Gemini pyridinium surfactants: Synthesis and conductometric study of a novel class of amphiphiles

Article abstract of DOI:10.1021/jo034602n

A new series of pyridinium cationic gemini surfactants was prepared by quaternization of the 2,2′-(α,ω-alkanediyl)bispyridines with N-alkylating agents, whose reactivity is briefly discussed. Particularly useful was the use of long-chain alkyl triflates (trifluoromethanesulfonates) for both overcoming the sterical hindrance in the pyridines and obtaining higher synthetic yields. Well-known 4,4′-(α,ω -alkanediyl)bis(1-alkylpyridinium) structures showed narrow temperature ranges for practical applications, due to their high Krafft points, while the new 2,2′-(α,ω)-alkanediyl)bis-(1-alkylpyridinium) series, accounted for good surface active properties. Due to the Krafft points below 0 °C, they could be exploited as solutions in water at any temperature. The characterization of the behavior of the series was performed by conductivity measurements. Some of the proposed structures exhibited unusual surface active behavior, which was interpreted in terms of particular conformational arrangements.

Full text of DOI:10.1021/jo034602n

Gem in i P yr id in iu m Su r fa cta n ts: Syn th esis a n d Con d u ctom etr ic
Stu d y of a Novel Cla ss of Am p h ip h iles1
Pierluigi Quagliotto,*,†,‡ Guido Viscardi,†,‡ Claudia Barolo,†,‡ Ermanno Barni,†,‡
Silvia Bellinvia,†,‡ Emilia Fisicaro,§ and Carlotta Compari§
Dipartimento di Chimica Generale ed Organica Applicata, Universita` di Torino, C.so Massimo d’Azeglio
48, 10125 Torino, Italy, Center for Nanostructured Surfaces and Interfaces, Universita` di Torino,
C.so Massimo d’Azeglio 48, 10125 Torino, Italy, and Dipartimento di Scienze Farmacologiche, Biologiche
e Chimiche Applicate, Universita` di Parma, Parco Area delle Scienze 27/ A, 43100 Parma, Italy
pierluigi.quagliotto@unito.it
Received May 8, 2003
A new series of pyridinium cationic gemini surfactants was prepared by quaternization of the 2,2′-
(R,ω-alkanediyl)bispyridines with N-alkylating agents, whose reactivity is briefly discussed.
Particularly useful was the use of long-chain alkyl triflates (trifluoromethanesulfonates) for both
overcoming the sterical hindrance in the pyridines and obtaining higher synthetic yields. Well-
known 4,4′-(R,ω-alkanediyl)bis(1-alkylpyridinium) structures showed narrow temperature ranges
for practical applications, due to their high Krafft points, while the new 2,2′-(R,ω-alkanediyl)bis-
(1-alkylpyridinium) series, accounted for good surface active properties. Due to the Krafft points
below 0 °C, they could be exploited as solutions in water at any temperature. The characterization
of the behavior of the series was performed by conductivity measurements. Some of the proposed
structures exhibited unusual surface active behavior, which was interpreted in terms of particular
conformational arrangements.
In tr od u ction
application, namely gene therapy, was the final aim of
this research topic.4 Even if several problems connected
with the complexity of the whole transfecting and thera-
peutic process were dealt with, a few results were
obtained. As an example, a first solution to the problem
of the rapid elimination of complexes from the blood-
stream was achieved, thus permitting a longer circulation
time for lipoplexes.5,6 However, a clear structure-activity
relationship about the transfecting performances of cat-
ionic surfactants is not available. In fact, a deeper study
about the aggregation behavior of these surfactants and
their DNA complexation ability, joined to a better char-
acterization of the surfactant/DNA complex, could give
important information on the different steps involved in
the transfection process.
The application of cationic surfactants in high technol-
ogy fields has become very important. Among these
applications, the use of surfactant aggregates, mainly
vesicles, to form complexes with DNA has impressively
grown. Starting from 1987 with the experiments of
Felgner and co-workers,2 these complexes, normally
known as lipoplexes, were used to introduce genetic
material into cells followed by DNA expression, showing
that the use of nonviral carriers was a possible alterna-
tive to normally used modified viruses.2-4 The study of
this application allowed for the use of commercially
available formulated cationic surfactants for the in vitro
transfection of cells. However, the more important in vivo
† Dipartimento di Chimica Generale ed Organica Applicata, Uni-
versita` di Torino.
Due to our experience in the synthesis and character-
ization of cationic surfactants,7-12 we envisaged the
‡ Center for Nanostructured Surfaces and Interfaces, Universita` di
Torino.
§ Dipartimento di Scienze Farmacologiche, Biologiche e Chimiche
Applicate, Universita` di Parma.
(5) Litzinger, D. C.; Brown, J . M.; Wala, I.; Kaufman, S. A.; Van, G.
Y.; Farrell, C. L.; Collins, D. Biochim. Biophys. Acta 1996, 1281, 139-
149.
(6) Hong, K.; Zheng, W.; Baker, A.; Papahadjopoulos, D. FEBS Lett.
1997, 400, 233-237.
(7) Fisicaro, E.; Compari, C.; Ro´øycka-Roszak, B.; Viscardi, G.;
Quagliotto, P. Curr. Top. Colloid Interface Sci. 1997, 2, 54-68.
(8) Viscardi, G.; Quagliotto, P.; Barolo, C.; Savarino, P.; Barni, E.;
Fisicaro, E. J . Org. Chem. 2000, 65, 8197-8203.
(1) Presented in part at the 5th World Surfactants Congress,
Florence, Italy, May 29-J une 2, 2000. Barni, E.; Barolo, C.; Compari,
C.; Fisicaro, E.; Quagliotto, P.; Viscardi, G. World Surfactants Congress,
5th 2000, 1044-1048.
(2) Felgner, P. L.; Gadek, T. R.; Holm, M.; Roman, R.; Chan, H. W.;
Wenz, M.; Northrop, J . P.; Ringold, G. M.; Danielsen, M. Proc. Natl.
Acad. Sci. U.S.A. 1987, 84, 7413-7417.
(3) (a) Meekel, A. A. P.; Wagenaar, A.; Kroeze, J . E.; Haadsma, P.;
Bosgraaf, B.; Stuart, M. C. A.; Brisson, A.; Ruiters, M. H. J .; Hoekstra,
D.; Engberts, J . B. F. N. Eur. J . Org. Chem. 2000, 665-673. (b) for an
extensive recent review, see: Kirby, A. J .; Camilleri, P.; Engberts, J .
B. F: N.; Feiters, M.; Nolte, R. J . M.; Soderman, O.; Bergsma, M.; Bell,
P. C.; Fielden, M. L.; Garcia Rodriguez, C.; Gue´dat, P.; Kremer, A.;
McGregor, C.; Perrin, C.; Ronsin, G.; van Eijk, M. C. P. Angew. Chem.,
Int. Ed. 2003, 42, 1448-1457.
(9) Fisicaro, E.; Pelizzetti, E.; Viscardi, G.; Quagliotto, P.; Trossarelli,
L. Colloids Surf. A: Physicochem. Eng. Aspects 1994, 84, 59-70.
(10) Fisicaro, E.; Ghiozzi, A.; Pelizzetti, E.; Viscardi, G.; Quagliotto,
P. J . Colloid Interface Sci. 1996, 182, 549-557.
(11) Fisicaro, E.; Ghiozzi, A.; Pelizzetti, E.; Viscardi, G.; Quagliotto,
P. J . Colloid Interface Sci. 1996, 184, 147-154.
(12) Ardizzone, S.; Bianchi, C. L.; Drago, P.; Mozzanica, D.; Qua-
gliotto, P.; Savarino, P.; Viscardi, G. Colloids Surf. A: Physicochem.
Eng. Aspects 1996, 113, 135-144.
(4) Miller, A. D. Angew. Chem., Int. Ed. 1998, 37, 1768-1785.
10.1021/jo034602n CCC: $25.00 © 2003 American Chemical Society
Published on Web 09/11/2003
J . Org. Chem. 2003, 68, 7651-7660
7651

Quagliotto et al.
CHART 1. P r od u cts P r ep a r ed in th e P r esen t
possibility of both in vitro and in vivo applications of our
already prepared structures. With this goal in mind, we
started a wide investigation involving organic synthesis,
physicochemical characterization, biological evaluations,
and molecular modeling. This article is the first report
on the organic synthesis and the early physicochemical
characterization of novel pyridinium gemini cationic
surfactants. Their unusual amphiphilic properties and
the tunability of structure, by variation of chain hydro-
phobicity, spacer length, and counterion, makes them
attractive for transfection purposes.
Wor k
In the past decade, a growing interest was devoted to
new structures called gemini surfactants. The challenge,
at first, was mainly directed to the synthesis and
fundamental surface activity characterization of these
structures, made up of two typical surfactant molecules
connected by a spacer at the headgroup level. Recently,
many studies concerning the practical use of gemini
molecules appeared in the literature, showing that the
first speculative interests were followed by practical
applications, due to their unusual properties:7,13,14 (i) the
lower critical micellar concentration (cmc) and C20 (the
concentration at which a reduction of the surface tension
of 20 mN/m is attained) values, (ii) the better adsorption
behavior at both the air/water and the solid/water
interfaces with respect to their monomers, and (iii) the
tendency to form micelles of different shapes and dimen-
sions (i.e., spherical, rodlike, threadlike, vesicles), even
at low concentration, when compared with similar normal
surfactants.
