Catalytic Opportunities in the Flavor and Fragrance Industry

UNIVERSITE DE LAUSANNE 620

CHIMIA 50(1996) Nr. 12 (Dczembcr)

We wish to thank the Swiss National Science ,2+ Foundation for continued and generous support of this work over the years.

Received: September 30, 1996

Me2SO"'jRh .•...... [I] a)M. Eigen, Elektroclzemie 1960,64, I 15; b) OSMe Me2S0• 2 M. Eigen, Pure Appl. Chem. 1963,6,97; c) M. Eigen, R.G. Wilkins, Adv. Chem. Ser. (O-bonded) bound-M~SO 1965,49,55. [2] a) S.F. Lincoln, A.E. Merbach, Adv. /norg. Chem. 1995, 42, I; b) D.H. Powell, A.E. T(K) k(S·I) Merbach, in 'Encyclopedia of Nuclear Magnetic Resonance', Eds. D.M. Grant and 238.0 11 R.K. Harris, Wiley, Chichester, 1996, Vol. 4, 257.0 95 p.2664. 264.0 185 [3] J.P. Hunt, H. Taube,J. Chem. Phys.1950, 18, 271.3 369 757. 277.5 630 [4] a) F.-C. Xu, H.R. Krouse, T.W. Swaddle, Inorg. Chem.1985, 24, 267; b)G. Laurenczy, 302.6 5340 I. Rapaport, D. Zbinden, A.E. Merbach, Magn. I I I Reson. Chem. 1991,29, S45. 3.2 2.8 2.4 [5] a) S.E. Castillo-Blum, A.G. Sykes, H. & IH(ppm) Gamsjager, Polyhedron 1987, 6,101; b) S.E. Castillo-Blum, D.T. Richens, A.G. Sykes, Fig. 3. ExperimentallH-NMR (400 MHz) spectra of [Cp*Rh(Me2S0h](PF6h (0.100 m) with free /norg. Chem. 1989, 28, 954. Me2S0 (0.304 m) in CD3N02 at various temperatures [6] A. Cusanelli, U. Frey, D.T. Richens, A.E. Merbach, J. Am. Chenl. Soc. 1996, 118,5265. [7] L. Dadci, H. Elias, U. Frey, A. Hornig, U. solvent exchange whereas increasing the These results suggest that increasing the Koelle, A.E. Merbach, H. Paulus, I.S. steric environment around the metal cen- n-acceptor ability of the bound solvent Schneider,lnorg. Chem. 1995, 34, 306. ter promotes the dissociation process and ligand creates a stronger metal-ligand bond [8] A. Cusanelli, L. Nicula-Dadci, U. Frey, A.E. thus leads to an increase in the rate of and thus decreases the exchange rate. Merbach, submitted for publication.

Cizimia 50 (1996) 620-625 strictive uses. Perfume producers are also © Neue Sclzweizerische Chemische Gesellschaft becoming more and more concerned with ISSN 0009-4293 the biodegradability and innocuity of their products. Catalysis is now seen as a substitute or Catalytic Opportunities in the a complement to conventional organic processes since it involves higher atom Flavor and Fragrance Industry economy, and therefore generates much less waste, as has already been shown in Hubert Mimoun* the bulk chemical industry [3]. A further important feature of catal ysis is enantioselectivity. Smell and taste in- Abstract. Because catalytic processes produce in general less residues than convention- volve chiral biological sensors that are al stoichiometric ones, they are advantageous for the environment. This short review sensitive to stereogenic centers in the will focus on new opportunities for catalysis in the field of terpene transformation, citral molecules, and the organoleptic proper- and ionones synthesis, and macrocyc1ic VS. aromatic musks. ties of two enantiomers often differ in intensity and quality [4]. For example, Introduction Friedel-Crafts acylation, etc. by alterna- (+)-(S)-carvone (caraway-like) and (-)- tive cleaner, safer, and environmentally (R)- carvone (spearmint-like) have differ- Like other chemical industries, the fla- friendly alternative technologies are un- entodors. Thus, enantioselectivecatalysis vor and fragrance (F&F) industry [1][2] der continuously development. Most of is the next frontier for the commercializa- has taken up the challenge to reduce cost the toxic metals such as chromium, lead, tion of chiral perfumes. This short review of waste treatment, thereby using process- selenium, mercury, cadmium have been will deal with some recent applications of es more friendly for the environment. Thus, removed from industrial processes. The catalysis in the F&F industry, and focus on the substitution of conventional organic use of reagents such as acrolein, methyl new opportunities in this area. chemistry processes such as stoichiomet- vinyl ketone, phosgene, and dimethyl sul- fate is severely limited because of trans- ric oxidation by permanganates and chro- *Correspondellce: Prof. Dr. H. Mimoun mates, reduction by dissolving metals and portation regulations, and/or have to be Firmenich S.A. metal hydrides, Wittig condensations (re- used at their production site. Chlorinated 45, Route de la Plaine cycling of triphenylphosphine oxide), and aromatic solvents also have very re- CH-1283 La Plaine-Geneva UNIVERSITE DE LAUSANNE 621

