CONTINUOUS FLOW PROCESS FOR MANUFACTURING SURFACE MODIFIED METAL OXIDE NANOPARTICLES BY SUPERCRITICAL SOLVOTHERMAL SYNTHESIS

Abstract

The invention concerns a continuous flow process for manufacturing surface modified metal oxide nanoparticles by supercritical solvothermal synthesis in an reaction medium flowing within a continuous flow chamber, said continuous flow chamber containing a hydrolysis area and a supercritical area, said process comprising the introduction of a flow of metal oxide precursor into the continuous flow chamber at a point P located in the hydrolysis area or in the supercritical area, and the introduction of a flow of is located downstream of P1 with respect to the flow direction, as well as the device for carrying out this process.

Claims

1. A continuous flow process for manufacturing surface modified metal oxide nanoparticles by supercritical solvothermal synthesis in a reaction medium flowing within a continuous flow chamber, said continuous flow chamber containing two areas: a hydrolysis area where the reaction medium is not in supercritical state and conditions are such that nucleation and growth of metal oxide nanoparticles can be initiated; and a supercritical area where the reaction medium is in supercritical state and the supercritical solvothermal synthesis of metal oxide nanoparticles can be performed, said process comprising the introduction of a flow of metal oxide precursor into the continuous flow chamber at a point P1 located in the hydrolysis area or in the supercritical area, and the introduction of a flow of surface modifier into the continuous flow chamber at a point P2 located in the hydrolysis area or in the supercritical area, wherein P2 is located downstream of P1 with respect to the flow direction.

2. The continuous flow process according to claim 1, wherein the reaction medium is an aqueous reaction medium and the solvothermal synthesis is a hydrothermal synthesis.

3. The continuous flow process according to claim 1, wherein said process further comprises the quench of the flow of surface modified metal oxide nanoparticles formed in the supercritical area at a temperature below the temperature of the supercritical area, preferably below the temperature of hydrolysis area, then the recovery of the surface modified metal oxide nanoparticles either in the form of liquid suspension or in dried form.

4. The continuous flow process according to claim 1, wherein several flows of surface modifier, identical or different, are independently introduced at the same injection point or at different injection points downstream of P1 with respect to the flow direction.

5. The continuous flow process according to claim 1, wherein the surface modifier is an organic ligand, thereby forming hybrid organic-inorganic nanoparticles.

6. The continuous flow process according to claim 1, wherein the metal oxide precursor is a metal salt, in particular an inorganic acid salt such as a nitrate, a chloride, a sulfate, an oxyhydrochloride, a phosphate, a borate, a sulfite, a fluoride or an oxyacid salt of Cu, Ba, Ca, Zn, Al, Y, Si, Sn, Zr, Ti, Sb, V, Cr, Mn, Fe, Co or Ni, or an organic acid salt such as an alkoxide, a formate, an acetate, a citrate, an oxalate or a lactate of Cu, Ba, Ca, Zn, Al, Y, Si, Sn, Zr, Ti, Sb, V, Cr, Mn, Fe, Co or Ni, more particularly a metal oxide precursor for manufacturing metal oxide nanoparticles chosen from TiO2, ZrO2, ZnO, BaTiO3, NiMoO3, NiWO3, Al2O3, Ga2O3, In2O3, SiO2, GeO2, V2O5, CeO2, CoO, α-Fe2O3, γ-Fe2O3, NiO, Co3O4, Mn3O4, γ-MnO2, Cu2O, CoFe2O4, ZnFe2O4, ZnAl2O4, Fe2CoO4, BaZrO3, BaFe12O19, LiMnO204, LiCoO2, La2O3.

7. The continuous flow process according to claim 6, wherein the metal oxide precursor is chosen from titanium (IV) isopropoxide, titanium (IV) propoxide, zirconium acetate, zirconium isopropoxide, zirconium propoxide or zirconium acetylacetonate.

8. The continuous flow process according to claim 1, wherein the concentration of the metal oxide precursor in the reaction medium is from 0.0001 mol/l to 1 mol/l, in particular from 0.001 mol/l to 0.1 mol/l, more particularly from 0.01 mol/l to 0.1 mol/l.

9. The continuous flow process according to claim 1, wherein the reaction medium is a mixture of water and ethanol or a mixture of water and isopropanol with a molar ratio water/alcohol from 1:5 to 5:2, in particular in from 1:4 to 2:1, in particular from 2:3 to 1:1, in particular around 4:5.

