Cerium acetate-modified aminopropylsilane triol: a precursor of corrosion-preventing coating for aluminum-finned condensers.

A poly-acetamide-acetoxyl methyl-propylsiloxane (PAAMPS) polymer containing a cerium (Ce) oxide derivative was synthesized through three spontaneous reactions: condensation, amidation, and acetoxylation. The PAAMPS synthesizing took place between the Ce acetate dopant and aminopropylsilane triol (APST) as the film-forming precursor aqueous solution at 150[degrees]C. PAAMPS was evaluated for use as a corrosion-preventing coating for air-cooled condensers consisting of two metal components: aluminum fins and carbon steel tube. The key factors to ensuring the maximum performance of the PAAMPS/Ce oxide coating system in mitigating the corrosion of both aluminum and steel included (1) minimizing the content of non-reacted water-soluble APST and Ce acetate remaining in the coating, (2) lowering the susceptibility of the coating's surface to moisture, (3) precipitating a passive [Ce.sup.3+] oxide ([Ce.sub.2][O.sub.3]) film insensitive to Cl dissolution over the metal's surfaces, and (4) possessing excellent adhesion to the metal's surfaces. The combination of these factors considerably decreased the rate of the cathodic oxygen reduction reaction at the corrosion site of the metals. In fact, the corrosion rate, measured in milli-inches per year (mpy), of bare steel was reduced more than one order of magnitude by applying this coating. Correspondingly, the coating extended the useful lifetime of steel exposed in a salt-fog chamber at 35[degrees]C from only ~10 hr to ~768 hr. Furthermore, this coating system far better protected an aluminum substrate from the corrosion than it did one of steel. Even a very thin nanoscale coating film deposited on the aluminum's surfaces reduced the corrosion rate, (mpy), by nearly two orders of magnitude over that of the bare aluminum. Also, the salt-spray resistance of aluminum panels covered by this nanoscale film was strikingly extended to more than 1440 hr, compared with ~40 hr for bare aluminum.

Keywords: Corrosion testing, ESCA, EXP, Auger, SIMS, FTIR, ATR, corrosion, corrosion protection, organosilane, waterborne, water-based, aluminum, iron, cast iron, steel

**********

One proposed method for improving the efficiency of the air-cooled condensers in geothermal binary power plants during the valuable summer season for selling electricity is to directly spray relatively clean geothermal brine or treated wastewater over the surfaces of the aluminum-finned steel condenser tubes. (1) As expected, the water-sprayed condensers generated almost the same output of electricity from the plant as they do under winter conditions. Although this technique was very attractive and effective in providing a more robust generation capacity, a serious concern noted in our post-test analyses of the sprayed condensers was their corrosion together with the deposition of many nonremovable silicate scales on their surfaces. These condensers were composed of the aluminum fins circulating the surfaces of steel tubes; the distance between fins was less than 3 mm, while the height of each fin attached to the tube was ~15 mm. Thus, massive amounts of scales filled some of these spaces between the fins. Further, we observed that the underlying fins surfaces beneath the scale layers underwent a localized corrosion. There is no doubt that such negative phenomena, corrosion and scale deposits, are detrimental to retaining the function and efficacy of the condensers.

In trying to resolve this problem, our previous study focused on assessing the ability of a polyaminopropylsiloxane (PAPS) coating derived from water-based aminopropylsilane triol (APST) precursor solutions to protect the condensers against brine-caused corrosion. (2) The outstanding wetting behavior of the APST solution on the aluminum surfaces ensured great coverage of the entire fin surfaces by a defect-free, continuous PAPS film only ~1.2 [micro]m thick. In addition; this film adhered well to the fins surfaces. Such properties adequately protected the fin against brine-induced corrosion. In contrast, when the PAPS coating was applied to steel surfaces, its effectiveness in reducing the corrosion rate of steel was much less.

Based upon this information, emphasis in this study was directed toward improving the corrosion-preventing performance of PAPS coating for the condensers. Ideally, not only would the improved coating adequately protect the steel tubes in the condensers against corrosion, but also its thin film would mitigate the corrosion of the aluminum fins. Several recent papers (3,8) documented that, among the rare-earth metal salts, the environmentally benign cerium (Ce) salt compounds displayed high potential as a promising alternative corrosion inhibitor for replacing the toxic hexavalent chromium ([Cr.sup.6+]) salt compounds used as a pigment in the polymer coatings and conversion coatings in anodizing baths. The formation of Ce oxide- and hydroxide-based passive layers over the aluminum and steel surfaces was thought to be the major contributor to mitigating the corrosion of these metals.

Accordingly, our particular interest is to incorporate this green Ce-based salt compound as a corrosion-inhibiting dopant into the APST and to investigate its ability to enhance the potential of a polysiloxane-based polymer as a corrosion-preventing thin barrier layer. From among the Ce-based salt compounds, Ce acetate was chosen for this study. Our study covers the changes in the extent of wettability and spreadability of the APST precursor solutions over both aluminum and steel substrates as a function of the concentration of Ce acetate dopant, the susceptibility to moisture of the surfaces of the polymer film derived from Ce acetate-doped APST, the conformation of the polymers yielded by the reaction between the Ce acetate and APST, the thermal stability of these polymers, the phase identification of Ce oxides, and the chemistry at the outermost surface sites of the coating films. All the data were integrated and correlated directly with information obtained from the corrosion-related studies including the potentiodynamic polarization curve and salt-spraying resistance.

EXPERIMENTAL

Materials

Aminopropylsilane triol [[H.sub.2]N-(C[H.sub.2])[.sub.3]-Si(OH)[.sub.3],APST] (25% in water), supplied by Gelest, Inc., was used as a network-forming monomeric organosilane precursor. A 25% APST solution was diluted by adding an appropriate amount of deionized water to prepare a 20% APST aqueous precursor. The cerium (Ce, III) acetate hydrate, (C[H.sub.3]C[O.sub.2]).sub.3]Ce.x[H.sub.2]O, as the corrosion-inhibiting dopant in APST, was obtained from Aldrich. The 0.2, 0.5, 1.0, 2.0, 3.0, 6.0, and 10.0 wt% Ce acetate was added to the 20 wt% APST solution to prepare the standard Ce acetate-doped APST solutions. To fabricate a thin film, the standard solutions were then further diluted with deionized water to make 0.5, 1.0, and 3.0 wt% doped solutions. The metal substrates, supplied from Advanced Coating Technologies, Inc., were 3003 aluminum (Al) and AISI 1008 carbon steel test panels.

Coating Method

The surfaces of these substrates were coated with the nondoped and doped APST precursor solutions in the following sequence; First, to remove any surface contaminants, the test panels were immersed for 20 min at 80[degrees]C in an alkaline solution consisting of 0.4 wt% NaOH, 2.8 wt% tetrasodium pyrophosphate, 2.8 wt% sodium bicarbonate, and 94.0 wt% water. The alkali-cleaned panels were washed with deionized water at 25[degrees]C for one min, and dried for 15 min at 100[degrees]C. Then, the panels were dipped into a soaking bath of precursor solutions at room temperature, and withdrawn slowly. The wetted substrates were heated in an oven for 120 min at 150[degrees]C to yield solid films.