CHIMIA 50(1996) NT. 12 (Dcl.cmbcr)

1. Fragrances from a-Pinene H and j3-Pinene +Me Q e a-Pinene (ca. 18000 t/year), j3-pinene R' (RObV=O (R012~O~ ,\ (ROh-{,.°1 (ca. 12000 t/year) and .1rcarene (ca. 1000 ·ROH 02.fR' ° fR' R' OH (I) t/year) are the main inexpensive mono- Me Me terpenes that are extracted from turpen- tine, a very important raw material in the R'=~ F&F industry.

1.1. Fragrances from a-Pinene its Colonel's Island plant (Georgia, USA) greatly improved by the use ofhydropho- Scheme 1 shows some of important [6]. Pinanol is then pyrolyzed to Iinalool, bic Beta and USY zeolites with high Si/Al fragrances manufactured from a-pinene. which can be rearranged to geraniol in the ratios. Thus, a-pinene and camphene were Hydration of a-pinene affords terpine hy- presence of trialkyl oxovanadates hydrated to a-terpineol and isoborneol, drate which is dehydrated to a-terpineol VO(ORh- Catalytic hydrogenation of gera- respectively, with selectivities as high as (+)-isomers). Epoxidation by peracetic niol affords citronellol. 70 and 90% [9]. acid gives a-pinene epoxide which rear- The allylic rearrangement of linalool ranges to campholenic aldehyde in the to geraniol deserves particular attention, 1.2. Fragrances from {3-Pinene and presence of catalytic amounts ofZn salts. since it is an ene-like reaction involving Myrcene Campholenic aldehyde is a very important the vanadium-oxo bond V=O as a partner Except for the synthesis of Nopol and intermediate used for the manufacture of (Eqn. 1) [7]. Nopyl acetate by a Prins condensation sandalwood fragrances, e.g., Sandalore® a-Pinene rearranges to camphene in with formaldehyde, {3-pinene is mainly (Givaudan), Bacdanol® (/FF), Brahma- the presence of Ti02. Camphene can be used in industry as its rearrangement prod- nol® (Dragoco), Polysantol® (Firme- acetylated to isobornyl acetate, or con- uct, Myrcene, which is obtained by pyrol- nich)[5]. Catalytic hydrogenation of a- densed with guaiacol and catalytically hy- ysis (Scheme 2). Diels-Alder reaction of pinene gives mainly cis-pinane which can drogenated to give isocamphylcyclohex- myrcene with acrolein gives Acropal®, be pyrolyzed to dihydromyrcene, from anol a sandal wood fragrance commercial- whereas reaction with 3-methylpent-3-en- which dihydromyrcenol can be obtained ized under the trade name of Sandela® 2-one followed by cyclization gives iso- by hydration in the presence of H2S04, (Givaudan), Santalex-7® (Takasago), or cyclenone, also known as Iso E Super® Autoxidation of pinane to the hydroperox- Sandel H&R® (Haarmann & Reimer) [8]. (/FF), which displays a pleasant woody ide, followed by hydrogenation to pina- Nonselective and polluting hydration amber note [10]. Myrcene can also be nol, is carried out in large scale by SCM in of terpenes by H2S04 has been recently hydrated to myrcenoJ through its cyclic

