Direct Force Measurements between Adlayers Consisting of Poly(amidoamine) Dendrimers with Primary Amino Groups or Quaternary Ammonium Groups

Journal of Colloid and Interface Science 254, 406–409 (2002) doi:10.1006/jcis.2002.8601

NOTE Direct Force Measurements between Adlayers Consisting of Poly(amidoamine) Dendrimers with Primary Amino Groups or Quaternary Ammonium Groups Adsorption of poly(amidoamine) generation 3 (PAMAM G3) dendrimer with surface amino groups or PAMAM G0 dendrimer with quaternary ammonium groups (C8qbG0) onto glass has been studied by colloidal probe atomic force microscopy. The adlayer–adlayer interactions for these adsorbates are quite different despite the fact that they are almost equal in the hydrodynamic radius. In aqueous PAMAM G3 dendrimer solutions the electrostatic repulsion is predominant. The conformation of the adsorbed layer is flat and the protrusion of the individual dendrimers is negligible. On the other hand, C8qbG0 behaves as a surfactant and the layered structure of C8qbG0 is expected to be a patchy bilayer. Dispersion stability of silica suspensions with the adsorption of these dendrimers can be correlated with the force data obtained. C 2002 Elsevier Science (USA) Key Words: colloidal probe atomic force microscopy; poly (amidoamine) dendrimer; quaternized dendrimer; glass surface.

INTRODUCTION Atomic force microscopy (AFM), which acquires topological mappings of nonconductive surfaces with subnanometer resolution, has been widely used to clarify adsorption phenomena at the solid/liquid interface in situ. As AFM images surface topology based on the force between the probe and the substrate, AFM can be used to make direct measurements of such forces. In particular, the interaction forces between a colloidal probe attached to the cantilever and the sample surface can be measured using the colloidal probe technique developed by Ducker et al. (1, 2). This technique has been utilized to study the adsorption of surfactants (3, 4), linear polymers (5–10), or both of them (11–14). However, few reports have been published for the adlayer–adlayer interaction consisting of dendritic macromolecules until now (15). Dendrimers, being highly branched polymers, have become the subject of extensive studies, because their functional groups and specific shape have unique properties compared with those of conventional linear polymers (16). We have studied adsorption characteristics of poly(amidoamine) or PAMAM dendrimers with surface carboxyl groups on α-alumina and with surface amino groups on silica (17). As was found in this work, the dendrimers with the earlier generation behave as simple electrolytes while those with the later generation behave as ionic surfactants or polyelectrolytes. A dispersion– flocculation–redispersion sequence of α-alumina and silica with adsorption of dendrimers with various concentrations has been observed only in the case of the later generation. In addition, coadsorption behaviors have also been investigated, including PAMAM dendrimers with ionic surfactants (18, 19), sugar-persubstituted PAMAM dendrimers with anionic surfactants (20), and PAMAM dendrimers with poly(ethylene oxide) (21). To identify the adsorption 0021-9797/02 $35.00

 C 2002 Elsevier Science (USA)

All rights reserved.

characteristics of dendrimers in detail, it is necessary to collect more data from a nanoscopic standpoint. In the present study, the interaction forces between glass surfaces in aqueous solutions of PAMAM G3 dendrimer were measured by colloidal probe AFM. Sedimentation tests of silica suspensions with the adsorption of the dendrimer were carried out to understand the stabilization and flocculation mechanisms of dispersions. Furthermore, adsorption of quaternized dendrimer (C8qbG0) was also investigated for comparison.