Measurements

The surface tension of doped APST solutions was determined with a Cenco-DuNouy Tensiometer Model 70535. The thickness of coating films deposited on the metal substrates was measured with the Dektak Surface Profile Measurement. Using a contact angle analyzer (Model CAA-3, Imass Instruments), the extent of wettability and spreadability of the nondoped and Ce acetate-doped APST solutions over the aluminum and steel surfaces was determined by measuring the contact angle within the first 30 seconds after dropping the solution onto the metal surfaces. Similar measurements were made of the contact angle to evaluate the susceptibility to moisture of the solid film's surfaces derived from the nondoped and doped APST solutions. In this case, the changes in the contact angle were made by dropping water onto the surfaces of the coating films. A thermogravimetric analyzer (TGA, Model 951, TA Instruments) was used to obtain information on the loss in weight and thermal stability of the films. Fourier Transform Infrared (FTIR, Model 1600, Perkin Elmer) was employed to identify the polymer structure of the solid film formed by the reaction between the APST and Ce acetate. The chemical state and compositions at the outermost surface sites of the coating were surveyed by X-ray photoelectron spectroscopy (XPS, Model CLAM 100, VG Scientific Ltd). DC electrochemical testing for corrosion was performed with the EG & G Princeton Applied Research Model 362-1 corrosion measurement system. The electrolyte was a 1.0 M sodium chloride. The specimen was mounted in a holder, and then inserted into an EG & G Model K47 electrochemical cell. The test was conducted under an aerated condition at 25[degrees]C, on an exposed surface area of 1.0 [cm.sup.2]. The polarization curves were measured at a scan rate of 0.5 mV[s.sup.-1] in the corrosion potential range from -1.35 to -0.75 V. The salt-spray testing of the coated Al and steel panels (75 mm X 75 mm size) was carried out in accordance with ASTM B 117.

RESULTS AND DISCUSSION

Properties of Ce Acetate-Doped APST Solutions

Table 1 gives the pH values and surface tensions of the standard 20 wt% APST solutions modified with 0.2, 0.5, 1.0, 2.0, 3.0, 6.0, and 10.0 wt% Ce acetate. As is seen, the value of pH of the APST solution declines with an increasing amount of Ce acetate, from a strong base of pH 10.94 for unmodified APST to mild base of pH 9.49 for 10 wt% Ce acetate. FTIR spectra (not shown) for all Ce acetate-dissolved APST solutions revealed the formation of Ce cation-bridged carboxylate complex, [Ce.sup.3+] (-OCOC[H.sub.3])[.sub.3], coexisting with the APST, suggesting that this complex may act as a mild acid. The surface tension of an unmodified APST solution gradually rose with increase in the proportion of Ce acetate to APST. In fact, its value of 58.7 dynes/cm for 10 wt% Ce acetate was equivalent to an increase of 7.5% over that of unmodified APST.

Solutions with a low surface tension display a better performance in wetting and spreading over the metal surfaces than those having a high surface tension. (9) Since good wetting by the solution contributes to a good coverage of solid film in terms of a defect-free, continuous barrier layer over the metal's surfaces, it is very important to know the magnitude of wetting and spreading behaviors of modified APST solutions on the metal surfaces. To obtain this information, the changes in contact angle, [theta], (in degrees) of solutions dropped onto the surfaces of aluminum and steel substrates were monitored as a function of the concentration of Ce acetate in the APST solution (Figure 1). The data illustrated that the value of [theta] conspicuously increased within the range of Ce acetate concentration from 0 to 3 wt%; beyond that, the extent of increase gradually rose between 3 and 10 wt% for both the aluminum and steel surfaces. With a 3 wt% Ce acetate-modified APST solution, the [theta] values for the aluminum and steel surfaces were 22.2 and 35.4 degrees, corresponding to an increase by as much as 71% and 30%, respectively, compared with those from the unmodified solution. By comparison, the values of 24 and 37 degrees, determined from dropping a 10 wt% acetate solution onto aluminum and steel surfaces, were only 8.1 and 4.5% increases, respectively, suggesting that there was no significant change in the wetting behavior on the metals of solutions modified with a Ce acetate concentration between 3 to 10 wt%. Relating this data to the surface tensions of the modified solutions, it appears that the [theta] value depends on surface tension; namely, an increase in surface tension results in an increment of the [theta] value, demonstrating that modified solutions with a high surface tension exhibit poor wetting ability on these metal surfaces. The data also reveals that the aluminum surfaces in either the modified or unmodified solutions had a higher wettability than steel surfaces. For instance, using the APST modified with 10 wt% Ce acetate, the [theta] value of 24.0 degrees for the aluminum surfaces was as much as 35% lower than that of the steel surfaces.

Characteristics of Solid Films

We previously reported that heating the APST solution transformed it into a solid film, which led to polycondensation reactions between the silanol groups within the APST, thereby forming the network structure of the polyaminopropylsiloxane (PAPS) polymer. (2) Thus, our attention particularly centered on whether the Ce acetate dopant reacts with APST during its polycondensation reactions. If there is a reaction between them, then the question raised is: what type of polymer structure is thereby assembled?

[FIGURE 1 OMITTED]

To respond to this intriguing question, we investigated the FTIR spectra of the solid films made by heating the unmodified, and 3, 6, and 10 wt% Ce acetate-modified 20 wt% APST solutions at 150[degrees]C. Figure 2 gives the FTIR spectra for these films over the frequency range from 2000 to 500 [cm.sup.-1]. According to the literature, (10,11) the film (Figure 2a) derived from the unmodified APST had the following absorption bands: 1654 and 1600 [cm.sup.-1], revealing the N-H bending vibration in the N[H.sub.2] group in conjunction with the bending mode of the H-O-H bond; 1443 [cm.sup.-1], that can be ascribed to C[H.sub.2] stretching in the propyl groups; 1120 [cm.sup.-1], reflecting the stretching of the Si-C bond in the Si-joined propyl groups; at 1020 [cm.sup.-1], corresponding to the antisymmetric bond stretching of the oxygen-bridged Si-O-Si linkage in the polysiloxane structure; at 929 [cm.sup.-1], representing the Si-O stretching mode of non-bridging oxygen; and 771 and 691 [cm.sup.-1], associated with the symmetric stretching and bending modes, respectively, of the Si-O-Si linkage. These data clearly revealed that during the transformation of the APST solution into a solid film at 150[degrees]C, dehydrating condensation reactions between the SiOH silanol groups within the APST create the polysiloxane structure that is represented by the formation of the Si-O-Si linkages, [equivalent to]Si-OH + HO-Si[equivalent to] [right arrow] [equivalent to]Si-O-Si[equivalent to] + [H.sub.2]O. Thus, the chemical structure of formed solid films was the polyaminopropylsiloxane (PAPS) polymer with the following network structure:

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FORMULA NOT REPRODUCIBLE IN ASCII]