In the field of cationic surfactants, the structures
bearing an ammonium headgroup were more widely
studied because of their fast and simple synthetic
pathway.13,15-19 However, different and more tailored
structures were prepared in recent years, showing the
great interest connected with these kinds of molecules.20
Due to our activity in the field of cationic surfactants and,
in particular, of pyridinium ones,8-12 we started a syn-
thetic project about the gemini pyridinium surfactants,
reported in the present paper. Many papers are present
in the literature that refer to viologen-type structures,21,22
similar to 1-3 (Chart 1). Their positional isomers dif-
fering for the connection site of the spacer on the pyridine
ring are lacking.
terms belonging to the 1,1′-dialkyl-4,4′-bispyridinium or
4,4′-(R,ω-alkanediyl)bis(1-alkylpyridinium) series. The
results showed low surface activity and a low tendency
to form micelles at room temperature or below, thus
reducing the range of applicability. In view of these
preliminary results, the preparation of a 2,2′-(R,ω-al-
kanediyl)bis(1-alkylpyridinium) series (structures 11-
15), differing for the positional isomerism of the spacer
on the pyridinium rings, came out. However, their
synthesis was not easy, as for former structures, owing
to steric hindrance of the pyridines in the 2-position,
which opposes the quaternization reaction that was
studied using different alkylating agents: iodide, meth-
anesulfonate, and trifluoromethansulfonate esters. The
triflates appeared the best for both the lower reaction
temperatures and the higher yields attainable. The
amphiphilic characterization of the products showed
peculiar behavior of some structures which, in our
opinion, is due to a particular conformational arrange-
ment.
The information about the behavior of these surfactant
molecules3,21 is scarce. Thus, the study was focused on
the preparation and preliminary characterization of a few
(13) Zana, R. Curr. Opin. Colloid Interface Sci. 1996, 1, 566-571
and references therein.
(14) (a) Rosen, M. J .; Tracy, D. J . J . Surfactants Deterg. 1998, 1,
547-554 and references therein. (b) For a more recent review, see:
Menger, F. M.; Keiper, J . S. Angew. Chem., Int. Ed. 2000, 39, 1906-
1920.
(15) Zana, R.; Benrraou, M.; Rueff, R. Langmuir 1991, 7, 1072-
1075.
Exp er im en ta l Section
Gen er a l P r oced u r es a n d Ma ter ia ls. See the Supporting
Information.
Syn t h esis. 1,1′-Did od ecyl-4,4′-d ip yr id in iu m ch lor id e
(1). In a three-necked round-bottom flask, purged with argon,
were introduced 4,4′-dipyridyl (15.6 g, 0.1 mol), dodecyl
chloride (204.57 g, 1.0 mol), and DMF (100 mL). The solution
was refluxed for 8 h and cooled at 0 °C with formation of a
beige solid, which was crystallized from acetone/methanol:
yield 34.55 g (62%); mp undetermined, dec above 280 °C; Rf
0.1 on silica (eluent methanol/acetic acid/chloroform, MAC 20:
10:70); UV (ethanol) λmax 266 nm, logꢀ 4.10; 1H NMR (DMSO-
d6) δ 0.87 (t, 6H, 2CH3), 1.26 (m, 36H, 18CH2), 1.99 (m, 4H,
N+CH2CH2), 4.71 (t, 4H, N+CH2), 8.82 (d, 2H, H′′), 9.42 (d,
2H, H′); FT-IR (KBr) 3048, 2916, 2848, 1636, 1570, 1512, 1468,
(16) Alami, E.; Levy, H.; Zana, R. Langmuir 1993, 9, 940-944.
(17) Alami, E.; Beinert, G.; Marie, P.; Zana, R. Langmuir 1993, 9,
1465-1467.
(18) Frindi, M.; Michels, B.; Levy, H.; Zana, R. Langmuir 1994, 10,
1140-1145.
(19) Danino, D.; Talmon, Y.; Zana, R. Langmuir 1995, 11, 1448-
1456.
(20) Pe´rez, L.; Torres, J . L.; Manresa, A.; Solans, C.; Infante, M. R.
Langmuir 1996, 12, 5296-5301.
(21) Lee, D. K.; Kim, Y. I.; Kwon, Y. S.; Kang, Y. S.; Kevan, L. J .
Phys. Chem. B. 1997, 101, 5319-5323.
(22) Vassiltsova, O. V.; Chuvilin, A. L.; Parmon, V. N. J . Photochem.
Photobiol. A 1999, 125, 127-134.
7652 J . Org. Chem., Vol. 68, No. 20, 2003

Gemini Pyridinium Surfactants
1174, 842, 786 cm-1. Anal. Calcd for C34H58Cl2N2: C, 72.18;
H, 10.33; N, 4.95. Found: C, 72.09; H,10.37; N, 4.93.
1,1′-Did od ecyl-4,4′-tr im eth ylen ebisp yr id in iu m Ch lo-
r id e (2).
1,3-Bis(2-p yr id yl)p r op a n e (5).24 In a three-necked flask,
previously purged with argon, was introduced 2-methylpyri-
dine (431 g; 458 mL; 4.64 mol), followed by the addition of
sodium (2.4 g; 0.104 mol) in small portions under magnetic
stirring. After the sodium dissolution (which took about 1 h),
the temperature was raised to 120 °C and 2-vinylpyridine (97.5
g; 100 mL; 0.93 mol) was added dropwise to the brown solution.
After 4 h, the reaction was stopped and cooled at 0 °C.
Methanol (50 mL) was added to quench the reaction giving a
red solution which was distilled to remove the excess of
methanol and 2-methylpyridine. Distillation of the resulting
oil gave a pale yellow oil: yield 85.05 g (46%); bp 97 °C/3 ×
10-5 Torr (lit.25 bp 135 °C/25 Torr); Rf 0.36 on silica (eluent:
butanol/acetic acid/water BAW 4:1:5, organic phase); UV
(ethanol) λmax 206, 263, 331 nm, log ꢀ 4.14, 3.90, 2.82; 1H NMR
(DMSO-d6) δ 2.07 (m, 2H, CH2), 2.76 (t, 4H, 2CH2), 7.19 (d,
2H, Hmeta), 7.26 (d, 2H, H′meta), 7.62(dt, 2H, Hpara), 8.47 (d, 2H,
Horto); FT-IR (KBr) 3008, 2924, 2858, 1590, 1568, 1474, 1434,
766 cm-1; GC-MS (m/z) 198 (M+). Anal. Calcd for C13H14N2:
C, 78.75; H, 7.12; N, 14.13. Found: C, 78.70; H, 7.14; N, 14.11.
1,4-Bis(2-p yr id yl)bu ta n e (6). The procedure of Richards
et al.26 was followed. The addition of lithium metal was
accomplished under argon in a glovebag to ensure the best
anhydrous conditions: yield 35.5 g (46%); bp 109-110 °C/1.5
× 10-4 Torr (lit.26 bp 140 °C/1 Torr, lit.25 bp 125 °C/0.2 Torr);
Rf 0.28 on silica (eluent BAW 4:1:5); UV (ethanol) λmax 257,
262, 269 nm, log ꢀ 3.76, 3.83, 3.69; 1H NMR (DMSO-d6) δ 1.70
(m, 4H, PyCH2CH2CH2CH2Py), 2.75 (t, 4H, PyCH2CH2CH2CH2-
Py), 7.18 (dd, 2H, Hmeta), 7.23 (d, 2H, H′meta), 7.67 (dt, 2H, Hpara),
8.47 (d, 2H, Horto); FT-IR (film) 2945, 2920, 2852, 1582, 1564,
1475, 1458, 1429, 1147, 1049, 771, 750, 734 cm-1; GC-MS
(m/z) 212 (M+). Anal. Calcd for C14H16N2: C, 79.21; H, 7.60;
N, 13.20. Found: C, 79.22; H, 7.57; N, 13.15.
The 1,3-bis(4-pyridyl)propane (25.00 g, 0.126 mol) was
dissolved in DMF (10 mL) in a three-necked round-bottom
flask. The temperature was raised to 150-155 °C, and the
dodecyl chloride (258.04 g, 297 mL, 1.26 mol) was slowly
dropped into the solution. The reaction was stopped after 12
h, and the DMF was evaporated under vacuum. The residue
was filtered, suspended in ethyl ether under stirring, obtaining
a brown solid, which was crystallized two times from hot
acetonitrile/toluene to give white-beige crystals: yield 63.0 g
(82%); mp 65-66 °C; Rf 0.1 on silica (eluent MAC 20:10:70);
UV (ethanol) λmax 230, 257 nm, log ꢀ 4.21, 3.88; 1H NMR
(DMSO-d6) δ 0.85 (t, 6H, 2CH3), 1.23 (m, 36H, 18CH2), 1.89
(m, 4H, N+CH2CH2), 2.12 (m, 2H, PyCH2CH2CH2Py), 2.96 (t,
4H, PyCH2CH2CH2Py), 4.54 (t, 4H, N+CH2), 8.04 (d, 2H, H′′),
8.97 (d, 2H, H′); FT-IR (KBr) 2954, 2918, 2848, 1640, 1570,
1516, 1468, 1176, 842, 726, cm-1. Anal. Calcd for C37H64
-
Cl2N2: C, 73.11; H, 10.61; N, 4.61. Found: C, 73.01; H, 10.57;
N, 4.63.
