pH-Responsive, Thermoset Polymer Coatings for Active Protection against Aluminum Corrosion

This paper describes the synthesis and use of multifunctional methacrylic monomers, which contain basic (amine) functional groups, including an example in which an acid-labile tert-butylcarbamate-protected glycine is used to form a novel methacrylic monomer. The “protected” amino acid-derived functional monomer (BOC-Gly-MA) is copolymerized with an epoxide functional methacrylic monomer (GMA), to deliver novel multifunctional polymers, which are processed into powder coatings and used to study filiform corrosion at the surface of an aluminum substrate. The BOC-Gly-MA-containing copolymers were shown to improve a coating’s anticorrosion performance, presenting the lowest average filiform corrosion (FFC) track length, total FFC number, and total corroded surface area (CSA) of the coatings investigated. Further to this, a mode of action for the role of BOC-Gly functional polymers in corrosion protection is proposed, supported by both solution and polymer–aluminum interface studies, delivering new insights into the mode of action of pH-responsive polymer coatings.


Powder coating manufacture
The acrylic polymers were manufactured into powder coatings, according to the standard powder formulation shown below.

Analytical methods and equipment 1.3.1 Mass Spectroscopy (MS)
Samples were acquired using a Thermo RSLC coupled to an ABSciex 6600 QTof.The equivalent of 1-5 µM (1-5 pmol/µl) was injected onto the system per run.The loading pump was used at high flow to run linear chromatography gradients.All samples were acquired in positive mode.The buffer system comprised of buffer A 0.1% (v/v) formic acid and buffer B 99.9% acetonitrile, 0.1% (v/v) formic acid.Chromatographic separations were achieved using a Fortis C8 column, 100 mm x 2.1 mm, 3 µM particle size, 45 °C at a flow rate of 250 µl/min.Samples were loaded onto the column and desalted online for 1 minute, with eluent diverted to waste.A valve switch directed flow to the mass spectrometer.Samples were then eluted from the column using a linear gradient from 3-70% Buffer B over 12 min.Total run time was 17 min.The eluent was directed into an ABSciex 6600 QTof, operated in positive mode.Source conditions were; temperature 180 °C, GS1 25, GS2 15, ISFV 4500v.Data was acquired in MS scan mode between 100-1000 m/z.Resolution of the instrument was 45,000 at m/z 829.54

High-resolution mass spectra
Results recorded using a Vanquish liquid chromatography front end connected to IDX high resolution mass spectrometer system.The chromatographic separation was achieved using a Waters Acquity UPLC BEH amide column (2.1 x 150 mm with particle size of 1.7 μm) part no 186004802, operating at 45 °C with a flow rate of 200 μL/min.MS data were acquired using the AcquieX acquisition workflow.

Differential Scanning Calorimetry (DSC)
Glass transition (Tg) analyses were conducted on all polymeric materials using a Perkin-Elmer Pyris DSC 8500 and analyzed on Pyris software (version 11.1.1.0492).Samples had previously been thermally cured in an oven at 170 °C.Sample masses of 3-6 mg were weighed into standard aluminum DSC pans with perforated lids and were heated from -50 °C to 200 °C at a constant rate of 20 °C per minute (unless otherwise stated) and the Tg reported as the midpoint of the endothermic step in the heat flow signal output (Tg onset and endpoint was also recorded).Melting point analysis was conducted using a PerkinElmer Pyris DSC 8500 with an intercooler II and analyzed on Pyris software (version 11.1.1.0492).
Solid samples between 3-6 mg were weighed into standard aluminum DSC pans and heated from -50 °C to 300 °C at a constant rate of 20 °C per minute and the melting point taken from the peak of the endothermic event that corresponds to the melting transition on the heat flow signal output.

SEM imaging
Was recorded using a field emission TESCAN MIRA 3 with gigantic chamber at 5kV, analyzed using Alicona 3d imaging and AztecEnergy.The samples analyzed were prepared by first cooling a 10 x 100mm section of coating free-film in liquid nitrogen, the cooled sample was then fractured.The edge of the samples fracture point was coated in silver to increase sample conductivity.

Microscopic infra-red spectra
Were recorded from 4000 to 650 cm -1 in 16 scans using a PerkinElmer Spotlight 150i FT-IR microscope, in ATR mode with an aperture of 100 x 100µm.

Statistical analysis
Was performed using JASP software, further details of the results produced are later discussed.

Monomer experimental
To a solution of BOC-GLY (2) (14.73 g, 84 mmol) and K 2 CO 3 (0.97 g, 7 mmol) in THF (20 mL, degassed using N 2 ) was added GMA (1) (9.95 g, 70 mmol) before heating to 50 °C under stirring and N 2 .After 24 hours the reaction mixture was cooled to room temperature and diluted with ethyl acetate (100 mL).The reaction mixture was washed with a saturated NaHCO 3 solution (3 x 100 mL), the organic phase was collected, and dried over MgSO 4 .The organic phase was then concentrated under reduced pressure, affording the resulting clear oil 3 as a mixture of isomers (3 a & 3 b ) (19.98 g 75%), as determined by 1 H-NMR and mass spectrometry.

