A Synthetic Route to Quaternary Pyridinium Salt-Functionalized Silsesquioxanes

A synthetic route to potentially biocidal silsesquioxanes functionalized by quaternary pyridinium functionalities has been developed. N-Alkylation reactions of the precursor compounds 4-(2-(trimethoxysilyl)ethyl)-pyridine (5) and 4-(2-trichlorosilylethyl)pyridine (6) with iodomethane, n-hexylbromide, and n-hexadecylbromide cleanly afforded the corresponding Nalkylpyridinium salts (7–10). The synthesis of a 4-(2-ethyl)pyridine POSS derivative (2) was achieved by capping of the silsesquioxane trisilanol Cy7Si7O9(OH)3 (1) via two different preparative routes. Attempts to use compound 2 as precursor for quaternary pyridinium salt-functionalized POSS derivatives were met with only partial success. Only the reaction with iodomethane cleanly afforded the new N-methylpyridinium salt 12 in high yield, whereas n-hexylbromide and nhexadecylbromide failed to react with 2 even under forcing conditions.


Introduction
Over the past fifty years a broad variety of new classes of polymers have been prepared and studied which should provide advances in developing a new family of compounds for antibacterial surface treatments [1].Such polymeric materials or films which kill or inactivate microorganisms upon their direct contact are known as biocidal (antimicrobial) polymers or also polymeric biocides.During the 1990s the interest in biocidal polymers arose rapidly due to their potential ability to keep surfaces and materials permanently antiseptic.This continues to be of current importance for a wide range of applications.Biocidal polymers are used, for example, in cartridge filters for the disinfection of potable and recreational water supplies, in filter units for air disinfection, as sterile bandages, clothing, surgical gloves for medical uses, as biocidal polymeric coatings on surfaces of ship hulls, shower walls and many other kinds of tubing.The ideal biocidal polymer should possess at least the following characteristics: (1) it should be easily and inexpensively synthesized; (2) it should be stable in long-term usage and storage at the temperature of its intended application; (3) it should be not soluble in water in the case of water disinfection applications; (4) it should not decompose to and emit toxic products; (5) it should not be toxic or irritating to those handling it; (6) it should be regenerable upon loss of activity; and (7) it should be biocidal to a broad spectrum of pathogenic microorganisms in brief times of contact [1][2][3][4].
By now various biocidal polymers have been produced and tested in different fields, but the achievement of a polymer which combines all of these characteristics continues to be elusive.In accordance with literature reviews there are several classes of biocides which possess great potential for the development of the ideal biocidal polymers for their sufficiently high activity against the two major classes of bacteria and fungi, Gram-positive and Gram-negative.They will not be removed from surfaces on washing, they remain capable of continually acting against the bacteria, they are not toxic or irritating, and one of the very attractive advantages is that they cause no antibiotic resistance [1][2][3][4].Their approximate composition is represented in Scheme 1.
Such biocidal polymers always comprise three essential structural parts: the carrying surface, usually the International Journal of Polymer Science carbohydrate-base (cotton cloth, wood, paper, or bulk cellulose, etc.), and also glasses and silica, and the polymer films (in Scheme 1 polyethylene, polyurethane, and polysiloxane) which contain anchored polyquat moieties (the quaternary nitrogen) and which directly contact to a cell membrane and cause its disruption.It has been experimentally proven that low antiseptic activity already reveals for the moieties with N-alkyl chains from three to eight carbon units in length and is very high for the moieties containing a 16-carbon lipophilic chain [1][2][3][4].
Polyhedral oligomeric silsesquioxanes (POSS) form an exciting class of hybrid organic-inorganic filler materials receiving considerable attention in recent years [5][6][7][8][9][10][11][12][13][14][15][16][17].They represent several elements of novelty, for example, molecular diameters between 1 and 3 nm, low density, high thermal stability, and an array of side-chain functionalities [18] which accounts for compatibility with various host polymers.Thus is seemed of interest to synthesize and characterize a new class of biocidal materials based on polyhedral oligomeric silsesquioxanes by attaching distant quaternary ammonium functional groups to the POSS cage.In this contribution, we report the first synthetic approach eventually leading to POSS derivatives comprising a pendant quaternized 4-(2ethyl)pyridyl group.
The key precursor 1 and the target molecules 2 and 3 are illustrated in Scheme 2. Trisilanol 1 was chosen as the best model for silica that has been developed to date for its close-range geometric similarity to known SiO 2 structures.It is a very useful model for both spectroscopic comparisons and chemical reactivity studies for it structurally resembles specific surface structures that occur on silica [19][20][21][22][23]. Alkylated quaternary 4-(2-ethyl)pyridyl groups containing 6-and 16-carbon chains are of considerable interest for their significant activity against a wide range of bacteria (3a, b, Scheme 2).The intended synthetic route involved capping of 1 with either 4-(2-(trimethoxysilyl)ethyl)-pyridine and 4-(2-trichlorosilylethyl)pyridine to give the 4-(pyridine)ethyl derivative 2 which could then be alkylated by treatment with appropriate alkyl halides to give the target compounds 3a and 3b.
The original preparation of 4-(2-(trimethoxysilyl)ethyl) pyridine ( 6) calls for treatment of the trichlorosilyl precursor 5 with trimethyl orthoformate in the presence of catalytic amounts of aluminum trichloride according to Scheme 4 [25].
International Journal of Polymer Science 3 Although the reaction outlined in Scheme 4 had been described only on a 1 g-scale, we found that by employing longer heating times up to ca. 10 g of 6 could be prepared without difficulties.As an alternative route to 6, the direct hydrosilylation of vinylpyridine (4) with trimethoxysilane, HSi(OMe) 3 , has also been investigated (Scheme 5).Such reactions of unsaturated substrates with trialkoxysilanes have been shown to work well in the presence of suitable hydrosilylation catalysts such as H 2 PtCl 6 × H 2 O (Speier's catalyst) or Karstedt's catalyst (= Pt(LL) 2 , LL = (CH 2 =CHSiMe 2 ) 2 O) [26][27][28][29][30][31].Reactions were carried out under various conditions (e.g., without solvent or in acetonitrile solution, without catalyst or in the presence of Speier's catalyst or Karstedt's catalyst), but surprisingly all these reactions failed to produce appreciable amounts of 6.Thus the original preparation (Scheme 4) remains the best access to the trimethoxysilyl derivative 6.