EXPERIMENTAL

Materials PAMAM dendrimer was synthesized by means of a procedure described by Tomalia et al. (22), where ethylenediamine was used as a nitrogen core. The dendrimer used in this study was PAMAM generation 3 (G3). Molecular weight of this dendrimer and the number of terminal amino groups are 6909 and 32, respectively. On the other hand, C8qbG0 was synthesized as follows (23): (i) a quaternized dendron was obtained by the reaction of N ,N -dimethyl-n-octylamine with epibromohydrin at 50◦ C for several hours; (ii) the excess epibromohydrin was removed completely by distillation at 50◦ C in vacuo; (iii) the dendron was reacted with PAMAM G0 dendrimer in chloroform at 40◦ C under nitrogen atmosphere for 7 h; (iv) the solvent was removed by distillation under reduced pressure; (v) as the recrystallization was performed several times from acetone the final product (quaternized PAMAM dendrimer having four octyl chains, C8qbG0, as shown in Scheme 1) was obtained in a good yield. The chemical structure of C8qbG0 was confirmed by FT-IR and 1 H NMR spectra and these data showed that the product is not contaminated by some impurities. IR: 3000, 3280 (υC−H ), and 3400 cm−1 (υN−H ). 1 H NMR (CDCl3 , tetramethylsilane): δ 0.9 (12H, CH3 –C–), 1.2–1.3 [40H, –C–(CH2 )5 –C–], 1.6 (8H, –C–CH2 –C–N+ –), 2.1–2.7 (64H, methylene groups), and 3.1 pm [24H, (CH3 )2 –N+ –]. Glass spheres with an average diameter of 20 µm and glass plates were obtained from Polyscience Inc. (USA) and Matsunami Glass Co. (Japan), respectively. The materials were ultrasonicated for 10 min in a concentrated H2 SO4 , rinsed with water, immersed in an aqueous 0.1 mol dm−3 KOH solution overnight, and finally rinsed thoroughly with water again. Fine silica particles used in the sedimentation test were supplied by Japan Aerosil Co. (Japan). Their specific surface area and average diameter are 50 m2 g−1 and 0.03 µm, respectively. All other reagents were of analytical grade. Water was deionized using a Milli-Q Plus system.

Methods Surface force measurements were carried out using a TMX2100 atomic force microscope (TMmicroscopes Inc., USA). A detailed description of the force measurement technique is shown elsewhere (14, 24). Data were collected using 406

407

NOTE

N+(CH3 )2 C8 H17 Br-

NH

G0

OH

4

C8qbG0 O O NH2 G0 NH2

NH NH

NH2

NH NH

N N

NH2

O O

PAMAM G0 dendrimer SCHEME 1.

Chemical structure of C8qbG0.

commercial silicon nitride cantilevers (Digital Instrument Inc., USA) with a spring constant of 0.58 N m−1 , modified by attaching a glass sphere, as described by Ducker et al. (1, 2). Prior to force measurements, the colloidal probe attached to the cantilever was immersed in 0.1 mol dm−3 KOH overnight, followed by a thorough deionized water rinse. Glass plates were washed following the same procedure. After washing, the probe and the plate were both set at their fixed positions on the AFM and immediately immersed in the sample solution. The system was allowed to equilibrate for 30 min prior to the initial injection and force runs. The temperature was maintained constant at 25◦ C using a thermomodule controller (MT86204C12, Netsu Denshi Co. Ltd., Japan). Dispersion stability of silica suspensions was evaluated by using a Turbiscan MA2000 (Formulaction, France). The suspension (0.01 g silica/10 cm3 ) was shaken in a water bath for 3 h at 25◦ C and then transferred to a test tube. Change in transmission of the test tube was monitored as a function of elapsed time.

RESULTS Figure 1 shows the normalized force values between PAMAM G3 dendrimer adlayers on glass. The initial dendrimer concentrations were set to 0, 10, and 50 ppm, and the pH value of each sample solution was adjusted to 5.0 by using HCl. The total concentration of background salts was held constant at 1 mmol dm−3 by the addition of NaCl. In the absence of PAMAM dendrimer, repulsion was observed in the range of 30 nm. The Decay length determined by best curve fitting is 9.4 nm, which is in good agreement with the theoretical Debye length (9.6 nm). This indicates that the repulsion is induced by the electric double-layer interaction. The interaction force was dramatically changed by the addition of PAMAM G3 dendrimer; that is, an attractive interaction was detected in the 10 ppm dendrimer solution while a repulsive force reappeared at 50 ppm. The range of repulsion is 18 nm. The empirically derived decay length is 8.9 nm, which is in agreement with the Debye length at the effective electrolyte concentration under this experimental condition (8.7 nm). Thus, it is found that the repulsion between the adlayers is also originated from the electric double-layer interaction. Interaction forces between glass surfaces with the adsorption of C8qbG0 are shown in Fig. 2. The initial C8qbG0 concentration was fixed at 0, 5, 10, and

0.3

0.2

0 ppm 5 ppm 10 ppm 50 ppm

0.2

F/R (mN/m)

F/R (mN/m)

0.3

0 ppm 10 ppm 50 ppm

0.1

0.1

0.0

0.0 -0.1

-0.1 0

10

20

30

40

Distance (nm)

-0.2 0

10

20

30

40

Distance (nm) FIG. 1. Compression force data between glass surfaces in aqueous PAMAM G3 dendrimer solutions containing 1 mmol dm−3 background salts (NaCl and HCl) at pH 5.

