Applied Catalysis A: General 250 (2003) 151–159

Potassium fluoride doped fluorapatite and hydroxyapatiteas new catalysts in organic synthesis

Abdellatif Smahi, Abderrahim Solhy, Hanane El Badaoui, Abderrahim Amoukal,Abdellatif Tikad, Mostafa Maizi, Saıd Sebti∗

Laboratoire de Chimie Organique Appliquée et Catalyse, Université Hassan II, Faculté des SciencesBen M’Sik B.P. 7955, 20702 Casablanca, Morocco

Received 17 January 2003; received in revised form 25 March 2003; accepted 28 March 2003

Abstract

Fluorapatite (FAP) and hydroxyapatite (HAP) were prepared and doped with potassium fluoride. KF/FAP and KF/HAPwere prepared by usual impregnating method. The doped materials were characterised by X-ray diffraction, BET surfacearea, BJH total pore volume and scanning electron microscopy (SEM). All this data were compared to that of unimpregnatedapatites. Some modifications in the particles morphology were observed by doping apatites with KF. The comparison of thecatalytic activity of KF, FAP, HAP, KF/FAP and KF/HAP indicate clearly the positive effect of the doping apatites by KF inthe Knoevenagel condensation used as model reaction. In this work, KF/FAP and KF/HAP were prepared and used for thefirst time as new and efficient catalysts in heterogeneous solid–liquid synthesis.© 2003 Elsevier Science B.V. All rights reserved.

Keywords: Heterogeneous catalysis; Fluorapatite; Hydroxyapatite; Potassium fluoride

1. Introduction

High surface area supports such as silica, aluminaand active carbon are commonly used as supports formetal salts. Recently, there have been a considerablenumber of papers dealing with the use of supportedmaterials in catalysis, particularly the KF/Al2O3. Theactivity of supported metal salts is strongly affectedby the interaction with the specific support and by thetextural and chemical properties of the support. Re-cently, potassium fluoride has been used to promotethe synthesis of 1,2-oxazoline derivative[1], fluorina-

∗ Corresponding author. Tel.:+212-61-46-48-19;fax: +212-22-24-96-72.E-mail address: saidsebti@yahoo.fr (S. Sebti).

tion [2], Suzuki coupling[3] and synthesis of thio-phene derivatives[4].

Supported potassium fluoride on alumina (KF/Al2O3) has been used in several organic reactions.The application and the advantage of the stronglybasic nature of KF/Al2O3 have been recently re-viewed [5]. Others papers have been reported in theKF/Al2O3 mediated organic synthesis. Thus, thissolid has been used in heterogeneous catalysis ofsome reactions as theN-dodecylimide synthesis[6],desulfonylation[7], cross-coupling reaction[8], de-protection of aryl silyl ether[9], cinnamaldehydesynthesis[10], ester hydrolysis[11], Suzuki coupling[12] and palladium-catalysed reactions[13].

On the other hand, we have reported the use ofthe doped potassium fluoride on natural phosphate(KF/NP) as a strongly base catalyst for flavanones

0926-860X/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved.doi:10.1016/S0926-860X(03)00254-0

152 A. Smahi et al. / Applied Catalysis A: General 250 (2003) 151–159

Scheme 1.

synthesis[14], Michael addition of mercaptans[15]and nitroalkanes[16], Knoevenagel condensation[17], nitrile hydration[18] and hydroxyphosphonatesynthesis[19].

In this paper, we report the first use of potassiumfluoride doped fluorapatite (FAP) and hydroxyapatite(HAP) in the heterogeneous liquid phase catalysis. TheKnoevenagel condensation (Scheme 1) has been usedas model reaction to study the activity of doped mate-rials in comparison with KF, FAP and HAP. We haverecently reported the use of FAP[20] and HAP[21]as new catalysts or supports in heterogeneous organicsynthesis free of solvent or in liquid phase.

2. Experimental

2.1. Preparation and characterisationof FAP and HAP

The preparation and characterisation of the fluora-patite [20] and the hydroxyapatite[21] were alreadydescribed. These syntheses were carried out by theco-precipitation method using diammonium phos-phate, calcium nitrate and ammonium fluoride inpresence of ammonia for FAP (Scheme 2) and di-ammonium phosphate, calcium nitrate and ammoniafor HAP (Scheme 3). Both FAP and HAP were cal-cined at 900◦C for 1 h. The structure of FAP and HAP

Scheme 2.

were confirmed by X-ray diffraction and infraredspectra. The lattice parameters of these apatites arein excellent agreement with standard data:a = 9.364and 9.422 Å andc = 6.893 and 6.883 Å for FAP andHAP, respectively. Surface area for calcined FAP andHAP were determined by the BET method from theadsorption–desorption isotherm of nitrogen at its liq-uid temperature (77 K) and were found to beS = 15and 38 m2 g−1, respectively. The total pore volumewas calculated by the BJH method atP/P0 = 0.98,VT = 0.0576 and 0.157 cm3 g−1 for FAP and HAP,respectively.