10. The continuous flow process according to claim 1, wherein the temperature of the reaction medium in the hydrolysis area is at least 100° C., in particular from 130° C. to 250° C., more particularly from 150° C. to 200° C.

11. The continuous flow process according to claim 1, wherein the temperature of the reaction medium in the supercritical area at least 240° C., in particular from 280° C. to 400° C., more particularly from 300° C. to 380° C.

12. The continuous flow process according to claim 1, wherein the pressure of the reaction medium in the continuous flow chamber is from 10 MPa to 30 MPa, in particular from 15 MPa to 25 MPa, more particularly around 22 MPa.

13. The continuous flow process according to claim 1, wherein the surface modifier is an organic ligand comprising an acid group, such as a carboxylic acid group, a phosphonic acid group or a sulfonic acid group, a silane group, an amine group, a thiol group, in particular a carboxylic acid group or a phosphonic acid group.

14. The continuous flow process according to claim 1, wherein the molar ratio of surface modifier/metal oxide precursor in the reaction medium is from 0.05 to 10, in particular from 0.1 to 1, more particularly from 0.15 to 0.2.

15. The continuous flow process according to claim 1, wherein both the injection points P1 and P2 are located in the hydrolysis area.

16. The continuous flow process according to claim 1, wherein the injection point P1 is located in the hydrolysis area and the injection point P2 is located in the supercritical area.

17. A device for carrying out the process according to claim 1, comprising a continuous flow chamber (1) heated with a heater (2a, 2b) which heats the continuous flow chamber (1) with an increasing gradient of temperature along the flow direction, said continuous flow chamber (1) having: an inlet (3) for introducing the flow of metal oxide precursor into the continuous flow chamber (1) at an injection point P1, one or several inlets (4a, 4b) for introducing the flow of surface modifier into said continuous flow heated chamber (1) at an injection point P2 which is different than and downstream of P1.

18. The device according to claim 17, wherein said continuous flow chamber (10) is a tube reactor.

19. The device according to claim 17, further comprising a filter (7) for recovering the surface modified metal oxide nanoparticles in dried from.

Description

[0110] The invention will now be further described in the following examples. These examples are offered to illustrate the invention and should in no way be viewed as limiting the invention.

[0111] FIG. 1 shows a schematic diagram of the continuous flow reactor system according to the invention as used in Example 1.

[0112] FIG. 2 represents the XRD pattern of TiO2 nanoparticles obtained by supercritical hydrothermal synthesis in a mixture of water and ethanol under supercritical conditions.

[0113] FIG. 3 represents the HR-TEM micrograph of TiO2 nanoparticles obtained by supercritical hydrothermal synthesis in a mixture of water and ethanol under supercritical conditions.

[0114] FIG. 4 represents the particle size distribution of the TiO2 nanoparticles obtained by supercritical hydrothermal synthesis in a mixture of water and ethanol under supercritical conditions.

[0115] FIG. 5 represents the particle size of surface modified TiO2 prepared with the process of the invention as a function of the injection point of the surface modifier.

[0116] FIG. 6 shows a schematic diagram of the continuous flow reactor system according to the invention.

EXAMPLE 1: FUNCTIONALIZATION OF TIO.SUB.2 .NANOPARTICLES

[0117] FIG. 1 shows a schematic diagram of the continuous flow reactor system.

ROH=ethanol
HPP=High pressure pump
P=Pressure gauge

V=Valve

[0118] Vr=Regulation Valve, also called back-pressure regulator

F=Filter

C=Condenser

[0119] The system comprises four modules R1 to R4 connected in series. R1 and R2 are hydrolysis modules for performing the hydrothermal synthesis under subcritical conditions. R3 and R4 are supercritical modules for performing the hydrothermal synthesis under supercritical conditions.

[0120] The injection points of the surface modifier are positioned before the reactor R1, between the different modules (R1-R2, R2-R3, R3-R4) and after the reactor R4.

