Controlled Attachment of PAMAM Dendrimers to SolidSurfaces

C. A. Fail, S. A. Evenson, L. J. Ward, W. C. E. Schofield, and J. P. S. Badyal*

Department of Chemistry, Science Laboratories, Durham University,Durham DH1 3LE, England, United Kingdom

Received July 25, 2001. In Final Form: September 18, 2001

Amine-terminated polyamidoamine (PAMAM) dendrimers can be immobilized onto anhydride-functionalized pulsed plasma polymer surfaces via amide linkage formation. The packing density ofdendrimers at the surface can be tailored by programming the pulse duty cycle parameters during plasmapolymerization. These PAMAM dendrimer layers are shown to be useful for a variety of surface-relatedphenomena, for example, fluorination, adhesion, and gas barrier enhancement.

1. IntroductionDendrimers are highly ordered, three-dimensional,

treelike, functional polymers comprising branched repeatunits emanating from a central core.1 Their high densityof terminal groups provides a large number of reactivesites2 for potential application as nanoscale catalysts,micelle mimics, drug delivery agents, chemical sensors,high-performance polymers, and adhesives.3-6 A numberof studies have appeared in the literature describing theimmobilization of polyamidoamine (PAMAM) dendrimersonto solid surfaces (predominantly multistep wet chemicalstrategies); these include silica,7-13 gold,14-17 and mica18

substrates. In the case of covalent attachment of den-drimers to solid surfaces, an intermediate coupling layeris normally required.18-22 However, the substrate-specificnature of such reactions prohibits their more widescaleapplicability.

Here, we describe an alternative methodology based onthe aminolysis reaction23,24 between amine-terminateddendrimers and maleic anhydride pulsed plasma polymersurfaces. In this case, the concentration of anhydridefunctional groups present at the surface (and henceaminolysis with amine-terminated dendrimers) can becontrolled by programming the electrical discharge pa-rameters (e.g., duty cycle, peak power, etc.), Scheme 1.Furthermore, the reactivity of such surface-immobilizedPAMAM dendrimer structures toward trifluoroacetic acidas well as their adhesive and gas barrier performancehave been investigated.

* To whom correspondence should be addressed.(1) Tomalia, D. A.; Naylor, A. M.; Goddard, W. A., III Angew. Chem.,

Int. Ed. Engl. 1990, 29, 138.(2) Tomalia, D. A. Macromol. Symp. 1996, 101, 243.(3) Dagani, R. Chem. Eng. News 1996, 74 (June 3), 30.(4) Yoon, H. C.; Hong, M. Y.; Kim, H. S. Langmuir 2001, 17, 1234.(5) Service, R. F. Science 1995, 267, 458.(6) Bosman, A. W.; Janssen, H. M.; Meijer, E. W. Chem. Rev. 1999,

99, 1665.(7) Fujiki, K.; Sakamoto, M.; Sato, T.; Tsubokawa, N. J. Macromol.

Sci., Pure Appl. Chem. 2000, 37, 357.(8) Baker, L. A.; Zamborini, F. P.; Sun, L.; Crooks, R. M. Anal. Chem.

1999, 71, 4403.(9) Rubin, S.; Bar, G.; Taylor, T. N.; Cutts, R. W.; Zawodzinski, T.

A., Jr. J. Vac. Sci. Technol., A 1996, 14, 1870.(10) Bar, G.; Rubin, S.; Cutts, R. W.; Taylor, T. N.; Zawodzinski, T.

A. Langmuir 1996, 12, 1172.(11) Evenson, S. A.; Badyal, J. P. S. Adv. Mater. 1997, 9, 1097.(12) Tsukruk, V. V.; Rinderspacher, F.; Bliznyuk, V. N. Langmuir

1997, 13, 2171.(13) Bliznyuk, V. N.; Rinderspacher, F.; Tsukruk, V. V. Polymer 1998,

21, 5249.(14) Hierlemann, A.; Campbell, J. K.; Baker, L. A.; Crooks, R. M.;

Ricco, A. J. J. Am. Chem. Soc. 1998, 120, 5323.(15) Li, J.; Piehler, L. T.; Qin, D.; Baker, J. R.; Tomalia, D. A.; Meier,

D. J. Langmuir 2000, 16, 5613.(16) Yoon, H. C.; Kim, H. S. Anal. Chem. 2000, 72, 922.(17) Tokuhisa, H.; Zhao, M.; Baker, L. A.; Phan, V. T.; Dermody, D.