2.2. Preparation and characterisation of KF/FAPand KF/HAP

The preparation of KF/FAP and KF/HAP was car-ried out by usual impregnation 8 g of the apatite wasadded to a solution of 100 ml of water contained 1 gof KF. The solution was mixed vigorously and evapo-rated to dryness and dried for 2 h at 150◦C before use.

The analysis of the new catalysts KF/FAP andKF/HAP with X-ray diffraction gives the diffractionpatterns almost identical to that of FAP and HAP,respectively as shown inFigs. 1 and 2. These resultsindicate that the structures of the apatites remainunaltered and no crystalline KF phase is observed.Its suggest that the KF is dispersed throughout thesurface of the support.

Scheme 3.

A. Smahi et al. / Applied Catalysis A: General 250 (2003) 151–159 153

Fig. 1. Comparison of the X-ray diffraction patterns of (a) FAP and (b) KF/FAP.

The surface areas of the non-doped and dopedmaterials are not significantly different. Thus, forKF/FAP and KF/HAP the surface area were 15 and31 m2 g−1, respectively, when the values obtained forFAP and HAP were 15 and 38 m2 g−1, respectively.The basic properties have been estimated by the ad-sorption of phenol on the catalyst after 2 h as 0.44;0.49; 0.69 and 0.80 mmol g−1 for FAP, HAP, KF/FAP

Fig. 2. Comparison of the X-ray diffraction patterns of (a) HAP and (b) KF/HAP.

and KF/HAP, respectively. These results are surpris-ing and seem to be in contrast with the important dif-ference of activity between the non-doped and dopedapatites observed in the Knoevenagel reaction studiedhere.

Interestingly, some changes of surface structureof doped catalysts were observed in scanning elec-tron micrograph (SEM). Thus, the comparison of the

154 A. Smahi et al. / Applied Catalysis A: General 250 (2003) 151–159

Fig. 3. Scanning electron micrograph (SEM) images of (a) FAP and (b) KF/FAP.

Fig. 4. Scanning electron micrograph (SEM) images of (a) HAP and (b) KF/HAP.

images of KF/FAP with FAP (Fig. 3) and KF/HAPwith HAP (Fig. 4) shows some modification in the par-ticles morphology of the doped materials. This resultsuggests that KF interacted with the surface of FAPand HAP, respectively. So, we can reasonably suspectthat the modification of the surface morphology ofthe apatites was to be responsible for the enhance-ment of the catalytic activity observed with the dopedmaterials.

2.3. Procedure of the catalytic tests

The general procedure is as follows: To a flask con-taining 1 ml of solvent was added 1.5 mmol of alde-hyde1, 1.5 mmol of activated methylene2 and 0.1 gof catalyst. The mixture was stirred at room temper-ature for the specified time. The catalyst was sepa-rated by filtration and washed with the solvent usedor dichloromethane. The solution was evaporated andthe product purified by distillation under vacuum orrecrystallisation.

3. Results and discussions

3.1. Heterogeneous catalysis with FAP and HAP

First of all, we have studied the use of FAP andHAP as catalysts of Knoevenagel reaction in solvent.The condensation of the benzaldehyde with methyl-cyanoacetate was chosen to optimise the reactionconditions. Thus, the synthesis of alkene3d using1.5 mmol of substrates, 0.1 g of catalyst and 1 mlof various solvents has been tested. After a time of60 min the best yields were obtained in methanol,ethanol and butanol (Table 1). Low yield of 3d wasobtained in isopropanol, ethylacetate and THF and no

Table 1Solvent effect in the FAP and HAP catalysed the condensation ofbenzaldehyde and methylcyanoacetate

Solvent MeOH EtOH BuOH iPrOH Ethylacetate THF

Yield of 3d (%)With FAP 25 20 18 16 7 8With HAP 50 46 40 32 14 15

A. Smahi et al. / Applied Catalysis A: General 250 (2003) 151–159 155

Fig. 5. Weight effect of the catalysts FAP and HAP in the condensation of benzaldehyde with methylcyanoacetate.

reaction was observed in dioxane, acetone, toluene,dichloromethane, hexane and cyclohexane. This re-sult confirms the crucial role played by the solventused to carry out the reaction.