[0121] The supercritical hydrothermal synthesis of TiO.sub.2 nanoparticles is performed with a mixture of water and ethanol (molar ratio water/ethanol=0.8) under the following conditions: [0122] Titanium precursor: Ti(O-iC.sub.3H.sub.7).sub.4 in an aqueous solution with a Water/Ethanol molar ratio of 8, with a concentration in the stock solution=4.10.sup.−2 mol.Math.L.sup.−1, [0123] pressure P within R1-R4=22 MPa, [0124] total flow Q within R1-R4=11.6 g.Math.min.sup.−1 [0125] Type of flow: turbulent (Re=3287), [0126] R1 and R2: [0127] Tube reactor of stainless steel with a total length of 12 m, composed of two modules each with a length of 6 m, [0128] Temperature of 150° C., [0129] R3 and R4: [0130] Tube reactor of stainless steel with a total length of 12 m, composed of two modules each with a length of 6 m, [0131] Temperature of 380° C.

[0132] After the synthesis, TiO.sub.2 nanoparticles (bare or functionalized) are recovered as solutions in water and ethanol. They are centrifuged and washed with ethanol 5 times to remove the unreacted surface modifier.

[0133] FIG. 2 represents the X-ray diffraction (XRD) pattern of the obtained TiO.sub.2 nanoparticles without no addition of surface modifiers to the reaction system. It can be attributed to the ICDD-PDF card 00-021-1272 which corresponds to the body-centered tetragonal lattice (Anatase phase, space group I41/amd, a=3.785 Å, c=9.514 Å). From the Debye-Scherrer equation applied to peaks (101) at 25.326° and peak (200) at 48.1°, it is estimated that the mean size of crystallites is 7.3 nm.

[0134] FIG. 3 represents the HR-TEM micrograph (High Resolution Transmission Electron Microscopy) of the TiO2 nanoparticles. It shows their monocrystalline state. Thus it can be considered that the mean size of the crystallites is equivalent to the mean size of the particles. The estimation of particle size and particle size distribution was done from the counting of about 200 nanoparticles observed on TEM micrographs. Aggregated particles range from 6 nm to 15 nm. The maximum population has a size around 10 nm, and the mean size correlates with the one estimated with XRD.

[0135] FIG. 4 represents the particle size distribution of the TiO.sub.2 nanoparticles.

[0136] The same experiment as above is performed but with addition of a surface modifier at various injections points of the system during the hydrothermal synthesis: [0137] Between R1 and R2, [0138] Between R2 and R3, [0139] Between R3 and R4, or [0140] After R4.

[0141] The injected surface modifier is either hexanoic acid (ha) or octylphosphonic acid (oPa). The molar ratio of Ti atoms injected per second to grafting heads of hexanoic acid molecules injected per seconds (Ti/ha ratio) is 6 (ha 6) or 12 (ha 12). The amount of injected surface modifier is adjusted in order to have a molar ratio of Ti atoms injected per second to grafting heads of phosphonic acid molecules injected per seconds (Ti/oPa) of 6 (oPa6) or 12 (oPa12). The surface modifier is in solution in a water-ethanol mixture, of the same composition and same water/alcohol molar ratio than the solvent of the titanium precursor.

[0142] The interaction of the functionalizing agents with the nanoparticles of TiO.sub.2 is evidenced by evaluating its influence on the calculated crystallite size (regarded as the particle size).

[0143] Table 1 gives the average size of crystallites (calculated by Debye-Scherrer equation) depending on the injection point and on the molar ratio of surface modifier injected per Ti atoms in the precursor.

TABLE-US-00001 TABLE 1 Size of the Samples Parameters observed crystallites (nm ± 10%) TI002 Functionalization ex situ with DibuP 8 TI003 Functionalization ex situ with Bis2P 7.1 TI004 Functionalization ex situ with oPa 7.7 TI005 Functionalization ex situ with 3oP 7.5 TI009 Injection of ha6 after R.sub.4 8.1 TI010 Injection of ha12 after R.sub.4 7.7 TI011 Injection of oPa6 after R.sub.4 7.9 TI012 Injection of ha6 between R.sub.3 and R.sub.4 7.7 TI013 Injection of ha12 between R.sub.3 and R.sub.4 7.8 TI014 Injection of oPa6 between R.sub.3 and R.sub.4 7.5 TI015 Injection of ha6 between R.sub.2 and R.sub.3 7.6 TI016 Injection of ha12 between R.sub.2 and R.sub.3 8.2 TI017 Injection of oPa6 between R.sub.2 and R.sub.3 7 TI018 Injection of ha6 between R.sub.1 and R.sub.2 6.7 TI019 Injection of ha12 between R.sub.1 and R.sub.2 6.7 TI020 Injection of oPa6 between R.sub.1 and R.sub.2 5.4

[0144] The lines TI002 to TI005 correspond to experiments where TiO2 nanoparticles are first synthesized as bare nanoparticles and the functionalization are performed in a second time, after the recovery of the nanoparticles in solution, as expressed by the use of the word “ex situ”.