1,1′-Did od ecyl-4,4′-t r im et h ylen eb isp yr id in iu m Br o-
m id e (3). The 1,3-bis(4-pyridyl)propane (10.0 g, 0.05 mol) was
dissolved in DMF (15 mL) in a three-necked round-bottom
flask. The temperature was raised to 150-155 °C, and the
dodecyl bromide (125.61 g, 120 mL, 0.5 mol) was slowly
dropped into the solution. The reaction was stopped after 6 h,
and the DMF was evaporated under vacuum. The residue was
filtered, suspended in ethyl ether, and crystallized two times
from acetonitrile/toluene giving white-gray crystals: yield
25.80 g (73.5%); mp 65 °C; Rf 0.1 on silica (eluent MAC 20:
10:70); UV (ethanol) λmax 229, 257 nm, log ꢀ 4.28, 3.88; 1H NMR
(DMSO-d6) δ 0.85 (t, 6H, 2CH3), 1.26 (m, 36H, 18CH2), 1.92
(m, 4H, N+CH2CH2), 2.12 (m, 2H, PyCH2CH2CH2Py), 2.96 (t,
4H, PyCH2CH2CH2Py), 4.57 (t, 4H, N+CH2), 8.07 (d, 2H, H′′),
9.01 (d, 2H, H′); FT-IR (KBr) 2954, 2911, 2847, 1636, 1570,
1514, 1468, 1174, 844, 786 cm-1. Anal. Calcd for C37H64Br2N2:
C, 63.78; H, 9.26; N, 4.02. Found: C, 63.86; H, 9.22; N, 4.00.
Gen er a l P r oced u r e for th e Syn th esis of Ba ses 4, 7, a n d
8. In a three-necked round-bottom flask equipped with gas
insertion adapter were introduced under argon anhydrous
ethyl ether (300 mL) and a 2.5 M solution of butyllithium in
hexane (100 mL, 0.25 mol). The temperature was lowered to
-20 °C with a dry ice-acetone bath, and the 2-methylpyridine
was added dropwise under stirring, giving a red orange
solution. The proper R,ω-dibromoalkane was then added
dropwise, under stirring. The solution was allowed to react
for 50 min at -20 °C and allowed to return to room temper-
ature, reacting for further 2 h. The reaction was quenched by
the dropwise addition of water (30 mL). The ethereal solution
was extracted with 1:2 hydrogen chloride solution. The aque-
ous layer was separated and treated with a sodium hydroxide
aqueous solution, until basic reaction, when an orange-reddish
oil separated. The oil was extracted three times with ethyl
ether. The ether solution of the base was dried with anhydrous
sodium sulfate, filtered, and evaporated, leaving a crude
product which was purified by distillation or column flash
chromatography (FC).23
1,8-Bis(2-p yr id yl)octa n e (7). The crude product obtained
following the general procedure was purified by flash chro-
matography on alumina with petroleum ether/ethyl acetate
95:5, obtaining a pale oil: yield 15.02 g (70%); Rf 0.41 on silica
(eluent BAW 4:1:5); UV (ethanol) λmax 257, 262, 269 nm, log ꢀ
1
3.81, 3.87, 3.73; H NMR (DMSO-d6) δ 1.28 (bs, 8H, 4 CH2),
1.65 (m, 4H, PyCH2CH2), 2.70 (t, 4H, PyCH2), 7.18 (td, 2H,
H
H
meta), 7.23 (d, 2H, H′meta), 7.68 (td, 2H, Hpara), 8.46 (d, 2H,
orto); FT-IR (film) 2928, 2854, 1591, 1475, 1435, 1265, 1150,
1051, 897, 742, 704, 626 cm-1; GC-MS (m/z) 268 (M+). Anal.
Calcd for C18H24N2: C, 80.55; H, 9.01; N, 10.44. Found: C,
80.32; H, 8.97; N, 10.41.
1,12-Bis(2-p yr id yl)d od eca n e (8). The crude product ob-
tained following the general procedure was purified by flash
chromatography on alumina with petroleum ether/ethyl ac-
etate 90:10, obtaining a pale oil: yield 22.33 g (55%); mp 39-
42; Rf 0.44 on silica (eluent BAW 4:1:5); UV (ethanol) λmax 257,
262, 269 nm, log ꢀ 3.76, 3.81, 3.68; 1H NMR (DMSO-d6) δ 1.25
(m 16H, 8CH2), 1.65 (m, 4H, PyCH2CH2), 2.70 (t, 4H, PyCH2),
7.17 (dd, 2H, Hmeta), 7.22(d, 2H, H′meta), 7.67 (td, 2H, Hpara),
8.46 (dd, 2H, Horto); FT-IR (film) 2935, 2850, 1587, 1561, 1471,
1436, 1276, 1259, 1145, 1051, 99, 750 cm-1; GC-MS (m/z) 324
(M+). Anal. Calcd for C22H32N2: C, 81.43; H, 9.94; N, 8.63.
Found: C, 81.34; H, 9.98; N, 8.59.
Dod ecylm eth a n su lfon a te (9). A general procedure27 was
applied for the synthesis of the dodecylmethanesulfonate. All
characterization data (mp, FT-IR, and NMR) agreed with
literature data.
1,2-Bis(2-p yr id yl)eth a n e (4). The crude product obtained
following the general procedure, using 1,2-dibromoethane as
the dibromoalkane reactant, was distilled under vacuum,
obtaining a pale oil: yield 25.78 g (56%); bp 73 °C/8 × 10-5
Torr; Rf 0.6 on silica (eluent MAC 20:10:70); UV (ethanol) λmax
Dod ecyl Tr iflu or om eth a n su lfon a te (10).28,29 To a solu-
tion of trifluoromethanesulfonic anhydride in dry dichlo-
1
263 nm, log ꢀ 3.81; H NMR (DMSO-d6) δ 3.13 (s, 4H, 2CH2),
(24) Sartoris, N. E.; Pines, H. J . Org. Chem. 1969, 34, 2119-2122.
(25) Bo¨nnemann, H.; Brinkmann, R. Synthesis 1975, 600-602.
(26) Richards, D. H.; Scilly, N. F.; Williams, F. J . Chem. Ind. 1970,
1298-1299.
(27) Marvel, C. S. and Sekera, V. C. Organic Syntheses; Horning,
E. C., Ed.; Wiley: New York, 1955; Collect. Vol. III, pp 366-369.
(28) Beard, C. D.; Baum, K.; Grakauskas, V. J . Org. Chem. 1973,
21, 3673-3677.
(29) Fife, W. K.; Ranganathan, P.; Zeldin, M. J . Org. Chem. 1990,
55, 5610-5613.
7.20 (dt, 2H, Hmeta), 7.25 (d, 2H, H′meta), 7.67 (dt, 2H, Hpara),
8.50 (dd, 2H, Horto); FT-IR (KBr) 3062, 2928, 2856, 1581, 1567,
1472, 1450, 787 cm-1; GC-MS (m/z) 184 (M+). Anal. Calcd for
C
12H12N2: C, 78.23; H, 6.57; N, 15.21. Found: C, 78.16; H,
6.60; N, 15.17.
(23) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 43, 2923-
2925.
J . Org. Chem, Vol. 68, No. 20, 2003 7653

Quagliotto et al.
romethane, at 0 °C under argon, was added dropwise an
equimolar solution of dodecyl alcohol and pyridine in dichlo-
romethane. The reaction was continued for 15 min and
successively the solution was quenched with water, extracted
repeatedly with water and separated. The organic phase was
then dried with Na2SO4 and evaporated. The resulting oil was
used immediately for the quaternization reactions. The product
was obtained in 85% yield: 1H NMR (CDCl3) δ 0.87 (t, 3H,
CH3), 1.26 (m, 16H, 8CH2), 1.41 (m, 2H, CF3SO3CH2CH2CH2),
1.82 (m, 2H, CF3SO3CH2CH2), 4.53 (t, 2H, CF3SO3CH2).
Gen er a l P r oced u r e for th e Qu a ter n iza tion of Ba ses
(t, 4H, N+CH2), 7.98 (t, 2H, Hmeta), 8.05 (d, 2H, H′meta), 8.48 (t,
2H, Hpara), 9.00 (d, 2H, Horto); FT-IR (KBr)3088, 2926, 2854,
1630, 1580, 1514, 1466, 1260, 1224, 1160, 1032, 782, 640 cm-1
.
Anal. Calcd for C44H74F6N2O6S2: C, 58.38; H, 8.24; N, 3.09.
Found: C, 58.42; H, 8.27; N, 3.06.
1,1′-Did od ecyl-2,2′-d od eca m eth ylen ebisp yr id in iu m Di-
tr iflu or om eth a n esu lfon a te (15a ). The solvent was evapo-
rated under vacuum and the resulting oil was washed several
times with petroleum ether. A white crystalline solid was
obtained: yield 2.30 g (94%); mp 93-96 °C; Rf 0.31 on silica
1
(eluent BAW 4:1:5); UV (ethanol) λmax 269 nm, log ꢀ 4.09; H
4-8 w ith th e Dod ecyl Tr iflu or om eth a n esu lfon a te.29,30
A
NMR (DMSO-d6) δ 0.85 (t, 6H, 2CH3), 1.24 (m, 52H, 26CH2),
1.70 (m, 4H, PyCH2CH2), 1.84 (m, 4H, N+CH2CH2), 3.08 (t,
4H, PyCH2), 4.54 (t, 4H, N+CH2), 7.98 (t, 2H, Hmeta), 8.04 (d,
2H, H′meta), 8.49 (t, 2H, Hpara), 8.98 (d, 2H, Horto); FT-IR (KBr)
3040, 2912, 2852, 1627, 1576, 1516, 1468, 1266, 1225, 1157,
1034, 640 cm-1. Anal. Calcd for C48H82F6N2O6S2: C, 59.97; H,
8.60; N, 2.91. Found: C, 56.01; H, 8.57; N, 2.94.
solution of an excess of the dodecyl trifluoromethanesulfonic
ester (1.25 mol per mol of nitrogen to be quaternized) in
anhydrous chloroform, kept under argon, was stirred at reflux,
and a solution of the proper base in chloroform was added
dropwise. The reaction was continued for 1.5 h. The reaction
was worked up as described in each of the following entries.