Scheme I. Fragrancesfrom a-Pinene

Dihydromyrcene Dihydromyrcenol Camphene

11 GUaiaC:1 / \H30+ or 1 21 H2 Niri .;' AcOH ~ cY~~1 Sandela R Isoborneol Isobornyl acetate

Geraniol Citronellol UNIVERSITE DE LAUSANNE 622

CHIMIA 50 (1996) Nr. 12 (Dcwnbcr) sulfone. Myrcenol is condensed with ac- NAP). Tagasago also commercializes oth- La Roche (HLR), BASF, and Rhone-Pou- rolein to give Lyra[® (IFF) [11], which has er fragrances derived from this BINAP lenc (RP) to the synthesis of citral and f3- a lily-of-the-valley odor. An alternative technology such as (+)- and (- )-citronel- ionone are shown in Scheme 3. synthesis of geraniol involves the 1,4 101 and (- )-7-hydroxycitronellal [14]. The HLR process involves the ethy- addition of HCI to myrcene to give gera- A further application of catalysis in the nylation of methyl heptenone to dehy- nyl chloride which is then hydrolyzed by transformation of myrcene is the industri- drolinalool, followed by Saucy-Marbet alkali. al synthesis by Rhone Poulenc (RP) of addition of methyl isopropenyl ether and (-)-Menthol is a widely used F&F geranylacetone in which methyl aceto- rearrangement to pseudoionone [18]. Ci- ingredient (ca. 5000 t/year) which has acetate is added to myrcene in the pres- tralcan be made via a Meyer-Schuster [19] both a natural (from peppermint and spear- ence of a H20-soluble rhodium tris(m- or trialkyl-vanadate rearrangement of de- mint oil) and a synthetic origin. It is indus- sulfonatoylphenyl)phosphine trisodium hydrolinalool. trially produced, inter alia, by hydrogen- salt (TPPTS) complex [IS]. The biphasic The BASF and RP processes differ in ation of thymol, followed by separation liquid-liquid process allows for an easy the choice of the starting materials used and resolution of the menthyl benzoates separation of the organic products by de- for the synthesis of pre noI and prenal, the enantiomers [12]. Takasago has recently cantation and the recycling of the aqueous two building blocks used in the manufac- disclosed a process for the manufacture of catalytic solution without significant loss ture of citra!. (- )-menthol (1000 t/year) involving a) the of activity. Geranylacetone is an impor- The BASF process is a triumph of ca- addition of diethylamine to myrcene cata- tant intermediate in the F&F industry but talysis in the synthesis of fine chemicals lyzed by lithium; b) the enantioselective is mainly used by RP for the synthesis of and involves both atom economy and en- isomerization of diethylgeranylamine to isophytol and tocopheryl acetate (vitamin vironmentally friendly technology. Ene an enamine in the presence of [Rh(BI- E, ca. 10000 t/year world market). condensation of isobutene and formalde- NAPh]+ catalyst; c) the hydrolysis of the hyde in the gas phase at 2800 affords dieneamine to (R)-citronellal; d) the cy- isoprenol [20]. In a continuous process, clization of (R)-citronellal to (- )-isopule- 2. Citral and Ionones the isoprenol is first isomerized to prenol gol in the presence of catalytic quantities over a supported palladium catalyst, then ofZnBr2; and e) the hydrogenation of (-)- Citral [16] is a key intermediate in the oxidized to prenal with a 95% yield at isopulegol to (-)-menthol (Scheme 2) [13]. F&F industry as well as in the vitamin A 5000 and a millisecond residence time The key step of this process is the isomer- synthesis. It allows the synthesis of a- and over a fixed bed containing a supported ization of diethylgeranylamine to the di- f}-ionone and methylionones after aldol Ag catalyst [21] (Eqn. 2). Two moles of ethyl enamine of citronellal which is condensation with acetone or methyl ethyI prenol and one mole of prenal condense to achieved with 96% enantiomeric excess ketone followed by acid-catalyzed cycli- prenal diprenyl acetal, which loses one (ee) and a turnover number (TON) of zation [17].lt is also used as intermediate mole of prenol, and rean'anges in a Clais- 400000. This is carried out in the presence in the synthesis of Rose ketones (a-, f}-, en-Cope sigmatropic reaction to give cit- of a chiral rhodium complex [Rh«S)- and 'fDamascones, Damascenone) and nu- ral (Eqn. 3) [22][23]. BINAPh]+ containing the atropisomeric merous woody and amber fragrances [5]. The Rhone-Poulenc route involves chiral diphosphine ligand 2,2'-bis(di- The different approaches by the three ethynylation of acetone to 2-methylbu- phenylphosphino)-I, I '-binaphtyl (BI- main vitamine A producers, F. Hoffmann- tyn-2-ol, followed by rearrangement of