[0145] FIG. 5 represents the crystallite size according to the injection point of the surface modifier. It clearly shows that the crystallite size decreases as the surface modifier is injected earlier in the synthesis process, especially with octylphosphonic acid. The earlier the injection of surface modifier, the smaller the crystallites. This is evidence of the interaction between the TiO.sub.2 nanoparticles and the surface modifier, and that the grafted surface modifier impedes the growth of the nanoparticles. This also indicates that octylphosphonic acid seems to have a greater interaction with TiO.sub.2 crystallites than hexanoic acid since its influence on crystallite size is stronger.

[0146] Furthermore, at least for hexanoic acid, a ratio of 1 grafting head of hexanoic acid molecule per 12 Ti atoms (Ti/ha ratio of 12) seems to not be enough to have an effective grafting on the crystallite without an hydrolysis step, as the size of the nanoparticles are unchanged if the hexanoic acid ha12 is injected after the hydrolysis step.

[0147] Furthermore, those results show that positioning the injection point so that the injection is done during the hydrolysis step of the process ensures a greater effect on the crystallite size.

[0148] FTIR (Fourier transform infrared spectroscopy) analyses performed for TiO.sub.2 functionalized with octylphosphonic acid by injection of the surface modifier between R.sub.2 and R.sub.3 shows three bands corresponding to an alkyl chain [2960 cm.sup.−1: v.sub.as(-CH.sub.2—CH.sub.3), 2925 cm.sup.−1: v.sub.as(-CH.sub.2—), 2850 cm.sup.−1: v.sub.s(-CH.sub.2—)], which is evidence of the presence of a functionalizing agent at the surface of TiO.sub.2 nanoparticles. Moreover, the band at 1100-1000 cm.sup.−1: v.sub.s(-P—O.sub.3) is well visible, assessing the grafting of the modifier at the surface of the nanoparticles via the P—O functions. It can be concluded that from this injection point, TiO.sub.2 nanoparticles are functionalized with octylphosphonic acid.

[0149] The same FITR analysis performed for TiO.sub.2 functionalized with octylphosphonic acid by injection of the surface modifier between R.sub.1 and R.sub.2 shows the presence of an alkylene band at 1460 cm.sup.−1: δ.sub.sc(—CH.sub.2—) and a stronger evidence of grafting with octylphosphonic acid at 1100-1000 cm.sup.−1: v.sub.s(-P—O.sub.3) than when the injection point is between R.sub.2 and R.sub.3.

[0150] TGA-MS (ThermoGravimetric Analyzer using a Mass Spectrometer) analysis carried out on the TiO.sub.2 functionalized with octylphosphonic acid by injection of the surface modifier after R.sub.4 show a mass loss higher than the bare nanoparticles: 7.5% against 2.9%. Moreover, the gas outputted by the loss is analyzed by the TGA-MS and is found attributable to organic fragments that correspond with octyl part of the octylphosphonic acid. Therefore, even though FTIR cannot pin-point the amount of functionalization, TGA-MS confirms that the TiO.sub.2 particles obtained by injecting a oPa modifier are functionalized by octylphosphonic acid or one of its derivative, even when the injection point is situated after the supercritical tunnel (immediately after R4).

[0151] The same analysis for TiO.sub.2 functionalized with octylphosphonic acid by injection of the surface modifier between R.sub.3 and R.sub.4 shows a 10% mass loss from which 7.1% can be attributed to organic parts which have a signal corresponding to the octyl part of the octylphosphonic acid.

[0152] The same analysis for TiO.sub.2 functionalized with octylphosphonic acid by injection of the surface modifier between between R.sub.1 and R.sub.2 shows a mass loss of 20%, with only 2.9 corresponding to the bare particle, thus 17.1% can be attributed to organic parts which have a signal corresponding to the octyl chain of the phosphonic acid used.

