New York, 1964. 21. Loewus, D. I., and Ecksyein, F., J. Am. Chem. Soc. 105, 3287 (1983). 22. Guinier, A., and Fournet, G., “Small Angle Scattering of X-Rays,” Wiley, New York, 1955. 23. Roman, J. P., Hoonaert, P., Faure, D., Bivier, C., Jaquet, F., and Martin, J. M., J. Colloid Interface Sci. 144, 324 (1991).