Scheme 2. Fragrallcesfrom {3-Pillene alld Myrcelle Q~ . OH Takasago ~ (-)-Isopulegol (-)-Menthol U'NEb~Cc°~Br2

enamine (Rl-Citronellal

502/ ~OMe OH C02Me

+ Rh(tppts) CHO 1/H30 Co-.?' Rhone Poulenc 50 O I 2"f:SO; I 2 Co 31 acrolein ecc-.?' . -C02 y~ Isophytone ~R : • -- Vitamin E IFF •.... C02Me - MeOH ~. - Geranylacetone UNIVERSITE DE LAUSANNE 623

CHIMIA 50 (19%) Nr. 12 (Dc/clllhcr)

Scheme 3. Citra I and f3-lonone >= + HCHO - BASF cl~'OH 4o~ HOY· oY 0 Dehydrolinalool Me Heptenone f.rMQ.! E.mna.I - + ~ ROCHE 1VO(ORb j RHONE-POULENC 0 1

0 A 0 U: ~o U: Pseudo Ion one Citral

Vitamin A

Perfumes

Beta - lonone

methyl butynol to prenal in the presence of Scheme 4. Aromatic and Macrocyetic Musks trialkyl vanadate or Ti(ORkCuCI catalysts [11][24]. Prenal can then be selec- tively hydrogenated to prenol in the pres- ~ ~ ence of H20-soluble catalyst made from RuCI3 and TPPTS (Eqn. 4) [25]. ~ ~ Tonalide R Galaxolide R Versalide R

3. Macrocyclic vs. Aromatic Musks

Aromatic musks (Scheme 4) such as Tonalide® (PFW, Givaudan), Galaxolide® (IFF), Versalide® (Givaudan), Phan- tolide® (PFW), Celestolide® (IFF), and

Vulcanolide® (Firmenich) are widely used Phantolide R Celestolide R Vulcanolide R (world market; ca. 7 000 t/year) both in fine perfumery and household soaps and detergents. Their success can be attributed to their excellent odor and fixative proper- ties, as well as their moderate cost and outstanding stability [26][5]. They are generally synthesized by acid-catalyzed cycloalkylation of petrochemical aromatic starting materials, p-cymene, a-meth- Civetone Muscone Exaltone ylstyrene, or (tert-butyl)benzene with ole- fins such as isoamylene, neohexene, dimethylbutenes, and isoprene, followed by Friedel-Crafts acylation [21]. These alkylations are carried out in the presence of AICl3 or H2S04/AcOH [27]. Better yields have been recently obtained in the presence of sulfonated resins or polysi- loxanes Deloxan® ASP catalysts [28]. As Ethylene Brassylate Exaltolide R Ambretolide an illustrative example, the synthesis of UNIVERSITE DE LAUSANNE 624

CHIMIA 50(1996) Nr. 12 (Del<111ber)

turnover numbers (> 106) in the presence + HCHO I (PdJ I 02ltAg) I 2) of a mixture of [NiBr( 173-allyl)P(iPrh(1Bu)] ~OH ~OH ~O and EtAICI2 as catalyst [30]. The high stability and lack of biodegradability of aromatic musks makes their I ~O future development questionable for envi- ~OH •. (3) ronmental reasons [31], and they are pro- gressively sought to be substituted by natural-identical macrocyclic musks such as Civetone (civet cat), Muscone (musk- deer), Exaltone® (muskrat), Exaltolide® (angelicarootpil), and Ambrettolide (am- brette seeds) [32]. Since their discovery by Ruzicka some 70 years ago [33], these macrocyclic musks have been the object of considerable studies, and many syn- KOH VO(OR13 thetic routes have been published [26] [34] + - [35]. Although most industrial syntheses >-0 NH! Uq. or CuC1IT1(OR). ~O involve ring-enlargement reactions from Prenal Methylbutynol cyclododecanone [26][36], ring-closing metathesis (RCM) has recently emerged (~) as an attractive alternative catalytic route for the synthesis of macrocyclic musks HafRu(TPPTS) [37]. Civetone can be prepared from me- ~OH H~ tathesis of inexpensive methyl oleate in Prenol Clalsen-Cope the presence ofWCIJSnMe4 (50% based Cltral on the theorical yield of 50%) [38] or WOCliCp2 TiMe2 (87% yield) [39] catalyst, followed by Dieckmann condensation and decarbomethoxylation. In an alternative route, methyl oleate is converted ( to Oleon (yield 80%) which is metathe- sized to Civetone in the presence ofRe207/ SiOz-A1203 catalyst [40] (Eqn. 8). An elegant synthesis of Exaltolide® from undec-lO-enoyl chloride, an inex- 2 pensive derivative of castor-oil pyrolysis, (6 recycle and hex-5-enol was carried out by Villem- in [41]. The resul ting dienic ester was closed by RCM with the elimination of ethylene in the presence of WCIJSnMe4' The yield of the RCM step was recently 7 substantially improved to 79% by Fiirst- neretal. [42] in the presence of the Grubbs > 106 TON [43] functional-group-tolerant ruthenium- carbene catalyst (Eqn. 9). ReM access to macrocyclic musks ap- co,Me • pears, therefore, to be very promising de- CO,Ma spite insufficient turnover numbers, and the fact that high dilution conditions are 11/DlK M , v-co. required to avoid intermolecular reactions.