[0153] It can be concluded that the continuous multi-injection process of the invention allows the in situ grafting of phosphonic acid molecules on TiO.sub.2 crystallites in one step. Small and very well crystallized nanoparticles of TiO.sub.2 are thus easy to obtain, especially by using a supercritical water/ethanol system. The position of the injection point of the surface modifier with respect to the flow direction has an influence on the amount of grafted surface modifier and on the size of the resulting nanoparticles. An early injection permits a higher functionalization and reduction of the crystallite size (and most probably the particle size too). However, the inventors have found that it is important to let the nucleation of the nanoparticles occurs before injecting the functionalizing surface modifier, otherwise the formation of TiO.sub.2 crystallites is polluted by wastes. Indeed, in those cases, the resulting product has a very complex and poorly resolved XRD pattern. This means that part of the material seems to be amorphous. Further, the species produced are not pure TiO.sub.2 particles functionalized by, for example, oPa chains, but probably particles of Ti—O.sub.x—P.sub.y materials. This is due to the high reactivity of P with metals, higher than O with metals and adding P modifier too early prevents TiO.sub.2 from being formed.

EXAMPLE 2: FUNCTIONALIZATION OF ZRO.SUB.2 .NANOPARTICLES

[0154] The same system as the one used in Example 1 was used to prepare ZrO.sub.2 crystallites with the same operating conditions.

Reactants:

[0155] Zr precursor: zirconium acetylacetonate, zirconium acetate, zirconium propoxide or zirconium isopropoxide.

[0156] Surface modifiers: hexanoic acid, octylphosphonic acid, phenylphosphonic acid, phosphorous acid or SIK7709-10 (12-Dodecylphosphonic acid)triethylammonium bromide).

[0157] Solvent: water and ethanol or isopropanol.

[0158] In each case, the amount of injected surface modifier was adjusted to have a molar ratio acid molecule/zirconia of 0.16, which corresponds to the Ti/ha or Ti/P of 6 in the TiO.sub.2 example.

[0159] After the synthesis, ZrO.sub.2 nanoparticles (bare or functionalized) are recovered as solutions in water and ethanol or isopropanol. They are centrifuged and washed with ethanol 5 times to remove the unreacted surface modifier.

[0160] A peak corresponding to P—O-metal bound can be found on ZrO.sub.2 crystallites under FTIR observation of the residue after the TGA analysis. Moreover, the associated Mass Spectroscopy of the gas emitted during the calcination at 1000° C. of the TGA analysis could not detect released fragments containing phosphorus.

[0161] These combined results mean that the phosphonic grafting heads, i.e. at least the phosphorous atoms, are still chemisorbed on the surface of ZrO.sub.2 after TGA analysis at 1000° C. and they do not take part in the mass loss of the sample during TGA analysis.

[0162] It is to be noted that the same peak was observed for FTIR analysis of the residues of TiO.sub.2 nanoparticles functionalized with oPa after the TGA analysis.

[0163] The results are provided in tables 2 and 3.

M=Monoclinic

T=Tetragonal

[0164] W/E=water/ethanol
W/iP=water/isopropanol
X means no dispersion in the medium of synthesis
Δ means acceptable but not so good dispersion
PA=Phosphorous acid
PPA=Phenylphosphonic acid

TABLE-US-00002 TABLE 2 Medium Water/Ethanol (molar ratio = 0.8) Precursor Zirconium acetylacetonate Zirconium (molar ratio propoxide P/Zr = 0.16) Surface Hexanoic Octyl Phosphorous Phenylphosphonic acid modifier acid phosphonic acid acid Injection Before R.sub.1 Between Between R.sub.2 and R.sub.3 point R.sub.1 and R.sub.2 Crystal M T M/T T T M M/T structure Size no data no data no data no data 4 nm ± 2 7.5 nm ± 2 9 nm ± 4 distribution Dispersion X X X X X X X

TABLE-US-00003 TABLE 3 Medium Water/Ethanol or Water/Isopropanol Water (molar ratio = 0.8) Precursor Zirconium propoxide or Zirconium Zirconium Zirconium acetate (molar ratio Zirconium isopropoxide acetate isopropoxide P/Zr = 0.16) Surface SIK7709-10 PA then PPA PPA then PA modifier Injection Between Between R.sub.2 and R.sub.3 Between First modifier between R.sub.1 and point R.sub.1 and R.sub.3 and R.sub.4 R.sub.2, then second modifier R.sub.2 between R.sub.2 and R.sub.3 Crystal M/T M/T M/T M/T M/T M/T M structure Medium W/iP W/E W/iP W/E W/iP Size 8.5 nm ± 4 8.5 nm ± 3 9 nm ± 3 6.5 nm ± 4 8.5 nm ± 4 no data no data distribution Dispersion Δ Δ Δ Δ Δ Δ X

[0165] The results show that for ZrO.sub.2 nanoparticles the structure of the functionalized nanoparticles depends on the nature of the surface modifier. Indeed, the XRD pattern is different whether hexanoic acid, octylphosphonic acid, phenylphosphonic acid or phosphorous acid is used.

