Hydrophobically modified poly(amidoamine) (PAMAM) dendrimers: their properties at the air–water interface and use as nanoscopic container molecules Yasmin Sayed-Sweet, David M. Hedstrand, Ralph Spinder* and Donald A. Tomalia* MichiganMolecular Institute, 1910 W. St. Andrews Road,Midland,MI 48640, USA Tri- and tetra-dendron poly(amidoamine) (PAMAM) dendrimers were converted into hydrocarbon-soluble polymers and used as hydrophobic nanoscopic scaolding by reacting their primary amino chain ends with various epoxyalkanes.These hydrophobically modified modules performed well as nanoscopic transport molecules. They mimicked classical inverse micelle behaviour by transporting copper(II) sulfate from an aqueous solution into an organic phase to form homogeneous, transparent, intensely blue toluene solutions.The modified dendrimers were examined at the air–water interface both with and without copper guest molecules. A number of critical macromolecular design parameters (CMDPs) such as generation (size), core (shape, topology) and surface groups were varied to determine their influence on Langmuir film properties. Nanostructures are macromolecular assemblies containing which is associated with most classical covalent polymerization methodology.The major shortcoming of these approaches is from 103 to 109 atoms with molecular masses of 103 to 1010 Da. Their dimensions vary from 1 to 102 nm. These nano- the inability to readily isolate precise macromolecular structures or control critical macromolecular design parameters dimensions and masses have very broad implications in such traditional disciplines as polymer science,1a–b catalysis,2 interfa- (CMDPs) such as: (a) size, (b) shape, (c) surface chemistry and (d) topology.On the other hand, nature solved this problem cial/colloid science,3 supramolecular chemistry,4 electronic microfabrication5a –b and molecular biology.6 It is quite apparent some 4.5 billion years ago during a critical phase in our natural molecular evolution from atoms to very complex molecular that the development of viable synthetic and characterization methodology which will allow systematicexamination of ‘struc- structure.The reproducible synthesis of proteins, DNA, RNA and bio-assemblies possessing precise CMDP controlled fea- ture-controlled nanostructures’ would be of significance to these evolving areas.7–9 For that reason, considerable inter- tures are prime examples of this success.Using strictly abiotic synthetic methods, it has been widely national interest has focused on these objectives. Synthetic polymer chemists routinely produce nano-struc- demonstrated over the past decade that dendrons, dendrimers10 –13 and more recently dendrigrafts3,14 can be routinely tures as part of a statisticalmacromolecular product continuum Plate 1 Scaled comparison of tri-dendron (NH3 core) poly(amidoamine) dendrimers; generation=4–7, sizes and shapes with various proteins, DNA and bio-assemblies J. Mater.Chem., 1997, 7(7), 1199–1205 1199constructed with CMDP control that rivals the structural dendrimers: a G2(NH3) dendrimer (0.25 g, 2.5 mmol) was dissolved in 10 ml of methanol and 0.35 g (4.0 mmol) of 1,2- regulation found in biological systems.Acomparison of various nanoscopic biostructures with tri-dendron poly(amidoamine) epoxyhexane was added. The reaction solution was heated for five days at 40°C. Additional methanol was added to the (PAMAM) dendrimers clearly illustrates the close scaling of size and shape that is possible (see Plate 1).reaction solution to give a final dendrimer concentration of 5 mass%, followed by filtering through a 0.2 micron Teflon filter. The ability to prepare well-defined dendrons and dendrimers15 leads to the concept of using these materials as funda- This solution was ultrafiltered using a flatstock Amicon unit employing a YM3 membrane with a MWCO of 1000.Fourteen mental nanoscopic building blocks, suitable for the construction of nanoscopic compounds16a,b and other complex recirculations were required to completely remove the excess epoxide. Methanol was removed by rotoevaporation and the assemblies. More recently, these dendritic materials have been identified as nanoscopic scaolding for catalysis,17,18 gene dendrimer was dried at room temperature under vacuum to give 0.37 g (80% yield) of modified product.dc (CDCl3 ): vectors,19–21 magnetic resonance imaging agents,22 electron conduction,23 photon transduction,24a–b as well as nanoscopic 172.6–172.9 (m), 69.8, 68.1, 67.8, 67.7, 64.3, 