Ion ic Exch a n ge Gen er a l P r oced u r e for Obta in in g
P r od u ct s 11-15 w it h Differ en t Cou n t er ion s. The ionic
exchange was performed using a strong anionic exchange
resin, conditioned by suspension in water for a few hours,
packing in a glass column, and treating in the order with a
solution of NaOH 4%, water until neutral reaction of the
effluent, aqueous 10% solution of sodium salt of proper
counterion, and water to eliminate the residual salt.
Before use, the resin was unpacked, filtered on a funnel,
and twice suspended in methanol for 0.5 h. Finally, the
methanolic suspension was packed in the column and washed
with 3 volumes of methanol. A methanolic solution of the
product to be subjected to the ionic exchange was put onto the
column and eluted with methanol. After evaporation of
methanol, the product with desired counterion was recovered
and purified by crystallization from acetonitrile/ethyl acetate
mixture.
1,1′-Did od ecyl-2,2′-d im et h ylen eb isp yr id in iu m
Di-
m eth a n esu lfon a te (11b). An excess of dodecylmethane-
sulfonate (28.7 g; 0.18 mol) was heated at 140 °C under
magnetic stirring, and the base 1a (5 g; 2.7‚10-2 mol) was
slowly added in small portions. After the additions were
completed, the mixture was reacted for 10 h. The resulting
dark mixture was suspended a few times in ethyl acetate, until
a brown powder separated. The powder was purified by column
chromatography on alumina (eluent: ethyl acetate/ethanol)
and finally by crystallization from acetonitrile/ethyl acetate.
The crystallization seemed to suffer from the possible partial
instability of the compound in that solvent system: yield 12.95
g. (67%); mp 189-190 °C. Rf 0.13 on silica (eluent BAW
4:1:5); UV (ethanol) λmax 269, 316 nm, logꢀ 4.09, 2.49; 1H NMR
(DMSO-d6) δ 0.86 (t, 6H, 2CH3), 1.25 (m, 36H, 18CH2), 1.89
(m, 4H, N+CH2CH2), 2.30 (s, 6H, CH3SO3), 3.64 (t, 4H, PyCH2),
4.65 (t, 4H, N+CH2), 8.09 (t, 2H, Hmeta), 8.21 (d, 2H, H′meta),
8.60 (t, 2H, Hpara), 9.11 (d, 2H, Horto); FT-IR (KBr) 3040, 2918,
2850, 1620, 1580, 1208, 1192, 786, 718 cm-1. Anal. Calcd for
1,1′-Did od ecyl-2,2′-t r im et h ylen eb isp yr id in iu m Dit r i-
flu or om eth a n esu lfon a te (12a ). A white solid, which sepa-
rated from the solution, was filtered off, giving the pure
product as a white powder: yield 2.73 g (80%); mp 200-202
°C; Rf 0.31 on silica (eluent BAW 4:1:5); UV (ethanol) λmax 269,
C
38H68N2O6S2: C, 64.01; H, 9.61; N, 3.93. Found: C, 63.97; H,
9.65; N, 3.95.
1,1′-Did od ecyl-2,2′-t r im et h ylen eb isp yr id in iu m
Di-
1
log ꢀ 4.10; H NMR (DMSO-d6) δ 0.85 (t, 6H, 2CH3), 1.24 (m,
m eth a n esu lfon a te (12b). In a three necked round-bottom
flask was introduced the dodecylmethanesulfonate 9 (26.99 g;
0.102 mol), and the mixture was warmed at 140 °C under
magnetic stirring. The base 5 (5 g; 0.0252 mol) was introduced
by adding small quantities every 10 min. The viscosity of the
mixture became greater as the reaction proceeded, preventing
the ease of stirring. After 6 h, the reaction was stopped. The
dark crude solid mixture was suspended with ethyl acetate
under vigorous stirring. The procedure was repeated a few
times, giving a gray-brownish powder. The product was finally
purified by crystallization with acetonitrile/ethyl acetate.
Alternatively, the product was obtained by anionic exchange
from 12a , giving a white product with the same analytical
data: yield 15.39 g (84%); mp 225-227 °C; Rf 0.19 on silica
36H, 18CH2), 1.86 (m, 4H, N+CH2CH2), 2.19 (m, 2H, PyCH2CH2-
CH2Py), 3.28 (t, 4H, PyCH2CH2CH2Py), 4.57 (t, 4H, N+CH2),
8.02 (t, 2H, Hmeta), 8.10 (d, 2H, H′meta), 8.54 (t, 2H, Hpara), 9.03
(d, 2H, Horto); FT-IR (KBr) 3088, 2922, 2852, 1634, 1576, 1514,
1482, 1468, 1256, 1226, 1166, 1032, 786, 644 cm-1. Anal. Calcd
for C39H64F6N2O6S2: C, 56.10; H, 7.72; N, 3.35. Found: C,
55.98; H, 7.70; N, 3.38.
1,1′-Didodecyl-2,2′-tetr am eth ylen ebispyr idin iu m Ditr i-
flu or om eth a n esu lfon a te (13a ). The reaction mixture was
evaporated, and the resulting oil was treated with ethyl acetate
obtaining a white solid, which was filtered off, giving the pure
product as a white powder: yield 2.10 g (70%); mp 129-131
°C; Rf 0.20 on silica (eluent BAW 4:1:5); UV (ethanol) λmax 269
nm, log ꢀ 4.15; 1H NMR (DMSO-d6) δ 0.85 (t, 6H, 2CH3), 1.23
(m, 36H, 18CH2), 1.87 (m, 8H, N+CH2CH2 + PyCH2CH2), 3.18
(t, 4H, PyCH2), 4.56 (t, 4H, N+CH2), 8.00 (t, 2H, Hmeta), 8.05
(d, 2H, H′meta), 8.50 (t, 2H, Hpara), 9.01 (d, 2H, Horto); FT-IR
(KBr) 3086, 2920, 2851, 1629, 1578, 1513, 1469, 1260, 1226,
1159, 1034, 642 cm-1. Anal. Calcd for C40H66F6N2O6S2: C,
56.58; H, 7.83; N, 3.30. Found: C, 56.50; H, 7.81; N, 3.33.
1,1′-Did od ecyl-2,2′-octa m eth ylen ebisp yr id in iu m Ditr i-
flu or om eth a n esu lfon a te (14a ). The solvent was evaporated
under vacuum and the resulting oil was purified by FC on
alumina with ethyl acetate/methanol 95:5 and 80:20. The
product was an oil, which crystallized on standing: yield 11.76
g (83%); mp 50-52 °C; Rf 0.20 on silica (eluent BAW 4:1:5);
1
(eluent BAW 4:1:5); UV (ethanol) λmax 269 nm, log ꢀ 3.97; H
NMR (DMSO-d6) δ 0.86 (t, 6H, 2CH3), 1.25 (m, 36H, 18CH2),
1.87 (m, 4H, N+CH2CH2), 2.21 (m, 2H, PyCH2CH2CH2Py), 2.34
(s, 6H, CH3SO3), 3.29 (t, 4H, PyCH2CH2CH2Py), 4.59 (t, 4H,
N+CH2), 8.01 (t, 2H, Hmeta), 8.12 (d, 2H, H′meta), 8.54 (t, 2H,
H
para), 9.02 (d, 2H, Horto); FT-IR (KBr) 3040, 2920, 2852, 1630,
1516, 1472, 1208, 1192, 786, 722 cm-1. Anal. Calcd for
39H70N2O6S2: C, 64.42; H, 9.70; N, 3.85. Found: C, 64.40; H,
C
9.66; N, 3.87.
1,1′-Did od ecyl-2,2′-t et r a m et h ylen eb isp yr id in iu m Di-
m eth a n esu lfon a te (13b). A procedure similar to that for 12b
was applied. The suspension of the crude product in ethyl
acetate gave a gray-beige solid, which was crystallized from
acetonitrile/ethyl acetate. The product was also obtained by
anionic exchange from 13a , giving a white product with the
same analytical data: yield 69%; mp 72-73 °C; Rf 0.22 on
silica (eluent BAW 4:1:5); UV (ethanol) λmax 268 nm, log ꢀ 4.15;
1H NMR (DMSO-d6) δ 0.86 (t, 6H, 2CH3), 1.25 (m, 36H,
1
UV (ethanol) λmax 269 nm, log ꢀ 4.13; H NMR (DMSO-d6) δ
0.85 (t, 6H, 2CH3), 1.23 (m, 44H, 22CH2), 1.73 (m, 4H,
PyCH2CH2), 1.85 (m, 4H, N+CH2CH2), 3.08 (t, 4H, PyCH2), 4.54
(30) Ranganathan, P.; Fife, W. K.; Zeldin, M. J . Polym. Sci. A 1990,
28, 2711-2717.
7654 J . Org. Chem., Vol. 68, No. 20, 2003

Gemini Pyridinium Surfactants
18CH2), 1.87 (m, 4H, N+CH2CH2), 1.91 (m, 4H, PyCH2CH2),
2.32 (s, 6H, CH3SO3), 3.22 (t, 4H, PyCH2CH2), 4.60 (t, 4H, N+-
CH2), 8.02 (t, 2H, Hmeta), 8.11 (d, 2H, H′meta), 8.55 (t, 2H, Hpara),
9.07 (d, 2H, Horto); FT-IR (KBr) 3040, 2920, 2852, 1630, 1516,
was freeze-dried from which a white powder was obtained.
