Printer Friendly

Reduced indentation recovery temperature at the surface of a crosslinked epoxy coating in humid conditions.

Abstract Relaxation of surface defects shows that the surface of a crosslinked epoxy system, under normal laboratory conditions, apparently has a softening transition temperature that is approximately 20[degrees]C lower than calorimetric measures of [T.sub.g] in the bulk. Physical aging data here confirms that this transition does resemble a glass transition. This observation is significant for properties that are determined by the surface of a coating. When ambient humidity is such that an epoxy may absorb significant quantities of moisture, it is plasticized, reducing its [T.sub.g]. When the relaxation of nanoindentations was measured in a humid environment, the value of this softening surface transition temperature was further reduced by 10[degrees]C. Thus, the surface properties of a polymer coating, such as wear, durability, friction, will depend not only on the ambient temperature, but also on whether the environment is humid. This is very important when such a coating responds to an environment that changes, e.g., natural weather.

Keywords Surface [T.sub.g], AFM, Humidity, Plasticization, Nanoindentation


A reduced surface glass transition temperature ([T.sub.g]) is now a well recognized finding in thermoplastics. A reduction of polymer density and enrichment of chain ends is expected at the polymer--air interface(1) and one can expect a lower [T.sub.g] there. The enhancement of dynamics at a surface and the confinement effect at polymer--substrate interface can change [T.sub.g] relative to its bulk value, over a length scale of the order of 100 nm from the interface.(2)

In prior work on crosslinked epoxy coatings,(3) the softening transition temperature, measured by the relaxation of roughness caused by photodegradation or nanoindentations at the surface of the material, was found to be significantly lower than the bulk Tg. Barring accidents of curing chemistry, typical cross-linking processes involve evaporation of solvent and constrained migration of monomers, which cause the accumulation of chain ends and limit the crosslink density near the surface.(4) In any case, the three-dimensional, constrained network structure of a cross-linked coating is inevitably limited near the surface to be a two-dimensional state. Thus, for a number of reasons one might anticipate [T.sub.g] of the material at the surface of a crosslinked polymer to be less than in the bulk, as seems to be the case.(3) It is natural to think about the likely response of a coating at a particular ambient temperature by referring to its [T.sub.g], but there may be a discrepancy between [T.sub.g] as it might be easily (and conventionally) measured in the bulk, and the value of [T.sub.g] that occurs where the performance is determined, i.e., at the surface.

Most coating polymers are somewhat polar and used in conditions where there is ambient humidity and thus they absorb a little moisture. It is well known that moisture will plasticize epoxies,(5) reduce their glass transition and make them softer. The work reported here to determine any additional effect of moisture on the surface [T.sub.g] was done, as before, by tracking the recovery of surface deformations to determine the surface [T.sub.g].


As before, a diglycidyl ether of bisphenol A (DGEBA), Epon[R] 828, was cured with Epikure[R] 3115, a polyamide based on dimerized fatty acid and polyamines and the coating samples were made and cured as before.(3) Apart from the solvent xylene, no other ingredients were present.

Near a visible registration mark, scratches or indentations were created on coating surfaces using Hysitron TriboIndentor[R] with a Berkovich diamond indenter of tip radius R [Less than] 100 nm, all with a peak force of 400 [mu]N. Indentations were very similar in initial depth and size. All controlled surface defects were made at room temperature and humidity and then the samples were stored in desiccators at room temperature before annealing on subsequent days.