L.; Garcia, M. E.; Peez, R. F.; Crooks, R. M.; Mayer, T. M. J. Am. Chem.Soc. 1998, 120, 4492.

(18) Wells, M.; Crooks, R. M. J. Am. Chem. Soc. 1996, 118, 3988.(19) Tokuhisa, H.; Crooks, R. M. Langmuir 1997, 13, 5608.(20) Liu, Y.; Zhao, M.; Bergbreiter, D. E.; Crooks, R. M. J. Am Chem.

Soc. 1997, 119, 8720.(21) Liu, Y.; Bruening, M. L.; Bergbreiter, D. E.; Crooks, R. M. Angew.

Chem., Int. Ed. Engl. 1997, 36, 2114.(22) Miksa,B.;Slomkowski,S.;Chehimi,M.M.;Delamar,M.; Majoral,

J.-P.; Caminade, A.-M. Colloid Polym. Sci. 1999, 277, 58.

(23) Evenson, S. A.; Badyal, J. P. S. J. Phys. Chem. B 1998, 102,5500.

(24) Evenson, S. A.; Fail, C. F.; Badyal, J. P. S. Chem. Mater. 2000,12, 3038.

Scheme 1. Immobilization of PAMAM Dendrimersonto Maleic Anhydride Pulsed Plasma Polymer

Surfaces

264 Langmuir 2002, 18, 264-268

10.1021/la0111598 CCC: $22.00 © 2002 American Chemical SocietyPublished on Web 12/06/2001

2. Experimental Section

Briquettes of maleic anhydride (Aldrich, 99% purity) wereground into a fine powder and loaded into a monomer tube.Plasma polymerization was carried out in an electrodelesscylindrical glass reactor (4.5 cm diameter, 460 cm3 volume, basepressure of 5 × 10-3 mbar, and with a leak rate better than1.0 × 10-10 kg s-1) enclosed in a Faraday cage. The chamber wasfitted with a gas inlet, a Pirani pressure gauge, a 30 L min-1

two-stage rotary pump attached to a liquid nitrogen cold trap,and an externally wound copper coil (4 mm diameter, 9 turns,spanning 8-15 cm from the gas inlet). All joints were greasefree. An L-C network was used to match the output impedanceof a 13.56 MHz radio frequency (rf) generator to the partiallyionized gas load. In the case of pulsed plasma depositionexperiments, the rf power supply was triggered by a signalgenerator. The pulse width and amplitude were monitored withan oscilloscope. The continuous wave power output of the rf supplycould be set between 5 and 90 W, while pulse on- (ton) and off-periods (toff) were varied between 5-800 and 5-1200 µs,respectively. For each set of parameters, the average power ⟨P⟩delivered to the system was calculated using the followingexpression: ⟨P⟩ ) Pp{ton/(ton + toff)}, where ton/(ton + toff) is definedas the duty cycle and Pp is the peak power transmitted by therf generator.25

Prior to each experiment, the reactor was cleaned by scrubbingwith detergent, rinsing in propan-2-ol, oven drying, and thenrunning a 30 min high-power (50 W) air plasma treatment. Next,the system was vented to air and the substrate (glass unlessotherwise stated) was inserted into the center of the reactor,followed by evacuation back down to base pressure. Subsequently,monomer vapor was introduced into the chamber at a constantpressure of 0.26 mbar (the vapor pressure of maleic anhydrideat 20 °C) and a flow rate of approximately 1.6 × 10-9 kg s-1 for5 min prior to plasma ignition. Following 10 min of deposition,the rf generator was switched off, and the monomer feed wasallowed to continue to pass through the system for a further 5min before venting up to atmospheric pressure.

Drops of polyamidoamine Starburst dendrimer solution (Al-drich, supplied as 10% w/v in methanol) were pipetted onto theplasma polymer surfaces under a nitrogen atmosphere. Anyremaining unreacted dendrimers were removed by subsequentrinsing with methanol. Solutions containing different dendrimergenerations and concentrations were employed in order to studythe changes in dendrimer packing density at the surface. Forinstance, PAMAM dendrimer generation 2 has the followingstructure: [-CH2N[CH2CH2CONHCH2CH2N[CH2CH2CONHCH2-CH2N(CH2CH2CONHCH2CH2NH2)2]2]2]2 giving rise to 16 surfaceprimary amine groups. Similarly, generations 3 and 4 contain32 and 64 surface primary amine groups, respectively.