Crystallization from acetonitrile/ethyl acetate gave white
crystals: yield 92%; mp 87-90 °C; Rf 0.31 on silica (eluent
BAW 4:1:5); UV (ethanol) λmax 269 nm, log ꢀ 4.04. 1H NMR
(DMSO-d6) δ 0.85 (t, 6H, 2CH3), 1.24 (m, 52H, 26CH2), 1.69
(m, 4H, N+CH2CH2), 1.81 (m, 2H, PyCH2CH2), 3.08 (t, 4H,
PyCH2CH2), 4.55 (t, 4H, N+CH2), 7.98 (t, 2H, Hmeta), 8.05 (d,
2H, H′meta), 8.48 (t, 2H, Hpara), 9.04 (d, 2H, Horto); FT-IR (KBr)
3030, 2917, 2851, 1628, 1511, 1474, 800, 719 cm-1. Anal. Calcd
for C46H82Cl2N2: C, 75.27; H, 11.26; N, 3.82. Found: C, 75.29;
H, 11.21; N, 3.79.
1,1′-Did od ecyl-2,2′-tr im eth ylen ebisp yr id in iu m Dibr o-
m id e (12d ). The product obtained by ionic exchange of 12a
was freeze-dried from which a white powder was obtained.
Crystallization from acetonitrile/ethyl acetate gave white
crystals: yield 88%; mp 153-155 °C; Rf 0.04 on silica (eluent
MAC 20:10:70); UV (ethanol) λmax 208, 272 nm, log ꢀ 4.06, 4.08;
1H NMR (DMSO-d6) δ 0.86 (t, 6H, 2CH3), 1.25 (m, 36H,
18CH2), 1.88 (m, 4H, N+CH2CH2), 2.21 (m, 2H, PyCH2CH2-
CH2Py), 3.40 (t, 4H, PyCH2CH2CH2Py), 4.68 (t, 4H, N+CH2),
8.04 (t, 2H, Hmeta), 8.23 (d, 2H, H′meta), 8.57 (t, 2H, Hpara), 9.14
(d, 2H, Horto); FT-IR (KBr) 3040, 2920, 2852, 1628, 1574, 1514
1466, 790, 722 cm-1. Anal. Calcd for C37H64Br2N2: C, 63.78;
H, 9.26; N, 4.02. Found: C, 63.70; H, 9.29; N, 3.99.
1472, 1208, 1192, 786, 722 cm-1. Anal. Calcd for C40H72
-
N2O6S2: C, 64.82; H, 9.79; N, 3.78. Found: C, 64.76; H, 9.74;
N, 3.80.
1,1′-Did od ecyl-2,2′-oct a m et h ylen eb isp yr id in iu m Di-
m eth a n esu lfon a te (14b). The ionic exchange from 14a gave
a white powder, crystallized from acetonitrile/ethyl acetate:
yield 88%; mp 108-110 °C; Rf 0.27 on silica (eluent BAW 4:1:
5); UV (ethanol) λmax 269 nm, logꢀ 4.20. 1H NMR (DMSO-d6) δ
0.84 (t, 6H, 2CH3), 1.23 (m, 44H, 22CH2), 1. 70 (m, 4H, N+-
CH2CH2), 1.82 (m, 4H, PyCH2CH2), 2.30 (s, 6H, CH3SO3), 3.09
(t, 4H, PyCH2), 4.56 (t, 4H, N+CH2), 7.99 (t, 2H, Hmeta), 8.05
(d, 2H, H′meta), 8.51 (t, 2H, Hpara), 9.02 (d, 2H, Horto); FT-IR
(KBr) 2926, 2854, 1629, 1465, 1210, 1191, 1059, 784 cm-1
.
Anal. Calcd for C44H80N2O6S2: C, 66.29; H, 10.11; N, 3.51.
Found: C, 66.21; H, 10.08; N, 3.49.
1,1′-Did od ecyl-2,2′-d od eca m eth ylen ebisp yr id in iu m Di-
m eth a n esu lfon a te (15b). The ionic exchange from 15a gave
a white powder, crystallized from ethanol/ethyl acetate: yield
94%; mp 119-120 °C; Rf 0.33 on silica (eluent BAW 4:1:5);
1
UV (ethanol) λmax 269 nm, log ꢀ 4.10; H NMR (DMSO-d6) δ
0.85 (t, 6H, 2CH3), 1.25 (m, 52H, 26CH2), 1.70 (m, 4H, N+-
CH2CH2), 1.82 (m, 4H, PyCH2CH2), 2.30 (s, 6H, CH3SO3), 3.08
(t, 4H, PyCH2CH2), 4.55 (t, 4H, N+CH2), 7.98 (t, 2H, Hmeta),
8.05 (d, 2H, H′meta), 8.50 (t, 2H, Hpara), 9.01 (d, 2H, Horto); FT-
1,1′-Did od ecyl-2,2′-t r im et h ylen eb isp yr id in iu m Diio-
d id e (12e). The product obtained by ionic exchange of 12a
was freeze-dried from which a white powder was obtained.
Crystallization from acetonitrile/ethyl acetate gave white
IR (KBr) 2926, 2854, 1629, 1465, 1210, 1191, 1059, 784 cm-1
.
crystals: yield 93%; mp 190-192 °C; R
f 0.04 on silica (eluent
MAC 20:10:70); UV (ethanol) λmax 218, 269 nm, log ꢀ 4.52, 4.21;
1H NMR (DMSO-d6) δ 0.87 (t, 6H, 2CH3), 1.26 (m, 36H,
18CH2), 1.89 (m, 4H, N+CH2CH2), 2.22 (m, 2H, PyCH2CH2-
CH2Py), 3.36 (t, 4H, PyCH2CH2CH2Py), 4.63 (t, 4H, N+CH2),
8.04 (t, 2H, Hmeta), 8.18 (d, 2H, H′meta), 8.57 (t, 2H, Hpara), 9.09
(d, 2H, Horto); FT-IR (KBr) 3034, 2920, 2852, 1628, 1572, 1512,
1468, 784, 721 cm-1. Anal. Calcd for C37H64I2N2: C,56.20; H,
8.16; N, 3.54. Found: C, 56.23; H, 8.12; N, 3.51.
1,1′-Did od ecyl-2,2′-t r im et h ylen eb isp yr id in iu m Di-p -
tolu en esu lfon a te (12f). The product obtained by ionic ex-
change of 12a was freeze-dried from which a white powder
was obtained. Crystallization from acetonitrile/ethyl acetate
gave white crystals: yield 91%; mp 245-248 °C; Rf 0.04 on
silica (eluent MAC 20:10:70); UV (ethanol) λmax 218, 269 nm,
log ꢀ 4.52, 4.21; 1H NMR (DMSO-d6) δ 0.87 (t, 6H, 2CH3), 1.26
(m, 36H, 18CH2), 1.90 (m, 4H, N+CH2CH2), 2.20 (m, 2H,
PyCH2CH2CH2Py), 2.31 (s, 6H, CH3C6H5SO3), 3.30 (t, 4H,
PyCH2CH2CH2Py), 4.59 (t, 4H, N+CH2), 7.12 (d, 4H, H′′), 7.48
(d, 4H, H′), 8.03 (t, 2H, Hmeta), 8.13 (d, 2H, H′meta), 8.55 (t, 2H,
Anal. Calcd for C48H88N2O6S2: C, 67.56; H, 10.39; N, 3.28.
Found: C, 67.51; H, 10.43; N, 3.31.
1,1′-Did od ecyl-2,2′-tr im eth ylen ebisp yr id in iu m Dich lo-
r id e (12c). The product obtained by ionic exchange of 12a was
freeze-dried from which a white powder was obtained. Crystal-
lization from acetonitrile/ethyl acetate gave white crystals:
yield 92%; mp 162-164 °C; Rf 0.15 on silica (eluent BAW 4:1:
1
5); UV (ethanol) λmax 269 nm, log ꢀ 4.17; H NMR (DMSO-d6)
δ 0.86 (t, 6H, 2CH3), 1.25 (m, 36H, 18CH2), 1.86 (m, 4H, N+-
CH2CH2), 2.21 (m, 2H, PyCH2CH2CH2Py), 3.40 (t, 4H, PyCH2-
CH2CH2Py), 4.70 (t, 4H, N+CH2), 8.04 (t, 2H, Hmeta), 8.26 (d,
2H, H′meta), 8.57 (t, 2H, Hpara), 9.19 (d, 2H, Horto); FT-IR (KBr)
3040, 2920, 2852, 1630, 1516, 1472, 786, 722 cm-1. Anal. Calcd
for C37H64Cl2N2: C, 73.11; H, 10.61; N, 4.61. Found: C, 73.05;
H, 10.58; N, 4.58.
1,1′-Did od ecyl-2,2′-t et r a m et h ylen eb isp yr id in iu m Di-
ch lor id e (13c). The product obtained from the ionic exchange
of 13a was freeze-dried from which a white powder was
obtained. Crystallization from acetonitrile/ethyl acetate gave
white crystals: yield 98%; mp 69-71 °C; Rf 0.18 on silica
H
para), 9.06 (d, 2H, Horto); FT-IR (KBr) 3060, 2922, 2852, 1630,
1574, 1514, 1458, 1200, 786 cm-1. Anal. Calcd for C51H78
-
1
(eluent BAW 4:1:5); UV (ethanol) λmax 268 nm, log ꢀ 4.16; H
N2O6S2: C, 69.66; H, 8.94; N, 3.19. Found: C, 69.59; H, 8.97;
N, 3.22.
NMR (DMSO-d6) δ 0.85 (t, 6H, 2CH3), 1.24 (m, 36H, 18CH2),
1.86 (m, 4H, N+CH2CH2), 1.90 (m, 4H, PyCH2CH2), 3.20 (t,
4H, PyCH2CH2), 4.59 (t, 4H, N+CH2), 7.99 (t, 2H, Hmeta), 8.08
(d, 2H, H′meta), 8.52 (t, 2H, Hpara), 9.02 (d, 2H, Horto); FT-IR
(KBr) 3046, 2922, 2852, 1628, 1512, 1464, 784, 722 cm-1. Anal.
