Catalysis by Hydrophobically Modified Poly(propylenimine

Aug 13, 2004 - Chemical Sciences, University of Madras, Guindy Campus, Chennai 600025, TN, India ... Department of Chemistry, Oklahoma State Universit...
1 downloads 0 Views 69KB Size
Download PDF
Langmuir 2004, 20, 8307-8312

8307

Catalysis by Hydrophobically Modified Poly(propylenimine) Dendrimers Having Quaternary Ammonium and Tertiary Amine Functionality Eagambaram Murugan,†,‡ Robert L. Sherman, Jr.,‡ H. Olin Spivey,§ and Warren T. Ford*,‡ Departments of Chemistry and of Biochemistry and Molecular Biology, Oklahoma State University, Stillwater, Oklahoma 74078, and Department of Physical Chemistry, School of Chemical Sciences, University of Madras, Guindy Campus, Chennai 600025, TN, India Received March 5, 2004. In Final Form: June 30, 2004 Four different quaternary ammonium chloride-modified poly(propylenimine) (PPI) dendrimers were synthesized by alkylation of a PPI dendrimer having eight dimethylamino end groups with 1-bromooctane or 1-bromododecane. By varying the mole ratio of alkyl bromide to dendrimer, averages of 4-10 quaternary ammonium groups were formed. The new amphiphilic dendrimers are surface active and are micellar catalysts in water. The dendrimers have critical aggregation concentrations between 8.5 × 10-4 and 9.0 × 10-5 M. Decarboxylation of 6-nitrobenzisoxazole-3-carboxylate at 25 °C was 650 times faster than in water alone in the presence of a dendrimer quaternized with eight dodecyl chains at a concentration of 2.45 mM in quaternary ammonium groups. The order of the catalytic efficiency of the new dendrimers decreased with the length and number of hydrophobic alkyl groups in the order (C12)8 > (C12)4 > (C8)10 > (C8)5. The pseudo-first-order rate constants for basic hydrolysis of p-nitrophenyl hexanoate in pH 9.4 buffer at 30 °C using the (C12)8 and (C12)4 dendrimers were 26 and 13 times higher than those for hydrolysis with no dendrimer. The kinetic data were fit to a single-site binding model to evaluate the contributions of binding constants of reactants to the dendrimers and catalytic rate constants of the bound species to the overall catalytic activity.

* To whom correspondence should be addressed. E-mail: [email protected]. † Visiting scientist from the University of Madras. ‡ Department of Chemistry, Oklahoma State University. § Department of Biochemistry and Molecular Biology, Oklahoma State University.

anion-exchange resins,6 and latexes.7,17-19 These catalysts do not assist in the breaking and making of chemical bonds. Instead, as phase-transfer catalysts they concentrate organic compounds and reactive anions into a small volume fraction of an aqueous mixture, where higher local concentrations of reactants and solvent effects accelerate the reaction. One advantage of polymers over low molar mass surfactants is that polymers retain catalytic activity at low concentrations, whereas surfactants are active only above their critical micelle concentrations. Dendritic polycations are a special class of phasetransfer catalytic media.20-25 Amphiphilic dendrimers act as unimolecular micelles that attract nonpolar compounds to their hydrophobic regions and counterions to their hydrophilic charged regions.26-29 The hydrophobic regions are useful for molecular encapsulation23,29-33 and for drug

(1) Fendler, J. H.; Fendler, E. J. Catalysis in Micellar and Macromolecular Systems; Academic Press: New York, 1975. (2) Fendler, J. H. Membrane Mimetic Chemistry; Wiley-Interscience: New York, 1982. (3) Bunton, C. A.; Savelli, G. Adv. Phys. Org. Chem. 1986, 22, 213. (4) Bunton, C. A. Surfactant Sci. Ser. 1991, 37, 323. (5) Kunitake, T.; Shinkai, S. Adv. Phys. Org. Chem. 1980, 17, 435. (6) Ford, W. T.; Tomoi, M. Adv. Polym. Sci. 1984, 55, 49. (7) Ford, W. T.; Miller, P. D. In Functional Polymer Colloids & Microparticles; Arshady, R., Guyot, A., Eds.; Citus Books: London, 2002; Vol. 4, p 171. (8) Menger, F. M.; Portnoy, C. E. J. Am. Chem. Soc. 1967, 89, 4698. (9) Bunton, C. A.; Nome, F.; Quina, F. H.; Romsted, L. S. Acc. Chem. Res. 1991, 24, 357. (10) Romsted, L. S.; Bunton, C. A.; Yao, J. Curr. Opin. Colloid Interface Sci. 1997, 2, 622. (11) Mackay, R. A. J. Phys. Chem. 1982, 86, 4756. (12) Menger, F. M.; Elrington, A. R. J. Am. Chem. Soc. 1991, 113, 9621. (13) Chaimovich, H.; Cuccovia, I. M. Prog. Colloid Polym. Sci. 1997, 103, 67. (14) Klotz, I. M. Ann. N. Y. Acad. Sci. 1984, 434, 302. (15) Fife, W. K. Trends Polym. Sci. 1995, 3, 214. (16) Yamazaki, N.; Nakahama, S.; Hirao, A.; Kawabata, J.; Noguchi, H.; Uchida, Y. Polym. Bull. 1980, 2, 269.

