Langmuir 2007, 23, 9282-9286
Nonsolvent-Induced Dewetting of Thin Polymer Films Lin Xu, Tongfei Shi,* and Lijia An* State Key Laboratory of Polymer Physics and Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, P. R. China ReceiVed March 19, 2007. In Final Form: June 3, 2007 The process of nonsolvent-induced dewetting of thin polystyrene (PS) films on hydrophilic surfaces at room temperature has been studied by using water as a nonsolvent. It is observed that the process of nonsolvent-induced dewetting is greatly different from other previous dewetting processes. The PS film is found in nonviscous state in our study. A mechanism of nonsolvent-induced dewetting is deduced in an order of penetration, replacement, and coalescent, and it is different from other previous dewetting mechanisms. The results of experiments are analyzed from thermodynamics and dynamics to support the hypothetical mechanism.
1. Introduction Over the past several decades, thin organic or polymeric films have drawn considerable attention for numerous technological applications in the fields of coating, adhesives flotation, multilayer adsorption, wetting, tear film breakup, etc.1,2 The dewetting of thin liquid films from a flat substrate is a common phenomenon in terms of its crucial impact on various technological processes, and it is also important in view of its fundamental scientific interests. In this direction, a lot of research papers on the dewetting processes through experiments, simulation, and theory have been reported.3-16 Most of the studies on dewetting are through thermal dewetting processes, in which instabilities are induced by spinodal dewetting or nucleation and growth dewetting, as well as other side mechanisms,17-23 whereas solvent-induced dewetting has received very little attention.24-29 Generally the main difference * To whom correspondence should be addressed. E-mail: [email protected]
; [email protected]
Tel: +86-431-85262137; +86-431-85262206. Fax: +86431-85262969. (1) de Gennes, P. G. ReV. Mod. Phys. 1985, 57, 827. (2) Jones, R. A. L.; et al. Europhys. Lett. 1990, 12, 41. (3) Redon, C.; Brochard-Wyart, F.; Rondelez, F. Phys. ReV. Lett. 1991, 66, 715. (4) Gabriele, S.; Sclavons, S.; Reiter, G.; Damman, P. Phys. ReV. Lett. 2006, 96, 156105. (5) Sharma, A.; Reiter, G. J. Colloid Interface Sci. 1996, 178, 383. (6) Becker, J.; Gru¨n, G.; Seemann, R.; Mantz, H.; Jacobs, K.; Mecke, K.; Blossey, A. Nat. Mater. 2003, 2, 59. (7) Kargupta, K.; Sharma, A. J. Colloid Interface Sci. 2002, 245, 99. (8) Krishnan, R. S.; Mackay, M. E.; Hawker, C. J.; Van Horn, B. Langmuir 2005, 21, 5770. (9) Masson, J.-L.; Green, P. F. Phys. ReV. E 2002, 65, 031806. (10) Masson, J.-L.; Green, P. F. Phys. ReV. Lett. 2002, 88, 205504. (11) Limary, R.; Green, P. F. Phys. ReV. E 2002, 66, 021601. (12) Mu¨eller-Buschbaum, P. J. Phys.:Condens. Matter 2003, 15, R1549. (13) Karapanagiotis, I.; Gerberich, W. W. Surf. Sci. 2005, 594, 192. (14) Xie, R.; Karim, A.; Douglas, J. F.; Han, C. C.; Weiss, R. A. Phys. ReV. Lett. 1998, 81, 1251. (15) Yang, M. H.; Hou, S. Y.; Chang, Y. L.; Yang, A. C-M. Phys. ReV. Lett. 2006, 96, 066105. (16) (a) Reiter, G. Phys. ReV. Lett. 2001, 87, 186101. (b) Shenoy, V.; Sharma, A. Phys. ReV. Lett. 2002, 88, 236101-1. (c) Damman, P.; Baudelet, N.; Reiter, G. Phys. ReV. Lett. 2003, 91, 216101. (17) Brochard-Wyart, F.; Daillant, J. Can. J. Phys. 1990, 68, 1084. (18) Seemann, R.; Herminghaus, S.; Jacobs, K. Phys. ReV. Lett. 2001, 86, 5534. (19) Wensink, K. D. F.; Je´roˆme, B. Langmuir 2002, 18, 413. (20) Sharma, A.; Mittal, J. Phys. ReV. Lett. 2002, 89, 186101-1. (21) Sharma, A.; Mittal, J.; Verma, R. Langmuir 2002, 18, 10213. (22) Suh, K. Y.; Lee, H. H. Phys. ReV. Lett. 2001, 87, 135502-1. (23) Yerushalmi-Rozen, R.; Kerle, T.; Klein, J. Science 1999, 285, 1254. (24) Lee, S. H.; Yoo, P. J.; Joon Kwon, S.; Lee, H. H. J. Chem. Phys. 2004, 121, 4346. (25) Thiele, U.; Mertig, M.; Pompe, W. Phys. ReV. Lett. 1998, 80, 2869. (26) Fondecave, R.; Brochard-Wyart, F. Macromolecules 1998, 31, 9305.