Calcd for C38H66Cl2N2: C, 73.40; H, 10.70; N, 4.50. Found: C,
73.45; H, 10.67; N, 4.49.
Con d u ct ivit y Mea su r em en t s. Conductivity measure-
ments were performed on a conductivity meter equipped with
a conductivity cell having cell constant of 0.943 cm-1, as
already reported.8 The addition of concentrated surfactant
solutions by a titrator and the collection of the conductivity
data were performed by using a computer controlled auto-
mated system, working with a homemade program, written
in Quick Basic, available from the author.
Kr a fft P oin t Mea su r em en ts. The Krafft points were first
estimated by visual inspection of the clearing point of surfac-
tant suspensions (50 mg in 2 mL of water). When values
different from 0 °C were evidenced, a careful procedure was
applied using conductivity measurements.31,32
1,1′-Didodecyl-2,2′-octam eth ylen ebispyr idin iu m Dich lo-
r id e (14c). The product obtained from the ionic exchange of
14a was freeze-dried from which a white powder was obtained.
Crystallization from acetonitrile/ethyl acetate gave white
crystals: yield 91%; mp 88-90 °C; Rf 0.22 on silica (eluent
BAW 4:1:5); UV (ethanol) λmax 269 nm, log ꢀ 4.07; 1H NMR
(DMSO-d6) δ 0.85 (t, 6H, 2CH3), 1.24 (m, 44H, 22CH2), 1.70
(m, 4H, N+CH2CH2), 1.82 (m, 4H, PyCH2CH2), 3.10 (t, 4H,
PyCH2CH2), 4.56 (t, 4H, N+CH2), 7.98 (t, 2H, Hmeta), 8.05 (d,
2H, H′meta), 8.51 (t, 2H, Hpara), 9.04 (d, 2H, Horto); FT-IR (KBr)
2920, 2850, 1629, 1576, 1520, 1474, 793, 723 cm-1. Anal. Calcd
for C42H74Cl2N2: C, 74.41; H, 11.00; N, 4.13. Found: C, 74.39;
H, 10.94; N, 4.16.
Resu lts a n d Discu ssion
Syn th esis. The synthesis of some gemini structures,
4,4′-(1,1′-dialkyl)bispyridinium salts with and without
(31) Kamenka, N.; Burgaud, I.; Treiner, C.; Zana, R. Langmuir 1994,
10, 3455-3460.
(32) Zana, R. J . Colloid Interface Sci. 2002, 252, 259-261.
1,1′-Did od ecyl-2,2′-d od eca m eth ylen ebisp yr id in iu m Di-
ch lor id e (15c). The product obtained by ionic exchange of 15a
J . Org. Chem, Vol. 68, No. 20, 2003 7655

Quagliotto et al.
SCHEME 1. P r ep a r a tion of Com p ou n d s 1-3
SCHEME 2. P r ep a r a tion of Ba ses 4-8
SCHEME 3. P r ep a r a tion of th e Gem in i
P yr id in iu m Su r fa cta n ts 11-15
spacer, was reported previously,21,22 but no surface activ-
ity data are available. In the first step of the present
study, we prepared some compounds of the above family
to evaluate their amphiphilic properties. Structures 1-3
were obtained by reacting the proper halide and base in
DMF at reflux (Scheme 1). The relative Krafft points
were determined obtaining values of 50 °C or higher. This
parameter means the minimum temperature at which
micellization occurs; below that temperature, a surfactant
is poorly soluble and it only adsorbs at the surfaces. The
values obtained for surfactants 1-3 suggest that they
are scarcely utilizable for usual purposes. Moreover, as
J ampolsky et al.33 previously stated, the preparation of
starting R,ω-bis(4-pyridyl)alkane bases appeared par-
ticularly difficult. Low yields characterize the preparation
of bases having a spacer equal to eight methylene groups
or longer.
The previous considerations induced us to modify the
spacer linking point on the pyridine ring, and 2,2′-(R,ω-
alkanediyl)bispyridines (structures 4-8) were prepared
on a 50 g minimum scale, following easy and well-stated
methods. The bases having 2, 8, and 12 methylene groups
as spacer were prepared by using n-BuLi, 2-methylpy-
ridine, and an R,ω-dibromoalkane (Scheme 2).34 1,3-Bis-
(2-pyridyl)propane25,35-41 and 1,4-bis(2-pyridyl)butane25,26
were prepared according to the literature, reacting the
sodium salt of 2-methylpyridine with 2-vinylpyridine at
120 °C, and by head-head dimerization of 2-vinylpyri-
dine with lithium metal in THF/hexane 1:4 respectively
(Scheme 2).
quaternization of 2-hindered pyridines with long chain
alkyl halides are reported in the literature.41 In particu-
lar, if the substituent in position 2 is branched, only
methyl was inserted on the nitrogen atom,42,43 and for
longer alkyl moiety more active reactants44 (e.g., fluoro-
sulfates or triflates) or higher pressures were applied.44-46
The polymethylene chain, bearing a pyridine at its end,
has a higher steric hindering than methyl and a reason-
able high degree of conformational freedom, especially
when the number of methylene groups is raised. Prob-
ably, the side chain in the 2-position of each pyridine ring
makes the nitrogen atom hindered to the SN2 Menschut-
kin reaction. Besides, the rate of the reaction is severely
affected by the sterical hindrance exerted by the alkyl
chain of the quaternizing reagent on the pentacoordinate
center which is normally postulated for the SN2 reac-
tion.46,47
Alkyl mesylates and triflates were considered whose
reactivity is higher than halides.48 The alkyl mesylates
(Scheme 3) were adopted successfully to quaternize the
shorter spacer bases 4-6 (n ) 2-4) with dodecyl chain
in solvent-free reactions at 140 °C, while they failed, in
the same conditions, for bases having longer spacers and
hexadecyl chain of alkylating agent. An accurate tem-
perature control and the addition of proper base to
alkylating agent were decisive factors for good yields.
When dodecyl methanesulfonate was added to base or
both reactants were added and warmed, the yield de-
The quaternization reaction with alkyl halides, as used
for obtaining surfactants 1-3, was unsuccessful. Only
monoquaternized product was obtained using alkyl iodide
as alkylating reagent. Analogously, few examples of
(42) Harris, G. H.; Shelton, R. S.; Van Campen, M. G.; Andrews, E.
(33) J ampolsky, L. M.; Baum, M.; Kaiser, S.; Sternbach, L. H.;
Goldberg, M. W. J . Am. Chem. Soc. 1952, 74, 5222-5224.
(34) Bianchetti, G. Il Farmaco 1956, 11, 346-356.
(35) Magnus, G.; Levine, R. J . Am. Chem. Soc. 1956, 78, 4127-4130.
(36) Magnus, G.; Levine, R. J . Org. Chem. 1957, 22, 270-273.
(37) Michalski, J .; Zajac, H. J . Chem. Soc. 1963, 593-597.
(38) Parks, J . E.; Wagner, B. E.; Holm, R. H. J . Organomet. Chem.
1973, 56, 53-66.
R.; Schumann, E. L. J . Am. Chem. Soc. 1951, 73, 3959-3963.
(43) Kosower, E. M.; Skorcz, J . A. J . Am. Chem. Soc. 1960, 82, 2195-
2203.
(44) Okamoto, Y.; Lee, K. I. J . Am. Chem. Soc. 1975, 97, 4015-
4018.
(45) Le Noble, W. J .; Asano, T. J . Am. Chem. Soc. 1975, 97, 1778-
1782.
(46) Le Noble, W. J .; Ogo, Y. Tetrahedron 1970, 26, 4119-4124.
(47) Gallo, R.; Roussel, C.; Berg, U. Adv. Heterocycl. Chem. 1988,
43, 173-299.
(48) Stang, P. J .; Hanack, M.; Subramanian, L. R. Synthesis 1982,
85-126.
(39) Ames, D. E.; Archibald, J . L. J . Chem. Soc. 1962, 1475-1481.
(40) Anderson, J . D.; Baizer, M.; Prill, E. J . J . Org. Chem. 1965, 30,
1645-1647.
(41) Edwards, P. N. Pyridine derivatives. US 3,875,175, 1975.
7656 J . Org. Chem., Vol. 68, No. 20, 2003

Gemini Pyridinium Surfactants
creased to about 15% and tars were produced. Nitroben-
zene was used for the reaction of dodecyl mesylate with
the 1,8-bis(2-pyridyl)octane 7 at 140 °C, but the yields
were low and the quantity of tars was substantial.
Probably, in these conditions the elimination reaction on
the alkylating reagent, promoted by the presence of the
bispyridine base, is competitive and gives the corre-
sponding alkene product. The polymerization of such
alkene could be the source of the observed tars.
The use of triflates appeared more general as an
approach. All the bases 4-8 were quaternized using
dodecyl triflate and the preliminary test with the hexa-
decyl one appeared successful.49 It is noticeable that in
the literature the use of long alkyl triflates is limited to
unhindered systems.28,29,48,50,51 This is the first time that
the sterical hindrance at the R-position of the pyridine
is overcome with long alkyl triflates.
The dodecyl triflate (10) was prepared by modifying
the method used for the preparation of alkyl triflates,
obtaining very good yields (80-94%).29 The literature
procedure appeared to be quite tedious and time wasting.
The product was recovered by simply washing the organic
solution with water to eliminate pyridinium triflate,
giving a straightforward and efficient protocol. The
quaternization step (Scheme 3) was performed by adding
a solution of the base to the alkyl triflate in chloroform
at 60 °C for about 1.5 h. The separation step was easier,
and the products were all solid white crystals except for
the compound 14a , which crystallized only after pro-
longed standing (a few months).
