The dendrimers presented so far have been synthesized through a convergent strategy. It was thought to make use of divergent methodologies- as an alternative approach to build up dendrimer-fullerenes. Since in the divergent approach the reactions take place at the surface of the growing dendrimer, the steric inhibition encountered when using the convergent procedure may be avoided thus allowing for higher generations. However, the rapid increase in the number of reactive groups at the periphery of the growing macromolecule, which is characteristic for divergent methods, may cause difficulties. Potential problems which may arise as growth is pursued include incomplete reaction of the terminal groups, which would lead to imperfections in the next generation
Due to the own characteristics of the divergent method, surface functionalities able to be deprotected must be present in the growing macromolecule (figure 3.31). It was decided to develop this concept using a variation of the new dendritic system presented in the last chapter. The synthetic plan would include the introduction of protecting groups in the phenolic hydroxyls of the building block. These protecting groups should be stable during the construction of the dendrimer and easily removable once attached to the fullerene.
Figure 3.31. Divergent approach. First generation fullerene-dendrimer A contains several protecting groups. The attachment of first generation dendrons D to the surface coupling sites of the deprotected fullerene derivative B results in the formation of a second generation dendrimer C (fp: protecting group; c: coupling site; fr: reacting group).
Removal of the protecting groups is not only necessary to develop a divergent approach, but is interesting in itself. Of special importance would be the formation of water-soluble dendrimers with potential abilities as molecular micelles or as hosts for the transport of biological guests.
The protection of the hydroxyl groups of the starting compound methyl-3,5-dihydroxy-benzoate (11) was studied. The protecting groups must be stable under the reductive, oxidative, and basic conditions employed during the synthesis of the dendrimer. The protection of these hydroxyl groups as tert-butyl, tetrahydropyranyl and 2-methoxyethoxy-methyl ethers was studied (figure 3.32).
Figure 3.32. Synthons 74-76.
Tert-butyl ethers should be stable under the reaction conditions and easily removable under acidic conditions. Unfortunately, the attempts to synthesize 74 failed. This included alkylations with 2-methylpropene under acid catalysis and alkylation with tert-butyl bromide in refluxing pyridine. The formation of tetrahydropyranyl ethers was tested. Despite of the acceptable yields in the formation of 75 under classical conditions, this strategy was immediately rejected. The generation of a chiral center per THP group attached results in the formation of a diasteroisomeric mixture of products. Undoubtedly, this would complicate the isolation and identification of the dendritic system in the following steps.
Finally, it was decided to use 2-methoxyethoxymethyl ethers (MEM) as protecting group. This kind of ethers are stable under dendrimer formation conditions. A priori, MEM ethers should be easily cleaved under acidic conditions, while the rest of the dendrimer should be unchanged.
Ester 76 was obtained in a yield of about 92 % by alkylation of 11 with 2-methoxyethoxy-methyl chloride in a slurry of NaH in THF (scheme 3.16). Starting from 76, the synthesis of malonate 80 was carried out in a way analogous to the preparation of 64 described in the previous chapter with comparable results. Different mixtures of solvent always containing
2 % of triethylamine were employed to the purify targets 76-80 by flash chromatography on silica gel in order to avoid the cleavage of MEM ethers.
Scheme 3.16. Synthesis of malonate 80: i) MEM-Cl, NaH/THF (91.6 %); ii) LiAlH4/Et2O (93 %); iii) allyl bromide, NaH/THF (91 %); iv) (a) 9-BBN/THF; (b) EtOH, H2O2, NaOH (92.6 %), v) (a) NaH/THF, (b) malonyl chloride (48.6 %).
The fullerene monoadduct 81 was obtained by treating fullerene with the corresponding malonate 80 in the presence of CBr4 and DBU (scheme 3.17).
Scheme 3.17. Fullerene monoadduct 81.
The compound was completely characterized, and all the spectroscopic data recorded were in good agreement with a bridged methanofullerene structure. The 13C-NMR spectrum (figure 3.33) reveals sixteen different types of sp2-carbons (145-139 ppm) and one type of
sp3-carbon (72.8 ppm) of the fullerene cage according to the C2v-symmetry of 81. The rest of the peaks in 13C-NMR and 1H-NMR spectra appear at about the same position as in the parent dendritic malonate 80. MALDI-TOF Mass spectrometry gave the molecular ion at
m/z = 1536.
Figure 3.33. 13C-NMR (100.5 MHz, 310C, CDCl3) spectrum of 82.
It was tried to cleave the four protecting groups present in 81. This would result in the formation of a molecule with four free hydroxyl groups (82) ready to undergo reaction with the bromide 83, thus yielding the second generation dendrimer 84 (scheme 3.18).
Scheme 3.18. Planed divergent synthesis of the second generation dendrimer 84.
Unfortunately all the methods of cleavage applied failed. With 10 equivalents of trifluoroacetic acid no reaction took place, whereas with 50 equivalents, the formation of a dark brownish precipitate without defined composition was obtained. With TiCl4 a very slow progress was observed together with the formation of a precipitate that also could not be identified.