61.9, 56.6, 55.2, 54.9, 52.3, 50.0, 49.1, 38.5, 37.5, 37.2, 34.6, 34.3, 33.7, 33.2, 27.8, host compartments or ‘unimolecular regular micelle mimics’ suitable for the containment of pharmaceuticals,agrochemicals, 27.7, 22.6, and 13.9. ESI–MS: theoretical MW=4817; observed MW=4818.5 (MH+).dyes and other guest molecules.25a–c,32,48 In this paper, we describe the facile conversion of amine terminated hydrophobic poly(amidoamine) dendrimers into Copper(II ) transport studies—‘blue toluene-chloroform ‘unimolecular inverse micelle’ modules by surface modification experiments’ with hydrophobic epoxy reagents.We use the term ‘unimolecu- A 3.9 mass% solution of epoxydecane-functionalized G4- lar inverse micelle’ to dierentiate these covalently fixed macro- (EDA) C10 PAMAM dendrimer in toluene was carefully lay- molecular structures from the dynamic equilibrium structure ered on top of a 0.1 M CuSO4 solution.After standing for of a ‘classic’ inverse micelle. The resulting hydrocarbon-soluble several hours at room temperature, a dark-blue colour emerged assemblies functioned as nanoscopically sized container mol- above the aqueous–organic interface. The experiment was ecules as evidenced by the transport of copper(II) salts into continued for several days over which time the toluene solution various hydrocarbon solvents.In this fashion, transparent, assumed a very intense, transparent blue colour. intensely blue toluene and chloroform solutions were formed. Copper(II)-containing hydrophobically modified dendrimers In order to understand better the phase transfer properties of for Langmuir film studies were prepared by the following these non-classical micelles, both tri- and tetra-dendron dendri- procedure: an 1,2-epoxyoctane-modified G5(NH3) C8 mers were examined with and without copper(II) guest mol- PAMAM dendrimer (0.30 g) was dissolved in 2 ml of chloro- ecules as Langmuir films at the air–water interface.form, and aqueous copper sulfate (30 mg in 2 ml water) was placed on top of the chloroform solution.The mixture was Experimental agitated on an orbital shaker for 24 h. After that time, no blue copper colour remained in the aqueous phase. The aqueous General procedures phase was decanted and the blue chloroform solution was The 13C (75.4 or 90.5 MHz) NMR spectra were recorded on dried with anhydrous sodium sulfate. Filtration and removal either a Varian Unity 300 or a Bruker WM360 SF spec- of the solvent by distillation in vacuo gave 0.24 g of a blue oil: trometer.The 13C NMR spectra were recorded in CDCl3 using dc (CDCl3): 173, 69, 51, 35, 32.0, 29.7, 25.8, 22.8, 14.2. the solvent line as the standard (77 ppm) or in D2O using 1,4- Over the time frame of these experiments (ca. one month), dioxane as an internal standard (66.3 ppm).Electrospray ioniz- there was no indication that the incorporation of copper ions ation mass spectroscopy was obtained using a Finnigan TSQ- into these dendrimers resulted in degradation of the dendrimer 700 spectrometer. The spectra were deconvoluted using the structure. BIOMASS software. Visible spectra were recorded on a Varian Cary 1 spectrometer using matched cells with pure solvent in Monolayer studies the reference beam.The PAMAM dendrimers were prepared ALauda FW-2 film balance was used to examine the properties as previously described.26a–c The conditions used to prepare of the dendrimer monolayers. Distilled water from a Millipore the PAMAM dendrimers employed reaction stoichiometries system (18 MV) was used as the subphase and maintained at and rigorous purification procedures which minimize both 23°C.Millimolar solutions of the dendrimers were prepared intramolecular (missing repeat units, intramolecular loops) and in CHCl3. In a typical experiment, dilute dendrimer solutions intermolecular (dimers, lower generation dendrimers) defects were added dropwise to the top of the subphase, while the that have been described and characterized previously.27 The solvent was allowed to evaporate before another drop was intramolecular defects contained in these samples can be added.This procedure was repeated until all the solution was thought of as ‘quantized’ coproducts, since they are exact delivered. In order to ensure that residual solvent was not multiples of the molecular mass of the desired structure.This present, dendrimer was equilibrated for 10–20 min before the is in contrast to defects found in conventional polymers which experiment was initiated. Surface pressure vs. area isotherms cover a broad Gaussian distribution, even in the case of were measured at a compression rate of 25–30 cm2 min-1. polymers prepared by living polymerization techniques. The The limiting areas reported in this paper are an average of at epoxyalkanes [1,2-epoxyhexane (99%), 1,2-epoxyoctane least three measurements.