Powder coating curing conditions
Powders were sprayed to a total adhered powder mass of 1.3 ± 0.1 g onto a 70 mm by 150 mm aluminum Q-panel for corrosion testing, or 100 mm X 150 mm for free film creation.The coated panels were then heated in a conventional fan-oven to 205 °C for 5 minutes, followed by 185 °C for a further 25 minutes.

Polymer synthesis
1.6.1 General method of small-scale polymer synthesis (solvent mass equal to 20% of the total monomer mass).The resulting solution of monomer and initiator was added drop-wise to a 100 mL 3-neck round bottom flask containing butyl acetate (HT) or isopropyl acetate (LT) (equal to 50% of the monomer mass) at an internal flask temperature of either 125 °C (HT method) or 85 °C (LT method) by use of a syringe pump at a rate of 0.5 mL/min, whilst under N 2 and constant stirring provided by a magnetic stirrer.Once solution addition was complete, the mixture was stirred for a further 1 hour at 125-130 °C (HT) or 85-90 °C (LT), after which time the hot polymer mixture was poured into foil lined trays, and then placed into a vacuum oven (-1020 mbar) at 100 °C (HT) or 85 °C (LT) for 3 hours to remove residual solvent and non-polymerized (residual) monomer.

High temperature (HT) and low temperature (LT) large scale polymerisation reaction conditions
A solution of mixed acrylic and methacrylic monomers (according to Table 1, main text) was combined with tert-butyl peroxy-2-ethylhexanoate (Trigonox 21S) (high-temperature, HT method or azobisisobutyronitrile (AIBN) (low-temperature, LT method) and diluted with butyl acetate (HT and LT) (solvent mass equal to 20% of the total monomer mass).This mixture was added dropwise to a 2 L preheated flange neck flask containing butyl acetate (HT and LT) (equal to 50% of the monomer mass) When formulating polymerisation reaction compositions, the initiator used, and its mole percentage was selected so that Mn and Mw were matched (within experimental error) under the differing processing conditions by i) varying the initiator:monomer ratio and ii) using a lower decomposition temperature initiator for the LT conditions.
The structures of BOC-Gly-MA, DPA-MA, and control (CTL) polymers are shown in Figure S1 to Figure S5.
Table S4 to Table S8 provide the weight percent of the monomers, and mole percent of the radical initiator present in each polymer formulation.Where a (T) or an (A) has been used to denoted whether Trigonox 21s or AIBN was used as the radical initiator respectively.
Due to the complex structure of the statistical (random) co-polymers synthesized during this research, steps were taken to simplify the 1 H-NMR analysis.While the polymer chains present in each sample will vary in molecular weight, the molar percentage of the co-monomers present will be representative of the whole polymer sample.As such, each 1 H-NMR has been integrated assuming 100 moles of the sample were present, so the consistent peaks such as that of the -OCH 3 of MMA could be used as a reference for further integrations.As it was known that the -OCH 3 would produce an integration of 3 per repeating MMA unit, and in a representative formulation MMA makes up 62.8 mol% of the co-polymer formulation, the relative integration for the peak representing -OCH 3 would therefore be 188.4(62.8 x 3).-------------- 5-DPA-HT* abundance -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 X : parts per Million : Proton  Figure S8: 1 H-NMR (crude) of the mixed aromatic (colored) impurity from the BOC-Gly-MA synthesis, when using tertbutylhydroquinone as a radical scavenger to suppress polymerisation.

Analytical experiments and calculations
The integration of the BOC-NHCH 2 -R (B) against the unprotected NH 2 -CH 2 -R (A) allowed for an estimation of the degree of BOC protection group removal when heated using equation S1 below.

Model acidic corrosion study
Functional co-polymer (500 mg, 50-BOC-HT) was placed in a glass vial containing 8mL of 4M HCl(aq)/THF (1:1 v/v).IR (Bruker alpha Platinum-ATR, 32 scans, 4000 to 650 cm -1 ) scans were performed on extracted polymer samples once every 24h for a total of 7 days.To obtain each polymer sample a 0.5 mL aliquot of the reaction solution was diluted in 5mL of DCM, this solution was then washed with 5mL of saturated NaHCO 3(aq) .The organic phase was collected and dried over MgSO 4 , solvent was subsequently removed under reduced pressure, to afford a dried polymer sample for IR analysis. 1H-NMR analysis was also performed on a sample taken prior to emersion in the aqueous solution, and one on the 7 th day extraction in the aqueous solution, the spectra are shown in Figure S10.

BOC functionality at the coating-metal interface study
To directly measure the impact of filiform corrosion on a functional thermoset coating, a new experimental method was developed.It was designed to allow for direct access to the coating aluminum interface without the requirement for destructive techniques (i.e., removal of coating from the aluminum substrate by scalpel).High weight percent (50% w/w) polymers capable of curing (50-BOC-HT) were synthesized to provide increased ease of detection of functional group transformation.The polymer coating was also adhered to a separate aluminum substrate to maintain structural integrity during the removal of corrosion products.