Quaternization of 4-(2-Trichlorosilylethyl)pyridine (5) and 4-(2-(Trimethoxysilyl)ethyl)-pyridine (6) with Different
Alkyl Halides.Next, quaternization reactions of the precursors 5 and 6 with different alkyl halides were studied.In quaternization reactions of tertiary amines with alkyl halides the reactivity of the latter generally decreases in the series RI > RBr > RCl, for example, MeBr > EtBr > n-C 6 H 13 Br > C 16 H 33 Br.Although for application reasons we were particularly interested in the reactivity of alkyl halides with longer chains, for example, with six or sixteen carbon atoms, reactions with methyl iodide as the simplest and most reactive alkyl halide were also included.Usually N-alkylation reactions of tertiary amines are performed in polar solvents (acetonitrile, methanol, ethanol, etc.).The highest reaction rates are normally observed in bipolar aprotic solvents (DMF, DMSO), but when using these solvents difficulties in product isolation and purification can became apparent.Quaternization reactions of 5 and 6 were carried out in anhydrous organic solvents (e.g., methanol, acetonitrile, THF) as illustrated in Scheme 6. Owing to the very high solubility of 4-(2-trichlorosilylethyl)pyridine (5) and iodomethane in all organic solvents, the N-alkylation with iodomethane was possible to perform in pentane, diethyl ether, THF, DMSO, acetonitrile, ethanol, and methanol.The highest reaction rates were observed in the more polar solvents and the slowest in nonpolar solvents.The pure product, the Nmethylpyridinium salt 7, was isolated as a white solid after recrystallization from diethyl ether in nearly quantitative yield (98%).
As expected, the analogous quaternization reactions with n-hexyl and n-hexadecyl bromide (Scheme 7) were much slower than the CH 3 I reactions.N-alkylation of 6 with n-hexylbromide could readily be performed in those polar solvents (acetonitrile, methanol) in which the highest reaction rates were observed for iodomethane, but the Nalkylation with n-hexadecylbromide could only be carried out in diethyl ether or THF because of its poor miscibility with the more polar solvents.Accordingly, in both cases the reaction rates and isolated yields were so low that even after 5-7 days of refluxing the reactions were less than 30% completed.
Much better yields (69-78%) were obtained in a second series of N-alkylation experiments which were performed under solvent-free conditions by just stirring a mixture of the reagents at temperatures of 100-130 • C for 48-72 h.Under these conditions the pure N-n-hexylpyridinium salt 9 was isolated from diethyl ether as a pale greenish oil and the Nn-hexadecylpyridinium salt 10 as a very pale greenish solid.The structures of all prepared pyridinium salts 7-10 were confirmed by 1  coupling between the protons in the positions 1,1 and 4 in the HQBC NMR experiment (Scheme 8).(11) by deprotonation of 1 with 3 equivelant of LiN(SiMe 3 ) 2 according to the literature [32,33].This was immediately followed by treatment with equimolar amounts of 4-(2-trichlorosilylethyl)pyridine ( 5).This method was highly efficient and gave the highest yields of the target product 2 (up to 92-96%).