(99.9%), 1,2-epoxydecane (99.9%) and 1,2-epoxydodecane (99.9%)] were purchased from either Lancaster or Aldrich. Methyl acrylate (99.9%), 1,4-diaminobutane (99.9%), 1,8- Results and Discussion diaminooctane (99.9%), and 1,12-diaminododecane (99.9%) Hydrophobic modification of dendrimers were all purchased from Aldrich. Ethylenediamine was received from Fisher and distilled before use.Perhaps one of the unique features of dendrimeric architecture, compared to classical random coil polymers, is the large Surface modification of PAMAM dendrimers number of well-defined chain ends. Several studies have shown that changing the chemical nature of the surface groups can Modification of generation 2, PAMAM dendrimers, ammonia core, G2(NH3), with 1,2-epoxyhexane is representative of the dramatically aect the physical properties of these polymers.For example, Fre� chet and co-workers have shown that poly- general method used to prepare hydrophobically modified 1200 J. Mater. Chem., 1997, 7(7), 1199–1205(ether) dendrimers can be transformed from organic solventsoluble polymers to water-soluble materials by a simple surface group transformation.28a–b It is important to note that other physical properties such as glass transitions are also influenced by the nature of the surface groups.29a–b Critical architectural components of the dendrimers such as the nature of the surface groups/chemistry, repeat unit composition and degree of branching all undoubtedly influence their physical properties and hence many of their eventual applications.At this point, the interplay between these parameters is not totally understood. In this study, we focused on the surface modification of water-soluble, amine-terminated PAMAM dendrimers to form ‘unimolecular inverse micelle’ prototypes which would be soluble in organic solvents such as toluene or chloroform.This transformation was easily accomplished by reacting the Plate 2 PAMAM dendrimer’s terminal primary amino groups with a variety of hydrophobic epoxyalkanes. These reactions were typically run in methanol, although modifications employing solved in toluene to form a clear solution.This solution was long chain epoxyalkanes, such as 1,2-epoxydodecane, required layered on a 0.1 M aqueous CuSO4 solution. During this the addition of a cosolvent (toluene) to maintain homogeneous quiescent experiment, a gradient of a dark-blue colour slowly reaction conditions. We found the reaction could be run either diused up from the organic–aqueous interface into the toluene with an excess of epoxide, which was then removed through layer.After a period of several days, the entire organic layer ultrafiltration of the product, or by simply utilizing stoichio- took on a dark-blue colour (see Plate 2). As expected, control metric amounts of the epoxide. By either method, we saw no experiments showed that copper ions were not transported indication that an ‘all or nothing’ distribution of modified and into the organic phase in the absence of dendrimer.Fig. 1 unmodified dendrimers formed, as has recently been reported compares the UV–VIS spectrum of the blue toluene layer with by Meijer and co-workers for the reaction of alkyl acid that of the starting aqueous CuSO4 solution. The blue-shift of chlorides with POPAM dendrimers.30 Even in experimental the absorption maximum in the visible region is characteristic runs where sub-stoichiometric quantities of epoxyalkane were of copper coordinated by amine ligands,40 which in this case used, only a modified product with a statistical distribution of could only come from the interior of the dendrimer.The ability surface alkyl groups centred around the stoichiometric value of the interior of a dendrimer to coordinate copper ions has calculated for the reaction was observed.31 The products were been extensively probed recently by EPR spectroscopy.41 This characterized by 13C NMR spectroscopy, and for the low example clearly shows that the dendrimer acts as a covalently molecular mass products, electrospray ionization mass spec- fixed phase transfer agent, with sucient interaction at the troscopy was also used.27b–c interface between the organic and aqueous solution to allow the