(17) Ford, W. T.; Yu, H. Langmuir 1993, 9, 1999. (18) Lee, J. J.; Ford, W. T. J. Org. Chem. 1993, 58, 4070. (19) Ford, W. T. React. Funct. Polym. 2001, 48, 3. (20) Cherestes, A.; Engel, R. Polymer 1994, 35, 3343. (21) Lee, J. J.; Ford, W. T.; Moore, J. A.; Li, Y. Macromolecules 1994, 27, 4632. (22) Piotti, M. E.; Rivera, F., Jr.; Bond, R.; Hawker, C. J.; Frechet, J. M. J. J. Am. Chem. Soc. 1999, 121, 9471. (23) Pan, Y.; Ford, W. T. Macromolecules 2000, 33, 3731. (24) Goetheer, E. L. V.; Baars, M. W. P. L.; van den Broeke, L. J. P.; Meijer, E. W.; Keurentjes, J. T. F. Ind. Eng. Chem. Res. 2000, 39, 4634. (25) Kreider, J. L.; Ford, W. T. J. Polym. Sci., Part A: Polym. Chem. 2001, 39, 821. (26) Newkome, G. R.; Yao, Z.; Baker, G. R.; Gupta, V. K. J. Org. Chem. 1985, 50, 2003. (27) Newkome, G. R.; Moorefield, C. N.; Baker, G. R.; Saunders: M. J.; Grossman, S. H. Angew. Chem., Int. Ed. Engl. 1991, 30, 1178. (28) Tomalia, D. A.; Berry, V.; Hall, M.; Hedstrand, D. M. Macromolecules 1987, 20, 1164. (29) Hawker, C. J.; Wooley, K. L.; Frechet, J. M. J. J. Chem. Soc., Perkin Trans. 1 1993, 1287. (30) Jansen, J. F. G. A.; de Brabander van den Berg, E. M. M.; Meijer, E. W. Science 1994, 266, 1226.

Introduction Cationic polymers and aggregates of quaternary ammonium ions accelerate the rates of chemical reactions of anionic compounds and of reactions involving anionic catalysts in aqueous solutions.1-7 The kinetics of these reactions have been described by enzyme-like and pseudophase ion-exchange models.1-3,8 Among the catalytically effective polycations are association colloids such as surfactant micelles,1,3,4,9,10 microemulsions,11,12 bilayer vesicles,2,5,13 both linear and branched polyelectrolytes,1,14-16

10.1021/la049420i CCC: $27.50 © 2004 American Chemical Society Published on Web 08/13/2004