It would appear that the best solvent for this reactionis the methanol. The same result has been observedwith NP[17]. Thus, methanol was chosen as a solventfor further study.

Thereafter, we examined the weight effect of thecatalyst in this reaction (Fig. 5). The yield of prod-uct 3d increased effectively as the catalyst weight in-creased to 0.1–0.3 g and increased slightly above 0.3 gof catalyst. Thus, we have chosen a weight of 0.1 g ofcatalyst for further study.

It should be pointed out that the yields obtained withHAP were higher than that of FAP. Consequently, theobserved differences in catalytic behaviour betweenFAP and HAP can be in part explained by the highersurface area of HAP (38 m2 g−1) comparatively withFAP (15 m2 g−1).

3.2. Heterogeneous catalysis with doped materialsKF/FAP and KF/HAP

As described in the introduction, the KF dopedAl2O3 can promote several organic synthesis. It was

already reported by us that for catalytic synthesis ofsome reactions, KF doped natural phosphate (NP) wassuperior to various solid catalysts. Analogously, wesuspect that the interaction between KF and FAP orHAP can enhance the activity of these solids. Underthe same conditions used with both FAP and HAP, wetested the activity of KF/FAP and KF/HAP in the con-densation of benzaldehyde with ethylcyanoacetate atroom temperature.

For an initial evaluation of activity of these dopedmaterials two series of kinetics experiments were car-ried out using varying values of catalyst weight. Theyield of product3g increased as catalyst weight in-creased. To evaluate the influence of the doping FAPand HAP by KF, we have chosen 0.1 g of the catalystfor further study as used with both FAP and HAP.

To evaluate the catalytic activity of the doped ma-terials in comparison with FAP and HAP, we carriedout the condensation of benzaldehyde with methyl-cyanoacetate, ethylcyanoacetate and malononitrile,respectively. The kinetic curves of these reactionsas shown inFigs. 6–8, indicate clearly the promot-ing effect of KF. In all the cases, KF/HAP appearto be more active than KF/FAP. This behaviour canbe attribute to the initial higher activity of HAP incomparison with FAP as observed below in addition

156 A. Smahi et al. / Applied Catalysis A: General 250 (2003) 151–159

Fig. 6. Kinetic curves of the alkene3d synthesis catalysed by FAP, HAP, KF/FAP, KF/HAP.

to the higher surface area of KF/HAP (31 m2 g−1)comparatively to KF/FAP (15 m2 g−1). This impor-tant result confirms the interaction effect between KFand FAP or HAP. The positive effect of doping FAP

Fig. 7. Kinetic curves of the alkene3g synthesis catalysed by FAP, HAP, KF/FAP, KF/HAP.

and HAP by KF observed in this work is similar tothat of KF/Al2O3 and KF/NP.

We have found that the best activity of our dopedcatalysts is observed in alcohols, which might hint at

A. Smahi et al. / Applied Catalysis A: General 250 (2003) 151–159 157

Fig. 8. Kinetic curves of the alkene3a synthesis catalysed by FAP, HAP, KF/FAP, KF/HAP.

dissolution of the impregnated KF. To indicate whetherthere is residual activity after catalyst removal, wehave stirred 0.1 g of KF/FAP or KF/HAP in 1 ml ofmethanol during 1 h. Then, the doped material wasseparated and the resulting solution was added to alde-hyde 1 and active methylene2. Only small trace ofproducts3d and 3i are obtained after 60 and 15 minof reaction, respectively. This result indicates that thereaction take place at the surface of the catalyst in het-erogeneous media. It is likely therefore that there issome sort of surface activation between fluoride andROH at the surface.

Another interesting fact is the recycling of the cata-lyst. Thus, we have carried out the synthesis of alkene3d using KF/FAP or KF/HAP in methanol and 60 minof reaction time. The isolated catalyst was washedwith 5 ml of dichloromethane and dried at 150◦C for1 h before reuse. The results obtained are reported inTable 2. It seen that the doped materials can be used atleast three times without significantly lost of activitywhich suggest low leaching of impregnated KF duringthe separating process.