interior of the dendrimer to coordinate copper ions.Further studies will be needed to determine the potentially complex Transport of copper(II ) salts into an organic solvent to form solution structure of these dendrimer unimolecular inverse ‘blue toluene’ micelles.The ability to use dendrimers as container molecules has continued to excite wide scientific interest. Encapsulation of Characterization of dendrimer monolayers at the air–water guest molecules into the interior of dendrimer hosts was first interface demonstrated by the incorporation of acetylsalicylic acid or Properties of synthetic polymers at the air–water interface 2,4-dichlorophenoxyacetic acid into ester-terminated PAMAM have been studied for many years.42a–b Generally, polymers dendrimers as early as 1989.32 We have referred to this have the ability to act as amphiphiles either through the phenomenon as ‘unimolecular encapsulation.’33 As an elegant interaction of polar groups on the chain of the polymer,43 continuation and expansion of this theme, Meijer and co- through the special construction of block/star polymers con- workers have described a so-called ‘dendrimer box’ phenom- sisting of both hydrophilic and hydrophobic chains,44a–b or by enon.25 It involves the trapping of certain guest molecules in preparing polymer surfactants with hydrophilic chain ends.45 the interior void spaces by post reaction with bulky surface group reagents.Dendritic micellar behaviour has been demonstrated on numerous occasions by the dissolution of organic molecules in dendrimers34a–e or by the polymerization of water-insoluble monomers in the interiors of carboxy-terminated dendrimers to form novel linear-dendritic composites.35 On the other hand, some dendrimers have been employed as micellar structures in electrokinetic capillary chromatography.34f,h Considerable experimental and characterization work by Turro et al.36–39 has unequivocally demonstrated the ‘unimolecular micelle’ features of dendrimers.In all of these cases, dendrimers are regarded as regular unimolecular micelles, which consist of a non-polar core and a polar outer shell. Evidence that modified PAMAM dendrimers behaved as unimolecular inverse micelles, and that the interior space of the dendrimer remains functionally active was clearly demonstrated by the following experiment: a G4(EDA) PAMAM Fig. 1 Visible spectrum of (a) an aqueous 0.1 M CuSO4 solution and dendrimer which was obtained from the exhaustive reaction (b) the complex formed betweean epoxydecane-modified G4 (EDA) PAMAM dendrimer and CuSO4 in toluene of amine-terminated dendrimer with 1,2-epoxyoctane was dis- J.Mater. Chem., 1997, 7(7), 1199–1205 1201Polymers that have no amphiphilic character at all can still form films at the air–water interface.46 Recently, first reports of the properties and structure of poly(ether) dendrimers at the air–water interface were published.47a–b Only dendrons of generation 4 and below, with hydroxy groups at the focal point, were found to act as surfactant-like molecules, while the higher generation dendrons and all the poly(ether) tri-dendron dendrimers above generation 1 did not act as surfactant-like polymers or show the ability to form multilayer structures.For these experiments, PAMAM dendrimers, prepared from both ammonia and various alkylenediamine core molecules, were exhaustively modified with several epoxyalkanes.These cores are illustrated in Scheme 1. The modified dendrimer samples were carefully placed on the water surface of a Langmuir trough as dilute chloroform solutions. In order to ensure total removal of the solvent, the dendrimer was equilibrated at the interface for 10–20 min.A typical isotherm Fig. 2 Langmuir isotherm of a G3(EDA) C8 PAMAM dendrimer obtained for all of the hydrophobically modified dendrimers modified with epoxyoctane we investigated is shown in Fig. 2. As the area available to the molecules decreases, the surface pressure increases until a plateau is reached. Decreasing the area available to the dendrimers beyond this point presumably causes the monolayer film to collapse and form multilayer structures, while the surface pressure remains constant upon further compression. This response is quite dierent from that observed by Fre� chet for poly(ether) dendrimers,47a–b where a nucleation phenomenon was observed for the low generation dendrons.The general shape of the isotherms was shown not to be dependent on the compression rate.The isotherms were found to be reversible, with