8308

Langmuir, Vol. 20, No. 19, 2004

delivery, including gene therapy.34-36 Previously, we tested several types of cationic dendrimers as catalysts for the unimolecular decarboxylation of 6-nitrobenzisoxazole-3carboxylate ion (NBIC).21,23,25 Hydrophilic dendrimers with alkyltrimethylammonium ions as end groups had low activity, while dendrimers with both octyl groups and triethylene glycol methyl ether groups at the chain ends were 1-2 orders of magnitude more active, approaching the activity of the best surfactant micelles and polymer latexes. Preparation of the most active dendritic catalysts required five synthetic steps starting from commercial poly(propylenimine) (PPI) dendrimers37,38 and several chromatographic purifications, making them generally impratical. In this paper we report a new family of quaternary ammonium dendrimers that are prepared in two simple steps from PPI dendrimers and have still higher catalytic activities for decarboxylation of NBIC and for basic hydrolysis of p-nitrophenyl hexanoate. Experimental Section Materials. Dendrimer DAB-Am-8, formaldehyde, formic acid, alkyl halides, and solvents from Aldrich Chemical Co. were used without further purification. 6-Nitrobenzisoxazole-3-carboxylic acid was synthesized by a literature method39,40 and was 95% pure by 1H NMR analysis. 4-Nitrophenyl hexanoate was used as received from TCI Inc. Triply deionized water had a conductivity less than 1 × 10-6 Ω-1 cm-1. Amberlite IRA-95 and Amberlite IRA-402 were conditioned as described previously.25 DAB-Am-8(CH3)16 (2). In a 100 mL round-bottom flask equipped with a reflux condenser and nitrogen protection 10.00 g (12.9 mmol) of DAB-Am-8 (1) and 30 g (370 mmol) of 37% formaldehyde reacted to form a solid. Then 60 g (2.18 mol) of 88% formic acid was added. The solid dissolved and began to evolve CO2. The mixture was kept at 90 °C for 24 h, cooled in air, placed in an ice bath, and made basic with 50% NaOH. The cloudy mixture was extracted three times with 50 mL of CH2Cl2. The combined CH2Cl2 solution was dried with Na2SO4. Solvent was removed under vacuum, and the residual oil was dried in a vacuum overnight at 56 °C to yield 8.40 g (67%) of a thick light yellow oil. 1H NMR (300 MHz, CDCl3, δ): 2.5-2.3 (br), 2.2-2.1 (s), 1.7-1.5 (br), 1.5-1.4 (br). 13C NMR (75 MHz, CDCl3, δ): 58.0, 55.9, 52.4, 52.3, 52.0, 45.5 (6.3 N(CH3)2), 42.2 (1.7 NHCH3), 25.6, 25.3, 24.5. DAB-Am-8(CH3)16(C8H17)5 (3). In a 100 mL round-bottom flask equipped with a reflux condenser and nitrogen protection 1.074 g (1.11 mmol) of 2 was dissolved in 50 mL of N,Ndimethylformamide (DMF), and 1.105 g (5.76 mmol) of 1-bromooctane was added. The solution was stirred at 80 °C for 48 h. The DMF was removed under vacuum. Final traces of DMF were removed by adding 100 mL of water, evaporating most of the water, adding 100 mL of ethanol, and removing the ethanol under vacuum. The residual oil was dissolved in 10 mL of water and passed through a column of Amberlite IRA-95 in free base form. The solution was then passed through a column of Amberlite IRA-402 in chloride form. After evaporation of water the product was dried in a vacuum at 56 °C to yield 1.018 g (53%) of a yellow/ orange gum. 1H NMR (300 MHz, CDCl3, δ): 3.9-3.7 (br), 3.7(31) Baars, M. W. P. L.; Froehling, P. E.; Meijer, E. W. Chem. Commun. 1997, 1959. (32) Sayed-Sweet, Y.; Hedstrand, D. M.; Spinder, R.; Tomalia, D. A. J. Mater. Chem. 1997, 7, 1199. (33) Sideratou, Z.; Tsiourvas, D.; Paleos, C. M. Langmuir 2000, 16, 1766. (34) Baker, J. R., Jr.; Quintana, A.; Piehler, L.; Banazak-Holl, M.; Tomalia, D.; Raczka, E. Biomed. Microdevices 2001, 3, 61. (35) Esfand, R.; Tomalia, D. A. Drug Discovery Today 2001, 6, 427. (36) Liu, M.; Kono, K.; Frechet, J. M. J. J. Controlled Release 2000, 65, 121. (37) de Brabander-van den Berg, E. M. M.; Meijer, E. W. Angew. Chem., Int. Ed. Engl. 1993, 32, 1308. (38) Bosman, A. W.; Jansen, H. M.; Meijer, E. W. Chem. Rev. 1999, 99, 1665. (39) Borsche, W. Ber. Dtsch. Chem. Ges. 1909, 42, 1310. (40) Clinton, R. O.; Laskowski, S. C. J. Am. Chem. Soc. 1952, 74, 2226.