A particular comment has to be devoted to compound
11b , obtained by synthesis with dodecyl methane-
sulfonate. It appeared particularly unstable during the
purification step, by either flash chromatography and
crystallization, with substantial loss of yield. However,
by crystallization a small sample useful for amphiphilic
characterization was obtained. Its instability prevented
the exchange of counterion by usual resin, in either water
or methanol. Due to this problem, the other products of
the series 11 were not prepared. The instability could be
due to the short spacer. The inductive effect of the two
positive charges of the pyridinium rings, increasing the
acidity of the methylene protons, could promote the
relative deprotonation by reaction with residual amine
centers of the ionic exchanger resin. Analogously, Ko-
sower et al.43 invoked the acidity of these protons to
explain the formation of 4-ethylidene-1-methyl-1,4-dihy-
dropyridine as a byproduct during the preparation and
crystallization of 4-ethyl-1-methylpyridinium iodide.
Attempts to protonate these centers with a proper acid
(HCl for exchange to chloride anion) did not give suc-
cessful results.
Am p h ip h ilic Ch a r a cter iza tion . The gemini surfac-
tants 1-3 and 11-15 from now on will be referred to by
an acronym, like the simple “n-s-n” recently used by
Zana et al.13,15 for the ammonium gemini surfactants. As
an example, the compound 3 should be named 12-Py(4)-
3-(4)Py-12 2Br, where the first number describes the
number of carbon atoms of alkyl chain to pyridinium
nitrogen, Py stands for the pyridinium ring, (4) indicates
the spacer connection position on the pyridinium ring, 3
is the number of methylenes in the polymethylenic
spacer, and 2Br stands for the two bromide counterions.
In the same way, compounds 11-15 should be identified
by the general name: n-Py(m)-s-(m)Py-n 2X, where n,
m, s and X have the already specified general meaning.
Kr a fft P oin ts. The Krafft points of all of the products
were determined, showing that the insertion of spacer
moiety in position 2 of piridinium ring increases the
solubility of gemini surfactant in water. An exception was
observed for surfactants 12a -15a having trifluoro-
methanesulfonate anion as counterion, for which the
Krafft point ranged between 80 and 95 °C. Different
counterions were considered in the series 12, having
three methylenes as spacer. An increase of halide polar-
izability determines a raising of the Krafft point, while
in the presence of p-toluenesulfonate the value of this
parameter raised to 79 °C. The last effect could be due
to both hydrophobic character and easier stacking of
benzene moiety between two pyridinium headgroups.
This preliminary study showed that when it is advisable
to work at low temperature and in the presence of
micelles, the chloride or the methanesulfonate counter-
ions are needed, while for different applications other
counterions can be suitable, e.g. bromide, starting just
above room temperature.
CMC Mea su r em en ts. The cmc and the degree of
counterion binding of the compounds under exam were
determined by conductivity (see Table 1) by using a
computer-controlled automated system to improve both
the data collection and the data analysis steps. In fact,
by the use of this system, it was possible to control when
the conductivity reached the equilibrium value for any
surfactant addition. Besides, when working with gemini
surfactants, it was common to obtain very smooth
conductivity vs C plots, where the transition between the
pre- and post-micellar regimes, was occurring on a wide
concentration range (see examples in the Supporting
Information) and difficult to be precisely determined. In
such a case, the classical method,56 in which the cmc is
determined by the intersection of the lines fitted in the
diluted and concentrated regions before and after the
cmc, is sometimes difficult to apply, since the transition
between the two linear regimes occurs gradually. In fact,
when trying to find the linear regimes in the classical
method, one assume that some points can be used or not
The ionic exchange step in methanol was performed
to obtain the surfactants 12-15 having different coun-
terions. The experimental procedure here described
proved them pure enough to be characterized even by
surface tension,49 which is commonly known as a tech-
nique severely affected by very small amounts of
impurities.52-55
(53) Rosen, M. J .; Dahanayake, M.; Cohen, A. W. Colloids Surf.
1982, 5, 159-172.
(54) Rosen, M. J .; Cohen, A. W.; Dahanayake, M.; Hua, X. Y. J . Phys.
Chem. 1982, 86, 541-545.
(55) Mysels, K. J . Langmuir 1996, 12, 2325-2326.
(56) Carpena, P.; Aguiar, J .; Bernaola-Galva´n, P.; Carnero Ruiz, C.
Langmuir 2002, 18, 6054-6058 and references therein. For the
determination of counterion binding, see also: (a) Bakshi, M. S. J .
Chem. Soc., Faraday Trans. 1993, 89, 4323-4326. (b) Zana, R.; Levy,
H.; Danino, D.; Talmon, Y.; Kwektat, K. Langmuir 1997, 13, 402-
408. (c) De Lisi, R.; Inglese, A.; Milioto, S.; Pellerito, A. Langmuir 1997,
13, 192-202.
(49) Quagliotto, P. et al. Manuscript in preparation.
(50) Ross, S. A.; Pitie´, M.; Meunier, B. J . Chem. Soc., Perkin Trans.
1 2000, 571-574.
(51) Howells, R. D.; Mc, C. Chem. Rev. 1977, 77, 69-92.
(52) Rosen, M. J . J . Colloid Interface Sci. 1981, 79, 587-588.
J . Org. Chem, Vol. 68, No. 20, 2003 7657

Quagliotto et al.
TABLE 1. Am p h ip h ilic P r op er ties of th e Su r fa cta n ts 11-15 a t 25 °C
a
spacer
no. of methylenes
counterion
X
Tk
cmcb
(mmol/L)
âb
(%)
cmcc
(mmol/L)
âc
(%)
compd
(°C)
1
2
3
11b
12b
13b
14b
15b
12c
0
3
3
2
3
4
8
12
3
Cl
Cl
Br
CH3SO3
CH3SO3
CH3SO3
CH3SO3
CH3SO3
Cl
50.5
50.9
63.6
<0
2.07
2.09
1.93
1.32
0.750
1.51
1.86d
1.28
1.11
0.22
0.837d
74
72
40
57
57
69
64d
67
44
48
82d
2.08
1.94
1.93
1.23
0.681
1.45
1.71d
1.07
1.06
0.21
0.634d
70
71
40
57
57
68
64d
69
44
70
80d
<0
<0
<0
<0
<0
13c
14c
15c
12d
12e
12f
16e
4
8
12
3
3
3
Cl
Cl
Cl
Br
<0
<0
<0
27.9
65.3
78.8
<0
I
p-CH3C6H4SO3
Cl
17.1
60
a
b
d
Krafft Point. From conductivity: classical method. c From conductivity: nonlinear fit. At 30 °C. e lit.53 1.78 × 10-2 M.
in the fit, making it matter of taste and sensibility.
Besides, the evaluation of the degree of counterion
dissociation R, taken as the ratio of the slopes after and
before the cmc respectively (R ) Smicellar/Spremicellar), and
of counterion binding â (â ) 1 - R) could be affected by
substantial errors.
Recently, a study concerning the evaluation of the cmc
from conductivity data was reporting an interesting
procedure, using a nonlinear fit.56 The fit of the whole
data set to a function, which represents the integral
function of a sigmoid (equation 1), was proposed:
From Table 1 it is evident that the gemini pyridinium
surfactants showing no Krafft Points, namely the chlo-
rides and methanesulfonates, have a low cmc. The
1-dodecylpyidinium chloride (16), taken as a standard,
shows a cmc of 1.71 × 10-2 M (lit.53 1.78 × 10-2 M). The
cmc of 12Py(2)-3-(2)Py-12 2Cl (12c) is more than 10 times
lower. This is in agreement with results obtained for
other series of gemini surfactants.15 The presence of two
hydrophobic chains cause an effect on the cmc and on
the micellization free energy (∆Gmic) whose relation with
the thermodynamic parameters was recently studied by
Zana57 since the introduction of gemini surfactants
revealed that a new, more general, treatment was
needed. This peculiarity discloses to gemini surfactants
wide ranges of applicability in a number of fields. The
structural features, that give the gemini surfactants a
high opportunity of modulation of properties, are a key
point for applications and, in particular, for formulators.
As a general trend, the chlorides (12c-15c) show
slightly lower cmc than the methanesulfonates (11b-
15b). When the cmc data were plotted against the
number of carbon atoms in the spacer length (Figure 1),
the methanesulfonates show a slight enhancement from
n ) 2 to n ) 3 and a following decrease that is more
pronounced for n being equal to or greater than 8 carbon
atoms. The chlorides show nearly the same general
behavior, but the decrease of the cmc is clearly evident
for n ) 12. Besides, in this series the compound having
n ) 2 (11c) is lacking, since it was impossible to produce
it either by direct synthesis or by ionic exchange, and
we could not notice the slight increase evidenced for
methanesulfonates going from n ) 2 to n ) 3.
1 + e(x-x )/∆x
0
F(x) ) F(0) + A1x + ∆x(A2 - A1)ln
(1)
(
)
1 + e-x /∆x
0
In this equation, the transition from two linear regimes
at low and high concentration is described. The param-
eters have the following meaning: F(0) is the initial
conductivity of water, A1 and A2 are the limiting slopes
for low and high concentration respectively, x0 is the
central point of the transition; i.e., the cmc and ∆x is the
width of the transition. The R value can be deduced from
the ratio A2/A1.
It was evident that the suggested function would
represent exactly the physical model to which the con-
ductivity data would adhere. The nonlinear fitting method
avoids any treatment of the data and the consequent
introduction of noise. Moreover, by the use of the non
linear fit which simply searches for the best correlation,
the evaluation of the degree of counterion binding â could
be performed, avoiding artifacts due to the personal
choice of points to be included in the linear fit of the
linear regimes.