Capping of Trisilanol
The second route was performed in toluene and involved the reaction of 1 with 4 in the presence of triethylamine (Scheme 10).This method was also found to be quite straightforward and afforded the desired compound 2 in good yields around 68-72%.
The pure product 2 was isolated as a white solid by the recrystallization from a toluene/acetonitrile mixture.The crystalline material is moderately soluble in pentane, toluene, diethyl ether and highly soluble in THF, but insoluble in DMSO, acetonitrile, methanol, and water.The constitution of 2 was confirmed by its 1 H, 13 C, and 29 Si NMR spectra and elemental analysis.

Attempted Quaternization Reactions of Compound 2.
As mentioned above, the final step towards the target compounds 3a and 3b would be the quaternization reaction of 2 with the appropriate alkyl bromides.While such reactions using the model compounds 5 and 6 were successful (cf.Section 2.2), nearly all attempts to carry out N-alkylation reactions with the pyridine-4-ethyl-functionalized POSS derivative 2 failed.Only with iodomethane it was possible to isolate the new N-methylpyridinium iodide 12. Since compound 2 is insoluble in DMSO, acetonitrile and methanol, its N-alkylation with iodomethane (Scheme 11) could not be carried out in these very polar solvents which would have been desirable.While in less polar THF and nonpolar toluene solubility of 2 is very high, the quaternization rates with iodomethane were so low that even under reflux conditions over 5-7 days the product (methylpyridinium salt 12) content in the reaction mixture did not exceed 5-8%.It was, however, found that the yield of 12 could be drastically improved (up to 92%) by prolonged heating of the precursor 2 in neat iodomethane.The resulting white solid was characterized by NMR spectroscopy.It dissolves in CDCl 3 but is insoluble in diethyl ether and hydrocarbons.Unfortunately, n-hexyl bromide and n-hexadecylbromide did not react with 2 even under forcing reaction conditions (e.g., extended heating of 2 in the neat alkylbromide).Thus the target compounds 3a and 3b thus far remain elusive.It also remains to be examined in the course of a future study if compounds like 12 exhibit biocidal properties.

Conclusions
In summarizing the results reported here, a possible synthetic route to new quaternary pyridinium salt-functionalized silsesquioxane (POSS) derivatives has been outlined.Such compounds could be of interest as potential POSS-based biocides.Judging from the initial results reported here it appears that the proposed synthetic route to N-alkylpyridiniumfunctionalized POSS derivatives such as 3a and 3b is principally feasible.Thus far, however, only with iodomethane a clean reaction to give the quaternized product 12 has been achieved, while under the chosen reaction conditions n-hexyl bromide and n-hexadecylbromide failed to react with 2 to give the corresponding N-alkylpyridinium salts.

Preparation of 4-(2-(Trimethoxysilyl)ethyl)pyridine (6)
(Modified from [25]).A 100 mL Schlenk-flask was charged with a small amount (ca.20 mg) of powdered anhydrous AlCl 3 and a stirring bar.Then, 4-(2-(trichlorosilyl)ethyl)pyridine (5, 12.39 g, 51.5 mmol) was added.Trimethylorthoformate (22.5 mL, 21.86 g, 0.206 mol) was then added in small portions of 3-5 mL, each time waiting 25-30 min until the vigorous gas evolution had stopped and the reaction mixture had cooled down to room temperature.(Caution: a more rapid addition of trimethylorthoformate causes intensive warming of the system which could lead to instantaneous splashing).At first, the yellow reaction mixture developed a pink and then an orange-brown color.After that the reaction mixture was refluxed at +80

Preparation of N-Methyl-4-(2-(Trichlorosilyl)ethyl) pyridinium Iodide (7).
In a 100 mL Schlenk flask equipped with a stirring bar, ) was dissolved in 5 mL of acetonitrile and iodomethane (0.5 mL, 0.63 g, 4.4 mmol) was added.The reaction mixture was stirred at room temperature for 1 h, the solvent was removed under vacuum, and the powdery yellow residue was twice washed with 10 mL of diethyl ether to give the pure product.Yield 1.66 g (98%