Film thicknesses were measured using an nkd-6000 spectro-photometer (Aquila Instruments Ltd), where the obtainedtransmittance-reflectance curves (over the 350-1000 nm wave-length range) were fitted to a Cauchy material model using amodified Levenburg-Marquadt method.

X-ray photoelectron spectroscopy (XPS) analysis was carriedout using a VG ESCALAB II electron spectrometer equippedwith a Mg KR1,2 X-ray source (1253.6 eV) and a concentrichemispherical analyzer. Photoemitted electrons were collectedat a takeoff angle of 30° from the substrate normal, with electrondetection in the constant analyzer energy mode (CAE, pass energy) 20 eV). Core-level XPS spectra were referenced to the C(1s)CxHy peak at 285.0 eV and fitted with Gaussian components allhaving equal full width at half-maximum (fwhm) using Mar-quardt minimization computer software.26 Instrumental sensi-tivity (multiplication) factors determined from chemical stan-dards were taken as C(1s)/O(1s)/F(1s)/N(1s) equals 1.0:0.36:0.23:0.72, respectively. Complete coverage by the plasma polymerlayer was checked for by verifying the absence of any Si(2p) signalin the case of the glass substrate (the smooth surface seen byatomic force microscopy (AFM) precluded any shadowing effects).No surface X-ray damage was observed.

Maleic anhydride plasma polymer films deposited onto siliconwafers were characterized by infrared spectroscopy using aGraseby Specac Golden Gate ATR accessory fitted to a MattsonPolaris instrument. FTIR spectra were acquired at 4 cm-1

resolution over the 400-4000 cm-1 range.A Digital Instruments Nanoscope III atomic force microscope

was used to identify individual dendrimers immobilized ontothe maleic anhydride plasma polymer surfaces. The microscopewas operated in Tapping Mode, where changes in oscillationamplitude of the cantilever tip provide a feedback signalcorresponding to variations in height across the underlyingsurface.27

Fourth-generation PAMAM dendrimers immobilized ontomaleic anhydride plasma polymer surfaces were chemicallyfunctionalized by placing the dendrimer-coated substrate into aglass vacuum apparatus and evacuating to a pressure of 5 ×10-3 mbar. The rotary pump was then isolated from the system,and the dendrimer layer was exposed to trifluoroacetic acid vaporfor 30 min. The whole apparatus was then evacuated back downto its initial base pressure prior to surface analysis.

Adhesion performance was explored by placing a 0.01 mL dropof 10% w/v fourth-generation PAMAM dendrimer solutionbetween two maleic anhydride plasma polymer coated strips ofpolypropylene film (ICI, 50 mm × 10 mm × 0.80 µm) to makea 1 cm2 overlap joint. This was then cured overnight at 120 °C.Subsequently, single lap adhesion tests were performed on theselaminates using an Instron 5543 tensilometer operating at acrosshead speed of 10 mm min-1.

Finally, the gas permeation characteristics of immobilizedfourth-generation PAMAM dendrimer layers were evaluated byreacting one piece of maleic anhydride plasma polymer coatedpolypropylene film with dendrimer solution followed by rinsingin methanol to remove excess dendrimer. Next, a second pieceof maleic anhydride plasma polymer coated polypropylene wasplaced face down on top, to sandwich the PAMAM dendrimerlayer between anhydride functionalities. This composite structurewas then clamped using two pieces of glass and placed into anoven to cure at 120 °C for 12 h. Gas permeation through thiscomposite layer assembly was measured using a mass spectro-metric sampling device.28 This entailed placing the substratebetween two drilled-out stainless steel flanges which wereconnected to an ultrahigh vacuum (UHV) chamber via a gatevalve (base pressure of 5 × 10-10 mbar). One face of the sandwichstructure was then exposed to oxygen gas (BOC, 99.998%) at apressure of 1000 mbar. Gas permeation across the substrate wasmonitored by a UHV ion gauge (Vacuum Generators, VIG 24)and a quadrupole mass spectrometer (Vacuum GeneratorsSX200). The quadrupole mass spectrometer’s response per unitpressure was calibrated by introducing oxygen gas directly intothe UHV chamber and recording the mass spectrum at apredetermined pressure of 4 × 10-7 mbar (taking into accountion gauge sensitivity factors). This value was then used to quantifythe mean equilibrium permeant partial pressure (MEPPP) ofgas permeation through the film in the steady-state flow regime.29

The barrier improvement factor (BIF) was then calculated byreferencing to the MEPPP measured for two untreated pieces ofpolypropylene film loaded into the gas permeability apparatus.