By comparison with the spectral features of the unmodified APST-derived film, the spectrum (Figure 2b) of the film derived from the 3 wt% Ce acetate-modified APST was characterized by the appearance of four new bands at 1659, 1554, 1380, and 1302 [cm.sup.-1], along with a considerable decrease in the absorbance of the N[H.sub.2]-related bands at 1654 and 1600 [cm.sup.-1] and all of the Si-C and Si-O-Si linkage-associated bands at 1120, 1020, 929, 771, and 691 [cm.sup.-1]. Among these new bands, the 1659, 1554, and 1302 [cm.sup.-1] frequencies are associated with the stretching mode of C-N bond in the amide group, -CO-NH-R (12). The remaining new band at 1380 [cm.sup.-1] is assignable to the symmetric bending mode of acetoxyl methyl group linked to the Si element, Si-O-CO-C[H.sub.3] (13). Thus, it is possible to assume that the amide group was formed by an amidation reaction between the amine group in APST and the Ce acetate: Ce(-O-COC[H.sub.3])[.sub.3] + 3[H.sub.2]N- [right arrow] 3C[H.sub.3]CO-NH- + Ce(OH)[.sub.3]. On the other hand, the acetoxyl methyl group was derived through the acetoxylation reaction between the silanol in APST and Ce acetate [Ce(-O-COC[H.sub.3])[.sub.3] + 3[equivalent to]Si-OH [right arrow] 3C[H.sub.3]CO-O-Si[equivalent to] + Ce(OH)[.sub.3]], underscoring that the Ce acetate had a chemical affinity for two groups in APST, amine and silanol.

[FIGURE 4 OMITTED]

A decline in the signal intensity of all Si-C and Si-O-Si linkage-related bands can be seen in the spectrum (Figure 2c) of the film containing 6 wt% Ce acetate. Correspondingly, the spectrum also showed a marked growth in the peak at 1380 [cm.sup.-1], attributed to the formation of acetoxyl methyl group, while the prominence of the amide-related peaks remained unchanged. In addition, it was very difficult to identify the presence of the N[H.sub.2]-related peak at 1600 [cm.sup.-1] in this spectrum, signifying that all the N[H.sub.2] groups in APST were depleted by the amidation reaction with Ce acetate, and then were substituted by the acetamide groups, -CO-NH-C[H.sub.3]. The spectrum (Figure 2d) of 10 wt% Ce acetate exhibited a further decay of peak intensity of the Si-C and Si-O-Si linkage bands as well as a striking growth of the acetoxyl methyl peak.

To visualize the changes in quantity of these new and original linkages or groups as a function of Ce acetate content, we quantitatively computed the absorbance of the four peaks at 1020, 1600, 1659, and 1380 [cm.sup.-1], respectively, attributed to the Si-O-Si linkage, N[H.sub.2] group, -CO-NH-C[H.sub.3] group, and Si-O-CO-C[H.sub.3] group. Figure 3 plots the absorbance of these peaks versus the Ce acetate contents of 0, 3, 6, and 10 wt%. The data revealed that the numbers of the Si-O-Si linkages and N[H.sub.2] groups tended to decline with an increasing content of Ce acetate; in particular, the latter group vanished when 6 wt% Ce acetate was added. In contrast, the quantity of a new linkage related to Si-O-CO-C[H.sub.3] was enhanced with the increased Ce acetate. For the other new -CO-NH-C[H.sub.3] group, its peak absorbance was reached when the Ce acetate content was 6 wt%, as the curve illustrates; beyond this, it seemed to level off.

Based upon the previous information, the alteration in the APST-derived polymer structure by incorporating Ce acetate is as follows: When the Ce acetate-dissociated APST solution was heated at 150[degrees]C, the three chemical reactions took place: condensation, amidation, and acetoxylation. The condensation reaction between the silanols in APST led to the formation of the polypropylsiloxane polymer network structure. The two remaining reactions occurred between Ce acetate and the amine or silanol groups within APST, leading to the formation of the acetamide and acetoxyl methyl moieties as the end reaction products in the polypropylsiloxane network; meanwhile, the Ce hydroxide derivative was introduced into this polymer. Since the increase in number of acetoxyl methyl moieties caused a decrease in densification of Si-O-Si linkages, we believe that the silanol preferentially reacted with Ce acetate, rather than with other silanols. Thus, incorporating an excessive amount of Ce acetate into the APST appears to be detrimental to the formation of an ideal polysiloxane network structure containing a high density of Si-O-Si linkages because some silanols in APST were eliminated by their preferential reaction with Ce acetate. This was the primary reason why solid films made with the increased content of Ce acetate showed fewer Si-O-Si linkages, along with more acetoxyl methyl moieties. Consequently, we represent in Structure 2 the hypothetical conformation of polymer derived from Ce acetate-modified APST. It is categorized as the poly-acetamide-acetoxyl methyl-propylsiloxane (PAAMPS) polymer.

[FORMULA NOT REPRODUCIBLE IN ASCII]

Our focus next shifted to gaining information on the thermal stability of PAAMPS polymers. Figure 4 depicts the TGA curves of PAAMPS polymers over the temperature range of 30[degrees] to 700[degrees]C. For comparison purposes, the PAPS polymer derived from APST without Ce acetate was tested. The curve for the latter, denoted as "0% Ce acetate," revealed a gradual loss in weight between 30[degrees] and 350[degrees]C; beyond this temperature, a large reduction took place. We assumed that the weight loss incurred up to 350[degrees]C was due mainly to evaporation of all the nonpolymerized APST monomers. The features of the curve of the PAAMPS polymer made with 2 wt% Ce acetate were similar to that of the PAPS polymer, although there was a lower weight loss between 30[degrees] and 350[degrees]C. The loss in weight in this temperature range for the 3 wt% Ce acetate-made polymer was even less, while the overall features of its curve closely resembled those of the 0% and 1% Ce acetate-containing polymers. These findings suggest that adding a proper amount of Ce acetate minimizes the amount of thermally unstable APST monomers in the polymer, and also demonstrate that there was no significant difference in the onset temperature of the major decomposition between the PAPS polymer and the 3 wt% Ce acetate-made PAAMPS polymer. In contrast, the curve's features for the 6 wt% Ce acetate polymer were very different. In particular, there were two pronounced decomposition states, the first beginning at the onset temperature near 260[degrees]C, and the second as the primary phase of decomposition occurring around 330[degrees]C. The latter temperature was ~20[degrees]C lower than that for the 0, 2, and 3 wt% Ce acetate-made polymers. The curve also revealed that in the range of 30[degrees] to 260[degrees]C, there was a small weight loss of 0.5%, reflecting the removal of most of the nonpolymerized APST monomers. The weight loss of ~2.4% occurring between 260[degrees] and 350[degrees]C was due to the thermal decomposition of nonreacted Ce acetate remaining in the polymer. In fact, the TGA curve (not shown) of Ce acetate itself showed that its decomposition temperature began around 230[degrees]C. Hence, a Ce acetate content of 6 wt% appears to be excessive for tailoring an advanced polymer structure from the APST precursor. As expected, a greater weight loss in this temperature range was observed with a further increase in the amount of Ce acetate to 10 wt%. The curve from that sample also signified that the beginning of secondary decomposition takes place at ~310[degrees]C, corresponding to a temperature as much as 40[degrees]C lower than that of the 3 wt% Ce acetate. This finding strongly suggested that adding > 6 wt% Ce acetate is detrimental to retaining good thermal stability of PAAMPS. As described earlier from the study on the alterations in polymer conformation, an excessive amount of Ce acetate led to the formation of a PAAMPS polymer with a low degree of densification of siloxan, Si-O-Si, linkages because of the uptake of the silanol in APST by Ce acetate. Thus, such a decline in its thermal stability was due to an insufficient densification of siloxane linkages in the polymer network.