Murugan et al. 3.6 (br), 3.4-3.0 (br), 2.5-2.0 (br), 1.9-1.6 (br), 1.6-1.3 (br), 1.3-1.0 (br), 0.7 (t). 13C NMR (75 MHz, CDCl3, δ): 63.9, 62.3, 57.5, 57.2, 55.5, 54.0, 51.8, 50.8, 50.4, 45.4, 42.1, 31.8, 29.4, 29.2, 26.2, 25.1, 24.4, 22.7, 22.4, 20.6, 13.9. DAB-Am-8(CH3)16(C8H17)10 (4). Using the method for 3, reaction of 2.000 g (2.06 mmol) of 2 and 4.115 g (21.4 mmol) of 1-bromooctane in 50 mL of DMF followed by ion exchange produced 3.512 g (68%) of a yellow/orange gum. 1H NMR (300 MHz, CDCl3, δ): 3.9-3.7 (br), 3.7-3.6 (br), 3.4-3.0 (br), 2.5-2.0 (br), 1.9-1.6 (br), 1.6-1.3 (br), 1.3-1.0 (br), 0.7 (t). 13C NMR (75 MHz, CDCl3, δ): 64.0, 63.2, 57.2, 52.0, 51.2, 50.6, 50.4, 31.5, 29.0, 28.9, 28.8, 26.2, 26.0, 22.7, 22.5, 22.4, 20.5, 13.9. DAB-Am-8(CH3)16(C12H25)4 (5). Using the method for 3, reaction of 3.000 g (3.10 mmol) of 2 and 3.086 g (12.4 mmol) of 1-bromododecane in 50 mL of DMF followed by ion exchange produced 4.102 g (72.5%) of a thick yellow oil. During rotary evaporation foaming was hard to control. Ethanol was added to reduce foaming. 1H NMR (300 MHz, CDCl3, δ): 4.0-3.8 (br), 3.8-3.6 (br), 3.5-3.1 (br), 2.6-2.0 (br), 2.0-1.7 (br), 1.7-1.4 (br), 1.4-0.9 (br), 0.7 (t). 13C NMR (75 MHz, CDCl3, δ): 63.7, 62.3, 57.8, 57.4, 55.6, 53.7, 51.8, 51.1, 50.4, 45.5 42.1, 31.7, 29.4, 29.3, 29.1, 26.2, 25.1, 24.9, 22.7, 22.4, 20.6, 13.9. DAB-Am-8(CH3)16(C12H25)8 (6). Using the method for 3, reaction of 3.000 g (3.10 mmol) of 2 and 6.170 g (24.7 mmol) of 1-bromododecane in 50 mL of DMF produced a thick yellow oil. The oil was dissolved in a mixture of 5 mL of water and 10 mL of methanol for the ion-exchange processes. During rotary evaporation foaming was hard to control. Ethanol was added to reduce foaming. After drying, 7.012 g (87%) of a fluffy/crunchy yellow solid (which becomes sticky after exposure to air) was recovered. 1H NMR (300 MHz, CDCl3, δ): 4.0-3.7 (br), 3.7-3.5 (br), 3.5-3.1 (br), 2.6-2.0 (br), 2.0-1.7 (br), 1.7-1.4 (br), 1.41.0 (br), 0.7 (t). 13C NMR (75 MHz, CDCl3, δ): 64.2, 63.6, 62.3, 53.7, 51.8, 51.1, 50.4, 45.4, 31.7, 29.4, 29.3, 29.1, 26.2, 25.1, 22.7, 22.5, 20.7, 13.9. Critical Aggregation Concentrations (CACs). Conductivities were measured at 25 °C using a YSI model 31 conductivity bridge with a YSI model 3402 conductivity cell. Twelve different aqueous solutions of concentrations ranging from 5 × 10-3 to 7 × 10-6 M in long alkyl groups were prepared from each dendrimer. A plot of concentration vs conductivity produced two intersecting lines. The intersection was taken as the CAC. Moisture Absorption. A 0.25 g sample of 6 was weighed into a tared vial, which was covered with a beaker and kept in ambient air. The sample initially gained 2 mg h-1. After 2 days the sample changed from a dry solid to a viscous oil. The sample was dried in a vacuum at 56 °C and reweighed. A total of 13% water by weight was absorbed during 48 h. Kinetics. Standard solutions of dendrimer 6 were prepared in nitrogen-purged aqueous NaOH solution at pH 11.4. To a 1 cm polystyrene cuvette was added 2.98 mL of the dendrimer solution. The cuvette was equilibrated in the spectrometer at 25.0 ( 0.1 °C for at least 20 min. The background dendrimer spectrum was recorded. A 10.6 mM solution of 6 in ethanol, 22 µL, was added and mixed by inverting the cuvette three times. (The ethanolic solution of 6 can be stored for only 1-2 days under nitrogen at 5 °C because of decarboxylation.) The average absorbance of the product 8 at 390-410 nm over time was recorded using a Hewlett-Packard model 8452A diode array spectrophotometer using HP 89532K UV/vis kinetics software. Infinite absorbance after complete reaction was measured after the sample was left at room temperature for at least 24 h. Rate constants were calculated from data acquired during the first 10% conversion using the first-order rate equation kobsd ) ln[(A∞ - Ao)/(A∞ - At)]/t, where t is the time of reaction and Ao, At, and A∞ are the absorbances at times 0, t, and infinity. Rate constants were measured two to five times at each dendrimer concentration. Average values are reported. The highest and lowest values of kobsd differed by less than 6% for 6, 19% for 5, 23% for 4, and 5% for 3. The pseudo-first-order hydrolysis of 9 at 30.0 ( 0.1 °C was measured by the same method in 20 mM borate buffer at pH 9.38. The buffer was prepared by adding NaOH to a solution of boric acid. The reactant 9 was added as 30 µL of a 2.5 mM solution in acetonitrile. The average UV/vis absorbance of p-nitrophenoxide ion (10) over 400-410 nm was recorded. The rate constants