~r' Conclusion 0••••" As can be seen the future of catalytic processes in the F&F industry looks very promising and development of such sys- Tonalide from p-cymene, neohexene, or 2-ene (diisobutene) with ethylene in the tems will be driven by the need for clean, 2,3-dimethylbutene, and acetyl chloride is presence ofW03/Si02 (Eqn. 6) [29]. safe, and environmentally friendly tech- shown in Eqn. 5. Neohexene is prepared Dimerization of propylene to 2,3- nologies. Both heterogeneous (selective on a large scale by Phillips Petroleum by dimethylbut-l-ene (Eqn. 7) can be achiev- oxidation, cyclization) and homogeneous cross-metathesis of 2,4,4-trimethylpent- ed with excellent selectivity and very high catalysis (metathesis, carbonylations, ox- UNIVERSITE DE LAUSANNE 625 CHI MIA 50 (1996) Nr. 12 (De,ember)

[24] P. Chabardes, Tetrahedron Lett. 1988,29, 6253. [25] 1.M. Grosselin, C. Mercier, Organometal- ~o [cat] ~ L , -HCI ExllllOllde lics 1991, 10, 2126;J. Mol. Cata!.1990,63, + Cl • C,H. L26 . ~OH [26] T.F. Wood, in ref. [15J, p. 509. (9) [27J T.F. Wood, W.M. Easter, M.S. Carpenter, CI~Ph 1. Angiolini, J. Org. Chen!. 1963,28,2248. [Cel]. WClolSnMe. (Villemln) or CI/~U - Ph (FOretnor) [28J S. Wieland, P. Panster, Proc. 4th Int. Sym- PCy, posium on Heterogeneous Catalysis and Fine Chemistry, Base], Switzerland, 1996, p.30. [29J R.L. Banks, D.S. Banasiak, P.S. Hudson, idations, enantioselective reactions, alky la- [12] K. Bauer, 1. F]eischer, R. Hopp, German 1.R. Norell, J. Mol. Cata!. 1982, 15, 35. tions) will be increasingly used in these Pat. 2109456 (1971) and 3 017 068 (1980) [30J G. Wilke, in 'Organometallics in Organic selective alternative processes. to Haarmann & Reimer. Synthesis 2', Eds. H. Werner and G. Erker, [13] S. Akutagawa, Appl. Catal. A 1995, 128, Springer-Verlag, Berlin, 1989, p. I. 171, and ref. cit. therein. [31] J.P. Rozat, F. Naef, Peljurner & Flavorist Received: September 30, 1996 [14] S. Akutagawa, K. Tani, in 'Catalytic asym- 1996,21,13. metric Synthesis', Ed. I. Ojima, VCH, New [32] B.D. Mookhergee, R.A. Wilson, in ref. [I] G. Ohloff, 'Scent and Fragrances', Sprin- York, 1993, p. 41. [15], p. 433. ger- Verlag, Berlin, 1994. [15] e.Mercier,P.Chabardes,PureAppl. Chem. [33] L. Ruzicka, Helv. Chirn. Acta 1926, 9,230, [2] K. Bauer, D. Garbe, 'Common Fragrances 1994,66, 1509, and ref. cit. therein. 249, 715, 1008. and F]avor Materials', VCH Verlag, Wein- [16] Citral is a common name to designate a [34J C.J.Roxburgh, Tetrahedron 1995,51,9767. heim, 1985. mixture of geranial (E)- and neral (2)- [35] H. Stach, M. Hesse, Tetrahedron 1988, 44, [3] R.A. Sheldon,J. Mol. Catal.1996, 107,75; Isomers. 1573. Chemtech 1994, 38. [17] P.Z. Bedoukian, in 'Fragrance Chemistry', [36] J. Becker, G. Ohloff, Helv. Chim. Acta [4] M.H. Boelens, Perfumer & Flavorist 1993, Ed. E.T. Theimer, Academic Press, New 1971,54,2889. 