between thermal dewetting and solvent-induced dewetting is that the cause of instability is the long-range force of van der Waals interactions in the thermal dewetting, whereas it is the short-range force of polar interactions in the solvent-induced dewetting. In fact, the long-range force in the solvent-induced dewetting becomes stabilized rather than a destabilizing factor due to the presence of a solvent.24 In these reports on solventinduced dewetting, nearly all solvents, which are chosen, can dissolve the polymers. However, we find that cracks appear in the coating layer and finally result in the coating layer peeling up in a damp environment, and in this process, water cannot dissolve the coating materials. In many practical applications, water vapor is a very important factor to influence the stability of the coating layer and functional thin films, which are stable in dry conditions. It is not clear why and how the nonsolvent influences the stability of film. In this paper, we use water as a nonsolvent to study the process of nonsolvent-induced dewetting of thin polystyrene (PS) films, which is greatly different from the process of thermal dewetting and good-solvent-induced dewetting. 2. Experimental Section The system under investigation is a polymeric film of polystyrene (supplier TCI), and we use water as nonsolvent to induce the thin PS film. PS (Mw ) 3600 g/mol, Mw/Mn ) 1.06, the bulk glass transition temperature Tg ) 71 °C) was dissolved in toluene. Thin polymer films were prepared by spin-coating the toluene solution onto the Si wafers with 1.9 nm native oxide layer (Wafer Works (Shanghai) Corp; P, 0.3-50 ohm-cm). Prior to spin-coating, the wafers were cleaned by deionized water and acetone, then boiled in a 2/1(v/v) solution of 98%H2SO4/30%H2O2 for 30 min, later thoroughly rinsed with deionized water, and dried with compressed nitrogen. The thickness of film was measured by ellipsometry (Jobin Yvon S.A.S., France). The residual solvent was removed in a vacuum oven for 24 h at room temperature, and the surface roughness of the PS films (below 0.3 nm) was measured by atomic force microscopy (AFM). Then the samples were exposed to the saturated water vapor. After a few minutes, we took out the sample from the saturated water vapor. The surface morphology was observed by a commercial atomic force microscopy (AFM) operating in tapping mode using (27) Kim, H. I.; Mate, C. M.; Hannibal, K. A.; Perry, S. S. Phys. ReV. Lett. 1999, 82, 3496. (28) Bonaccurso, E.; Butt, H.-J.; Franz, V.; Graf, K.; Kappl, M.; Loi, S.; Niesenhaus, B.; Chemnitz, S.; Bo¨hm, M.; Petrova, B.; Jonas, U.; Spiess, H. W. Langmuir 2002, 18, 8056. (29) Govor, L. V.; Reiter, G.; Bauer, G. H.; Parisi, J. Appl. Phys. Lett. 2006, 89, 133126.
10.1021/la700805f CCC: $37.00 © 2007 American Chemical Society Published on Web 08/04/2007
Dewetting of Thin Polymer Films
Langmuir, Vol. 23, No. 18, 2007 9283
Figure 2. log-log plot of q* vs time shows the films breakup process under different humidity at 35 ( 1 °C. The black square symbols and the red circle symbols represent the breakup process of the films under saturated water vapor and saturated 10% NaOHwater solution vapor at 35 ( 1 °C, respectively.