Catalysis by Hydrophobically Modified Dendrimers

Langmuir, Vol. 20, No. 19, 2004 8309

Scheme 1

Scheme 2 Figure 1. Dependence of conductivity on the concentration of quaternary ammonium ions of 6. The intersection of the lines corresponds to CAC ) 9.0 × 10-5 M. Table 1. Compositions and CACs of Quaternary Ammonium Chloride Dendrimers no. of N+Cldendrimer groupsa CAC (M) 3 4 a

reported are averages of duplicate experiments from which the values differed by 5% or less. Regression analysis of observed rate constants and dendrimer concentrations to calculate equilibrium binding constants and catalytic rate constants was carried out by a nonlinear leastsquares method using the Marquardt-Levenberg algorithm.41

Results Dendrimer Syntheses. Reductive methylation of DAB-Am-8 (1) having eight primary amine end groups with formaldehyde and formic acid produced the tertiary amine dendrimer 2 as shown in Scheme 1. However, 2 had an average of only 6.3 tertiary amine and 1.7 secondary amine groups according to peak heights in the 13C NMR spectrum.25 The conversion to tertiary amine did not improve by treatment with more formaldehyde and formic acid or for longer reaction time. For conversion to quaternary ammonium ions, as shown in Scheme 2, the small amount of secondary amine groups was not a problem. The extent of formation of quaternary ammonium ions was controlled by the molar ratio of 1-bromooctane or 1-bromododecane to dendrimer. Most of the DMF was removed from the reaction mixtures by rotary evaporation. We tried to remove the last amount of DMF by codistillation with water, but aqueous solutions of dendrimer 6 foamed during attempted rotary evaporation. We also tried to extract dendrimer 6 from aqueous sodium hydroxide into dichloromethane, but an intractable emulsion formed. The best method found for removal of water and DMF from dendrimer 6 was to add excess ethanol to an aqueous solution of 6 and then rotary evaporate. Foaming of the ethanolic solution was reduced enough to permit rotary evaporation. Dendrimer 5 with four dodecyl chains foamed less than 6, and dendrimers 3 and 4 with octyl chains did not foam at all. A tertiary amine ion-exchange resin was used to convert the trialkylammonium bromide groups of the dendrimer (formed by the HBr byproduct of the alkylations of residual secondary amines) to tertiary amine groups. The qua(41) Miller, P. D.; Spivey, H. O.; Copeland, S. L.; Sanders, R.; Woodruff, A.; Gearhart, D.; Ford, W. T. Langmuir 2000, 16, 108.

4.8 9.7

8.5 × 10-4

no. of N+Cldendrimer groupsa CAC (M) 5 6

4.0 7.7

3.8 × 10-4 9.0 × 10-5

From titration of Cl-.

ternary ammonium bromide dendrimers were converted to the more stable chloride form using a quaternary ammonium chloride ion-exchange resin.25 The average number of quaternary ammonium ions per dendrimer was determined by titration of the chloride counterions, and the number of alkyl groups was determined by the relative molar amount of alkyl bromide used in the synthesis, assuming complete reaction. The products are mixtures; the 1H and 13C NMR spectra do not have unique peaks due to specific compounds. Methyl carbon peaks at 42 and 45.5 ppm were diagnostic for secondary methylamines and tertiary dimethylamines, respectively. The compositions of the new dendrimers are listed in Table 1. Quaternary ammonium ion dendrimers were prepared also by reaction of tertiary amine dendrimer 2 with 1-iodobutane, benzyl bromide, 2-ethylhexyl bromide, and 1-bromohexadecane by the same methods used to synthesize 3-6. The catalytic activities of the butyl-, benzyl-, and 2-ethylhexylammonium ion dendrimers for decarboxylation of NBIC were very low. The hexadecylammonium ion dendrimer was too insoluble in water to test its catalytic activity. Similar permethylation of the PPI dendrimers having 32 and 64 primary amine end groups followed by partial quaternization with 1-bromododecane also gave quaternary ammonium ion dendrimers that were much less active catalysts than 6. Critical Aggregation Concentrations. The foaming of aqueous solutions of 6 indicates surface activity. Consequently, CACs of the dendrimers were measured via the conductivity of dendrimer solutions. The measurements showed two linear regions of conductivity vs concentration as shown in Figure 1. The intersection of the two lines, which corresponds to a lower degree of dissociation of the counterions from the aggregated dendrimer at the higher concentrations than from the less aggregated or nonaggregated dendrimer at lower concentrations, was taken as the CAC. The CAC values, which are reported in Table 1, decrease with increasing number of alkyl groups and increasing charge on the dendrimer, and decrease with increasing length of the alkyl groups, as expected. Decarboxylation of 6-Nitrobenzisoxazole-3-carboxylate Ion (7). This reaction, shown in Scheme 3, is

8310

Langmuir, Vol. 20, No. 19, 2004

Murugan et al.