The recovery of the coating's surface defects at certain relative humidities (RH) and temperature was studied in an environmental chamber, Associated Environmental Systems BMA Company, Model LH1.5. Samples with previously made indentations were exposed to different combination of RH and temperatures, 50% RH, 75% RH, and 95% RH at 25[degrees]C, and 35, 40, and 45[degrees]C at 70% RH. Immediately after being withdrawn from the chamber, the indentations were scanned by atomic force microscopy, AFM, in tapping mode AFM (Veeco Dimension 3100 with a NanoScope Ilia controller). As will be seen later, Figs. 3 and 4 have indentation profiles showing how they had recovered at various annealing times which are given in the figure legend and the caption. After 60 h, there was no systematic change perceptible for any of the samples. The scan size was varied, depending on the dimension of deformations, but usually 10 x 10 [mu]m. Images were analyzed in Nanoscope 6.12r1 SPIP[R] software.(3) Moisture content at saturation was measured on free films exposed in the humidity chamber at set RH and temperature for 2 days, and then measured by thermogravimetric analysis (TGA). The water content was obtained by calculating the total weight loss from room temperature to 105[degrees]C when increasing the temperature at a rate of 5[degrees]C/min.

Surface modulus was determined using the Nano-DMA[TM] (dynamic mechanical analysis) mode, which is available on the Hysitron Tribolndentor[R] as well as the more conventional determination of the modulus from the indentation unloading curve. In NanoDMA mode, the indenter tip penetrated in (20 nm) increments up to 400 nm and used a frequency of 100 Hz. The static load increased from typically 10 to 2000 [mu]N with a dynamic load of 2 [mu]N. The harmonic amplitude was in the range 1.0-1.8 nm with phase angle that varied 2-3 degrees either side of zero. Only relative trends in modulus were sought, so knowing the exact profile of the indenter tip was not necessary. Moduli were calculated simply from the force curve and the tip contact area. NanoDMA measurements were done at room temperature on samples annealed for various periods at 55[degrees]C, since that was the temperature thought to be the surface transition temperature for this material.(3)

Bulk [T.sub.g] was measured previously(3) at different heating rates by differential scanning calorimetry (TA Instruments 01000), DSC.

Results and discussion

The [T.sub.g] behavior of the bulk was determined previously.(3) At a 5[degrees]C/min heating rate, the result was 77.4[degrees]C, used as the nominal bulk [T.sub.g] for discussion since it is common to use a heating rate of 5[degrees]C/min for such determinations. Compared to other techniques, DSC usually returns the lowest value for [T.sub.g] yet the surface transition temperature that was found before was lower than the bulk [T.sub.g] even at heating rates much lower than 5[degrees]C/min.

Surface transition temperature via indentation recovery

Figure 1(a) shows the final extent of recovery of indentations (100 nm deep initially) on the epoxy when annealed at temperatures below and above its bulk [T.sub.g]. Here, the depth of the indentation is taken from the lowest point that the AFM probe penetrates, to the position of the average surface(3) before indentation, i.e., not including the pileup around the indentation which varies in direction according to the shape of the three-cornered indentor. Figure 1(b) shows examples of the scratch profiles at the end of the annealing period. The inflection point in Fig. 1(a) would suggest that a transition in behavior occurs at approximately 50-60[degrees]C.

Surface recovery is driven by viscoelastic recovery after the indentation stress is removed, aided by surface energy minimization which tends to reduce the surface profile as much as the rheological properties at the surface permit.(6) Especially for deeper indentations, most of the stressed material underneath an indenter tip is beyond its yield stress and can be expected to respond faster while the stress is still applied. This accelerated response will persist for some period after the stress is removed. However, as before(3) these measurements were taken at various annealing temperatures, when the annealing did not start until some hours of storage elapsed after the indentations were made, thus it is not clear that the enhanced recovery noted here is connected to faster response of the "yielded" material. Further, the transition temperature measured previously for the dry epoxy films was the same value whether measured by nanoindentation or by recovery in surface degradation.(3) Explaining this behavior is beyond the scope of this brief note.

Figure 1 indicates that significant recovery of surface indentation occurs at temperatures lower than the bulk [T.sub.g]. If the inflection point at 50-60[degrees]C is taken as a surface transition temperature for the epoxy, it is [tilde]20[degrees] lower than the bulk [T.sub.g], consistent with other studies of surface Tg in polymer films,(7-11) but smaller than such effects seen in thermoplastics.