3. Results and Discussion

(a) Pulsed Plasma Polymerization of Maleic An-hydride. Plasma polymerization of maleic anhydrideproduced a well-adhered layer onto glass slides. Thehigh-resolution C(1s) XPS spectra of maleic anhydridepulsed plasma polymer film, Figure 1a, were fitted to fivedifferent carbon environments:30 hydrocarbon (CHx ∼285.0 eV), carbon singly bonded to an anhydride group

(25) Savage, C. R.; Timmons, R. B. Chem. Mater. 1991, 3, 575.(26) Evans, J. F.; Gibson, J. H.; Moulder, J. F.; Hammond, J. S.;

Goretzki, H. Fresenius Z. Anal. Chem. 1984, 319, 841.

(27) Babcock, K. L.; Prater, C. B. Phase Imaging: Beyond Topography;Digital Instruments Application Note; Digital Instruments: SantaBarbara, CA, 1995.

(28) Westover, L. B.; Tou, J. C.; Mark, J. H. Anal. Chem. 1974, 46,568.

(29) Crank, J.; Park, G. S. In Diffusion in Polymers; Academic Press:London, 1968; Chapter 1.

(30) Ryan, M. E.; Hynes, A. M.; Badyal, J. P. S. Chem. Mater. 1996,8, 37.

Controlled Attachment of PAMAM Dendrimers Langmuir, Vol. 18, No. 1, 2002 265

(C-C(O)dO ∼ 285.7 eV), carbon singly bonded to oxygen(-C-O ∼ 286.6 eV), carbon doubly bonded to oxygen(O-C-O/-CdO ∼ 287.9 eV), and anhydride groups(OdC-O-CdO ∼ 289.4 eV). Compared to continuouswave plasma polymerization, pulsing the electrical dis-charge on the millisecond to microsecond time scale atlow duty cycles provided far better control over the coatingcomposition (i.e., the concentration of anhydride func-tionalities).30 This can be attributed to there being lessfragmentation of the precursor molecule and reduced ion/photon damage of the growing plasma polymer layerduring the duty cycle on-period, combined with radical-initiated polymerization of maleic anhydride occurring inthe off-period.30 Optimum anhydride group retention(toff ) 1200 µs, ton ) 20 µs, Pp ) 5 W, 10 min deposition,and thickness ) 34 ( 5 nm) corresponded to 58% of thesurface carbon atoms belonging to cyclic anhydride repeatunits (as determined from the C(1s) XPS envelope).Infrared spectroscopy confirmed the presence of anhydridefunctionalities (1849 and 1780 cm-1) in the depositedfilm,31 Figure 2. The flat topography of the underlyingglass substrate seen by AFM was retained at the surfaceof the deposited pulsed plasma polymer layer, Figure 3.

(b)FunctionalizationofMaleicAnhydridePlasmaPolymer with PAMAM Dendrimers. Exposure of themaleic anhydride plasma polymer film to dendrimersolution followed by rinsing in methanol gave rise to thestable attachment of dendrimers to the surface. Themorphology, size, and intermolecular spacing of theimmobilized dendrimer moieties were examined by atomicforce microscopy, Figure 3. AFM micrographs of fourth-generation PAMAM dendrimers covalently bonded to themaleic anhydride plasma polymer layer appear as smalldots (height ) 2.5 ( 0.03 nm, width ) 5.3 ( 0.04 nm);these represent individual dendrimers fixed onto theunderlying (darker shading) plasma polymer surface.Assuming a spherical shape, the dendrimers should be4.5 nm wide.32,33 Therefore, it appears that the dendrimershave slightly flattened out on the maleic anhydride plasma

polymer surface (in order to maximize amide bondformation). This is consistent with previous computersimulation and AFM studies which have reported aflattening and spreading out of PAMAM dendrimermolecules over a surface.13-15,34

XPS analysis verified that reaction had indeed takenplace at the surface. The dendrimer species contributethree types of carbon functionality to the C(1s) spectra:35

carbon singly bonded to an amide carbon/amine nitrogen(C-C-NHR(dO)/C-N ∼ 285.7 eV), carbon singly bondedto an amide nitrogen (-CH2-NH-CdO ∼ 286.0 eV), andan amide group (RHN-CdO ∼ 287.9 eV), Figure 1. Thesegave rise to attenuation of the anhydride group C(1s) signal(OdC-O-CdO ∼ 289.4 eV) as the surface coverage ofPAMAM dendrimers increased. A corresponding rise inthe N(1s) peak at ∼400 eV was seen and can be taken tobe indicative of dendrimer attachment to the maleicanhydride plasma polymer surface via amide linkages16

(a weak N(1s) component at 401.9 eV could be attributedto either reaction of dendrimer amine groups withatmospheric CO2

36,37 or hydrogen bonding38). The packingdensity of dendrimers at the surface could be varied bydiluting the solution with methanol, Figure 4. For all threedendrimer generations under investigation, the surfaceconcentration of nitrogen (% N) was found to correlate tothe degree of dilution. Submonolayer coverages cor-responded to dilutions below 0.01% w/v and tied inwith the reappearance of the anhydride group peak(OdC-O-CdO ∼ 289.4 eV) in the C(1s) XPS envelope,

(31) Lin-Vien, C.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. TheHandbook of Infrared and Raman Characteristic Frequencies of OrganicMolecules; Academic Press: San Diego, CA, 1991; Chapter 9.

(32) Prosa, T. J.; Bauer, B. J.; Amis, E. J.; Tomalia, D. A.;Scherrenberg, R. J. Polym. Sci. 1997, 35, 2913.

(33) Grohn, F.; Bauer, B. J.; Akpalu, Y. A.; Jackson, C. L.; Amis, E.J. Macromolecules 2000, 33, 6042.

(34) Mansfield, M. L. Polymer 1996, 37, 3835.(35) Beamson, G.; Briggs, D. High-Resolution XPS of Organic

Polymers; Wiley: Chichester, 1992.(36) Yan, L.; Marzolin, C.; Terford, A.; Whitesides, G. M. Langmuir

1997, 13, 6704.(37) Sprik, M.; Delamarche, E.; Michel, B.; Rothlisberger, U.; Klein,

M. L.; Wolf, H.; Ringsdorf, H. Langmuir 1994, 10, 4116.(38) Kerber, S. J.; Bruckner, J. J.; Wozniak, K.; Seal, S.; Hardcastle,

S.; Barr, T. L. J. Vac. Sci. Technol., A 1996, 14, 1314.

Figure 1. C(1s) XPS spectra of (a) maleic anhydride pulsedplasma polymer layer, (b) the plasma polymer functionalizedwith a 0.00125% w/v solution of fourth-generation PAMAMdendrimers, (c) the plasma polymer functionalized with a 10%w/v solution of fourth-generation PAMAM dendrimers, and (d)following reaction of (c) with trifluoroacetic acid vapor. Figure 2. Infrared spectra of (a) maleic anhydride pulsed

plasma polymer layer (Pp ) 5 W; ton ) 20 µs; toff ) 1200 µs; 10min), (b) 10% w/v solution of fourth-generation PAMAMdendrimer solution, (c) fourth-generation PAMAM dendrimersdeposited from 0.01% w/v solution onto maleic anhydride pulsedplasma polymer surface, and (d) following annealing of (c) to120 °C.

266 Langmuir, Vol. 18, No. 1, 2002 Fail et al.

Figure 1. An alternate way of varying the number ofdendrimers attached to the surface is to change the

concentration of anhydride groups (i.e., alter the pulsedplasma deposition conditions30).

Infrared spectroscopy showed the appearance of amideabsorption bands39 (1650 and 1580 cm-1) characteristic ofthe dendrimer molecules (internal amide bonds and amicacid groups formed between terminal dendrimer aminegroups and the maleic anhydride plasma polymer surface),Figure 2. The absorbance at approximately 1450 cm-1

can be attributed to CH2 groups present in the dendrimermolecules. Heating at 120 °C caused a decrease in theamide band intensities relative to the peak at 1450 cm-1.This occurs as a consequence of the internal amide groupsin the PAMAM dendrimer molecules undergoing a retro-Michael reaction to form imide linkages (1710 cm-1);40

there should also be imide bond formation at the dendrimerbinding sites on the maleic anhydride pulsed plasmapolymer surface. The described approach for attachingPAMAM dendrimers onto solid surfaces is applicable toa whole variety of substrates and therefore offers a distinctadvantage compared to alternate methods (e.g., poly-(maleic anhydride)-c-poly(methyl vinyl ether) layers fixedonto aminosilane-functionalized silicon surfaces21).