Scheme 3

Table 3. Binding Constants and Catalytic Rate Constants for Decarboxylation of 7 no. of data dendrimer pointsa

Table 2. Kinetic Data for Decarboxylation of 7 in Aqueous NaOH at pH 11.4 and 25.0 °C dendrimer

103[N+] (M)

kobsd (s-1)

kobsd/kwa

3

32.2 10.7 32.8 10.9 2.45 0.0245 2.45 0.0245 24.0 2.40 2.37 2.83

1.36 × 10-4 3.43 × 10-5 2.09 × 10-4 2.98 × 10-5 1.24 × 10-3 3.04 × 10-5 2.02 × 10-3 1.01 × 10-4 2.58 × 10-5 1.36 × 10-5 1.51 × 10-3 6.47 × 10-5

43.7 11.0 67.5 9.6 401 9.8 650 32.7 8.32 4.39 487 20.9

4 5 6 11b 12c 13d

kw ) 3.1 × 10-6 s-1. b Reference 25. c Reference 23. d Reference 21. See the text for the description of structures 11-13. a

Scheme 4

3 4 5 6 11c 12d 13e

6 4 5 5

103kcb (s-1)

kc/kw

K b (M-1)

0.137 (0.155, 0.121) 44 300 (390, 230) 0.21 (0.24, 0.18) 68 860 (1100, 610) 2.00 (3.0, 1.4) 645 700 (1100, 400) 2.70 (3.0, 2.4) 871 1600 (1900, 1400) 0.028 9.3 290 1.75 564 2400 0.08 25 1700

a Number of data points in the regression analysis. b Standard deviation upper and lower error limits are in parentheses. c Reference 25. d Reference 23. e Reference 21. See the text for the description of structures 11-13.

Table 4. Kinetic Data for Hydrolysis of 9 in Aqueous 20 mM Borate Buffer at pH 9.38 and 30.0 °C catalyst 5 6

103[N+] 102kobsd kobsd/ (M) kwa 103kcb (s-1) (s-1) 6.15 0.0615 6.15 0.0615

0.96 0.10 1.85 0.163

K b (M-1)

13.6 15 (19, 12) 2800 (2900, 2700) 1.50 26.2 23 (24, 22) 720 (810, 640) 2.32

a The pseudo-first-order rate constant in the absence of dendrimer was kw ) 7.03 × 10-4 s-1. b Standard deviation upper and lower error limits are in parentheses.

Scheme 5

exquisitely sensitive to the environment of 7. The rate constants were measured by the appearance of the UV absorbance of the 2-cyano-5-nitrophenoxide ion (8) in pH 11.4 aqueous solutions. Rate constants for decarboxylation of 7 at fixed dendrimer concentrations are reported in Table 2. The largest rate enhancement, a factor of 650 times faster than in water, at a dendrimer concentration of 2.45 × 10-3 M in quaternary ammonium ion groups, was attained by dendrimer 6 having eight dodecyl groups. The activities of the other dendrimers decreased in the order 5 > 4 > 3. The decarboxylation kinetics were analyzed by the enzyme-like model shown in Scheme 4, which was first proposed by Menger and Portnoy for micellar catalysis.8 The substrate S is assumed to equilibrate between the aqueous phase and the dendrimer faster than decarboxylation occurs either in water or in the dendrimer pseudophase. Product P can be formed in either phase. The observed first-order rate constant kobsd depends on the rate constants kw and kc (the catalytic rate constant) and on the fractions of S free in the water and bound to dendrimer according to eq 1, where [S]w, [S]b, and [S]t are the concentrations of S in water and bound to dendrimer and the total concentration of S, respectively. The binding constant K is defined by eq 2. Equation 3 gives the observed

kobsd ) kw[S]w /[S]t + kc[S]b/[S]t

(1)

K ) [S]b/([S]w[N+])

(2)

kobsd ) (kw/K+ kc[N+]/(1/K + [N+])