To explore the scope of this methodology the con-densation of different substrates was investigated. Toappreciate the important enhancement of the activityof the doped materials, we have carried out this con-

densation with all substrates using the unsupportedKF powder catalyst. We have used amount (0.0125 g)of KF contained in 0.1 g of KF/FAP or KF/HAP(W/W = 1/8) used in this work. All results obtainedwith unsupported KF were poor (Table 3). The posi-tive effect of the doped solid catalysts can be observedin all cases. For example in entry 4, the yields ob-tained after 1 min of reaction time are 18, 17, 19, 78and 85% for KF, FAP, HAP, KF/FAP and KF/HAP, re-spectively. Thus, 78% yields obtained with KF/FAP issignificantly higher than 18%+ 17% yields obtainedwith both KF and FAP. Similar result is observedwith KF/HAP (85% yield) in comparison with bothKF and HAP (18%+ 19% yields). The results shownin Table 3 demonstrate that the procedure used bydoping FAP and HAP with KF can be successfully ap-plied to different kinds of substrates. The significant

Table 2Recycling of KF/FAP and KF/HAP catalysts in the synthesis ofalkene3d

Run Fresh Cycle 1 Cycle 2

Yield of 3d (%)With KF/FAP 80 78 77With KF/HAP 91 88 89

158 A. Smahi et al. / Applied Catalysis A: General 250 (2003) 151–159

Table 3Comparison of the catalytic activity of KF, FAP, HAP, KF/FAP and KF/HAP in the synthesis of several alkenes

Entry Products Yields (%)a

KF FAP HAP KF/FAP KF/HAP

1 3a 10 (10) 36 (10) 43 (10) 71 (10) 89 (10)

2 28 (15) 49 (15) 57 (15) 90 (15) 91 (15)

3 3b 23 (3) 47 (3) 50 (3) 87 (3) 90 (3)

4 3c 18 (1) 17 (1) 19 (1) 78 (1) 85 (1)

5 3d 0 (30) 10 (30) 32 (30) 59 (30) 80 (30)

6 5 (60) 25 (60) 50 (60) 80 (60) 91 (60)

7 3e 12 (5) 27 (5) 31 (5) 67 (1) 93 (1)

8 3f 17 (5) 13 (5) 25 (5) 79 (5) 95 (5)

9 3g 2 (60) 2 (60) 10 (60) 38 (60) 41 (60)

10 6 (120) 7 (120) 23 (120) 53 (120) 64 (120)

11 3h 17 (15) 40 (30) 42 (30) 93 (15) 98 (15)

12 3i 11 (15) 6 (30) 29 (30) 93 (15) 97 (15)

a Values given in parentheses denote the time in minutes.

comparison of activities and interest of our catalystswith other known materials is difficult because thedifference of reaction conditions used in each case asweight of catalyst, temperature of reaction, reusabil-ity, availability and others parameters. However, thecomparison was made on the basis of (yield (%), time(min) and reaction temperature (◦C)) in the synthesis

of 3a as KF/FAP (90, 15, rt), KF/HAP (89, 10, rt), NP(81, 15, rt)[17], KF/NP (82, 3, rt)[17], Na/NP (96,1, rt) [17], KF/Al2O3 (86, 150, rt)[22], AlPO4/Al2O3(86, 15, rt)[23], Al2O3 (96, 3, rt) [24], MgO (94, 5,rt) [25], hydrotalcite (54, 60, 160)[26], zeolite (91,60, 160)[26] and (78, 720, 60)[27], Na2CO3/zeolite(98, 15, rt)[28] and resine (92, 300, rt)[29].

A. Smahi et al. / Applied Catalysis A: General 250 (2003) 151–159 159

In conclusion, we have prepared and used for thefirst time KF/FAP and KF/HAP as new catalysts in het-erogeneous solid–liquid system. Some changes of sur-face of the fluorapatite and hydroxyapatite by dopingwith potassium fluoride have been observed in SEMimages. The comparison of the activity of KF/FAP andKF/HAP with KF, FAP and HAP has been studied us-ing the Knoevenagel condensation as reaction model.The results obtained pointed out clearly the positiveeffect of doping these phosphates by KF. This behav-ior is similar to that observed in several previouslyworks using KF/Al2O3 and KF/NP.

These results suggest that KF/FAP and KF/HAP canbe used as basic catalysts for other applications in theorganic synthesis.