FIG. 2. Compression force data between glass surfaces in aqueous C8qbG0 solutions containing 1 mmol dm−3 NaBr at pH 7.

408

Transmission (%)

NOTE

60

50 PAMAM G3 dendrimer

Transmission (%)

40 60

50 C8qbG0

40 0

10

20

30

40

50

60

Elapsed time (minutes) FIG. 3. Dispersion stability of silica suspensions with the adsorption of PAMAM G3 dendrimer at 0 ppm (), 10 ppm (), and 50 ppm () and of C8qbG0 at 0 ppm (), 5 ppm (), 10 ppm (), and 50 ppm (). All the samples contained 1 mmol dm−3 background salts in a similar manner as force measurements and the pH value was adjusted to 5 for PAMAM G3 dendrimer and to approximately 7 for C8qbG0, respectively.

50 ppm. All sample solutions contained 1 mmol dm−3 NaBr as a background electrolyte and the pH value of each solution was approximately 7. In the 5 ppm C8qbG0 solution a strong attraction was detected at a separation of 14 nm. By contrast, the interaction changed to repulsion at 10 and 50 ppm and their decay and Debye length values are in agreement with each other (the former length is 9.2 nm and the latter one is 9.5 nm at 10 ppm; 8.8 and 9.1 nm at 50 ppm). Thus, the repulsion is originated from the electric double-layer interaction. Here, our calculations assume C8qbG0 behaved as a simple 1:1 electrolyte capable of dissociating fully. It is notable that an obvious jump-in is observed at 50 ppm. This phenomenon corresponds to the squeeze-out of C8qbG0 adlayers under a limiting compression. The jump-in distance was 6.7 nm, which is four times longer than the hydrodynamic radius of C8qbG0 in pure water (1.31 nm (23)). Comparing the resultant forces with dispersion stability of suspensions is one of the aims of this investigation. Figure 3 shows the changes in transmission of silica suspensions as a function of elapsed time. The value of the vertical axis represents the mean value of transmission at each sample height. Unstable suspensions were obtained at 10 ppm for PAMAM G3 dendrimer and at 5 ppm for C8qbG0, respectively. The results obtained here match the prediction of the resultant forces in which attractive components play a key role at these dendrimer concentrations. Accordingly, it is obvious that the difference in the interaction forces between adlayers leads to the dispersion–flocculation– redispersion sequence of silica suspensions with the dendrimer concentration on a macroscopic level.

DISCUSSION Since PAMAM dendrimers with surface amino groups have an outer shell of primary amines with pK a values in the range of 7–9 (25), the terminal amino groups are proton-donated at pH 5. As a result, adsorption of the dendrimers on negatively charged glass occurs through the electrostatic attraction force between them. In fact, the adsorbed amount of the dendrimers increases with increasing concentration and then reaches a plateau value, regarded as a typical Langmuirian adsorption isotherm (17). Under this situation, the conformation of the dendrimers organized at the solid/solution interface has been of interest in recent years.