(3)

rate constant in terms of the binding constant, the microscopic rate constants, and the dendrimer concentra-

tion expressed as the concentration of quaternary ammonium ions. K and kc were determined by nonlinear regression analysis of observed rate constants at four to six dendrimer concentrations. The data covered a range from (2-3) × 10-5 M in long alkyl groups where the rate constants are close to linear in concentration to (2-3) × 10-3 M in long alkyl groups where there is little dependence of rate constants on concentration. For the calculations, the concentrations of quaternary ammonium ions were assumed to be equal to the concentrations of long alkyl groups. The results are reported in Table 3. The large error limits on kc and K are due to a strong inverse correlation of these quantities. (A fast rate could be due to a large binding constant, a high catalytic rate constant, or contributions from both equilibrium and rate constants). The data show that dendrimers 5 and 6, which are quaternized with dodecyl groups, give catalytic rate constants more than 10 times higher than dendrimers 3 and 4, which are quaternized with octyl groups. Modestly larger binding constants also make the dodecylammonium ion dendrimers more active catalysts than the octylammonium ion dendrimers. Hydrolysis of 4-Nitrophenyl Hexanoate (9). The most active dendrimers for decarboxylation of 7 were tested as catalysts for the basic hydrolysis of 4-nitrophenyl hexanoate, shown in Scheme 5. Rates of hydrolysis were followed by the pseudo-first-order appearance of the UV absorbance of 4-nitrophenoxide ion (10) in borate buffer solutions at pH 9.38. Observed rate constants are reported in Table 4 along with the catalytic rate constants and equilibrium binding constants calculated from data over

Catalysis by Hydrophobically Modified Dendrimers

wide ranges of dendrimer concentrations using the model of Scheme 4. Discussion The decarboxylation of 7 is an excellent probe for environmental effects on reaction rates. The rate is 108 times faster in hexamethylphosphoramide than in water, due both to stabilization of the carboxylate ion by hydrogen bonding and to stabilization of the transition state by dipolar aprotic solvents.42,43 The reaction is catalyzed in aqueous solutions by cationic surfactant micelles,44-47 bilayer vesicles,48-50 microemulsions,51 polyelectrolytes,52-55 cationically modified silica gel,56 cross-linked polymer gels,16,57 polymer latexes,18,21,58 and dendrimers.21,23,25 Table 2 shows that 2.45 mM quaternary ammonium ions of dendrimers 6 and 5, which are modified with eight and four dodecyl groups, respectively, gave observed rate constants 650 and 401 times the rate constant in the absence of dendrimer. There were significant rate enhancements even at a quaternary ammonium ion concentration of 2.45 × 10-5 M. The rate enhancements using 6 are about a factor of 3 less than rate enhancements attained with the most active quaternary ammonium ion latexes and the most active surfactant micelles.18,58 The dendrimers 4 and 3, which are modified with 10 and with 5 octyl groups were less active but still better than most of the dendrimers reported previously, as shown in Table 2. The earlier less active catalysts were a dendrimer with a polyether core based on pentaerythritol and 36 alkyltrimethylammonium iodide end groups (13)21 and a PPI dendrimer with 8 end groups and both end group and branch point nitrogen atoms fully quaternized with methyl groups (11).25 The only previous dendrimers having activity nearly as high as that of 6 were PPI dendrimers with 8 and with 32 end groups modified with both octyl and triethylenoxy methyl ether groups and then fully quaternized with methyl groups.23 Tables 2 and 3 include that octyl- and triethylenoxy-modified dendrimer with 32 end groups (12). Table 3 reveals why the dendrimers 5, 6, and 12, which have more hydrophobic alkyl groups, are more active than the more hydrophilic dendrimers 11 and 13, which lack long alkyl chains: The unimolecular rate constant kc for decarboxylation of bound 7 is 10-100 times greater using the more hydrophobic dendrimers. We (42) Kemp, D. S.; Cox, D. D.; Paul, K. G. J. Am. Chem. Soc. 1975, 97, 7312. (43) Kemp, D. S.; Paul, K. G. J. Am. Chem. Soc. 1975, 97, 7305. (44) Bunton, C. A.; Minch, M. J. Tetrahedron Lett. 1970, 3881. (45) Bunton, C. A.; Minch, M. J.; Hidalgo, J.; Sepulveda, L. J. Am. Chem. Soc. 1973, 95, 3262. (46) Rupert, L. A. M.; Engberts, J. B. F. N. J. Org. Chem. 1982, 47, 5015. (47) Nusselder, J. J. H.; Engberts, J. B. F. N. Langmuir 1991, 7, 2089. (48) Kunitake, T.; Okahata, Y.; Ando, R.; Shinkai, S.; Hirakawa, S. J. Am. Chem. Soc. 1980, 102, 7877. (49) Germani, R.; Ponti, P. P.; Savelli, G.; Spreti, N.; Cipiciani, A.; Cerichelli, G.; Bunton, C. A. J. Chem. Soc., Perkin Trans. 2 1989, 1767. (50) Scarpa, M. V.; Araujo, P. S.; Schreier, S.; Sesso, A.; Oliveira, A. G.; Chaimovich, H.; Cuccovia, I. M. Langmuir 2000, 16, 993. (51) Bunton, C. A.; De Buzzaccarini, F. J. Phys. Chem. 1981, 85, 3139. (52) Kunitake, T.; Shinkai, S.; Hirotsu, S. J. Org. Chem. 1977, 42, 306. (53) Shah, S. C.; Smid, J. J. Am. Chem. Soc. 1978, 100, 1426. (54) Shinkai, S.; Hirakawa, S.; Shimomura, M.; Kunitake, T. J. Org. Chem. 1981, 46, 868. (55) Suh, J.; Klotz, I. M. Bioorg. Chem. 1979, 8, 283. (56) Tundo, P.; Venturello, P. Tetrahedron Lett. 1980, 21, 2581. (57) Wang, G.-J.; Engberts, J. B. F. N. J. Org. Chem. 1994, 59, 4076. (58) Miller, P. D.; Ford, W. T. Langmuir 2000, 16, 592.