only slight changes in the collapse pressure, as long as the surface pressure applied to the film remained below the collapse point. Addition of sulfuric acid to the subphase (pH= 2) did not strongly aect the shape of the isotherms, although the collapse pressure increased 10–15%. One of our goals was to examine the eect of certain critical macromolecular design parameters, such as the core type (multiplicity of branching sites and size of the core), the nature of the surface groups (length of the alkyl chain) and the dendrimer generation (size), in influencing the Langmuir film data.For these experiments, the area occupied per dendrimer molecule was calculated by extrapolating to the x-axis from the point where the collapse curve and the linear region of the increasing pressure part of the isotherm meet.The dendrimers were assumed to be spherical molecules for these calculations. Table 1 summarizes the isotherm data for three surface modifications and five cores over a range of dendrimer generations. In Fig. 3 the surface radii obtained from the Langmuir data for the epoxyoctane-substituted PAMAM dendrimers are compared to the hydrodynamic radii of unsubstituted ammonia core dendrimers as determined by size exclusion chromatography (SEC).15b In general, the agreement is very good.The experimental areas determined from the isotherms increased as a function of generation. It is only at the highest generations that diameters determined from the isotherms diers significantly from the SEC data.Changing the hydrocarbon length on the dendrimer surface does not appear to have a significant eect on the surface area occupied by the dendrimer at the collapse point (Fig. 4). Even at the highest generation studied (generation 5), the limiting surface areas were all the same within the experimental errors associated with the measurements.We also examined the eect of the core length and multiplicity on the isotherm. Fig. 5 shows the surface area plotted vs. dendrimer generation for ammonia and ethylenediamine (EDA) core dendrimers. Both EDA and alkylenediamine core dendrimers build molecular mass 25% faster than dendrimers based on ammonia, since tetravalent cores produce tetra- Scheme 1 Tetra-dendron poly(amidoamine) (PAMAM) dendrimers amplified from various alkylenediamine cores dendron dendrimers, while ammonia core leads to tri-dendron 1202 J.Mater. Chem., 1997, 7(7), 1199–1205Table 1 Dendrimer isotherm dataa surface area/ surface area/ sample A° 2 per molecule sample A° 2 per molecule G0(NH3) C8 154 G2(EDA) C6 727 G1NH3) C8 284 G2(EDA) C8 877 G2(NH3) C8 756 G2(EDA) C12 798 G2(NH3) C12 702 G3(EDA) C6 1249 G3(NH3) C8 1211 G3(EDA) C8 1344 G3(NH3) C12 1147 G3(EDA) C12 1233 G4(NH3) C8 2804 G4(EDA) C6 2363 G4(NH3) C12 2848 G4(EDA) C8 2620 G5(NH3) C8 4369 G4(EDA) C12 3243 G5(NH3) C12 4742 G5(EDA) C6 5779 G2(Butane) C12 944 G5(EDA) C8 6449 G2(Octane) C12 889 G5(EDA) C12 5733 G2(Dodecane) C12 871 aNomenclature: the following nomenclature is used to describe the dendrimers prepared for this study.A fifth generation PAMAM dendrimer grown from an ethylenediamine core and modified with 1,2-epoxyhexane is noted as G5(EDA) C6. Fig. 5 A surface area vs. generation number plot for epoxydecane Fig. 3 Comparison of the radii of various generations of PAMAM dendrimers prepared from ammonia cores determined by size exclusion modified dendrimers grown from ammonia (#) and EDA ($) cores chromatography for the unmodified dendrimers ($) and by the limiting area measurements obtained from the Langmuir film studies for the epoxyoctane-modified dendrimers (#) Fig. 6 A plot of surface area vs. the number of carbons contained in the epoxyalkane chains for fifth generation dendrimers based on EDA ($) and ammonia (#) cores Fig. 4 Plot of surface area vs. generation number for PAMAM dendrimers grown from EDA cores and substituted with epoxyhexane (#), epoxyoctane ($) and epoxydecane (() Previous theoretical studies have suggested that the shape of PAMAM dendrimer becomes more highly spherical as a function of increasing generation number.32 Experimental con- dendrimers. There was no significant dierence in the limiting surface areas obtained by comparing the two types of cores firmation of the importance of this shape change has shown that a number of physical properties also exhibit dramatic for generations 2–4.At generation 5, Fig. 6 shows that dendrimers based on EDA cores have a surface area at collapse changes coincidental with the shape change.33 It may be