It is useful to seek confirmation that this transition can be thought of as a glass transition. Mechanical modulus increases due to physical aging (structural relaxation) when a polymeric coating is kept at temperatures lower than its [T.sub.g].(12) If the temperature is at, or above [T.sub.g], there is no aging effect. Aging at 55[degrees]C was studied, since it is about 20[degrees] lower than bulk [T.sub.g] of the epoxy, and thus bulk mechanical properties would show obvious signs of aging. However, 55[degrees]C is also at the center of this temperature range that resembles a surface glass transition. The surface modulus of an epoxy sample was measured with nanoDMA, and is shown as a function of annealing time in Fig. 2.


The surface modulus does not exhibit much change after extensive annealing at 55[degrees]C, suggesting a lack of physical aging here and is consistent with 55[degrees]C being very close to the Tg at the surface. The smooth overlapping modulus profiles also indicate that there are no additional changes occurring due to either post-curing or other chemical changes at the surface.

Humidity-induced recovery

Figures 3(a)-3(c) are the cross-section profiles of indentations when exposed in the chamber at different RH. The cabinet condition at 25[degrees]C and 50% RH is close to the typical laboratory ambient environment (53% RH and 21[degrees]C, at the time), where only limited indentation recovery had been observed previously. At higher RH, as seen in Figs. 3(b) and 3(c), the indentation recovery is accelerated similar to an increase of temperature. Figure 3(d) summarized the indentation depth changing with RH. The exposure to 95% RH produces much greater recovery than at the lower RH. Although it is not a focus here, the increased water content of the films may be in more than one form(13) and so affects the mechanical response differently.


Figure 4 shows the indentation recovery at different temperatures with a set RH. At 70% RH, in Figs. 4(a), 35[degrees]C, and 4(b), 40[degrees]C, it can be seen that indentations recover gradually with time, up to 40 h.


In Fig. 4(c), at 45[degrees]C and the same RH, indentation recovered almost completely in an hour. The large difference between Figs. 4(c) and 3(b) contrasts behavior at two temperatures but at the same RH. Comparing Figs. 4(b) and 4(c), the more rapid recovery indicates that the motion of surface layer molecules is enhanced much more between 40 and 45[degrees]C when the RH is 70%. Therefore, 45[degrees]C, about 10[degrees] lower than the surface transition temperature of the dry coating, must be the corresponding surface transition temperature of this wet sample.

A free film sample equilibrated in the same situation was found to contain 2.3% of water (by weight)(14) measured by TGA. However, a surface layer may contain a higher amount of moisture due to extra "free volume" in the less constrained polymer, and in any case, plasticization of surface polymer may well be different to plasticization of the bulk. Actual water content of samples at a certain RH but at different temperatures will be different, because the RH depends on temperature and is not an absolute value which would dictate the partition of water into the epoxy coating. For samples with higher water content, or completely saturated, the surface transition temperature might be further reduced.

If one is concerned with a property determined by the coating surface one must not only recognize that the behavior of the surface may be substantially different from that of the bulk, but if the ambient conditions contain moisture, the surface behavior may be even further from expectations. Performance where the service temperature and humidity change will depend on how those conditions change. It is very apparent why it is troublesome to correlate performance from one climatic exposure with that in a different climate.


Recovery of both nanoindentation and degradation roughness shows that there is a transition in surface recovery on dry epoxy at approximately 55[degrees]C, 20[degrees] lower than its bulk [T.sub.g] (determined by calorimetry). This surface recovery is enhanced substantially in a humid environment. When RH is 70%, surface indentation recovered significantly at 45[degrees]C, approximately 10[degrees] lower than the transition temperature for the dry material.

The lack of physical aging seen in mechanical modulus measured at 55[degrees]C at the surface of the dry epoxy indicates that the surface transition temperature found at that temperature has some similarity to a glass transition temperature. However, there are complications in using nanoindentation to study surface relaxation, so further clarification of the features observed here would require more subtle techniques.