This method could be very useful when applied to
gemini surfactants having the above-mentioned behavior,
especially when the hydrophobic chains are lengthened.
In this paper, we tried to analyze our data concerning
new gemini pyridinium surfactants, by the classical
method and by this above-mentioned new method. The
values of the cmc estimated by the two different data
analysis approaches differ very slightly for most of the
compounds, showing that the two methods are working
well.
The increase in the cmc going from n ) 2 to n ) 3 (12b)
is similar to the behavior shown by gemini ammonium
surfactants15 and it could be due to the small increase in
conformational freedom when the spacer is lengthened.
This fact could help the molecule to find a better
conformation in solution, giving rise to the small increase
in the cmc. This is only a hypothesis as far as our
compounds are concerned, since the difference in cmc
(11b vs 12b) is very small. However, this kind of behavior
(57) Zana, R. Langmuir 1996, 12, 1208-1211.
7658 J . Org. Chem., Vol. 68, No. 20, 2003

Gemini Pyridinium Surfactants
F IGURE 2. Degree of counterion binding â vs spacer length
for the compounds having a chloride 12c-15c (b) or meth-
anesulfonate 11b-15b (2) counterion.
F IGURE 1. cmc vs spacer length for the compounds having
a chloride 12c-15c (b) or methanesulfonate 11b-15b (2)
counterion.
The 12Py(2)-3-(2)Py-12 2Br compound (12d ), has a cmc
of 8.37 × 10-4 M, while the corresponding chloride (12c)
is micellizing at 1.86 × 10-3 M, so the cmc is about halved
by changing the counterion from chloride to bromide, as
expected. Since the iodide has a Krafft point of about 65
°C, its cmc was not measured.
is markedly enhanced in other series of surfactants such
as the gemini ammonium ones.
The cmc values for chlorides (12c-15c) are decreasing
when the spacer is lengthened and seem to evidence an
effect due to the increase of hydrophobicity of the spacer.
When the spacer length exceeds n ) 8, e.g., n ) 12,
the contribution of the spacer to the whole hydrophobicity
of the molecule has a dramatic effect on the cmc value,
higher than that shown by n-s-n ammonium surfac-
tants.15 In effect, the cmc of 12-12-12 2Br is about one-
third of the cmc of 12-3-12 2Br, while the cmc of 12Py(2)-
12-(2)Py-12 2Cl (15c) is less than one-sixth of the
12Py(2)-3-(2)Py-12 2Cl (12c) cmc. The greater extension
of this effect could be related to the different arrangement
imposed by the nature of the headgroups. In this case,
the pyridinium ring could decrease the conformational
freedom with respect to the corresponding gemini am-
monium surfactant. This hypothesis could explain why
this behavior on the cmc is clearly evident starting from
n ) 12 since, when the spacer is quite short, the degree
of conformational freedom is quite low and the head-
groups does not have enough freedom to find a better
arrangement, thus anticipating the aggregation. In fact,
the presence of two charges on the pyridinium rings
limits the ability to find better conformations since most
of the possible conformations are experiencing the Cou-
lombic repulsion. Another contribution to the lowering
of the cmc for the compounds having n ) 12, is probably
connected with the great hydrophobicity of the spacer
which, in some cases, was demonstrated to act as a
further “third” hydrophobic chain by folding into the core
of the monolayer or the micelle.17 The higher degree of
counterion binding of compounds having short spacers
(12c-13c) with respect to the monomer (16) is another
effect that could explain the lack of conformational
freedom, since the two charges of the pyridinium rings
are probably keeping one of the two counterion between
them, so displaying a stabilizing effect.
Degr ee of Cou n ter ion Bin d in g. The degree of coun-
terion binding â is an important parameter since it is
the expression of how many counterions are contained
in the Stern layer to counterbalance the electrostatic force
that opposes to the micelle formation. The evaluation of
â by using the classical method could result very difficult,
as described above, while the non linear fit give access
to more reliable values. It should be stressed that a more
reliable way to obtain the value of â is connected with
the Corrin-Harkins plot reporting log CMC vs log(CMC
+ concentration of added salt).58 However, since most of
the previous literature was essentially based on â values
obtained from the classical method, we prefer to report
here values obtained according to the two methods shown
above to permit an adequate comparison. The subject of
the strict reliability of the â values will be worthy of a
deeper study.
Normal micelles made up from 1-dodecylpyridinium
chloride (16) show a â value of 60% and it is evident that
the corresponding gemini surfactants having short spac-
ers (12c-15c) have particular behavior since they bind
their counterions more strongly than the “monomer” does.
We have already explained above the possible reasons
for this behavior. For longer spacers the degree of
counterion binding is decreasing and reaches lower
values than those shown by the monomer (see Figure 2).
This seems to be a general trend for gemini surfactants,
as the n-s-n ammonium gemini amphiphiles.15 The con-
formational freedom connected with 8 and 12 methylenes
is probably enough to leave the two headgroups far apart,
so that the interaction with the hydrated counterions is
substantially decreased. As a consequence, the short
spacer surfactants are probably aggregating in non
spherical micelles (at least spheroidal, elongated ones)
since their packing parameter should be higher than 1/3.
Finally, to see the effect of the different halide coun-
terions, a comparison was made between the compound
12c and 12d , working at 30 °C, since 12d is showing a
Krafft point at about 28 °C.
(58) Sugihara, G.; Nakamura, A. A.; Nakashima, T.-H.; Araki, Y.
I.; Okano, T.; Fujiwara, M. Colloid Polym. Sci. 1997, 275, 790-796.
J . Org. Chem, Vol. 68, No. 20, 2003 7659

Quagliotto et al.
As far as the long spacer surfactants are concerned, they
probably can aggregate in spherical micelles or nearly
spheroidal micelles, due to the higher conformational
freedom that the spacer gives to the molecules. Their
packing parameter should decrease with respect to that
of the short spacer surfactants, according to spherical
micelles. Besides, the low degree of counterion binding
is normally related with small, nearly spherical, micelles.
A peculiar behavior occurs with the methanesulfonates,
for which â follows a trend similar to those observed for
the chlorides, except for the compound having four
methylenes as spacer (13b) for which an “anomalous”
value of 40% is obtained.
This particular behavior is presently under analysis,
and we do not yet have clear background knowledge to
explain it. As above depicted, this anomalous â value
should evidence a particular nature of the aggregate. It
seems that the compound 13b forms small micelles with
a low charge density. This could be explained by a
molecular arrangement with the tetramethylenic spacer
in an all s-trans conformation, which maintain the two
rings and the two charges far apart. As a result of this
conformation, the counterion could not be tightly linked
between the two positive headgroups. In this situation,
if the aggregation number is low, the charge density is
low and water could enter deeper into the micellar core.
More work is needed, however, to get a deeper insight
on the micellization behavior, thus measurements with
several different techniques are currently in progress.
general synthetic method. For the first time, pyridines
bearing a substituent in the 2-position having a more
crowding effect than methyl or ethyl, were efficiently
alkylated with chains longer than ethyl or propyl. Surely,
a well-addressed study of the synthetic opportunities
offered by the long chain alkyl triflates in alkylating
sterically hindered pyridines is opportune. In fact, the
use of the hexadecyl trifluoromethanesulfonate could
allow to overcome the lower reactivity connected with the
longer alkyl chain, opening the way to the synthesis of
the series having hexadecyl or longer chains. In view of
the recent peculiar properties shown by pyridinium
amphiphiles, even in the biological domain for their use
as DNA carriers for transfection purposes, this possibility
becomes very attractive.
The amphiphilic characterization was performed with
conductivity measurements, obtaining the Krafft point,
the cmc and the degree of counterion binding.
The obtained results showed a dependence of the cmc
from the number of carbon atoms in the spacer, which is
decreasing when the spacer is lengthened. This effect
seems to be more pronounced than that found in the
literature for gemini ammonium surfactants. A similar
relationship was also found for the degree of counterion
binding vs the spacer length, evidencing that short spacer
surfactants give micelles with a high charge density. For
these aggregates a conformational arrangement in which
the headgroups bring one of the two counterions between
them is postulated. Surfactants having long spacers seem
to aggregate in micelles whose charge density is consis-
tently lower. It could be guessed that the degree of
counterion binding should probably indicate that the
spacer has a crucial effect on the shape of the micelles.
Con clu sion s
A novel series of gemini surfactants bearing the
pyridinium ring as the headgroup was prepared. During
the synthetic project it was found that the linkage point
of the spacer on the pyridinium ring was crucial for the
definition of the synthetic pathway and the surfactant
properties.
The easiest series to be prepared, namely n-Py(4)-s-
(4)Py-n 2X surfactants, revealed high Krafft points and
thus, low potential toward practical applications. The
second one, related to the n-Py(2)-s-(2)Py-n 2X, was a
source of synthetic challenges. The use of alkyl meth-
anesulfonates as quaternizing agents, even if successful
in several cases, turned out to be a method lacking of
generality. The resort to alkyl trifluoromethanesulfonates
as the alkylating reagents opened the way to a quite
Ack n ow led gm en t. This work was supported by
contributions from the Consiglio Nazionale delle
Ricerche (CNR) and from the Ministero dell’Istruzione,
dell’Universita` e della Ricerca (MIUR), Italy. We are
particularly grateful to Compagnia di San Paolo (Torino,
Italy) and the Fondazione Cassa di Risparmio di Torino
for having supplied laboratory equipment.
Su p p or tin g In for m a tion Ava ila ble: General experimen-
1
1
tal procedures, H NMR spectra for all new compounds, H-
1H COSY spectra for compound (13c), and typical conductivity
vs concentration plots. This material is available free of charge
via the Internet at http://pubs.acs.org.
J O034602N
7660 J . Org. Chem., Vol. 68, No. 20, 2003