possiblethat the increasing dierence seen in this study between 20–47% larger than that obtained for the ammonia core dendrimers depending on the length of the surface alkyl chain.the surface areas of the higher generation dendrimers may be J. Mater. Chem., 1997, 7(7), 1199–1205 1203influenced by this shape change. To further investigate the Model of hydrophobically modified PAMAM dendrimers at the eect of core length, a number of dendrimers with longer tetra- air–water interface dendron cores (1,4-diaminobutane, 1,8-diaminooctane and With these limited data, one can only speculate as to how 1,12-diaminododecane) were also examined.These data, pre- these hydrophobically modified dendrimers are organized at sented in Fig. 7, indicate that for the lower generation dendri- the air–water interface.Since the surface pressure vs. area mers there is no significant dierence as a function of the core curves do not show significant shape changes as a function of length. As was seen for the comparison between the ammonia the dendrimer size, the terminal hydrophobe length or the and EDA cores in Fig. 6, it is possible that a greater eect length/hydrophobicity of the core, it appears that the isotherms would be seen at higher generations for these dendrimers.are either not sensitive to these structural parameters, or that all of the dendrimers examined are interacting with the sub- Copper(II ) salt-containing dendrimers at the air–water interface phase in a very similar manner. One model which may be A key question was whether these copper-loaded dendrimers appropriate for the wer generation PAMAM dendrimers is could still be organized at the air–water interface.This issue that the accessible hydrophilic dendrimer interior interacts was examined by using dendrimer copper(II) complexes pre- with the aqueous subphase while the hydrophobically modified pared in chloroform solutions, which were introduced at the terminal groups reorganize to extend outward away from the air–water interface as described previously.Fig. 8 shows an air–water interface. Since the length of the hydrophobic chain isotherm obtained for a G5(NH3) C8 sample loaded with does not seem to impact the surface area occupied by the copper. In general, all the copper-loaded samples exhibit dendrimer, either the chains from adjacent dendrimers are able isotherms that are similar in shape to those containing no to interdigitate or they extend upward away from each other.copper. The collapse pressure increased slightly upon incorpor- For the low pH experiments (i.e. pH=2), the subphase was ating copper, which may indicate a slight stiening of the acidic enough to protonate the interior tertiary amines (pKa dendrimer interior due to copper complexation.Over the range ca. 4.5), which should help to increase the interaction between of dendrimers investigated [i.e. G2(NH3) C8 to G7(NH3) C8], the dendrimer interior and the subphase, thereby causing the no dierences were observed in the isotherms except for the observed larger collapse pressure. At higher generations, it higher collapse pressure noted above.This work clearly demon- would seem that a dendrimer reorganization of this type would strates that metal-loaded dendrimers can be readily organized become increasingly more dicult. A second model one must into two-dimensional layers. consider is that the dendrimers are simply acting like hydrophobic spheroids floating at the air–water interface.At this point, it is dicult to tell which of these models is the most probable. Perhaps there is a transition from the first model which may be operable for the lower generations to the floating hydrophobic spheroid model for the more congested higher generations. Conclusions Hydrophilic PAMAM dendrimer scaolds were readily converted to hydrophobic modules by facile reactions of amine groups with epoxyalkanes.The modified dendrimers perform as nanoscopic container molecules, reminiscent of ‘unimolecular inverse micelles’ as demonstrated by the transport of copper(II) ions from an aqueous solution into toluene or chloroform. These experiments, as well as other work,25,27,35 clearly show that the interior void spaces of dendrimers are Fig. 7 A plot showing the surface area versus the number of methylene available for the incorporation of guest molecules. The proper- units contained in the core for generation two PAMAM dendrimers ties of the modified PAMAM dendrimers at the air–water modified with epoxydodecane interface were most strongly