Langmuir, Vol. 20, No. 19, 2004 8311

attribute the larger kc in the hydrophobic dendrimers to less stabilization of 7 by hydrogen bonding to water. Moreover, the binding constants K of reactant anion 7 in most cases also are larger for the more hydrophobic dendrimers (with the exception of 13). The new dendrimers are mixtures, as are most synthetic polymers. Dendrimers 3-6 have the average compositions shown in Table 1. A statistical distribution of the number of alkyl groups, or even a tendency toward equal numbers of quaternary ammonium ions per molecule, is expected because the rates of alkylations, which form positive charge, probably decrease with increasing positive charge of the dendrimer. The dendrimer 6 with eight dodecyl groups also enhances the rate hydrolysis of 9 in weakly basic buffer solution by about the same amount as the most active quaternary ammonium ion latexes.41 The data in Table 4 show that the rate enhancement by 6 is due entirely to a greater catalytic rate constant of the bound 9. Although kobsd for the reaction is pseudo-first-order, the rate-limiting step is bimolecular attack of hydroxide ion at the ester carbonyl group, so the high catalytic rate constant may be due to a high local concentration of hydroxide ion, a high activity of hydroxide ion in the dendrimer pseudophase, or both. Under the conditions of the catalysis experiments, dendrimers 3-6 contain tertiary amine as well as quaternary ammonium ion functional groups. Even at pH 9.4, the tertiary amines should remain unprotonated because their basicity is reduced by the many proximate quaternary ammonium groups. In uncharged form the tertiary amines should not bind reactive anion 7. The data available do not allow determination of whether tertiary amines act as general-base catalysts for the hydrolysis of p-nitrophenyl esters in the presence of quaternary ammonium ions having hydroxide, chloride, or borate as counterions. In addition to catalytic activity, the new hydrophobic dendrimers have another major advantage over the previously most active dendrimer 12: They can be synthesized in two simple steps, whereas 12 requires five synthetic steps and several chromatographic purifications.23 The surface-active behavior of dendrimer 6 prompted us to measure critical aggregation concentrations. While 6 had the lowest CAC, 4 and 5 also exhibited CACs. Attempts to determine the sizes of aggregates of dendrimer 6 by dynamic light scattering gave irreproducible results. The product of the CAC and number of quaternary ammonium ions in each dendrimer in every case is markedly less than the critical micelle concentration (CMC) of the singly charged analogue dodecyltrimethylammonium chloride (CMC ) 16 mM), as expected for gemini surfactants.59-61 Univalent surfactant catalysts normally have very low activity at concentrations below the CMC, because organic compounds do not bind strongly to surfactant unimers. The kinetic data do not show any large change in the rate constants in dendrimer solutions at the CAC. Conclusions Dendrimer 6 with eight end groups and an average of eight dodecyl chains is the most active quaternary (59) Rosen, M. J.; Tracy, D. J. J. Surfactants Deterg. 1998, 1, 547. (60) Menger, F. M.; Keiper, J. S. Angew. Chem., Int. Ed. 2000, 39, 1907. (61) Zana, R. Surfactant Sci. Ser. 1998, 74, 241.

8312

Langmuir, Vol. 20, No. 19, 2004

ammonium ion dendrimer known for catalysis of the unimolecular decarboxylation of 6-nitrobenzisoxazole-3carboxylate and for the bimolecular hydrolysis of pnitrophenyl hexanoate. The only other quaternary ammonium ion dendrimers with comparable catalytic activity require arduous syntheses, whereas 6 is prepared in two simple steps from a commercial dendrimer. The catalytic activity of 6 is due mainly to higher catalytic rate constants of the bound reactants, and is slightly less than those of

Murugan et al.

the most active quaternary ammonium ion latexes and surfactant micelles. Acknowledgment. This research was supported by the Petroleum Research Fund of the American Chemical Society. E.M. thanks the Government of India for a BOYSCAST fellowship. LA049420I