In the 10 ppm PAMAM G3 dendrimer solution, at surface separations less than approximately 5 nm the interaction force becomes attractive, as shown in Fig. 1. In addition, sedimentation experiments show that dispersion stability of silica suspension with the adsorption of the dendrimer from this concentration is significantly low (Fig. 3). These results indicate that the electrostatic component of the force is minimized due to the neutralization of the negative surface charge through the adsorption of the oppositely charged dendrimer. Furthermore, the electrostatic patch theory is relevant as presented by Bremmell et al. (9), where bridging between surfaces could occur even when the overall surface charge is neutral because patches of positive and negative sites coexist and may attract each other at small surface separations. On the contrary to this case, in the 50 ppm PAMAM G3 dendrimer solution the repulsion in the range of 18 nm was observed. It is necessary to point out that the probe’s point of zero separation for a bare surface does not correspond exactly with the zero point for an adsorbed surface due to the incompressible adsorbate layer sandwiched between the solid surface and the probe in the latter case (5, 6). The obtained force curve showed a monotonous increase with decreasing apparent separation between two surfaces. This result means that the adsorbed dendrimer layers remain trapped even when the surfaces are closely approached with each other. In addition, the resultant force data suggest that the electric double-layer interaction is predominant, while the steric hindrance due to the protrusion of the individual dendrimers adsorbed on the surfaces is negligible. Thus, PAMAM dendrimers with surface amino groups form a flat adlayer on the negatively charged glass surfaces and then the electrostatistically stable suspension was successfully obtained (Fig. 3). A similar finding has been presented by the direct visualizations of the dendrimer film. Imae et al. (26) investigated the self-assembling of PAMAM G4 dendrimer with proton-donated surface amino groups on mica and found that the adsorbed layer is uniformly flat, accompanying no ordering of the dendrimers. On the other hand, individual dendrimers deposited on solid substrates are no longer spherical but instead oblate shape (27, 28). This is due to a strong interaction between terminal functional groups of dendrimers and activated adsorption sites on substrates. The compression of dendrimers at the interface has also been suggested by the analysis of the adsorption isotherms, i.e., the occupied areas of the adsorbed dendrimers at the adsorption plateau region are almost 3–4 times larger than the cross-sectional areas of the corresponding dendrimers in aqueous media (17). In spite of the fact that C8qbG0 and PAMAM G3 dendrimer are almost equal in the hydrodynamic radius (1.31 nm for C8qbG0 (23) and 1.75 nm for PAMAM G3 dendrimer (29)), the adlayer–adlayer interactions for these adsorbates were quite different. This can be attributed to a view that C8qbG0 behaves as a tetrameric surfactant (30) in which four quaternary ammonium species are linked at the level of the head groups by a spacer. In fact, the surface tension of C8qbG0 aqueous solution decreases with increasing concentration and then remains constant above 5500 ppm (critical micelle concentration, cmc) (23). This new type of surfactant, called gemini, has been attracted owing to its superior performance on interfacial orientation. We have proposed in our previous experiments that the charging-up is relevant in the adsorption of conventional ionic surfactants (14, 24). In the present study very similar force profiles were also obtained, as shown in Fig. 2: (i) that a strong attraction was detected at 5 ppm suggests the adsorption of C8qbG0 neutralizes the negative charge on the surface, accompanying the formation of a hydrophobic layer; (ii) repulsive forces induced by the electric double-layer interaction were observed at both 10 and 50 ppm. It is surprising that the charge reversal occurs at concentrations far below its cmc even in the presence of 1 mmol dm−3 background salts. This is probably because adsorbed C8qbG0 molecules are packed more tightly than conventional surfactant molecules, resulting in the formation of a patchy layer. With the advent of the AFM imaging technique, it is now possible to focus on the self-assembly of surfactant molecules at the solid/liquid interface. Manne et al. (31) investigated the layered structure of cationic gemini surfactants adsorbed on mica and found that the surfactant packing parameter (32) plays an analogous role in interfacial aggregation as it does in bulk solution. From the point of view, one would expect that the C8qbG0 adlayer consists of a bilayer: the adsorption proceeds with the orientation of its hydrocarbon tails toward

409

NOTE the aqueous solution and following the formation of the outer layer through hydrophobic interactions. In addition, the fact that an obvious jump occurs at a separation of 6.7 nm in the 50 ppm C8qbG0 solution also indicates the formation of such a bilayer.