influenced by the dendrimer generation. Even changing the length of the dendrimer surface hydrophobefrom hexyl to dodecyl did not cause any significant dierences in the limiting area of the dendrimer at the collapse point. Only for the highest generation dendrimers (generation 5) did the size of the core seem to influence the area taken up by the dendrimer at the collapse point. Further examination of these dendrimers structures at the air–water interfaces will be required to gain a complete understanding of these unique nanoscopic organizations.The authors wish to thank Drs June W. Klimash and Douglas R. Swanson for providing several of the dendrimers used in these studies. The Air Force Oce of Scientific Research (Contract No. AFOSR-91-0366) is gratefully acknowledged for their support of this work. D.M.H., R.S. and D.A.T. also acknowledge the support of Dendritech, Inc, the Army Research Laboratory as a sponsor of the MMI/ARL Dendritic Fig. 8 Langmuir isotherm of a G4 (NH3) C8 PAMAM dendrimer Polymer Center of Excellence, and the Army Research Oce surface modified with epoxyoctane and its interior void space loaded with copper ions (Grant No. DAAH04-95-1-0652). 1204 J. Mater. Chem., 1997, 7(7), 1199–120527 (a) P.B. Smith, S. J. Martin, M. J. Hall and D. A. Tomalia, in References Applied Polymer Analysis and Characterization, ed. J. Mitchell, 1 (a) P. R. Dvornic and D. A. Tomalia, Science Spectra, 1996, 5, 36; Hanser, 1987, pp. 357–385; (b) G. J. Kallos, D. A. Tomalia, (b) J.M. J. Fre� chet, Science, 1994, 263, 1710. D. M. Hedstrand, S. Lewis and J. Zhou, Rapid Commun. Mass 2 L. J. de Jongh, in Physics and Chemistry of Metal Cluster Spectrom., 1991, 5, 383; (c) P.R. Dvornic and D. A. Tomalia, Compounds, ed. F. Levy, Kluwer Academic Publishers, Dordrecht, Macromol. Symp., 1995, 98 (35th IUPAC Int. Symp. Macromol., 1994. 1994) 403. 3 P. R. Dvornic and D. A. Tomalia, Curr. Opin. Colloid Interface Sci., 28 (a) C. J. Hawker, K. L. Wooley and J. M. J. Fre� chet, J.Chem. Soc., 1996, 1, 221. Perkin T rans., 1, 1993, 1287; (b) C. J. Hawker and J. M. J. Fre� chet, 4 (a) J.-M. Lehn, Supramolecular Chemistry, VCH, New York, 1995; J. Chem. Soc., Perkin T rans., 1, 1992, 2459. (b) C. A. Mirkin, R. C. Mucic, J. J. Storho and R. L. Letsinger, 29 (a) K. L.Wooley, C. J. Hawker, J. M. Pochan and J. M. J. Fre� chet, Nature, 1996, 382, 607. Macromolecules, 1993, 26, 1514; (b) E.M. M. de Brabander-van 5 (a) A. Kumar, N. L. Abbott, E. Kim, H. A. Biebuyck and den Berg, A. Nijenhuis, M. Mure, J. Kuelen, R. Reintjens, G. M. Whitesides, Acc. Chem. Res., 1995, 28, 219; (b) R. Dagani, F. Vandenbooren, B. Bosman, R. de Raat, T. Frijns, S. v.d. Wal, ‘Nanoengineering’ in Chem. Eng. News, December 2, 1996, p. 20. M. Castelijns, J. Put and E.W. Meijer, Macromol. Symp., 1994, 6 P. R. Dvornic and D. A. Tomalia, Chem. Br., 1994, 30, 641. 77, 51. 7 G. M. Whitesides, J. P. Mathias and C. T. Seto, Science, 1991, 30 S. Stevelmans, J. C. M. van Hest, J. F. G. A. Jansen, D. A. F. J. van 254, 1312. Boxtel, E. M. M. de Brabander-van den Berg and E. W. Meijer, 8 J. Dutta, H. Hofmann and G. Schmid, Adv. Mater., 1996, 8, 555. J.Am. Chem. Soc., 1996, 118, 7398. 9 R. Dagani, ‘Nanostructured Materials’ in Chem. Eng. News, 31 L. T. Piehler and D. A. Tomalia, Michigan Molecular Institute, November 23, 1992, p. 18. unpublished results. 10 D. A. Tomalia and H. D. Durst, in T op. Curr. Chem. Vol. 165: 32 A. M. Naylor, W. A. Goddard III, G. E. Kiefer and D. A. Tomalia, Supramolecular Chemistry I—Directed Synthesis and Molecular J.Am. Chem. Soc., 1989, 111, 2339. Recognition, ed. E. Weber, Springer-Verlag Berlin Heidelberg, 33 D. A. Tomalia, A. M. Naylor and W. A. Goddard III, Angew. 1993, pp. 193–313. Chem., 1990, 102, 119; Angew. Chem., Int. Ed. Engl., 1990, 29, 138. 11 G. R. Newkome, C. N. Moorefield and F. Vo� gtle, in Dendritic 34 (a) D. A. Tomalia, V. Berry, M. Hall and D. Hedstrand, molecules, ed.G.Walter, VCH, New York & Weinheim 1996. Macromolecules, 1987, 20, 1164; (b) G. R. Newkome, 12 (a) N. Ardoin and D. Astruc, Bull. Soc., Chim Fr., 1995, 132, 875. C. N.Moorefield, G. R. Baker, M. J.Saunders and S. H.Grossman, (b) D. Astruc, C.R. Acad. Sci. Paris, 1996, 322, 757. Angew. Chem., 1991, 103, 1207; (c) K. L.Wooley, C. J. Hawker and 13 B. I. Voit, Acta Polymer., 1995, 46, 87.J. M. J. Fre` chet, J. Am. Chem. Soc., 1993, 115, 11496; (d) G. R. 14 R. Yin, D. R. Swanson and D. A. Tomalia, Polym. Mater. Sci. Eng., Newkome, J. K. Young, G. R. Baker, R. L. Potter, L. Audoly, 