Acknowledgements

Financial assistance of the “Ministère de l’Ensei-gement Supérieur” of Morocco (PROTARS, P2T3/59)is gratefully acknowledged. We thank also the “Of-fice Chérifien des Phosphates (OCP)” and “Centred’Etudes et de Recherches des Phosphates Minéraux(CERPHOS)” for their support. Thanks are expressedto Dr. D.J. Macquarrie, University of York, for coop-eration in some of work presented here.

References

[1] S. Shoda, R. Izumi, M. Suenaga, K. Saito, M. Fujita, Chem.Lett. 2 (2002) 150.

[2] G. Macfie, B.A. Brookes, R.G. Compton, J. Phys. Chem. B105 (2001) 12534.

[3] G.W. Kabalka, V. Namboodiri, L. Wang, Chem. Commun. 8(2001) 775.

[4] S.H. Mashraqui, H. Haribarasubrahmanian, S. Kumar,Synthesis (1999) 2030.

[5] B.E. Blass, Tetrahedron 58 (2002) 9301.[6] A. Zhao, B. Xu, Chem. World 43 (2002) 487.[7] C.C. Silveira, C.R. Bernardi, A.L. Braga, T.S. Kaufman,

Synlett 6 (2002) 907.

[8] J.-X. Wang, Y. Yang, B. Wei, Y. Hu, Y. Fu, Bull. Soc. Chem.Jpn. 75 (2002) 1381.

[9] B.E. Blass, C.L. Harris, D.E. Portlock, Tetrahedron Lett. 42(2001) 1611.

[10] Y. Zhou, Chem. World 42 (2001) 79.[11] G.W. Kabalka, L. Wang, R.M. Pagni, Green Chem. 3 (2001)

261.[12] G.W. Kabalka, R.M. Pagni, C.M. Hair, L. Wang, V.

Namboodiri, Spec. Publ. R. Soc. Chem. 266 (2001) 219.[13] D. Villemin, F. Caillot, Tetrahedron Lett. 42 (2001) 639.[14] J.M. Fraile, J.I. Garcia, J.A. Mayoral, S. Sebti, R. Tahir,

Green Chem. 3 (2001) 271.[15] Y. Abrouki, M. Zahouily, A. Rayadh, B. Bahlaouan, S. Sebti,

Tetrahedron Lett. 43 (2002) 8951.[16] S. Sebti, H. Boukhal, N. Hanafi, S. Boulaajaj, Tetrahedron

Lett. 40 (1999) 6207.[17] S. Sebti, A. Smahi, A. Solhy, Tetrahedron Lett. 43 (2002)

1813.[18] S. Sebti, A. Rhihil, A. Saber, N. Hanafi, Tetrahedron Lett.

37 (1996) 6555.[19] S. Sebti, A. Rhihil, A. Saber, M. Laghrissi, S. Boulaajaj,

Tetrahedron Lett. 37 (1996) 3999.[20] S. Sebti, R. Nazih, R. Tahir, L. Salhi, A. Saber, Appl. Catal.

A 197 (2000) L187;S. Sebti, R. Nazih, R. Tahir, A. Saber, Synth. Commun. 31(2001) 993.

[21] S. Sebti, R. Tahir, R. Nazih, S. Boulaajaj, Appl. Catal. A 218(2001) 25;S. Sebti, R. Tahir, R. Nazih, S. Boulaajaj, Appl. Catal. A 228(2002) 155;S. Sebti, A. Solhy, R. Tahir, A. Smahi, Appl. Catal. A 235(2002) 273.

[22] J. Yamawaki, T. Kawate, T. Ando, T. Hanafusa, Bull. Chem.Soc. Jpn. 56 (1983) 1885.

[23] J.A. Cabello, J.M. Campelo, A. Garcia, D. Luna, J.M.Marinas, J. Org. Chem. 49 (1984) 5195.

[24] F. Texier-Boullet, A. Foucaud, Tetrahedron Lett. 23 (1982)4927.

[25] H. Moison, F. Texier-Boullet, A. Foucaud, Tetrahedron 43(1987) 537.

[26] M.J. Climent, A. Corma, S. Iborra, A. Velty, J. Mol. Catal.182–183 (2002) 327.

[27] T.I. Reddy, R.S. Varma, Tetrahedron Lett. 38 (1997)1721.

[28] B. Siebenhaar, B. Casagrande, M. Studer, H.-U. Blaser, Can.J. Chem. 79 (2001) 566.

[29] T. Saito, H. Goto, K. Honda, T. Fujii, Tetrahedron Lett. 33(1992) 7535.