18,3. York, 1982, p. 285. [37] S. Warwel, H. Bachem, A. Deckers, N. [5] G.Ohloff, B. Winter, e. Fehr, in 'Perfumes, [18] G. Saucy, R. Marbet, Chimia 1960, 14, Doring, H. Kiitker, E. Rose, Seifen, Ole, Art, Science & Technology', Eds. 362. Fette, Wachse 1989,115,538. P.M. Muller and D. Lamparsky, Elsevier, [19] K.H. Meyer, K. Schuster, Ber. 1922,55, [38] P.B van Dam, M.C. Miuelmmeijer, e. Boel- New York, 1991, p. 287. 819; H. Rupe, E. Kambli, Helv. Chim. Acta houver, Fette, Seifen, Anstrichmittel1974, [6] J. Dorsky, in ref. [2], p. 399. 1926,9,672. 76,264. [7] P. Chabardes, E. Kuntz, J. Varagnat, Tetra- [20] P.R. Stapp, Ind. Eng. Chem., Prod. Res. [39J J. Tsuji, S. Hashiguchi, Tetrahedron Lett. hedron 1977, 33, 2799. Dev. 1976, 15, 189. 1980, 2955; J. Organomet. Chern. 1981, [8] G. Frater, D. Lamparsky, in ref. [2], p. 557. [21] W.F. Hoe1derich, in 'Studies in Surface 218,69. [9] J.e. Van der Waa], H. van Bekkum, J. Science and Catalysis', Eds. L. Guczi, F. [40J M.F.e. Plugge, J.e. Mol, Synlett. 1991, Vital, J. Mol. Catal. 1996, 105, 185; H. Solymosi, and P. Tetenyi, 1993, Vol. 75, 507. Valente, 1. Vital, Proc. 4th Int. Symposium Part A, p. 27. [41] D. Villemin, Tetrahedron Lett. 1980, 1715. on Heterogeneous Catalysis and Fine Chem- [22] N. Goetz, R. Fischer, German Pat. 2 157 [42] A. Fiirstner, K. Langemann, J. Org. Chern. istry, Base], Switzerland, 1996, p. 214. 035 (197]); P. Chabardes, J. Chasal, Eur. 1996, 61, 3942. [10] US Pat. 3 929 677 (1973) to IFF. Pat. 344043 (1988). [43] R.H. Grubbs, S.J. Miller, S.J. Fu, Ace. [II] US Pat. 4 007137 (1975) to IFF. [23] J. Paust, Pure Appl. Chem. 1991,63,45. Chern. Res. 1995, 28, 446.

Chimia 50 (1996) 625-629 1. Introduction © Neue Schweizerische Chemische Gesellschaft 1SSN 0009-4293 Because of the ever increasing demand on water resources, efforts have been di- rected during the last decades towards New Directions towards the water sanitation, and in the development of models for predicting the response of Understanding of Physico- water systems to anthropogenic con- straints. Chemical Processes in Aquatic This paper briefly describes our re- search, which is focused on the micro- Systems scale characterization of colloids in aquatic systems, and on the understanding of their transformation processes, their reac- Christine Couture, Charles-Philippe Lienemann, Denis Mavrocordatos, tivity, and their role in the scavenging of and Didier Perrett toxic trace elements and nutrients.

Abstract. Some applications of Analytical Electron Microscopies (AEM) for the characterization of colloids in natural waters are presented. These techniques go much *Correspondence: Dr. D. Perret beyond the basic level of morphological observation and render possible to gain lnstitut de chimie minerale et analytique detailed physico-chemical information difficult to obtain by conventional analytical Universite de Lausanne methods. Hence, AEM appear as a challenging necessity for understanding the colloid- BCH mediated transport of toxics in the environment. CH-IO 15 Lausanne