[FIGURE 5 OMITTED]

Nonetheless, the results from this TGA study clearly verified that adding a proper amount of Ce acetate plays an important role in minimizing the amount of thermally unstable nonpolymerized APST monomers remaining in the polymer and in retaining good thermal stability. In other words, the remaining APST monomers might have reacted with the Ce acetate to form the propylsilane structure containing the acetamide and acetoxyl methyl groups, which possess greater thermal stability than does APST.

In preparing water-repellent and -shedding coating films, the susceptibility of the coatings' surfaces to moisture is among the inevitable factors governing good corrosion-preventing performance. To obtain this information, the degree of water wettability of the PAPS and PAAMPS films' surfaces was estimated from the average value of the contact angle of a water droplet on their surfaces. Figure 5 plots the contact angle, [theta] (degree), versus the concentration of Ce acetate. At concentrations between 0 and 3 wt%, the [theta] value was markedly enhanced with an increased Ce acetate content; in fact, the [theta] value of PAAMPS' surface made with 3 wt% Ce acetate was more than twice as high as that of the PAPS' surface without Ce acetate. Since a high [theta] value correlates with a low rate of wetting, the surfaces of the films made with Ce acetate of up to 3 wt% became less susceptible to moisture. However, beyond this content, this value declined from 56.9[degrees] at 3 wt% to 48.0[degrees] at 10 wt%. Relating this finding to the alterations in polymer structure by adding Ce acetate, a certain amount of nonpolymerized APST monomers remaining in the film enhanced the extent of surface wettability because of their solubility in water. The APST remnants favorably reacted with Ce acetate to form water-insoluble acetoxyl methyl- and acetamide-linked propylsilane compounds. Thus, adding a proper amount of Ce acetate was responsible for developing the water repelling, water-insensitive film surfaces containing a minimum amount of APST. Since Ce acetate is also soluble in water, incorporating an excessive amount of it incurred a negative effect in preparing a film less susceptible to moisture.

[FIGURE 6 OMITTED]

Surface Chemistry

To support the information on the water-wetting behaviors of PAPS and PAAMPS coatings' surfaces previously described, we inspected the elemental compositions at the outermost surface site of PAPS and 3 wt% Ce acetate-derived PAAMPS coatings by X-ray photoelectron spectroscopy (XPS) (Table 2). Quantitative data for the respective atoms were obtained by comparing the XPS [C.sub.1s], [N.sub.1s], [O.sub.1s], [Si.sub.2p], and [Ce.sub.3d] peak areas, which then were converted into atomic concentrations using the differential cross section for core-level excitation. All the XPS measurements were made at an electron take-off angle of 45[degrees], which corresponds to an electron-penetration depth of ~5.0 nm. Thus, XPS data provide the atomic composition present in the superficial layer of ~5.0 nm thick. The atomic fraction of this layer in PAPS is composed of 63.1% C, 4.5% N, 26.5% O, and 5.9% Si. Comparing these values with those from the surfaces of the PAAMPS, there are three marked differences: (1) the incorporation of more carbon and oxygen atoms, (2) a decrease in N and Si contents, and (3) the detection of some Ce atoms. Differences (1) and (3) can be accounted for by the formation of moieties and derivatives yielded by the reactions between the Ce acetate and APST. Our study next shifted to identifying such reaction products and derivatives by surveying the XPS [C.sub.1s], [N.sub.1s], and [Ce.sub.3d] core-level excitations for both the PAPS and PAAMPS coatings' surfaces. In these core-level spectra, the scale of the binding energy (BE) was calibrated with the [C.sub.1s] of the principal hydrocarbon-type carbon peak fixed at 285.0 eV as an internal reference standard. A curve deconvolution technique was employed to substantiate the information on the carbon-, nitrogen-, and cerium-related chemical states from these core-level spectra. In the [C.sub.1s] region (Figure 6), the symmetrical single peak at the BE position of 285.0 eV for the PAPS was due to carbon in the propyl group. In contrast, the [C.sub.1s] curve of PAAMPS was characterized by appearing at three addition peaks at 286.2, 288.0, and 289.5 eV, while the C-H bond carbon at 285.0 eV still remained as the major component. The peak component at 286.2 eV is assignable to the C in C-O and C-N bonds. (14) The contributors to the peaks at 288.0 and 289.5 eV are the C in C=O and -COO bonds, respectively. (15) In the same figure, the [N.sub.1s] spectra for the PAPS and PAAMPS surfaces had the peaks at 399.8 and 400.8 eV, respectively. The former peak is ascribed to N in the amine group, (16) and the latter one reveals the N originating from the amide group. (16) Relating this finding to the chemical structure of PAAMPS illustrated earlier, it is possible to assume that among these carbon-associated bonds, three bonds, C-H, C-N, and C=O, belong to the acetamide moiety, and that three bonds, C-H, C=O, and -COO, are associated with the acetoxyl methyl moiety linked to Si. Since the acetamide and acetoxyl methyl moieties were introduced through the reactions between the Ce acetate and the N[H.sub.2] or SiOH within APST, we believe that these hydrophobic moieties present at the top surface of PAAPS play an important role in repelling the water. In contrast, the superficial layer of the APST-derived PAPS without Ce acetate contained nonreacted N[H.sub.2] and SiOH groups, suggesting that these hydrophilic groups might be the major reason why the PAPS film's surface was susceptible to moisture, reflecting a lack of water-repellent property.

Figure 7 illustrates the [Ce.sub.3d] core-level spectrum from the PAAMPS surface. The two peaks at 885.5 eV in the 3[d.sup.5/2] region and at 904.2 eV in the 3[d.sup.3/2] region can be attributed to the formation of [Ce.sup.3+] oxide as [Ce.sub.2][O.sub.3], (17) but not to the [Ce.sup.4+] oxide as Ce[O.sub.2] because of the absence of satellite peak near 916 eV. (18) Hence, upon heating the test panel at 150[degrees]C, the [Ce.sub.2][O.sub.3] was formed in the coating by the dehydration of Ce(OH)[.sub.3] as the by-product yielded through the reactions between Ce acetate and APST: 2Ce[(OH).sub.3] [right arrow] [Ce.sub.2][O.sub.3] + 3[H.sub.2]O.