[0166] With hexanoic acid, the monoclinic structure of the bare nanoparticles is maintained and with octylphosphonic acid the tetragonal structure of ZrO.sub.2 is obtained. With these two surface modifiers, a well crystallized material is obtained, whereas with phosphorous acid and phenylphosphonic acid, the final material is poorly crystalline and it is difficult to distinguish clearly some phases, even though the mixture of crystallites of the monoclinic phase and crystallites of the tetragonal phase for the phosphorous acid and the presence of crystallites of the tetragonal phase for the phenylphosphonic acid can be guessed.

[0167] Surface modifier SIK7709-10 contains two active sites: a phosphonic acid moiety and an ammonium bromide moiety.

[0168] Experiments were done with a mixture of phenylphosphonic acid and of (1-butyl)triethylammonium bromide to simulate both active sites and see whether there will be a competition between the two moieties and which one will take the advantage.

[0169] The surface modifiers are solubilized together in a water/ethanol solution of molar ratio of 0.8 with a P/Zr and a N/Zr molar ratio of 0.16. Zirconium acetylacetonate was at a concentration of 4.10.sup.2 mol.Math.L.sup.−1. Two injection points were tested: between R1 and R2 and between R2 and R3 with an injection flow of 10 mL.Math.min.sup.−1. The overall pressure is kept at 23 MPa. R1 and R2 were heated at 200° C. and R3 was heated at 380° C.

[0170] FTIR analysis of the obtained nanoparticles shows evidence of the presence of nitrogen containing compounds. Thus, it means that the phosphonic acid is preferentially grafted on the surface of the nanoparticles of ZrO.sub.2 over the ammonium bromide.

[0171] A similar test was done in order to compare the relative reacting strength of phosphonic acid and carboxylic acid, i.e. which molecule will preferentially graft over the nanoparticles surface between the two surface modifiers considered.

[0172] It was evidenced that phosphonic acids is preferentially grafted over carboxylic acids or bromide. Therefore, the functionalization of a crystallite with bromide ending or with carboxylic acid functions can be performed respectively with a surface modifier comprising both a phosphonic acid function and a bromide and with a surface modifier comprising the carboxylic function.

[0173] Surface modifier SIK7709-10 can be used to graft a crystallite with ending bromide functions without grafting dangling phosphonic groups, which in turn would have the un-wanted effect of bridging particles one to each other, thus leading to strong aggregation of the nanoparticles.

[0174] The multi-injection setup was also used to inject separately in the chamber at a distance from each other two modifiers, namely phenylphosphonic acid and phosphorous acid. The injection points were respectively situated between R1 and R2 for the first modifier and between R2 and R3 for the second one. The surface modifiers were both solubilized in water with a P/Zr molar ratio of 0.08 each (as opposed to a ratio P/Zr of 0.16 for single-modifier experiments).

[0175] The precursor used was zirconium acetate, dissolved in water at a concentration of 4.10.sup.2 mol.Math.L.sup.−1.

[0176] The separate injection of two different modifiers effectively leads to nanoparticles doubly grafted. The use of phenylphosphonic acid as a first surface modifier allows obtaining a doubly functionalized crystallite which has a mono-crystalline structure essentially composed of monoclinic crystals, while the use of phosphorous acid as first surface modifier created two types of crystallites: tetragonal and monoclinic crystallites.

[0177] Therefore, the nanoparticle size, structure and the amount of grafting can be controlled by adjusting the relative amounts and the order of injection of the surface modifiers into the reaction system.

[0178] Since some surface modifiers may graft preferentially over other surface modifiers, the arrangement of the surface modifiers grafted on the crystallites will depend on the order of injection of the surface modifiers.

[0179] The above results show that: [0180] if the injection of the surface modifier is done earlier, especially before having passed ⅔.sup.rd of the reaction time, the amount of surface modifier grafted over the crystallite is higher but the particle size is smaller. [0181] Phosphonic acids have a greater effect on particle size than carboxylic acids and bromide reactive groups. [0182] The nature of the precursor can have an influence on the crystalline structure for certain materials.