50-BOC-D0
50-BOC-D7 The coated aluminum panel was cut into 30 x 30 mm squares, which were then placed in-between a square of PTFE and a secondary thicker piece of cast aluminum (aluminum substrate B, see Figure S12), both cut to 35 x 35 mm.The polymer coated surface of the substrate samples was placed in contact with the cast aluminum substrate B, as illustrated below.The difference in area between the coated substrate (30 x 30 mm) and the aluminum substrate B (35 x 35 mm) was employed to produce an outer edge of uncovered aluminum which was susceptible to corrosive attack.As such filiform corrosion would initiate from the exposed edge, and propagate across the aluminum substrate B surface, whilst interacting with the coated substrate secured on top, simulating the formation of filiform corrosion in a coated product which has experienced damage.
The current study focuses on the interaction between the aluminum substrate B and the functional coating surface, and not the relationship between the functional coating and the aluminum (Q) panel.As such, the variation in coating thickness resulting from the addition of the tape was not theorized to have a significant effect on the results of the study.

Filiform corrosion test procedure (SAE J2635)
Quantification of filiform corrosion (FFC) present on all coated aluminum substrates followed the process set out by the SAE (society of automotive engineers) J2635 filiform corrosion test procedure for painted aluminum wheels and painted aluminum wheel trim.To initiate the accelerated corrosion process, each coated aluminum substrate was scribed to remove the protective polymer coating and expose the underlying aluminum substrate using a carbide tipped blade, for a length of 100 mm using a guide to reliably produce a straight line.Scribed samples are then placed into a copper accelerated salt spray (CASS) chamber at a 45° angle for 6 hours, upon completion the samples are washed with deionized water.
The coated aluminum substrates are then placed into a humidity chamber (60±1 °C, 85% relative humidity ±3%) for a period of 30 days, at an angle of 45° to allow for moisture run off.FFC length measurements are recorded perpendicular to the scribe, without including the scribe.FFC number was recorded from both sides of the scribe omitting the first and final 5 mm of the scribe.FFC length and number were recorded both manually, and automatically using Quantiz measurement techniques.Whereas corroded surface area and average FFC width measurements were only obtainable using Quantiz analysis. (

Statistical testing data
Analysis was conducted using median absolute deviation (MAD) for comparison of FFC length, as the results were not normally distributed.MAD was used as it is a more robust measurement of non-normally distributed data, compared to standard deviation (SD).The non-normal distribution is a result of the relatively low presence of longer FFC trails compared to a higher abundance of shorter trails, causing a tailing effect.
heated to 125 °C (HT) or 85 °C (LT), by use of a peristaltic pump at a rate of 7.4 mL/min over 3 hours, whilst under N 2 and constant stirring provided by an overhead stirrer.Once the solution addition was complete, the reaction was then stirred for a further 1 hour at 125-130 °C (HT) or 85-90 °C (LT).A solution of Trigonox 21S (HT) or AIBN (LT) equal to 10% mass of initial initiator weight was then added to the reaction, where stirring at a temperature of 125-130 °C (HT) or 85-90 °C (LT) was maintained for a further 45 minutes, for the purpose of reacting residual monomer.The reaction was then maintained at 100 °C and a vacuum distillation was performed to remove the majority of the solvent.After 40 minutes of distillation under vacuum the hot polymer mixture was poured into foil lined trays, the tray containing the polymer was then placed into a vacuum oven (-1020 mbar) at 100 °C (HT) or 85 °C (LT) for 3 hours to remove residual solvent and non-polymerized (residual) monomer.

Figure S1 :
Figure S1: General structure of the large-scale CTL functional polymers.

Figure S3 :
Figure S3: General structure of the large-scale BOC and GMA functional co-polymers.
Figure S6 a-h): DSC traces for polymers used in this study and used to provide T g data presented in main manuscript.

Figure
Figure S7 a-j: NMR spectra of BOC-Gly-MA (a) and polymers (b-j) used in this study.
Figure S9: 1 H-NMR Spectra analyzed to determine the percentage of BOC groups remaining after TGA experimentation.
Figure S10: 1 H-NMR Spectra analyzed to determine the percentage of BOC groups remaining after model corrosion study.

Figure S11 :
Figure S11: Drawdown bar and taped aluminum substrate before coating application.

Table S4 :
Monomer and initiator weight and mole percent values used to formulate the small-scale control IP polymers.

Table S5 :
Monomer and initiator weight and mole percent values used to formulate the large-scale control (CTL-HT) polymers.

Table S6 :
Monomer and initiator weight and mole percent values used to formulate the small-scale BOC functional polymers.

Table S7 :
Monomer and initiator weight and mole percent values used to formulate large scale BOC functional polymers.

Table S8 :
Monomer and initiator weight and mole percent values used to formulate large scale DPA functional polymers.

Table S9 :
Results presenting relative peak intensities for both BOC-LT and BOC-HT samples, with the calculated loss in mass presented in the final column.
/  )   =   Equation S2 was used to calculate the average FFC width ( ) present on an aluminum substrate, post   exposure to accelerated corrosion conditions.Where is the total corroded surface area, is the     total number of FFC and is the average length of a single FFC filament, where each measurement ( ,     , ) is recorded for every sample panel in a coating series.