Corrosion Protection

All the above data were correlated with the ability of the PAAMPS coatings to protect steel and aluminum against corrosion. In this part of our study, we used the PAAMPS polymers derived from the Ce acetate-modified 20 wt% APST precursor solutions as the coating films. The thickness of the coatings ranged from 2.1 to 2.5 [micro]m. The data on corrosion was obtained from the polarization curves of 150[degrees]C-baked coatings on metal panels exposed to an aerated 0.5 M NaCl solution at 25[degrees]C. Figure 8 illustrates the typical potentiodynamic polarization curves of log current versus potential for the steel panels coated with the unmodified polymers, and 1, 3, and 10 wt% Ce acetate-modified polymers. The shape of the curves shows the transition from cathodic polarization at the onset of the most negative potential to the anodic polarization curves at the end of positive potential. The potential axis at the transition point from cathodic to anodic curves is normalized as the corrosion potential, [E.sub.corr]. When compared with the uncoated steel, there are two noticeable differences in the curve's features for the unmodified PAPS polymer-coated steel marked as "0% Ce acetate." One was a large shift in [E.sub.corr] value to a less negative site and the other related to a marked reduction of cathodic current density (A/[cm.sup.2]) in the potential region between -0.75 and -0.95 V. Difference (1) directly reflects the degree of coverage of the coating over the entire steel surface; namely, the good coverage providing a continuous nonporous coating layer is responsible for moving the [E.sub.corr] value to a more positive site. For difference (2), the indication of low cathodic current density is attributed to inhibiting the cathodic reaction at the corrosion site of steel, particularly the oxygen reduction reaction, 2[H.sub.2]O + [O.sub.2] + 4[e.sup.-] [right arrow] 4O[H.sup.-]. Hence, such a reaction, which leads to the corrosion of steel, appears to be retarded by depositing a PAPS polymer coating on the steel's surfaces. Some investigators reported the strong chemical affinity of the silanol groups in the organosilanes for the oxide layer at the outermost surface site of metals, reflecting the coating's great adherence to the metal's surfaces. (19-22)

[FIGURE 7 OMITTED]

Correspondingly, our previous XPS study on the adherence of PAPS coating to metals revealed that a certain amount of PAPS was transferred to the metal's surfaces during tensile failure at the coating/metal joint, due to its excellent adhesion to the metal's surfaces. (2) Also, our data showed that the extent of the coverage of aluminum's surfaces by the PAPS' remnants was higher than that at the interfacial steel side, emphasizing that the adherence of PAPS to aluminum was much greater than to steel. Nevertheless, we assumed that one pivotal factor governing the retardation of cathodic corrosion reaction by PAPS coating was

its great adherence to steel's surfaces, due to the interaction at interfaces between the functional silanol group in polymer and the iron oxide at the top surface of steel. The interaction products formed at the interfaces probably acted as a barrier layer to limit the transport of water to the underlying steel. In contrast, when 1 wt% Ce acetate was incorporated into the coating, its curve showed a further shift in the cathodic current density to negative sites and a decrease in [E.sub.corr]. The former characteristic strongly verified that [Ce.sup.3+] oxide formed by the in-situ conversion of Ce acetate in the coating also served to inhibit the cathodic oxygen reduction reaction. The precipitation of this [Ce.sup.3+] oxide on the steel's surfaces might act as a passive oxide film. On the other hand, the chloride ions dissociated from NaCl progressively react with the passive iron oxide film, causing its dissolution in the first step of the corrosion of steel. (23) In contrast, the passive titanium oxide film had considerable resistance to the breakdown and dissolution caused by Cl ions. Thus, like the titanium oxide, the passive [Ce.sup.3+] oxide film precipitated over the steel appears to be insensitive to reactions with Cl ions. If this interpretation is valid, it is reasonable to presume that the passive [Ce.sup.3+] oxide film further inhibits the cathodic reaction of steel in an aqueous chloride environment. Regarding the shift of [E.sub.corr] to a negative site, the in-situ conversion of Ce acetate into [Ce.sup.3+] oxides may cause the formation of a porous coating layer, thereby increasing the rate of the permeability of electrolytes through the coating film. As seen in curve 4 (Figure 8), the extent of cathodic protection by [Ce.sup.3+] oxide was greatly enhanced with an increased content of Ce acetate to 3 wt%. However, a slight shift of [E.sub.corr] to a more negative site can be taken as evidence that the porosity of the film was somewhat increased because of the incorporation of increased [Ce.sup.3+] oxides into the coating. As expected, a further shift of [E.sub.corr] to a negative site was observed in the curve from the coating containing 10 wt% Ce acetate. The curve also demonstrated that regardless of the incorporation of abundant [Ce.sup.3+] oxide in the coating, the beginning of cathodic current density was at a less negative site, compared with that of the 3 wt% Ce acetate. This phenomenon was related directly to the delamination of coating from the steel's surfaces, signifying poor adherence of the coating to steel. It can be accounted for by insufficient bonds at interfaces between the functional silanol groups in the coating and the iron oxide in steel. Such a poor interfacial bonding was thought to be due to the destitution of functional silanol groups in the coating in contact with the steel because of the uptake of silanol by Ce acetate.

[FIGURE 8 OMITTED]

Although the in-situ Ce acetate[right arrow][Ce.sup.3+] oxide conversion increased the porosity of the coating, the inhibition of the cathodic corrosion reaction of steel by a coating made with a proper amount of Ce acetate was due to the combined effect of the following three factors: (1) the hydrophobic characteristic of the coating's surface, (2) good bonding of the coating to the steel, and (3) precipitation of a passive [Ce.sup.3+] oxide film insensitive to Cl dissolution over the steel's surfaces.

Based upon these polarization curves, we determined the absolute corrosion rates of steel, expressed in the conventional engineering units of milli-inches per year (mpy). Equation (1), proposed by Sterm and Geary, (24) was used in the first step:

[I.sub.corr] = [[beta].sub.a] * [[beta].sub.c]/2.303([[beta].sub.a] + [[beta].sub.c])[R.sub.p] (1)

where [I.sub.corr] is the corrosion current density in A/[cm.sup.2], [[beta].sub.a] and [[beta].sub.c], having the units of V/decade of current, refer to the anodic and cathodic Tafel slopes, respectively, (which were obtained from the log I versus E plots encompassing both anodic and cathodic regions), and [R.sub.p] is the polarization resistance which was determined from the corrosion potential, [E.sub.corr]. When [I.sub.corr] was computed through equation (1), the corrosion rate (mpy) can be obtained from the following expression:

Corrosion rate = 0.13 [I.sub.corr](EW)/d (2)

where EW is the equivalent weight of the corroding species in g, and d is the density of the corroding species in g/[cm.sup.3].

