Polydopamine and Polydopamine-Silane Hybrid ... - ACS Publications

Author to whom correspondence should be addressed. Email: [email protected]. Polydopamine and Polydopamine-Silane Hybrid Surface Treatments in...
0 downloads 0 Views 4MB Size
Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Article

Polydopamine and Polydopamine-Silane Hybrid Surface Treatments in Structural Adhesive Applications Ngon T. Tran, David P. Flanagan, Joshua A. Orlicki, Joseph L. Lenhart, Kenneth Proctor, and Daniel Brainard Knorr Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b03178 • Publication Date (Web): 03 Jan 2018 Downloaded from http://pubs.acs.org on January 4, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Distribution A: Approved for Public Release

Polydopamine and Polydopamine-Silane Hybrid Surface Treatments in Structural Adhesive Applications Ngon T. Tran‡, David P. Flanagan†, Joshua A. Orlicki†, Joseph L. Lenhart†, Kenneth L. Proctor‡, Daniel B. Knorr Jr. †,* † ‡

U. S. Army Research Laboratory, Aberdeen Proving Ground, Maryland 21005 Oak Ridge Institute for Science and Education, Belcamp, Maryland, 21017

Keywords: Polydopamine, Lap Shear, Adhesive, Silane Coupling Agents Abstract Numerous studies have focused on the remarkable adhesive properties of polydopamine, which can bind to substrates with a wide range of surface energies, even under aqueous conditions. This behavior suggests that polydopamine may be an attractive option as a surface treatment in structural bonding applications, where good bond durability is required. Here, we assessed polydopamine as a surface treatment for bonding aluminum plates with an epoxy resin. A model epoxy adhesive consisting of diglycidyl ether of bisphenol A (DGEBA), and Jeffamine® D230 polyetheramine was employed and lap shear measurements (ASTM D1002 10) were made (i) under dry conditions to examine initial bond strength and (ii) after exposure to hot/wet (63 oC in water for 14 days) conditions to assess bond durability. Surprisingly, our results showed that polydopamine alone as a surface treatment provided no benefit beyond that obtained by exposing the substrates to an alkaline solution of tris buffer used for the deposition of polydopamine. This implies that polydopamine has a potential Achilles’ heel, namely, the formation of a weak boundary layer that was identified using X-ray photoelectron spectroscopy (XPS) of the fractured surfaces. In fact, for longer deposition times (2.5 h and 18h), the tris buffer treated surface outperformed the polydopamine surface treatments, suggesting that tris buffer plays a unique role in improving adhesive performance even in the absence of polydopamine. We further showed that the use of polydopamine-3-aminopropyltriethoxysilane (APTES) hybrid surface treatments provided significant improvements in bond durability at extended deposition times relative to both polydopamine and an untreated control. Introduction Epoxy-based adhesives are widely used to join aluminum parts in aerospace,1 automotive,2-3 and military applications. Adhesives used in military applications must function in a very broad range of conditions (e.g., static to high strain rates as well as extremes in temperature and humidity), and there is a continual effort to identify surface treatments that can provide superior performance under these conditions. Water adsorption is particularly problematic at organicinorganic interfaces, as it can lead to a deterioration in the performance of polymeric synthetic adhesives.4 In addition, the adsorption of water on a free surface and at an interface between the same surface (i.e., an adherend like aluminum) and a glassy polymer (e.g., adhesive) are nearly equivalent.5 Conventional silane coupling agents used in structural adhesives are very effective but may suffer from hydrolysis of the metal−O−Si−C bond in the presence of water, and this tendency varies with the type of metal oxide.6

a

Author to whom correspondence should be addressed. Email: [email protected] ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Distribution A: Approved for Public Release

Catechols are a class of aromatic moieties with two hydroxyl groups that are in the ortho position relative to one another. In nature, catechol-containing protein residues help provide mussels with strong adhesion in underwater environments.7-8 Catechols have shown high affinity for some metal oxides in the presence of water9 or under anhydrous conditions.10 As such, catechol-containing surface treatments may provide a bonding mechanism that could be used at interfaces in structural adhesive applications. Polydopamine (PDA) is a catechol-containing polymer that has the ability to coat a wide variety of surfaces, and can be easily synthesized in a mildly alkaline solution under ambient conditions.11 PDA has been extensively reviewed,12-17 and has found utility water treatment, energy applications, hybrid materials, and sensors.13 PDA has only rarely been used in structural applications where high interfacial strengths and load transfer are important. These applications have primarily been in composite materials. For example, polydopamine was used as a coating for carbon and glass fibers to improve the fiber-matrix interfacial shear stress in composites.18 PDA was also used to treat glass fiber posts for dental applications, and provided improvements in interfacial mechanical properties of these posts cured in epoxy resins.19 Polydopamine also showed the potential to replace silane-based coupling agents in water-based organic/inorganic composite materials.20 Yang et al. showed that hydrogen bonding improved interfacial stress transfer in between PDA-modified clay and an epoxy matrix.21 Many types of catechol containing polymers have been synthesized, and the use of these as adhesives has often focused on biological applications like tissue adhesives.22 Others have focused on improving adhesive strength by designing polymers with catechol containing moieties,23 but the lap shear measurements for these polymers are significantly lower than those realized for epoxy-based adhesives.24 PDA itself has been considered for use as a primer,14 but these applications largely involve immobilization of other polymers and biological applications, and are not focused on structural applications. To our knowledge, only one group has published the use of polydopamine in epoxy adhesives with aluminum adherends.25 These authors did not see significant improvements in lap shear when PDA was mixed into the epoxy (rather than applied as a surface treatment), though the reasons for this were not discussed extensively because the paper focused on carbon nanotube-epoxy composite materials. Previous studies have also not focused on the influence of the tris (itself a polyol amine) that is often used as a pH buffer during the deposition of PDA. Despite the fact that tris can covalently incorporate into PDA26, control experiments related to the influence of tris buffer are lacking and may be critical. We recently developed a strategy wherein conventional silane coupling agents can be covalently incorporated into polydopamine to form a silane-polydopamine hybrid coating.27 The structures of the constituent components, dopamine hydrochloride, 3-aminopropyltriethoxysilane (ATPES), and the structures of polydopamine and the hybrid material are provided in Fig. 1. The overall strategy of using these hybrid materials in structural adhesives is that APTES would provide covalent bonding between an aluminum adherend and the epoxy adhesive,28 while the catechol moieties in the polydopamine would improve adhesive bonding in the presence of moisture via hydrogen bonding and displacement of water.10 The overall goal is to provide a more diverse surface treatment than that of only a silane coupling agent, which we hypothesized would allow for better structural bond durability in the presence of water. The bond strength and durability of structural adhesives employing these types of hybrid strategies have not previously been evaluated, nor has the influence of the deposition solution in relevant control measurements.

ACS Paragon Plus Environment

Page 2 of 41

Page 3 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Distribution A: Approved for Public Release

In this paper, we utilized our hybrid coating shown in Fig. 1 in structural adhesive applications. Specifically, we performed single lap shear measurements on 2024-T3 aluminum that was surface treated with polydopamine and polydopamine-silane hybrids. We chose a model structural adhesive composed of DGEBA/D230 and we analyzed the pre-fracture surfaces to assess the type of failure (e.g., adhesive or cohesive). X-ray photoelectron spectroscopy (XPS) was used to monitor the effect of exposure of 2024-T3 to the deposition solution and evaluate the surface chemistry at the fractured interfaces.

Figure 1. (Top) Reaction of dopamine to form polydopamine, which may contain moieties from tris buffer (red). (Bottom) Reaction of dopamine and APTES (blue) to form polydopamineAPTES hybrids.