REFERENCES 1. Ducker, W. A., Senden, T. J., and Pashley, R. M., Nature 353, 239 (1991). 2. Ducker, W. A., Senden, T. J., and Pashley, R. M., Langmuir 8, 1831 (1992). 3. Rutland, M. W., and Senden, T. J., Langmuir 9, 412 (1993). 4. Bremmell, K. E., Jameson, G. J., and Biggs, S., Colloids Surf. A 146, 75 (1999). 5. Biggs, S., Langmuir 11, 156 (1995). 6. Biggs, S., J. Chem. Soc. Faraday Trans. 1 92, 2783 (1996). 7. Braithwaite, G. J. C., Howe, A., and Luckham, P. F., Langmuir 12, 4224 (1996). 8. Biggs, S., and Proud, A. D., Langmuir 13, 7202 (1997). 9. Bremmell, K. E., Jameson, G. J., and Biggs, S., Colloids Surf. A 139, 199 (1998). 10. Giesbers, M., Kleijn, J. M., Fleer, G. J., and Cohen Stuart, M. A., Colloids Surf. A 142, 343 (1998). 11. Bremmell, K. E., Jameson, G. J., and Biggs, S., Colloids Surf. A 155, 1 (1999). 12. Muir, I., Meagher, L., and Gee, M., Langmuir 17, 4932 (2001). 13. McNamee, C. E., Matsumoto, M., Hartley, P. G., Mulvaney, P., Tsujii, Y., and Nakahara, M., Langmuir 17, 6220 (2001). 14. Sakai, K., Yoshimura, T., and Esumi, K., Langmuir 18, 3993 (2002). 15. Imae, T., Funayama, K., Aoi, K., Tsutsumiuchi, K., Okada, M., and Furusaka, M., Langmuir 15, 4076 (1999). 16. Zeng, F., and Zimmerman, S. C., Chem. Rev. 97, 1681 (1997). 17. Esumi, K., and Goino, M., Langmuir 14, 4466 (1998). 18. Esumi, K., Fujimoto, N., and Torigoe, K., Langmuir 15, 4613 (1999). 19. Esumi, K., Fujimoto, N., Torigoe, K., and Koide, Y., J. Jpn. Soc. Colour Mater. 73, 290 (2000). 20. Esumi, K., Sakagami, K., Kuniyasu, S., Nagata, Y., Sakai, K., and Torigoe, K., Langmuir 16, 10264 (2000).

21. Esumi, K., Nakaie, Y., Sakai, K., and Torigoe, K., Colloids Surf. A 194, 7 (2001). 22. Tomalia, D. A., Baker, H., Dewald, J., Hall, M., Kallos, G., Martin, S., Roeck, R. J., and Smith, P., Macromolecules 19, 2466 (1986). 23. Yoshimura, T., Fukai, J., Mizutani, H., and Esumi, K., J. Colloid Interface Sci., to appear. 24. Sakai, K., Torigoe, K., and Esumi, K., Langmuir 17, 4973 (2001). 25. Tomalia, D. A., Baker, H., Dewald, J., Hall, M., Kallos, G., Martin, S., Roeck, J., Ryder, J., and Smith, P., Polym. J. 17, 117 (1985). 26. Imae, T., Ito, M., Aoi, K., Tsutsumiuchi, K., Noda, H., and Okada, M., Colloids Surf. A 175, 225 (2000). 27. Tsukruk, V. V., Rinderspacher, F., and Bliznyuk, V. N., Langmuir 13, 2171 (1997). 28. Li, J., Piehler, L. T., Qin, D., Baker, J. R., Tomalia, D. A., and Meier, D. J., Langmuir 16, 5613 (2000). 29. Uppuluri, S., Keinath, S. E., Tomalia, D. A., and Dvornic, P. R., Macromolecules 31, 4498 (1998). 30. Zana, R., In, M., L´evy, H., and Duportail, G., Langmuir 13, 5552 (1997). 31. Manne, S., Sch¨affer, T. E., Huo, Q., Hansma, P. K., Morse, D. E., Stucky, G. D., and Aksay, I. A., Langmuir 13, 6382 (1997). 32. Israelachvili, J. N., Mitchell, D. J., and Ninham, B. W., J. Chem. Soc. Faraday Trans. 1 72, 1525 (1976). Kenichi Sakai Saiki Sadayama Tomokazu Yoshimura Kunio Esumi1 Department of Applied Chemistry and Institute of Colloid and Interface Science Science University of Tokyo Kagurazaka, Shinjuku-ku Tokyo 162-8601, Japan Received April 10, 2002; accepted July 12, 2002

1

To whom correspondence should be addressed.