(c) Reaction of PAMAM Dendrimers with Tri-fluoroacetic Acid. Further verification of the im-mobilization of PAMAM dendrimers onto maleic anhy-dride pulsed plasma polymer surfaces was obtained byreacting the remaining terminal PAMAM dendrimeramine groups with trifluoroacetic acid vapor to produceamide linkages (solvent rinsing ruled out physisorption).In this case, the CF3 functionality in the C(1s) XPSspectrum at 293 eV served as a marker, Figure 1. A goodcorrelation was found between the amount of nitrogenmeasured at the surface (i.e., dendrimer density) and theF(1s) signal detected by XPS following exposure totrifluoroacetic acid, Figure 5. These results are consistentwith previous studies where 4-(trifluoromethyl)benzoylchloride was used to functionalize a PAMAM dendrimermonolayer adsorbed onto a flat gold substrate.17

(d) Adhesion. Adhesion studies were undertaken byplacing a drop of fourth-generation PAMAM dendrimersolution between two pieces of polypropylene coated withmaleic anhydride plasma polymer followed by heating to

(39) Williams, D. H.; Fleming, I. Spectroscopic Methods in OrganicChemistry, 4th ed.; McGraw-Hill Ltd.: London, 1989.

(40) Zhao, M. Q.; Liv, Y. L.; Crooks, R. M.; Bergbreiter, D. E. J. Am.Chem. Soc. 1999, 121, 923.

Figure 3. Tapping mode atomic force micrographs of (a) theglass substrate, (b) maleic anhydride pulsed plasma polymerdeposited onto glass (Pp ) 5 W; ton ) 20 µs; toff ) 1200 µs; 10min), and (c) fourth-generation PAMAM dendrimers (whitespots) attached to the maleic anhydride pulsed plasma polymerlayer (darker background).

Figure 4. Nitrogen concentration (% N) at the surface of themaleic anhydride plasma polymer layer following dendrimerfunctionalization as a function of solution dilution (% w/v).

Figure 5. XPS correlation between the amount of nitrogenN(1s) present at the surface and fluorine F(1s) detected followingexposure to trifluoroacetic acid vapor, in the case of fourth-generation PAMAM dendrimers immobilized onto maleicanhydride pulsed plasma polymer layers.

Controlled Attachment of PAMAM Dendrimers Langmuir, Vol. 18, No. 1, 2002 267

120 °C. A maximum force per unit area of 20 N cm-2 wasrecorded for this sandwich structure, Figure 6. Thiscompared favorably with the substrate failure value forthe parent polypropylene film of approximately 26 N. Noadhesion was observed in the absence of dendrimer. Itwas found that the heating step (imide formation) was aprerequisite for achieving good adhesion.

(e) Gas Barrier. Oxygen gas permeation measure-ments were also carried out for dendrimer layers sand-wiched between maleic anhydride plasma polymer coatedpolypropylene film. Thermal curing at 120 °C gave rise to

a significant improvement in the gas barrier, Table 1.This can also be attributed to imidization by retro-Michaelchemistry giving rise to a highly cross-linked impermeablestructure.40

4. Conclusions

Starburst PAMAM dendrimers can be chemically fixedonto a variety of solid substrates by predepositing a well-adhered maleic anhydride pulsed plasma polymer layer.The intermolecular spacing and concentration of den-drimer molecules attached to the surface can be controlledby either varying the level of anhydride group incorpora-tion during plasma polymer film deposition or changingthe dilution of the dendrimer solution. The external aminegroups associated with the fixed dendrimers are availablefor further chemical reaction, for example, fluorination orimidization (for adhesion and gas barrier).

Acknowledgment. S.A.E. thanks Smith & Nephewfor financial support.

LA0111598

Figure 6. Lap-shear adhesion test: (a) bulk failure ofpolypropylene substrate and (b) polypropylene/plasma polymer/fourth-generation PAMAM dendrimer/plasma polymer/polypro-pylene joint (heated to 120 °C).

Table 1. Gas Barrier Measurements for DendrimerMolecules Sandwiched between Polypropylene Film and

Then Cured at 120 °C

plasma polymer gas barriera

yes 34.9no 2.1

a Gas barrier improvement factor measured with respect to twopieces of polypropylene film sandwiched together (i.e., no maleicanhydride plasma polymer coating or dendrimer molecules).

268 Langmuir, Vol. 18, No. 1, 2002 Fail et al.