1995, 73, 277. D. Cooper and C. D. Weiss, Macromolecules, 1993, 26, 2394; 15 (a) D. A. Tomalia, J. R. Dewald, M. J. Hall, S. J. Martin and (e) C. J. Hawker, K.L.Wooley and J. M. J. Fre� chet, J. Chem. Soc., P. B. Smith, Preprints 1st SPSJ Int. Polym. Conf., Soc. of Polym. Perkin T rans. 1, 1993, 1287; ( f ) N. Tanaka, T. Tanigawa, Sci. Japan, Kyoto, 1984, p. 65; (b) D. A. Tomalia, H. Baker, K. Hosoya, K. Kimata, T. Araki and S. Teraba, Chem. L ett., 1992, J. Dewald, M. Hall, G. Kallos, S. Martin, J. Roeck, J. Ryder and 959; (g) S. A. Kuzdzal, C.A. Monning, G. R. Newkome and P. Smith, Polym. J. (T okyo), 1985, 17, 117. C. N. Moorefield, J. Chem. Soc., Ch 1994, 2139; 16 (a) D. A. Tomalia, Aldrichim. Acta, 1993, 26(4), 91, (h) P. G. H. M. Muijselaar, H. A. Claessens, C. A. Cramers, (b) D. A. Tomalia, Adv.Mater., 1994, 6, 529. J. F. G. A. Jansen, E. W. Meijer, E. M. M. de Brabander-van den 17 J. W. J.Knapen, A. W. van der Made, J. C. de Wilde, P. W. N. M. Berg and S. Van der Wal., J. High Resolut. Chromatogr., 1995, van Leeuwen, P. Wijkens, D. M. Grove and G. van Koten, Nature, 18, 121. 1994, 372, 659. 35 V. U. Wege and R. H. Grubbs, Polym. Prepr. Am. Chem. Soc., Div. 18 P. Bhyrappa, J. K. Young, J. S. Moore and K. S. Suslick, J. Mol. Polym. Chem., 1995, 36(2), 239. Catal. A: Chem., 1996, 113, 109. 36 M. C. Moreno-Bondi, G. Orellana, N. J. Turro and D. A. Tomalia, 19 M. X. Tang, C. T. Redemann and F. C. Szoka, Jr., Bioconj. Chem., Macromolecules, 1990, 23, 910. 1996, 7, 703. 37 K. R. Gopidas, A. R. Leheny, G. Caminati, N. J. Turro and 20 (a) J. F. Kukowska-Latallo, A. U. Bielinska, J. Johnson, D. A. Tomalia, J. Am. Chem. Soc., 1991, 113, 7335. R. Spindler, D. A.Tomalia and J. R. Baker, Jr., Proc. Natl. Acad. 38 M. F. Ottaviani, S.Bossman, N. J. Turro and D. A. Tomalia, J. Am. Sci. USA, 1996, 93, 4897; (b) A. Bielinska, J. Kukowska-Lattalo, Chem. Soc., 1994, 116, 661. L. T. Piehler, D. A. Tomalia, R. Yin and J. R. Baker, Jr., Polym. 39 M. F. Ottaviani, E. Cossu, N. J. Turro and D. A. Tomalia, J. Am. Mater. Sci. Eng., 1995, 73, 273. Chem.Soc., 1995, 117, 4387. 21 A. Bielinska, J. F. Kukowska-Latallo, J. Johnson, D. A. Tomalia 40 F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry, and J. R. Baker, Jr., Nucleic Acids Res., 1996, 24, 2176. Wiley, New York, 1980, pp. 815–817. 22 E. C. Wiener, M. W. Brechbiel, H. Brothers, R. L. Magin, 41 M. F. Ottaviani, F. Montalti, N. J. Turro and D. A. Tomalia, O. A. Gansow, D.A. Tomalia and P. C. Lauterbur, Magn. Reson. J. Phys. Chem. B, 1997, 101, 158. Med., 1994, 31, 1. 42 (a) G. L. Gaines, Insoluble Monolayers at L iquid–Gas Interfaces, 23 (a) L. I. Miller, T. Hashimoto, I. Tabakovic, D. R. Swanson and Interscience, New York, 1996, p. 264; (b) R. A. Hahn, in L angmuir– D. A. Tomalia, Chem.Mater., 1995, 7, 9; (b) R. G. Duan, L. I. Miller Blodgett Films, ed. G. Roberts, Plenum Press, New York, 1990, and D. A. Tomalia, J. Am. Chem. Soc., 1995, 117, 10783. pp. 68–73. 24 (a) C. Devadoss, P. Bharathi and J. S. Moore, J. Am. Chem. Soc., 43 R. H. G. Brinkhuis and A. J. Schouten, L angmuir, 1992, 8, 2247. 1996, 118, 9635; (b) Z. Xu and J. S. Moore, Acta Polym., 1994, 44 (a) X. Cha, R. Yin, X. Zhang and J. Shen, Macromolecules, 1991, 45, 83. 24, 4985; (b) S. Granick and J. Herz, Macromolecules, 1985, 18, 460. 25 (a) J. F. G. A. Jansen, E. M. M. de Brabander-van den Berg and 45 I. R. Gentle, P. M. Saville, J. W. White and J. Penfold, L angmuir, E. W. Meijer, Science, 1994, 266, 1226; (b) J. F. G. A. Jansen, 1993, 9, 646. E. W. Meijer and E. M. M. de Brabander-van den Berg, J. Am. 46 J. Kumaki,Macromolecules, 1988, 21, 749. Chem. Soc., 1995, 117, 4417; (c) J. F. G. A. Jansen, R. A. J. Janssen, 47 (a) P. M. Saville, J. W. White, C. J. Hawker, K. L. Wooley and E. M. M. de Brabander-van den Berg and E. W. Meijer, Adv. J. M. J. Fre� chet, J. Phys. Chem., 1993, 97, 293; (b) P. M. Saville, Mater., 1995, 7, 561. J. W. White, C. J. Hawker, J. M. J. Fre� chet, K. L. Wooley, 26 (a) D. A. Tomalia, H. Baker, J. Dewald, M. Hall, G. Kallos, J. Penfold and J. R. P.Webster, J. Phys. Chem., 1995, 99, 8283. S. Martin, J. Roeck, J. Ryder and P. Smith, Macromolecules, 1986, 48 R. Esfand, A. E. Beezer, J. C. Mitchell and I. J. Twyman, Pharm. 19, 2466; (b) D. A. Tomalia, V. Berry, M. Hall and Sci., 1996, 2, 1. D. M. Hedstrand, Macromolecules, 1987, 20, 1167; (c) D. A. Tomalia, M. Hall and D. M. Hedstrand, J. Am. Chem. Soc., 1987, 109, 1601. Paper 7/00860K; Received 6th February, 1997 J. Mater. Chem., 1997, 7(7), 1199–1205