Table 3 gives the [I.sub.corr] and corrosion rate from this Tafel calculation for steel panels coated with the unmodified and Ce acetate-modified polymers. The corrosion rate of the steel for the PAPS coating (0 wt% Ce acetate) was 7.407 X [10.sup.-1] mpy accompanying [I.sub.corr] of 1.621 X [10.sup.-6] A/[cm.sup.2]. These values were nearly one order of magnitude lower than those of the uncoated bare steel, suggesting that the PAPS coating adequately protected the steel against NaCl-associated corrosion. As expected, the coatings made with the Ce acetate, ranging from 1 to 3 wt%, further reduced the corrosion rate and [I.sub.corr] value; in particular, 3 wt% was the most effective content in improving its ability to protect the steel. However, when 6 wt% Ce acetate was used to modify the coating, it did not offer as effective corrosion prevention as did 3 wt%. In fact, the corrosion rate rose ~73% to 2.752 X [10.sup.-1] mpy; in addition, the [I.sub.corr] was two times more likely to be increased than that of 3 wt%. Adding 10 wt% Ce acetate further increased these values. Comparing these corrosion rates, the effectiveness of the content of Ce acetate in offering better corrosion protection was ranked in the following order: 3 wt% > 2 wt% > 1 wt% > 6 wt% > 10 wt% > 0 wt%. This finding strongly supported all the previous results and discussions; namely, the PAAMPS coating system made with 2 and 3 wt% Ce acetate had the following five important factors, which play the major roles in mitigating the corrosion: (1) minimum contents of water-soluble nonreacted APST monomers and Ce acetate dopant, (2) moderate densification of siloxane linkages, (3) good adherence to steel's surfaces, (4) hydrophobic characteristic of its surfaces, and (5) precipitation of a passive [Ce.sup.3+] oxide film over the steel. In contrast, less effective coating systems made with an excessive amount of Ce acetate or none at all lacked one or two of these five pivotal factors.

Table 3 also includes the results of Tafel analyses for the uncoated panels, and the PAPS- and 0.5 wt% Ce acetate/PAAMPS-coated aluminum panels. The data clearly revealed a great performance of Ce acetate in reducing considerably the corrosion rate and [I.sub.corr]. Although the content of Ce acetate was as low as 0.5 wt%, the corrosion rate was reduced by two and by one order of magnitude, respectively, compared with those of the uncoated bare aluminum and aluminum coated with PAPS without Ce acetate. Concurrently, the value of [I.sub.corr] decreased more than one order of magnitude over the bare aluminum, and also was more than twice as low as the PAPS-coated aluminum. It should be noted that no cathodic-anodic Tafel region was found for aluminum panels coated with the Ce acetate of [greater than or equal to] 1.0 wt%, emphasizing that these coatings do not allow the NaCl electrolyte to permeate through them, so they display an outstanding protection over the aluminum's surfaces. This information strongly demonstrated that the PAAMPS coating affords far better protection against corrosion for aluminum than for steel.

To support the electrochemical polarization data, we carried out salt-spray resistance tests, exposing uncoated and coated aluminum and steel panels in a 5% salt-fog chamber at 35[degrees]C. In trying to assess the maximum corrosion-preventing performance, coatings of ~8.5 [micro]m thickness were deposited on the metals' surfaces. To prepare such thick coatings, the dipping-withdrawing-baking process for these metal panels was repeated four times. Table 4 shows the results reported as the total exposure time needed to generate rust and oxide stains on the coated steel and aluminum surfaces. For the bare steel panels, their entire surface was corroded after exposure for as little as 10 hr. When the steel was coated with the PAPS polymer, it was protected for ~120 hr. As expected, the PAAMPS coatings made with the Ce acetate offered a great extension of the useful lifetime of the steel, the extent of which depended on the content of the Ce acetate. Thus, the lifetime of steel was increasingly extended with a rise in content from ~165 hr for 0.2 wt% to ~768 hr for 3.0 wt%. Beyond this, the efficacy of Ce acetate in extending its lifetime declined. In fact, the lifetime for the 6 wt% Ce acetate coating was ~600 hr, which was 168 hr shorter than that of 3 wt%. The lifetime was even less when the content of Ce acetate was increased to 10 wt%. Hence, the PAAMPS coating with 3 wt% Ce acetate conferred the maximum protection against the corrosion of the steel, extending its lifetime by as much as 768 hr, which was ~648 hr longer than that of the PAPS coating without Ce acetate.

For the aluminum panels, the rate of corrosion was dramatically reduced by depositing the PAAMPS coating on it, strongly emphasizing that the coating's efficacy for this metal was much greater than for steel. The corrosion of bare aluminum started a little more than 35 hr after it was exposed. An incredible lengthening of its useful lifetime to ~1248 hr was obtained from a coating with a Ce acetate content of as little as 0.2 wt%. This lifetime was ~27% longer than that of the PAPS coating. Adding more than 1.0 wt% led to a very long useful lifetime of >1800 hr, reflecting the excellent effectiveness of a passive Ce oxide layer in PAAMPS coating in inhibiting the corrosion of aluminum.

Thin PAAMPS Coatings

The above information urged one to investigate the ability of a coating film of nanoscale thickness to protect aluminum. Since the PAAMPS coating derived from 3 wt% Ce acetate-modified 20 wt% APST precursor solution showed the best performance as a corrosion-preventing barrier for steel, we adapted this formulation and diluted it with deionized water to make the thin-film forming precursor solutions. Three diluted solutions were prepared by mixing a 0.5, 1.0, and 3.0 g of 3 wt% Ce acetate-modified 20 wt% APST precursor solution with 99.5, 99.0, and 97.0 g deionized water, respectively. Using these solutions, the single-layered coating film was deposited over the aluminum's surfaces through the dipping-withdrawing-baking coating process. The thickness of deposited film was estimated by XPS. The resulting XPS spectra (not shown) of the films from 0.5 and 1.0 wt% solutions showed the presence of Al atoms attributed to the underlying aluminum substrate. Since the penetration depth of the photoelectrons was ~5 nm, the estimated thickness of these films was less than 5 nm. However, no Al atoms were detected in the film made from the 3.0 wt% solution, implying that the thickness of this film covering the aluminum's surfaces was more than 5 nm thick. Nevertheless, we conducted electrochemical polarization and salt-spray resistance tests for these thin film-covered aluminum panels.

Table 5 lists the corrosion rate and [I.sub.corr] obtained from the cathodic-anodic Tafel curves for these coated panels. The data revealed an outstanding reduction in corrosion-related factors by these thin films, including the rate of corrosion and the value of [I.sub.corr]. The coating derived from the 0.5 wt% solution decreased the corrosion rate and [I.sub.corr] by more than one order of magnitude over those of the bare aluminum. The values of these factors further declined from 9.829 X [10.sup.-3] mpy to 2.989 X [10.sup.-3] mpy for the corrosion rate and from 2.295 X [10.sup.-8] A/[cm.sup.2] to 6.979 X [10.sup.-9] A/[cm.sup.2] for [I.sub.corr], when the concentration of Ce acetate was increased to 1.0 wt%. With a 3 wt% solution, these values were nearly two orders of magnitude below those of the bare aluminum.

This information was strongly supported from the results of the salt-spray resistance test for these coated aluminum panels (Table 6). There is no doubt that these nanoscale thickness PAAMPS coatings significantly extended the useful lifetime of aluminum, corresponding to a salt-spray resistance of more than 1440 hr, compared with only 40 hr for bare aluminum.

CONCLUSIONS

In trying to improve the usefulness of aminopropylsilane triol (APST) solution as a precursor of corrosion-preventing coatings for aluminum-finned steel condenser tubes at geothermal power plants, a 20 wt% APST aqueous solution was modified with cerium (Ce) acetate dopant of up to 10 wt%. The incorporation of Ce acetate not only reduced the pH of APST solution, ranging from 10.94 for 0 wt% Ce acetate to 9.49 for 10 wt%, but also increased the surface tension of the APST solutions. The latter factor directly affected the changes in wettability and spreadability of the modified solutions on aluminum and steel surfaces; namely, the high surface tension reduced wetting performance. Although both metals were cleaned with the same manner, the surfaces of aluminum possessed better wetting behaviors than those of steel. The transformation of the Ce acetate-doped APST solution into a solid film by heating it at 150[degrees]C was carried out through the following three spontaneous reactions: (1) the condensation reaction between the silanol, SiOH, groups within the APST to assemble siloxane, Si-O-Si, linkages in the polypropylsiloxane polymer network, (2) the amidating reaction between Ce acetate and amine, N[H.sub.2], group in APST to form the acetamide moiety as the end reaction product in the polymer network, and (3) the acetoxylation reaction between the Ce acetate and the silanol group to make the acetoxyl methyl moiety. The last two reactions led to the in-situ conversion of Ce acetate into the Ce hydroxide, Ce(OH)[.sub.3], derivative as the reaction byproduct, followed by its in-situ transformation into [Ce.sup.3+] oxide, [Ce.sub.2][O.sub.3], during heating. Most of the amine groups were converted into the acetamide moiety when 3 wt% Ce acetate was added to the APST. The extent of densification of siloxane linkages depended primarily on the content of Ce acetate; an increase in its content caused a low densification of siloxane linkage because of the depletion of silanol groups in the APST by an increased rate of acetoxylation. Nevertheless, the polymer structure formed through these reactions was categorized as the poly-acetamide-acetoxyl methyl-propylsiloxane (PAAMPS) polymer.

It was found that 3 wt% of Ce acetate was the best amount for conferring the maximum performance of a PAAMPS coating in mitigating the corrosion of both aluminum and steel. Among the factors governing this maximum performance was (1) a minimum content of nonreacted water-soluble APST and Ce acetate remaining in the coating, (2) a lower susceptibility of the coating's surfaces to moisture, thereby conferring water-repellent and -shedding properties, (3) the precipitation of a passive [Ce.sup.3+] oxide film insensitive to Cl dissolution over the metal's surfaces, and (4) excellent adherence of the coating to metal's surfaces. The combination of all these factors contributed to decreasing significantly the rate of cathodic oxygen reduction reactions at the corrosion site of metals, thus extending their useful lifetime. With an ~8.5-[micro]m thick coating film, the useful lifetime of steel in a 5% salt-fog chamber at 35[degrees]C was prolonged from only ~10 hr to ~768 hr. This lifetime was more than six times as long as that of steel coated with an APST without Ce acetate, strongly demonstrating the great effectiveness of Ce acetate in improving the ability of APST to protect steel against corrosion. The PAAMPS coating displayed a far better corrosion-preventing performance for aluminum, compared with that for steel. In fact, when a coating with nanoscale thickness of [less than or equal to] 5 nm was deposited on the aluminum's surfaces, its corrosion rate, milli-inches per year (mpy), was nearly two orders of magnitude lower than that of bare aluminum. In addition, the salt-spray resistance of aluminum coated with this nanoscale film was strikingly extended from ~40 hr to > 1440 hr.

References

(1) Kutscher, C.F. and Dostenaro, D., "Assessment of Evaporative Cooling Enhancement Methods for Air-Cooled Geothermal Power Plants," Geotherm. Resour. Counc. Trans., 26, 775 (2002).

(2) Sugama, T., Gawlik, K., and Jung, D., "Polyaminopropylsiloxane Coatings for Geothermal Air-Cooled Condensers," Recent Res. Devel. Mat. Sci., 4, 695 (2003).

(3) Mansfeld, F., Breslin, C.B., Pardo, A., and Perez, F.J., "Surface Modifications of Stainless Steels: Green Technology for Corrosion Protection," Surf. Coat. Technol., 90, 224 (1997).

(4) Aballe, A., Bethencourt, M., Botana, F.J., and Marcos, M., "Ce[Cl.sub.3] and La[Cl.sub.3] Binary Solutions as Environment-Friendly Corrosion Inhibitors of AA5083 Al-Mg Alloy in NaCl Solutions," J. Alloys Compd., 323, 855 (2001).

(5) Dabala, M., Armelao, L., Buchberger, A., and Calliari, I., "Cerium-Based Conversion Layers on Aluminum Alloys," Appl. Surf. Sci., 172, 312 (2001).

(6) Forsyth, M., Forsyth, C.M., Wilson, K., Behrshing, T., and Deacon, G.B., "ATR Characterization of Synergistic Corrosion Inhibition of Mild Steel Surfaces by Cerium Salicylate," Corros. Sci., 44, 2651 (2002).

(7) Fahrenholtz, W., O'Keefe, M.J., Zhou, H., and Grant, J.T., "Characterization of Cerium-Based Conversion Coatings for Corrosion Protection of Aluminum Alloys," Surf. Coat. Technol., 155, 208 (2002).

(8) Yu, X. and Cao, C., "Electrochemical Study of the Corrosion Behavior of Ce Sealing of Anodized 2024 Aluminum Alloy," Thin Solid Films, 423, 252 (2003).

(9) Fox, H.W. and Zisman, W.A., "The Spreading of Liquids on Low Energy Surfaces. I. Polytetrafluoroethylene," J. Colloid Sci., 5, S14 (1950).

(10) Launer, R., "Infrared Absorption of Organosilicon Compounds: Spectras Structure Correlations," in Silicon Compounds Register and Review, Anderson, R., Arkles, B., and Larson, G.L. (Eds.), Petrarch Systems Corp., 69, 1987.

(11) Su, G.J., Borelli, N.F., and Miller, A.R., "An Interpretation of the Infrared Spectra of Several Glasses," Phys. Chem. Glasses, 3, 167 (1962).

(12) Bellamy, L.J., The Infra-Red Spectra of Complex Molecules, Chapman & Hall, London, p. 231, 1975.

(13) Nakanish, K. and Solomon, P.H., Infrared Absorption Spectroscopy, Holden-Day, Inc., San Francisco, p. 15, 1977.

(14) Toth, A., Bertoti, I., Szekely, T., Sazanov, J.N., Antonova, T.A., Shchukarev, A.V., and Gribanov, A.V., "XPS Study on the Thermal Degradation of Poly-N,N'/4,4'-Diphenylether/Pyromellitimide," Surf. Interface. Anal., 8, 261 (1986).

(15) Clark, D.T. and Thomas, H.R., "Applications of ESCA to Polymer Chemistry. X. Core and Valence Energy Levels of a Series of Polyacrylates," J. Polym. Sci., 14, 1671 (1976).

(16) Briggs, D., Brewis, D.M., and Konieczko, M.B., "X-ray Photoelectron Spectroscopy Studies of Polymer Surfaces," J. Mater. Sci., 14, 1344 (1979).

(17) Kasten, L.S., Grant, J.T., Grebasch, N., Voevodin, N., Amold, F.E., and Donley, M.S. "An XPS Study of Cerium Dopants in Sol-Gel Coatings for Aluminum 2024-T3," Surf. Coat. Technol., 140, 11 (2001).

(18) Barreca., D., Battiston, G.A., Gerbasi, R., and Tondello, E. "Study of Cerium Dioxide Thin Films by X-ray Photoelectron Spectroscopy," Surf. Sci. Spect., 7, 297 (2000).

(19) Gettings, M. and Kinloch, A.J. "Surface Analysis of Polysiloxane/Metal Oxide Interfaces," J. Mater. Sci., 12, 2511 (1977).

(20) Bailey, R. and Castle, J.E., "XPS Study of the Adsorption of Ethoxysilanes on Iron," J. Mater. Sci., 12, 2049 (1977).

(21) Leung, Y.L., Yang, Y.P., Wong, P.C., Mitchell, K.A.R., and Foster, T., "X-ray Photoelectron Spectroscopic Studies with a Bias-Potential Method for Studying Silane-Aluminum Interfaces," J. Mater. Sci. Lett., 12, 844 (1993).

(22) Fang, J., Flinn, B.J., Leung, Y.L., Wong, P.C., Mitchell, K.A.R., and Foster, T., "A Characterization of the [gamma]-glycidoxypropy-ltrimethoxysilane and Aluminum Interface by SIMS and XPS," J. Mater. Sci. Lett., 16, 1675 (1997).

(23) Townsend, H.E., Cleary, H.J., and Allegra, L.A., "Breakdown of Oxide Films on Steel Exposed to Chloride Solutions," Corrosion, 37, 384 (1981).

(24) Sterm, M. and Geary, A.L., J. Electrochem. Soc., 104, 56 (1957).

T. Sugama -- Brookhaven National Laboratory*

* Energy Resources Division, Energy Science & Technology Department, Upton, NY 11973.

This program report, issued by the DOE Office of Geothermal Technologies, was performed under the auspices of the U.S. Department of Energy, Washington, D.C., under Contract No. DE-AC02-98CH10866.

Table 1 -- The pH Values and Surface Tensions of the Unmodified and the
0.2, 0.5, 1.0, 2.0, 3.0, 6.0, and 10.0 wt% Ce Acetate-Modified 20 wt%
APST Solutions

Ce Acetate, wt%  pH               Surface Tension, dynes/cm

 0               10.94            54.6
 0.2             10.90            56.1
 0.5             10.83            56.9
 1.0             10.66            57.4
 2.0             10.46            57.8
 3.0             10.33            58.0
 6.0             10.00            58.6
10.0              9.49            58.7

Table 2 -- XPS Surface Chemical Composition for PAPS and 3 wt% Ce
Acetate-Derived PAAMPS Coatings' Surfaces

           Atomic Composition, %
Coating  C     N    O     Si   Ce

PAPS     63.1  4.5  26.5  5.9  0.0
PAAMPS   66.8  1.8  28.7  2.2  0.5

Table 3 -- Tafel Analyses for Potentiodynamic Polarization Curves of
Noncoated and 5 wt% PAPS Coatings Modified with Various Concentrations
of Ce Acetate for Steel and Aluminum Panels

                                                        [[beta].sub.a],
Substrate    Ce Acetate, wt%  [E.sub.corr](l = 0), (V)  (V/decade)

Uncoated     --               -0.6749                   0.1000
  steel
Steel         0               -0.4520                   0.0402
Steel         1               -0.5951                   0.0799
Steel         2               -0.5711                   0.0555
Steel         3               -0.5538                   0.0664
Steel         6               -0.5544                   0.0387
Steel        10               -0.5472                   0.0679
Uncoated Al  --               -0.5690                   0.0903
Al            0               -0.9321                   0.6639
Al            0.5             -0.8916                   0.4456

             [[beta].sub.c],  [I.sub.corr],        Corrosion Rate,
Substrate    (V/decade)       (A/[cm.sup.2])       (mpy) (a)

Uncoated     0.1000           1.189 X [10.sup.-5]  5.435
  steel
Steel        0.4746           1.621 X [10.sup.-6]  7.407 X [10.sup.-1]
Steel        0.7682           5.311 X [10.sup.-7]  2.428 X [10.sup.-1]
Steel        0.4597           4.403 X [10.sup.-7]  2.012 X [10.sup.-1]
Steel        0.5620           3.475 X [10.sup.-7]  1.589 X [10.sup.-1]
Steel        0.5350           7.020 X [10.sup.-7]  2.752 X [10.sup.-1]
Steel        0.4482           7.672 X [10.sup.-7]  3.507 X [10.sup.-1]
Uncoated Al  0.3792           2.497 X [10.sup.-7]  1.070 X [10.sup.-1]
Al           0.2091           3.104 X [10.sup.-8]  1.330 X [10.sup.-2]
Al           0.1440           1.260 X [10.sup.-8]  5.397 X [10.sup.-3]

(a) mpy: milli-inches per year.

Table 4 -- Salt-Spray Resistance Tests for Uncoated and Coated Steel and
Aluminum Panels

Substrate      Ce Acetate (wt%)  Salt-Spray Resistance (hr)

Bare steel     --                  ~10
Steel           0                 ~120
Steel           0.2               ~165
Steel           0.5               ~264
Steel           1.0               ~312
Steel           2.0               ~720
Steel           3.0               ~768
Steel           6.0               ~600
Steel          10.0               ~408
Bare aluminum  --                  ~40
Aluminum        0                 ~980
Aluminum        0.2              ~1248
Aluminum        0.5              ~1400
Aluminum        1.0              >1800
Aluminum        2.0              >1800
Aluminum        3.0              >1800

Table 5 -- Tafel Analyses for Potentiodynamic Polarization Curves of
Thin Films Derived from 0.5, 1.0, and 3.0 wt% (3 wt% Ce Acetate-Modified
20 wt% APST) Solutions for Aluminum Panels

            Ce Acetate-
            Modified     [E.sub.corr]  [[beta].sub.a],  [[beta].sub.c],
Substrate   APST, wt%    (l = 0), (V)  (V/decade)       (V/decade)

Uncoated    --           -0.5690       0.0903           0.3792
  aluminum
Aluminum    0.5          -0.5429       0.0275           0.1758
Aluminum    1            -0.5353       0.0358           0.2140
Aluminum    3            -0.4926       0.0619           0.2055

Substrate   [I.sub.corr], (A/[cm.sup.2])  Corrosion Rate, (mpy) (a)

Uncoated    2.497 X [10.sup.-7]           1.070 X [10.sup.-1]
  aluminum
Aluminum    2.295 X [10.sup.-8]           9.829 X [10.sup.-3]
Aluminum    6.979 X [10.sup.-9]           2.989 X [10.sup.-3]
Aluminum    2.810 X [10.sup.-9]           1.203 X [10.sup.-3]

(a) mpy: milli-inches per year.

Table 6 -- Salt-Spray Resistance Tests for the Aluminum Panels Covered
with the Nanoscale Thickness Coating Films Derived from 0.5, 1.0, and
3.0 wt% (3 wt% Ce Acetate-Modified 20 wt% APST) Solutions

Substrate   Ce Acetate-Modified APST (wt%)  Salt-Spray Resistance (hr)

Uncoated    --                                ~40
  aluminum
Aluminum    0.5                             ~1260
Aluminum    1.0                             >1440
Aluminum    3.0                             >1440