Experimental Materials DGEBA/D230 was chosen as the model adhesive because of its reasonable pot life, commercial availability, reasonable processing viscosity, and the fact that it has been well characterized via experiments29-31 and by molecular dynamics simulation.32 Diglycidyl ether of bisphenol A (DGEBA, Hexion Inc. EPONTM Resin 825), poly(propylene glycol) bis(2aminopropyl ether) (D230, Huntsman Jeffamine® D 230), dopamine hydrochloride (>98 %, Sigma Aldrich), 3-aminopropyltriethoxysilane (APTES, >98 %, Sigma Aldrich) were purchased and used without further purification. Tris buffer (pH = 8.5) was prepared from 2-amino-2(hydroxymethyl)-1,3-propanediol (>99.9 %, Sigma Aldrich), sulfuric acid (95-98 %, Sigma Aldrich), and Milli-Q water (18 MΩ·cm−1). 2024-T3 aluminum lap shear panels were cut to size based on ASTM-D1002-10. XPS and contact angle samples were cut from lap shear panels. Aluminum oxide blasting grit (60) was purchased from Grainger.

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Distribution A: Approved for Public Release

Preparation of Epoxy Adhesive D230 (12.23 g, 0.053 mmol, 1.00 equiv) was added to a 60 mL jar containing DGEBA (37.77 g, 0.106 mmol, 2.00 equiv) at 35 °C. After stirring for 5-10 min, the resulting solution was degassed for 10 min at −80 kPa and 35 °C in a vacuum oven (Yamato ADP 200C). At atmospheric pressure, the solution was heated at 35 °C for 2.3 h and then 40 °C for 1 h. The transparent, colorless, and viscous solution was allowed to reach room temperature over 0.5 h. At which point, the resulting adhesive was used in assembling lap shear specimens. Preparation of Substrate and Surface Treatment The aluminum 2024-T3 substrates were prepared according to the guidelines published previously.33 Oil, grease, dirt, and heavy surface oxidation were removed using acetone (>99.5%, Sigma Aldrich) and an abrasion pad, Scotch-Brite 048011-65530. Grit-blasting was performed using virgin 60-grit aluminum oxide abrasive blasting media (Grainger) and grit blaster (Econoline). Ultrahigh purity nitrogen gas was used to remove loose dust and grit blast particles. Surface treatments were performed within 4 h of surface cleaning. Dopamine hydrochloride (4.000 g, 21.09 mmol) was dissolved in either 50 mM tris buffer (pH = 8.5, 2 L) or ultrapure water (18 MΩ·cm, 2 L). In some depositions, the appropriate amount of APTES was added to achieve the desired A:D ratio (APTES:dopamine ratio) immediately after dissolution of dopamine hydrochloride. Aluminum panels were immediately suspended in the resulting 2 mg/mL dopamine hydrochloride solution, which was continuously stirred open to atmospheric oxygen for the specified times. Post-treatments included rinsing with ultrapure water and heating to 100 °C for 1 h under a nitrogen atmosphere or combinations thereof. A well-established industrial technique for silane deposition was employed to obtain the APTES control for comparison with hybrid APTES-dopamine surface treatments. A 1 wt% solution of APTES (4.32 mL, 18.5 mmol) was added to a solution of ethanol:water (9:1, 495.68 mL), and the resulting solution allowed to stir for 1 h at room temperature. The aluminum panels were immediately immersed in the solution for 2 min, dried with ultra-pure nitrogen gas, and heated for 1 h at 100 °C under a nitrogen atmosphere. Assembly and Fabrication of Single-Lap Joint A minimum of ten single-lap joint specimens were performed for each experiment. Previously published guidelines33 were used to prepare single-lap samples to be in accordance with ASTM D1002. Tooling fixtures included shim plates and 3 stacking steel plates (5.4 kg) to control bondline thickness (0.381 mm). Samples were cured at 100 °C overnight in a convection oven (Lindberg Blue) and under positive nitrogen pressure to avoid oxidation of the resin. Environmental Conditioning of Specimens Following curing and cooling to room temperature, the aluminum panels were divided into five lap shear specimens using a band saw, and the adhesive overflow were removed using a belt sander. As prescribed by ASTM D1002, the resulting lap shear specimens were either placed under vacuum prior to dry testing or completely immersed in a 63 °C deionized water bath for 2 weeks prior to hot/wet testing.

ACS Paragon Plus Environment

Page 4 of 41

Page 5 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Distribution A: Approved for Public Release

Testing of Specimens Mechanical tests were performed using Instron 5500R Model 1125, 0.0212 mm·s−1 crosshead speed, and a 5000 lb load cell. A pair of self-aligning grips held each lap shear sample on both ends with grip area of 25.4 x 25.4 mm2. Analysis of Fractured Specimens Modes of failure were assigned based on visual examination with a large magnifying glass. For failed joints to be considered cohesive failure or mixed mode, the adhesive must be on both sides of the joints. Hence, failing through the adhesive, which is quite common for many samples that were subjected to the hot/wet fatigue conditioning, is not necessarily cohesive failure because the adhesive may not be on both sides of the failed joint. Failing either at the metal oxide layer or surface treatment layer is considered as adhesive (or interfacial) failure because the adhesive is not on the adherend. Since polydopamine and polydopamine-silane surface treatments discolored Al 2024-T3 adherend over time (> 0.1 h), failure at the metal oxide can be differentiated from failure at the surface treatment layer because the former would be metallic gray whereas the latter would be brown due to the polydopamine. Please see supporting information for photographs of fractured single lap shear specimens. Also, analysis and interpretation of failure modes are discussed below and in the supporting information. Contact Angle Measurements Contact angle measurements were made on various samples using a Ramé-Hart Model 290 automated goniometer. During the measurement, advancing contact angles were measured by the addition of increments of 15 µL of DGEBA/D230 solution on the surface using the drop shape method. Contact angle measurements were performed in at least 3 separate locations on the samples. XPS Measurements XPS was performed on a Physical Electronics VersaProbeII instrument with an Al Kα source. Compositions were obtained from high resolution scans with a pass energy of 23.5 eV and a step size of 0.05 eV for C 1s and Al 2p regions. At least three separate regions were measured for each sample. XPS data were processed using CasaXPS software as described in the supplemental information. Results and Discussion Polydopamine Surface Treatments with Various Post-Deposition Treatments Our first task was to evaluate polydopamine depositions by themselves. The ratio “A:D” used herein refers to the molar ratio of APTES to dopamine used in the solution. So, a deposition of only polydopamine is “0:1”. We suspected that the performance of the depositions may be significantly affected by the post-deposition treatment for two reasons. First, silanes are often heated after deposition to drive the silane condensation reaction, but PDA films have shown cracking behavior after heating on certain substrates.34 Second, in our previous work, we observed a significant amount of poorly adhered polydopamine that was removed by water rinsing after deposition. Therefore, we tried different combinations of post deposition treatments after an 18 h deposition of only polydopamine out of a tris buffer (TB) solution. These included: a heat treatment to 100 oC under nitrogen (designated with an “h”), or a water rinse (designated

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Distribution A: Approved for Public Release

with a “w”), neither of these, or a combination of these. Our heat treatment was conducted under a nitrogen atmosphere because it was previously shown that the primary amines of APTES can oxidize to imines at higher temperatures,35 and the water rinse was conducted with the hope of removing weakly bound residual polydopamine material.

Figure 2. Lap shear results for dopamine-only (0:1) surface treatment depositions for (a) 0.1 h or (b) 18 h with various post-deposition treatments including heating at 100oC under nitrogen (h) and rinsing with water (w). APTES control regions (gray is dry, blue is hot/wet) encompass the standard deviation of measurements obtained for a standard APTES pre-treatment deposited out of an ethanol/water mixture as described in the experimental section. The lap shear results for dopamine-only (0:1) surface treatments are shown in Fig. 2. These PDA surface treatments were conducted for short deposition times (0.1 h, Fig. 2(a)) and long deposition times (18 h, Fig. 2(b)) to determine the effect of deposition time. The control (i.e., 2024-T3 substrates that had no surface treatment after grit blasting) is the left-most data set. Focusing on the short deposition time (0.1 h), we see that PDA surface treatments without rinsing or heating (0.1h) or with only heating (0.1h, h) showed much lower hot/wet shear

ACS Paragon Plus Environment

Page 6 of 41

Page 7 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Distribution A: Approved for Public Release

strengths than the control. In contrast, the dry shear strength for the sample that was only rinsed (0.1h,w) was not statistically different from that of the untreated control, while the hot/wet performance was significantly better than the control. The best performance observed for the PDA deposited for 0.1 h was observed for the adherends that were rinsed and then heated after deposition (0.1h,w,h). As shown, both the dry and hot/wet shear strengths for this sample were significantly higher than that of the untreated control, but were significantly lower than that of a standard APTES control. This may be due to a combination of removal of weakly bound polydopamine and thermal stabilization of the polydopamine surface via interchain and intrachain crosslinking that can occur in catechol-containing polymers. Results for the 18 h deposition (Fig. 2b) were all significantly lower than the 0.1 h depositions. However, of all the 18 h data, the samples that were water rinsed and heated (18h,w,h) provided the best hot/wet performance. We note that the 18h,w sample that was only water rinsed showed the best dry performance. However, error bar of the 18h,w,h sample was overlaps the error bar of the 18h,w value, and the wet performance of the 18h,w,h samples were better than that of the 18h,w samples. Based on these results, we determined that the posttreatment should include a water rinse followed by a heat treatment for the remaining experiments. Failure modes of fractured lap shear specimen in the post-deposition screening experiments above (Fig. 2) also provide support for the importance of post-deposition treatment (see Fig. 3 and supporting information). Without post-deposition washing, all adherends that were surface treated with polydopamine only (or 0:1) depositions at both 0.1 and 18 h exhibit near complete adhesive failure for both dry and hot/wet conditions. For the 18 h studies, the bonded joints failed either between the adhesive and the surface treated layer or in the surface treated layer. This was apparent because each fractured lap joint had an adherend that was not metallic grey (as would be expected for the aluminum adherend) but brown, and the adherend side contained aggregates of the brown polydopamine. Furthermore, longer depositions with post-deposition washing and heating [i.e., (0:1,TB,18h,w,h)] showed less cohesive failure compared to ones at 0.1 h (i.e., 0:1,TB,0.1h,w,h). These results suggest that thicker polydopamine depositions may create a weak interface, resulting in low lap shear performance, which will be discussed further below.

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Distribution A: Approved for Public Release

Figure 3. Representative dopamine-only fractured lap shear specimen for post-deposition screening experiments for (a) 1 h depositions and (b) 18 h depositions. At this point, we made additional controls for both the short and long deposition times that involved a surface treatment that was conducted in the tris buffer but contained no dopamine (i.e., the “0:0” datasets in Fig. 2). This is a critical control that is often neglected in the existing literature. As shown, this additional control dataset showed significant improvements in both dry and hot/wet performance relative to the untreated control. Furthermore, the performance of this sample for 0.1 h deposition (22.6±4.4 MPa dry, 16.4±1.2 MPa hot/wet) was statistically similar to that of the best performing dopamine-only deposition (0:1,w,h, i.e., 22.2±1.0 MPa dry, 17.9±2.3 MPa hot/wet). A similar result was observed for the 18 h 0:0 data (24.4±3.1 MPa dry, 14.1±1.1 MPa hot/wet), which was better than the best polydopamine treated samples. This surprising result suggests that exposure of the substrate to a slightly alkaline tris buffer solution (pH=8.5) provides a significant improvement in shear strength. Alkaline pretreatments for aluminum substrates have previously been investigated along with many other surface treatments.36 These treatments have included NaOH solutions where the alkaline treatment was used to remove unstable regions in the surface oxide of the aluminum.37 Saleema et al. 38 showed increases in dry lap shear strength using a weak NaOH solution that are comparable (on a relative basis) to the improvements that we observed here. While alkaline treatments can improve shear strength, Zain et al. showed that silane treatments provided better durability than alkaline treatments alone for polyurethane-based adhesives.39 Failure modes of fractured lap shear specimen of tris buffer controls also provide support for the advantageous effects of tris buffer as a surface treatment. More cohesive failure was observed for the tris buffer controls (0:0,TB,time,w,h) compared to the untreated controls for both dry and hot/wet conditions and for all times investigated (i.e., 0.1 h, 2.5 h, and 18 h) (see Fig. 4 and supporting information). These results indicate that tris buffer, either by promotion of aluminium oxide over hydroxide on adherened surfaces or by covalent linkage between the epoxy with the strongly adsorbed surface layer of tris(hydroxymethyl)aminomethane40 enhances the interfacial interactions. These results also provide evidence against significant detrimental effect of corrosion by the tris buffer (pH=8.5) because only improvement in interfacial interactions and lap shear data relative to the untreated controls were observed even with prolonged depositions.

ACS Paragon Plus Environment

Page 8 of 41

Page 9 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Distribution A: Approved for Public Release

Figure 4. Representative fractured lap shear specimen for untreated control and tris buffer studies.

Lap Shear Results for Polydopamine-APTES Hybrid Surface Treatments Lap shear results for the 0.1 h PDA-APTES hybrid surface treatment depositions for various A:D ratios are shown in Fig. 5. As shown, the depositions that included APTES with the polydopamine showed dry and hot/wet shear strengths that were better than that of the untreated control. However, only one of these treatments (the 7.5:1) was statistically higher than that of the “0:0” sample exposed only to tris buffer without dopamine or APTES. Our previous work showed that composition of the 7.5:1 sample with tris was approximately 75% APTES on a molar basis, and that the deposition rates for 7.5:1 were about the same as that of the 0:1, but about half as much as the 3.5:1 deposition rate. Overall, then the 7.5:1 deposition here is probably better than both controls because it contains mostly APTES. However, it is not better than a standard APTES only control prepared from 90/10 ethanol/water, which gives dry lap shear results of 34.1±1.5 MPa, and a hot/wet lap shear result of 22.7±0.9 MPa. This is probably because the PDA-APTES hybrid film thickness was very low at a deposition time of 0.1 h. Note that tensile tests of dry and hot/wet treated dogbones of the adhesive showed no statistical change in modulus and only a 10% drop in tensile strength due to the hot/wet treatment (see supplemental information). Therefore, much of the difference between dry and hot/wet specimens must be attributed to interface effects. It is possible that a 0.1 h deposition time is too short. Jiang et al. suggested that polydopamine deposition takes place in a sequential process wherein nanoaggregates are first formed and deposited.41 This is followed by the formation and deposition of PDA particles; these particles eventually aggregate at longer times. During the process of the synthesis of polydopamine, the solution is first colorless and transparent when the dopamine hydrochloride is added. The solution then changes from colorless to yellow, and then to amber, and eventually becomes opaque and turns black as aggregates of PDA form in the solution. We chose the 0.1 h deposition time because it is approximately when the color changes to amber and likely corresponds to when nanoaggregates of PDA are present rather than the large aggregates present at longer times. Therefore, we performed an additional set of measurements for a deposition time of 2.5 h (Figure 6).

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Distribution A: Approved for Public Release

Figure 5. Lap shear results for APTES-PDA hybrid surface treatments at various A:D ratios for a deposition time of 0.1 h. APTES control regions (gray is dry, blue is hot/wet) encompass the standard deviation of measurements obtained for a standard APTES pre-treatment deposited out of an ethanol/water mixture as described in the experimental section.

Figure 6. Lap shear results for APTES-PDA hybrid surface treatments at various A:D ratios for a deposition time of 2.5 h. APTES control regions (gray is dry, blue is hot/wet) encompass the standard deviation of measurements obtained for a standard APTES pre-treatment deposited out of an ethanol/water mixture as described in the experimental section. Results for depositions after 2.5 h are shown in Fig. 6 above. A deposition time of 2.5 h was selected because this time should produce pinhole-free films for the 0:1 polydopamine only treatment according to previous work.42 Interestingly, the sample exposed to only tris buffer

ACS Paragon Plus Environment

Page 10 of 41

Page 11 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Distribution A: Approved for Public Release

(0:0) showed a substantial improvement over the untreated control sample for both dry (~46%) and wet (~36%) conditions. The remainder of the dry results show that hybrid surface treatments with 1:1, 3.5:1 and 7.5:1 ratios were not statistically different from the sample exposed to only tris buffer (i.e., “0:0”). The 0:1 PDA only deposition was significantly lower, which suggests that deposition of PDA degrades bonding under dry conditions. The hot/wet data are interesting in that the 3.5:1 deposition showed the best performance, followed by both the 0:0 and the 7.5:1 deposition, which were similar. Interestingly, the 0:1 and 1:1 depositions were statistically no better than the untreated control. These results show that the addition of APTES to PDA improves bond durability beyond that of tris alone or PDA alone. Additional measurements were performed after depositions for 18 h.

Figure 7. Lap shear results for APTES-PDA hybrid surface treatments at various A:D ratios for a deposition time of 18 h. APTES control regions (gray is dry, blue is hot/wet) encompass the standard deviation of measurements obtained for a standard APTES pre-treatment deposited out of an ethanol/water mixture as described in the experimental section. The 18 h deposition results are shown in Fig. 7. As shown, the best surface treatment was the 1:1 deposition in tris buffer, which showed good dry performance comparable to the untreated control and hot/wet performance that was modestly better. Unfortunately, none of the surface treatments were statistically better than the control exposed only to tris buffer for 18 h (0:0). Under hot/wet conditions, all the treatments with APTES performed better than the deposition performed with only polydopamine. Previous work43 showed that silane coupling agents can improve corrosion resistance of 2024 aluminum, and this may be the reason why the APTES-containing formulations perform better under hot/wet conditions at long times. However, they still did not show high shear strengths, with the best hot/wet performer (3.5:1) having a strength of only 15.9±1.1 MPa relative to an untreated control strength of 13.0±0.8 MPa. All the data for the hybrid systems are summarized in Fig. 8. As shown, the addition of APTES to polydopamine generally improved the lap shear performance. However, 18 h depositions were generally poor as were the polydopamine only depositions at most deposition times. It is possible that 18 h deposition times were too long and only aggregates of the APTESPDA hybrid were present. Another plausible explanation for the poor 18 h results is surface

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Distribution A: Approved for Public Release

degradation due to corrosion over long deposition times, because the addition of APTES increases the alkalinity of the solution from pH=8.5 to pH=10.2, and this higher pH level is outside of the passivation window for aluminum and many of its alloys based on the Pourbaix diagram for aluminum.44 To investigate this aspect, we ran an additional control of APTES only depositions in tris buffer (i.e., 1:0,18h,w,h). These experiments were performed using the same aqueous conditions as those of the polydopamine. The lap shear results of these 1:0 APTES-only samples were poor for both dry and hot/wet conditions (i.e., 13.5±3.0 MPa and 9.6±1.3 MPa), making them worse than the untreated control. This result provides support for the idea that corrosion under high pH conditions negatively affected the lap shear strengths at long deposition times, which will be discussed further below.

Figure 8. Temporal effects on lap shear strength for polydopamine-only and hybrid polydopamine-silane surface treatments for (a) dry and (b) hot/wet data. Failure modes of specimens with hybrid polydopamine-silane surface treatments show that, despite the detrimental effects of polydopamine (0:1,TB,time,w,h) at the interface between Al2024-T3 adherend and DGEBA/D230 epoxy, some of the enhanced interfacial interactions of the APTES were maintained (see Fig. 9 and supporting information). At longer deposition times (i.e., 2.5 and 18 h), specimens of both 0:1 and 1:1 depositions exhibited essentially complete adhesive failure. However, depositions of 3.5:1 and 7.5:1 showed significant amounts of cohesive failure. These results suggest that the benefit of the silane coupling agent is partially offset by the adverse effects of polydopamine, resulting in enhanced interfacial adhesion compared to the untreated controls in both dry and hot/wet conditions.

ACS Paragon Plus Environment

Page 12 of 41

Page 13 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Distribution A: Approved for Public Release

Figure 9. Representative fractured lap shear specimen for polydopamine-only and hybrid polydopamine-silane surface treatments for (a) dry and (b) hot/wet conditions. Overall, our lap shear data and the failure modes of fractured lap shear specimen show that polydopamine adversely affected the interfacial region between the Al 2024-T3 adherend and the DGEBA/D230 adhesive and that it did not offer any additional resistance to hot/wet conditions. Favorable shear strengths for the hybrid depositions seem to be due to the individual contributions of tris buffer and/or APTES. The above results suggest that the majority of the benefit from the hybrid PDA-APTES films comes from APTES. Therefore, is important to reach a fundamental understanding of why polydopamine fails as a surface treatment in structural applications. Baldan45 notes that there are six main theories of adhesion including: (i) diffusion and interdiffusion theory, (ii) adsorption theory, (iii) chemical bonding theory, (iv) the mechanical interlocking model, (v) electrostatic attraction theory, and (vi) weak boundary layer theory. All but one of these theories may participate in our present system, i.e., diffusion and interdiffusion theory, which primarily deals with polymeric adhesives and a polymeric adherend. Adsorption theory is widely accepted, and it postulates that materials adhere due to attractive forces between an adhesive and a substrate in intimate contact.45 For our purposes, the adsorption theory and chemical bonding theory are essentially identical. The latter can be understood as a special case of the former, wherein the forces are due to covalent bonding. In our case, APTES or the secondary amines in PDA can covalently bind to the epoxies in DGEBA to increase the bond density between the adhesive and substrate. In mechanical interlocking theory, adhesion relies on surface irregularities (e.g., bumps and hook-type irregularities) to provide mechanical connections between the adhesive and the

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Distribution A: Approved for Public Release

surface. This theory has led to the use of surface patterning2 and surface roughening via grit blasting46-47 as ways of increasing adhesive strength. Grit blasting of aluminum is common in Army applications,33 and was therefore adopted in this work. It is important to note, however, that mechanical interlocking is only effective if the substrate is effectively wet by the adhesive,45 and can cause air voids at the adhesive/adherend interface if wetting is poor. Electrostatic attraction theory involves the formation of an electrical double layer at the interface between the surface and adhesive or shared electrons due to differences in the electronic band structure.45 This may occur for silane-treated aluminum interfaces, where hydrogen bonding48 can drive the coupling agents to the metal oxide interface, where they subsequently covalently bond via a condensation reaction, usually under high temperature conditions.49 Electrostatic interactions in polydopamine may also occur as is evidenced by its adhesion to proteins,50 but this effect is not entirely clear and is not likely to be responsible for the poor adhesion observed here. Weak boundary layer theory postulates the presence of a weak layer between the adhesive and adherend, which can involve things like surface contaminants (e.g., grease),51 weak surface oxides36 on the adherend, or deleterious reactions between the adhesive and adherend.52 In our case, there are at least three different ways in which a weak boundary layer may be responsible for the poor results of PDA. First, it is possible that the wetting between the DGEBA/D230 adhesive and the PDA-treated surfaces is poor. To investigate this possibility, we measured contact angles of these using the drop shape method as discussed below. Second, the aluminum surface itself might have been weakened due to corrosion upon exposure to the alkaline conditions, creating a weak oxide layer. To explore this option, we performed XPS on the 0:0 surfaces treated under alkaline conditions to gage the differences in surface chemistry. Finally, the polydopamine film itself may actually be quite weak structurally, as may be guessed based on its aggregate structure at long deposition times.41 In this case it would effectively operate as a contaminant and result in adhesive failure in the bond. To explore this option, we performed XPS on the fractured interfaces to look at the chemical nature of these interfaces. Contact Angle Measurements using Adhesive Solution We performed contact angle measurements with an uncured DGEBA/D230 solution to determine the effect of polydopamine depositions on adhesive wetting of the aluminum 2024 adherend. These adherends were surface treated according to the same procedure as the lap shear experiments for the following conditions: untreated, (0:1,0.1 h,w,h), (0:1,18 h,w,h), (0:1,18 h,h), (3.5:1,0.1 h,w,h), and (1:0). All surfaces exhibited good wetting, with contact angles of