If the use of a coating depends on the tribological or other surface properties, e.g., gloss via surface roughness, then any predictions of its performance should include allowance for the effects seen here.

Acknowledgments This work was funded by the Army Research Laboratory and NSF EPSCoR, North Dakota.


(1.) Jones, RAL, Richards, RW, Polymers at Surfaces and Interfaces. Cambridge University Press, UK (1999)

(2.) Torkelson, JM, The Distributions of T., Values and Physical Aging Across Thin and Ultrathin Polymer Films and Within Polymer Nanocomposites, American Physical Society, March 2005

(3.) Shi, X, Croll, SG, "Thermal-Induced Recovery of Small Deformations and Degradation Defects on Epoxy Coating Surface." J. Coat. Technol. Res., 7 (1) 73-84 (2010)

(4.) Kiil, S, "Quantification of Simultaneous Solvent Evaporation and Chemical Curing in Thermoset Coatings." J. Coat.. Technol. Res., 7 (5) 569-586 (2010)

(5.) Xian, G, Karbhari, VM, "Segmental Relaxation of Water-Aged Ambient Cured Epoxy." Polytn. Deg. Stab., 92 1650-1659 (2007)

(6.) Ranjbar, Z, Rastegar, S, "Evaluation of Mar/Scratch Resistance of a Two Component Automotive Clear Coat Via Nano-Indenter." Prog. Org. Coat., 64 4387-4391 (2009)

(7.) Jones, RAL, "The Dynamics of Thin Polymer Films." Curr. Opin. Colloid Interface Sci., 4 (2) 153-158 (1999)

(8.) Kerle, T, Lin, Z, Kim, HC, Russell, TP, "Mobility of Polymers at the Air/Polymer Interface." Macromolecules, 34 3484-3492 (2001)

(9). Zaitseva, AV, Zaitsev, VB, Rudoy, VM, "The Study of Polystyrene Surface Glass Transition by Luminescent Molecular Probes." Surf Sci., 566-568 821-825 (2004)

(10.) Fakhraai, Z, Forrest, JA, "Measuring the Surface Dynamics of Glassy Polymers." Science, 319 600-604 (2008)

(11.) Bliznyuk, VN, Assender, HE, Briggs, GAD, "Surface Glass Transition Temperature of Amorphous Polymers. A New Insight with SFM." Macromolecules, 35 6613-6622 (2002)

(12.) Shi, X, Fernando, BMD, Croll, SG, "Concurrent Physical Aging and Degradation of Crosslinked Coating Systems in Accelerated Weathering." J. Coat. Technol. Res., 5 (3) 299-310 (2008)

(13.) Cao-Paz, A, Covelo, A, Farina, J, Novoa, XR, Perez, C, Rodriguez-Pardo, L, "Ingress of Water into Organic Coatings: Real-Time Monitoring of the Capacitance and Increase in Mass." Prog. Org. Coat., 69 150-157 (2010)

(14.) Shi, X, Hinderliter, BR, Croll, SG, "Environmental and Time Dependence of Moisture Transportation in an Epoxy Coating and Its Significance for Accelerated Weathering." J. Coat. Technol. Res., 7 (4) 419-430 (2010)

X. Shi, S. G. Croll (*)

Department of Coatings and Polymeric Materials, North Dakota State University, Fargo, ND 58102, USA


COPYRIGHT 2011 American Coatings Association, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Shi, Xiaodong; Croll, Stuart G.
Publication:JCT Research
Date:Jul 1, 2011
Previous Article:Wear and corrosion properties of electroless nickel composite coatings with PTFE and/or Mo[S.sub.2] particles.
Next Article:A Monte Carlo determination of the effectiveness of nanoparticles as spacers for optimizing [TiO.sub.2] opacity.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters