On the Characteristics of Ion Implanted Metallic Surfaces Inducing

Mar 29, 2010 - Rose , J. W. In Heat Exchanger Design Handbook; Hewitt , G. F., Ed.; Begell House: New York, 1998; Chapter 2.6.5. There is no correspon...
0 downloads 0 Views 4MB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

On the Characteristics of Ion Implanted Metallic Surfaces Inducing Dropwise Condensation of Steam Michael H. Rausch, Alfred Leipertz, and Andreas P. Fr€oba* Lehrstuhl f€ ur Technische Thermodynamik (LTT), Universit€ at Erlangen-N€ urnberg, Am Weichselgarten 8, D-91058 Erlangen, Germany Received November 12, 2009. Revised Manuscript Received February 5, 2010 The present work provides new information on the characteristics of ion implanted metallic surfaces responsible for the adjustment of stable dropwise condensation (DWC) of steam. The results are based on condensation experiments and surface analyses via contact angle (CA) and surface free energy (SFE) measurements as well as scanning electron microscopy (SEM). For studying possible influences of the base material and the implanted ion species, commercially pure titanium grade 1, aluminum alloy Al 6951, and stainless steel AISI 321 were treated with Nþ, Cþ, Oþ, or Arþ using ion beam implantation technology. The studies suggest that chemically inhomogeneous surfaces are instrumental in inducing DWC. As this inhomogeneity is apparently caused by particulate precipitates bonded to the metal surface, the resulting nanoscale surface roughness may also influence the condensation form. On such surfaces nucleation mechanisms seem to be capable of maintaining DWC even when CA and SFE measurements indicate increased wettability. The precipitates are probably formed due to the supersaturation of ion implanted metal surfaces with doping elements. For high-alloyed materials like AISI 321 or Hastelloy C-276, oxidation stimulated by the condensation process obviously tends to produce similar surfaces suitable for DWC.

Introduction The modification of metallic surfaces by ion implantation (II) was proved to be able to induce stable DWC of pure steam in the late 1980s.1 It was the first time that this special condensation form implying very large heat transfer coefficients2 could be obtained without applying poorly wettable coatings. However, to date it is still not clear what characteristics of the ion implanted metal surfaces are responsible for the appearance of DWC. Approaches available in the literature are more or less speculative and based on a reduction of the SFE of the treated metal, implicating reduced wettability.3 In contrast, our recent experimental results4 showed that DWC can be achieved by implantation of Nþ ions in titanium surfaces in spite of reduced CAs for water and increased polar, dispersive, and cumulated SFEs measured after the surface modification. AFM analyses and condensation experiments performed in the same work indicated a connection between the formation of particulate precipitates on the treated surface and the observed change of the condensation form. Considering results from materials research for nitrogen implanted titanium surfaces,5-7 these precipitates are likely to be titanium nitrides. They are formed because II with sufficiently high ion doses introduces more doping elements into the titanium *Corresponding author: Tel þ49-9131-85-29789, Fax þ49-9131-85-29901, e-mail [email protected]. (1) Zhao, Q.; Zhang, D.; Lin, J. Proceedings of the 1st International Conference on Heat Transfer in Energy Conservation; Shenyang, 1988; pp 177-179. (2) Rose, J. W. In Heat Exchanger Design Handbook; Hewitt, G. F., Ed.; Begell House: New York, 1998; Chapter 2.6.5. (3) Zhao, Q.; Burnside, B. M. Heat Recovery Syst. CHP 1994, 14, 525–534. (4) Rausch, M. H.; Leipertz, A.; Fr€oba, A. P. Int. J. Heat Mass Transfer 2010, 53, 423–430. (5) Fouquet, V.; Pichon, L.; Drouet, M.; Straboni, A. Appl. Surf. Sci. 2004, 221, 248–258. (6) Kasukabe, Y.; Nishida, S.; Yamamoto, S.; Yoshikawa, M.; Fujino, Y. Appl. Surf. Sci. 2008, 254, 7942–7946. (7) Mu~noz-Castro, A. E.; Lopez-Callejas, R.; Granda-Gutierre, E. E.; ValenciaAlvarado, R.; Barocio, S. R.; Pe~na-Eguiluz, R.; Mercado-Cabrera, A.; de la Piedad Beneitez, A. Prog. Org. Coat. 2009, 64, 259–263.

Langmuir 2010, 26(8), 5971–5975

surface layer reachable by the ions than the material can dissolve in its structure.8 The resulting supersaturation of the titanium surface with the implanted element is degraded by the formation of pertinent chemical compounds. The precipitated titanium nitrides seem to induce a nucleation effect during condensation of steam. This can be deduced from our previous experiments4 with a titanium sample showing a mixed form of filmwise condensation (FWC) and DWC after applying a low Nþ ion dose. The DWC areas enlarged with increasing cooling power, suggesting increased activity of the nuclei. However, it remains unclear if titanium nitrides play a special role in inducing DWC or if there is a more general effect of the surface characteristics obtained by II. Following up this question, results from condensation experiments and surface analyses performed for several materials which were partially ion implanted with various ion species are discussed in this work.

Experimental Section Disks with a diameter of 60 mm and a thickness of 12 mm were prepared from commercially pure titanium grade 1, aluminum alloy Al 6951, stainless steel AISI 321 (X6CrNiTi18-10), and the nickel-based alloy Hastelloy C-276 (NiMo16Cr15W). All samples for studying the effects of II were polished in order to avoid interference by varying surface roughness. A mean surface roughness Ra of about 0.15 μm was obtained for these samples. The titanium surfaces, which tend to show strong oxidation effects during the condensation of steam, were additionally stabilized by a preoxidation procedure before II.4 After cleaning with ethanol and deionized water, the titanium, Al 6951, and AISI 321 samples were implanted using ion-selective ion beam technology.9 Separate samples were treated with Cþ, Oþ, Nþ, or Arþ with an ion dose of 1016 cm-2 at 20 keV. Each sample was partially masked (8) Zwicker, U. Titan und Titanlegierungen; Springer-Verlag: Berlin, 1974; pp 154-159. (9) Nastasi, M.; Mayer, J. W.; Hirvonen, J. K. Ion-Solid Interactions: Fundamentals and Applications; Cambridge University Press: Cambridge, 1996; Chapter 15.

Published on Web 03/29/2010

DOI: 10.1021/la904293f

5971

Article

Rausch et al. Table 1. Measured CAs and SFEs on Implanted and Unimplanted Sample Areas

material

ion

Θw (deg)

Θd (deg)

γp (mN m-1)

γd (mN m-1)

γ (mN m-1)

titanium

Arþ Cþ Oþ Arþ Cþ Oþ Nþ

69.1 ( 3.9 77.5 ( 2.2 61.0 ( 3.4 72.3 ( 3.0 62.4 ( 3.9 69.9 ( 0.7 94.6 ( 0.6 78.0 ( 4.6 96.8 ( 3.8 79.4 ( 2.3 96.4 ( 0.8 73.1 ( 4.8 97.2 ( 2.3 73.1 ( 2.0

38.9 ( 2.3 45.7 ( 3.0 40.2 ( 8.1 37.2 ( 2.1 37.2 ( 3.9 36.9 ( 2.0 56.4 ( 3.7 44.1 ( 1.6 58.8 ( 4.1 38.1 ( 3.4 57.9 ( 2.3 38.3 ( 1.8 57.6 ( 4.0 35.1 ( 2.1

15.9 ( 2.0 12.2 ( 0.8 20.9 ( 1.0 13.9 ( 1.4 19.3 ( 1.8 15.1 ( 0.2 5.2 ( 0.8 11.8 ( 2.3 4.4 ( 1.2 10.3 ( 0.9 4.7 ( 0.3 13.5 ( 2.6 0.7 ( 0.2 7.5 ( 1.6

30.2 ( 0.6 28.3 ( 1.0 28.5 ( 3.1 31.2 ( 0.6 29.9 ( 1.2 31.0 ( 0.7 27.3 ( 2.5 29.0 ( 0.5 26.8 ( 1.0 32.3 ( 1.0 27.0 ( 1.2 31.2 ( 1.5 30.8 ( 2.2 34.7 ( 0.2

45.7 ( 2.1 40.5 ( 1.8 48.8 ( 3.5 45.1 ( 1.8 49.3 ( 2.8 46.1 ( 0.9 32.4 ( 1.7 40.8 ( 1.9 31.2 ( 2.1 42.5 ( 1.9 31.5 ( 1.1 45.0 ( 1.6 31.5 ( 2.3 42.3 ( 0.8

titanium titanium Al 6951 Al 6951 Al 6951 Al 6951

during II. Thus, the effects on implanted surface parts could be compared with unimplanted parts of the same sample. For the characterization of the wettability of the studied surfaces, measurements of the CA and the SFE were performed with a surface analyzer (SURFTENS universal, OEG GmbH, Germany). The sessile drop method was applied for measuring the CAs for deionized water (Θw) and diiodomethane (Θd) at atmospheric conditions, using a drop volume of about 4 μL. From the resulting CAs, the cumulated SFE γ, as well as its polar (γp) and disperse (γd) components, were calculated according to the methods of Owens and Wendt10 or Wu.11 The latter method is preferred for samples with a cumulated SFE below 35 mN m-1. For surface topography analysis, a scanning electron microscope (JSM-6400, JEOL, Japan) was used. The condensation form of pure steam at a temperature of about 100 °C was tested in an air cooled condensation chamber equipped with optical accesses.12

Results and Discussion CAs and SFEs for the implanted and unimplanted areas of the titanium and Al 6951 samples were measured prior to the condensation experiments in order to evaluate their influence on the condensation form. The mean values of the obtained CAs and SFEs are listed in Table 1. The given uncertainties indicate the range of all measured values. For the aluminum alloy Al 6951, all implanted ion species induced distinctly decreased CAs and increased SFEs, indicating increased wettability for water and diiodomethane. In contrast, increased CAs and decreased SFEs predominated for titanium. Only for the titanium samples implanted with Cþ and Oþ, slightly reduced Θd and increased γd values were measured. Comparing the wettability parameters for the titanium samples presented in Table 1, the largest CAs and lowest SFEs were found for the surface implanted with Arþ. Nevertheless, a perfectly closed condensate film was observed here. This can be seen on the left half of the sample shown in Figure 1a. The unimplanted area was covered with a condensation pattern consisting of static film zones, streaks, and droplets, which was probably caused by surface inhomogeneities and can be interpreted as nonideal FWC. On the titanium surfaces implanted with Cþ and Oþ, stable DWC was obtained (see Figure 1b,c). These experiments confirm our previous results4 showing that the wettability of ion implanted metallic surfaces indicated by CAs and SFEs is not decisive for the resulting condensation form of steam. Instead, the presumed connection between the formation of precipitates and the appearance of DWC reported in (10) Owens, D. K.; Wendt, R. C. J. Appl. Polym. Sci. 1969, 13, 1741–1747. (11) Wu, S. J. Adhesion 1973, 5, 39–55. (12) Rausch, M. H.; Fr€oba, A. P.; Leipertz, A. Int. J. Heat Mass Transfer 2008, 51, 1061–1070.

5972 DOI: 10.1021/la904293f

Figure 1. Condensation forms after about 24 h of condensation for polished and preoxidized titanium samples implanted with (a) Arþ, (b) Cþ, and (c) Oþ on the left half.

the same work seems to be corroborated because the noble gas argon cannot form stable compounds with other elements except under extreme conditions.13 Only sputtering effects, interstitial inclusions, or accumulations of argon in pores and at lattice defects can be caused by II of Arþ in the present case. Judging from the closed water film on the implanted sample area, the removal of disordered surface layers by sputtering effects is presumable. Because of the low solubility of carbon in titanium,8 the implantation of Cþ at the ion dose applied in this work is likely to induce the formation of particulate TiC precipitates. This has been confirmed by several researchers investigating carbon ion implanted titanium or titanium alloy surfaces by various surface analysis methods, e.g., XPS, SIMS, or GIXRD.14-16 Although the solubility of oxygen in titanium is distinctly larger,8 implantation of additional Oþ can also contribute to oxide precipitation (13) Khriachtchev, L.; Pettersson, M.; Runeberg, N.; Lundell, J.; R€as€anen, M. Nature 2000, 406, 874–876. (14) Krupa, D.; Jezierska, E.; Baszkiewicz, J.; Wierzchon, T.; Barcz, A.; Gawlik, G.; Jagielski, J.; Sobczak, J. W.; Bilinski, A.; Larisch, B. Surf. Coat. Technol. 1999, 114, 250–259. (15) Soltani-Farshi, M.; Baumann, H.; R€uck, D.; Richter, E.; Kreissig, U.; Bethge, K. Surf. Coat. Technol. 1998, 103-104, 299–303. (16) Wenzel, A.; Hammerl, C.; K€oniger, A.; Rauschenbach, B. Nucl. Instrum. Methods Phys. Res., Sect. B 1997, 129, 369–376.

Langmuir 2010, 26(8), 5971–5975

Rausch et al.

Article

Figure 2. Condensation forms after about 24 h of condensation for polished Al 6951 samples implanted with (a) Arþ, (b) Cþ, (c) Oþ, and (d) Nþ on the left half.

Figure 4. Changing condensation pattern for polished AISI 321 implanted with Nþ on the left sample part.

Figure 3. Surface roughness of polished Al 6951 after DWC on left implanted (Cþ) and FWC on right unimplanted sample part after 5 days of condensation.

because of the saturation of the titanium surface with oxygen by the applied preoxidation procedure. After implantation of oxygen ions, different types of oxides were reported in previous works.17,18 All results obtained for titanium surfaces treated with II support the assumed role of precipitates in inducing DWC of steam.4 Additionally, they show that not only titanium nitrides but also other particulate titanium compounds can contribute to the adjustment of this condensation form. In contrast, the measured wetting parameters fail to predict clear tendencies. For the Al 6951 samples, a stationary condensation pattern of streaks and droplets was found on the unimplanted areas and on the surface implanted with Arþ (see Figure 2). This is a typical (17) Hammerl, C.; Bohne, Y.; Assmann, W.; Helming, K.; Rauschenbach, B. Nucl. Instrum. Methods Phys. Res., Sect. B 2003, 206, 1072–1076. (18) Li, J.; Sun, M.; Ma, X. Appl. Surf. Sci. 2006, 252, 7503–7508.

Langmuir 2010, 26(8), 5971–5975

observation for aluminum surfaces12 and can be attributed to the formation of a rough and porous oxide layer on the material wetted by the hot condensate. Also for this material, the implantation of Arþ has no influence on the condensation form whereas the surface modification with Cþ, Oþ, or Nþ induces stable DWC. The condensation forms are in contrast to the wettability indicated by the measured CAs and SFEs (Table 1), confirming that previous research3 assigning the DWC effect to reduced SFEs after II has to be discarded as already followed from our former results4 for titanium surfaces implanted with Nþ. Instead, the formation of particulate precipitates is likely to be responsible for the appearance of DWC for Al 6951 as well. Materials research groups found that nanocrystalline Al4C3,19,20 γ-Al2O3,21 or AlN22-24 are precipitated on aluminum and aluminum alloy surfaces treated with carbon, oxygen, or nitrogen ions, respectively. The resulting surface characteristics with precipitates (19) Baba, K.; Hatada, R.; Flege, S.; Kraft, G.; Ensinger, W. Surf. Coat. Technol. 2009, 203, 2617–2619. (20) Foerster, C. E.; da Silva, S. L. R.; Fitz, T.; Dekorsy, T.; Prokert, F.; Kreissig, U.; Richter, E.; M€oller, W.; Lepienski, C. M.; de M. Siqueira, C. J. Surf. Coat. Technol. 2005, 200, 5210–5219. (21) Bourcier, R. J.; Myers, S. M.; Polonis, D. H. Nucl. Instrum. Methods Phys. Res., Sect. B 1990, 44, 278–288. (22) Manova, D.; Huber, P.; M€andl, S.; Rauschenbach, B. Surf. Coat. Technol. 2000, 128-129, 249–255. (23) M€oller, W.; Parascandola, S.; Telbizova, T.; G€unzel, R.; Richter, E. Surf. Coat. Technol. 2001, 136, 73–79. € (24) Osterle, W.; D€orfel, I.; Urban, I.; Reier, T.; Schultze, J. W. Surf. Coat. Technol. 1998, 102, 168–174.

DOI: 10.1021/la904293f

5973

Article

Figure 5. Condensation patterns for unpolished (a) AISI 321 and (b) Hastelloy C-276 after about 20 h of condensation.

Figure 6. SEM analysis of the AISI 321 sample shown in Figure 5a after condensation.

existing next to the base material induce a change not only of the condensation form of steam but also of the oxidation behavior during the condensation process. Polished Al 6951 areas showing FWC rapidly oxidize, resulting in a rough and matte surface. On the contrary, polished and implanted areas inducing DWC retain their mirrorlike appearance as long as this condensation form remains stable. When DWC gradually turns into FWC, which is observed on implanted Al 6951 surfaces after several months of permanent condensation, the surface is simultaneously rendered to the rough and matte state. The altered oxidation behavior evoked by II is illustrated in Figure 3 and can be attributed to the modified surface chemistry because of the precipitated compounds. Erosion of the precipitates might be the reason for the changing condensation form in long-term experiments. 5974 DOI: 10.1021/la904293f

Rausch et al.

It should be noted that the increased oxidation resistance is obtained in spite of a relatively low ion dose compared to, e.g., hardening applications for metallic tools.25 It can be followed from this that no completely closed or thick layers of the produced compounds are necessary for obtaining DWC. This conclusion is supported by available literature indicating that closed precipitate layers are not likely to be formed for ion doses below 1017 cm-2.14,24 The findings for Al 6951 demonstrate that the adjustment of DWC, probably due to the formation of particulate precipitates as shown for titanium in our previous work,4 is also possible for other materials by implantation with various ion species. AISI 321 samples implanted with Arþ, Cþ, Oþ, and Nþ showed analogous condensation behavior to titanium and Al 6951, confirming this conclusion. Thus, the choice of doping elements capable of forming chemical compounds with the substrate material seems to be essential for obtaining the above surface characteristics suitable for DWC. Another interesting phenomenon, exemplarily illustrated in Figure 4 by a sample partially implanted with Nþ, was observed for the stainless steel. DWC was gradually spreading from the bottom of the unimplanted part of the sample. Here the sample was found to be rather clean after the first condensation experiment, but on the rest of the unimplanted area loose precipitates obviously formed by oxidation processes could be found and were partially wiped off. These loose particles may have been washed off by the condensate in the DWC area of the unimplanted sample part. The removal of the loose oxidation products seems to start at the bottom of the sample because the condensate flow is strongest here. Another possible reason for this behavior might be the condensate dripping off the sample at this place, resulting in increased interactions between condensate and sample surface. After restarting the condensation experiment, DWC could be observed on the clean and implanted sample parts, spreading all over the plate within about 3 days of permanent condensation. Because of thorough cleaning procedures of the apparatus before each condensation experiment, contamination can be discarded as possible reason for the spreading of DWC on the unimplanted material. In order to exclude influences of polishing effects, an unpolished AISI 321 sample was tested under condensation conditions. Starting from complete FWC, the condensation pattern shown in Figure 5a developed within about 20 h of condensation. In contrast to titanium and Al 6951, similar behavior was found for Hastelloy C-276 (see Figure 5b). After the condensation experiment, SEM analysis was performed for the areas where DWC and FWC occurred on the AISI 321 sample (see Figure 6). It shows that the DWC zone became very rough in a microscopic scale. Considering the results illustrated in Figure 4, this change is probably caused by oxidation effects. The strong impact on the topography also suggests a substantial modification of the surface chemistry due to the proceeding oxidation reactions. As spontaneous spreading of DWC was only observed for the high-alloyed materials AISI 321 and Hastelloy C-276, the presence of sufficiently large fractions of at least two material components seems to be a prerequisite for this phenomenon. The observation of FWC on the rather pure titanium and low-alloyed Al 6951 surfaces without II supports this assumption. Judging from all the discussed results for various metals, II seems to artificially induce surface modification effects which naturally appear on some materials during the condensation of steam. In both cases, chemical inhomogeneity caused by the (25) M€andl, S. Plasma Immersion Ion Implantation for Surface Modification of Materials; Mensch-und-Buch-Verlag: Berlin, 2002; Chapter 5.

Langmuir 2010, 26(8), 5971–5975

Rausch et al.

Article

coexistence of precipitates and original substrate material at the surface is likely to be instrumental in inducing DWC. Because of the particulate character of the precipitates and the microscopically roughened surface after oxidation, topographical effects must also be taken into consideration for the DWC mechanism taking place on such surfaces.

Conclusions This work shows that particulate precipitates formed due to the supersaturation of the metallic substrate with doping elements are likely to be responsible for the adjustment of DWC of steam by II. The choice of ion species which are able to form chemical compounds with the substrate material seems to be essential for obtaining surface characteristics appropriate for DWC. This was demonstrated by Arþ implantation in various metallic surfaces which, in contrast to II with Nþ, Cþ, or Oþ, failed to induce DWC. For Al 6951 also a drastic change of the oxidation could be observed after implanting ions suitable for achieving DWC, confirming a significant change of the surface chemistry. Spontaneously spreading DWC areas accompanied by strong oxidation effects with nanoscale surface roughening were observed on unimplanted AISI 321. The change of the surface properties is likely to be caused by the formation of particulate oxides of the alloy components, naturally producing a surface

Langmuir 2010, 26(8), 5971–5975

modification similar to that artificially obtained by II. In contrast to titanium and Al 6951, Hastelloy C-276 showed a likewise behavior, indicating that this phenomenon is only possible for high-alloyed materials. Our results suggest that DWC of steam on appropriately modified metallic surfaces originates from nucleation effects. For this, both chemical inhomogeneity and microscopic roughness of the surface resulting from the precipitation or oxidation processes may be decisive. Based on this assumption, a model for the microscopic condensation mechanism maintaining DWC in spite of reduced CAs and increased SFEs is being developed. In further investigations the chemical composition of relevant surfaces will be studied in more detail by XPS, SIMS, and AES in order to substantiate the available results. For these experiments, the choice of surface-sensitive methods is important because II only penetrates the metal surface by about 10-50 nm with the given implantation parameters. Acknowledgment. We gratefully acknowledge the financial support of this work by the German National Science Foundation (DFG, Deutsche Forschungsgemeinschaft). We are grateful to Karsten Durst and Richard Kosmala from the Institute for General Materials Properties (WW1) of the University of ErlangenNuremberg for the support in sample preparation and surface analysis by SEM.

DOI: 10.1021/la904293f

5975