Figure 1. Series of AFM images of PS films (in height mode) in different time interval under saturated water vapor at 35 ( 1 °C: (a) 50 min, (b) 60 min [2D FFT image shown in right inset], (c) 85 min, (d) 150 min, (e) 840 min, (f) 4800 min. a SPA300HV with a SPI3800N controller (seiko instruments inc, Japan) at room temperature. We obtain the contact angles of two different liquids on the solid surface by drop shape analysis (Kruss, Germany) at experimental temperature, and use the results to calculate the surface energy of the solid substrate and PS film by the method of harmonic mean.
3. Results and Discussion Figure 1 shows a series of AFM images of PS films (h ) 7 nm (∼4.5Rg), h is the thickness of films), which are under saturated water vapor at 35 ( 1 °C in different time interval. It can be observed that in the nonsolvent-induced dewetting process the holes forms, and the number of holes increases gradually, but the dimension of holes is homogeneous. It may be due to the fact that the velocity of formation of the holes is faster than the velocity of the holes growth, which results from the decrease of the drive force with the growth of the holes. Then the holes start to coalesce. Finally the ribbons rupture into small droplets. It is interesting that small droplets do not coalesce into big droplets but appear like fractal cluster aggregates (see Figures 1f and 6d). The phenomenon implies that PS cannot be in a viscous state in the process of aggregation. A more detailed dynamics study
Figure 3. AFM image and its line profile image show that PS film on Si substrate was treated for 45 h under saturated water vapor at 35 ( 1 °C.
of aggregation will be the subject of a future study. We propose that the formation of holes may be due to the fact that the water molecules penetrate into the polymer film,28 and then the capillary force drives the formation of holes and the growth of holes. It is expected that the capillary force must be strong enough to drive the process of formation of holes. The capillary force is approximately calculated by the following equation:30
2γ cos θ R
where P is the additional pressure, γ is the surface tension at the PS/water interface, θ is the contact angle of water on PS film, and R is the radius of holes, which is obtained from the holes forming in the early stage. According to the formula, the additional pressure can reach several Mpa in the early stage of forming holes, which is strong enough to drive the formation and growth of holes. This supports our hypothesis. Once water molecules reach the substrate, they will replace the nonpolar polystyrene from the substrate because water spreads more easily than polystyrene on the substrate as per our knowledge on thermo(30) Handbook of surface and colloid chemistry; Birdi, K. S., Ed.; CRC Press: Boca Raton, FL, 2002; p 48.
9284 Langmuir, Vol. 23, No. 18, 2007
Xu et al.
Figure 4. Series of AFM images of PS films (in height mode, 7 nm), which are treated at different temperatures for 12 h (except (a)) were shown. (a) Not treated after taking out it from the vacuum oven; (b) at 35 °C, (c) at 38 °C, and (d) at 41 °C.
dynamics.28 The replacement process makes the coalescent of holes. However, the irregular structure appears on the PS film in the process of holes coalescent (see Figure 1c,d). This may be because of the state of PS being not in viscous state. In order to investigate the dynamical behavior of the dewetting process, characteristic wave number (q*) is introduced. Figure 2 shows the time dependence of q* at 35 ( 1 °C on log-log scale. It is obvious that q* as the function of time exhibits four stages. In the first stage, q* increases with time due to the increase of the number density of holes. AFM images show that this process is similar to that of thermal nucleation dewetting, during which a continuous breakup of holes appears, but their drive forces are different.18 In the second stage, the coalescent process happens and q* follows the power law t-0.95 ( 0.058. This may be that the difference between the surface tension of water (72.8 mN/m) and the surface tension of PS (27.74 mN/m) induces pressure gradient, driving the coalescent process.31-33 In the third stage, the ribbons break up and form droplets. The pattern becomes “pinned”.31-33 Finally, the process of aggregation leads to the decrease of wave vector. Based on the above results, it can be seen that the mechanism of the nonsolvent-induced dewetting is different from that of the thermal dewetting and usual good-solvent-induced dewetting. We think that the process of the nonsolvent-induced dewetting may first undergo penetration, followed by replacement, coalescent, and finally aggregation. To prove the new mechanism in terms of thermodynamics, a simple calculation is shown. The spread coefficient of the replacement process S ) (γp/g + γp/s) - (γw/g + γw/s) where γp/g, γp/s, γw/g, and γw/s are the interfacial tension of the PS film/air, (31) Siggia, E. D. Phys. ReV. A 1979, 20, 595. (32) Koga, T.; Kawasaki, K. Physica A 1993, 196, 389. (33) Sung, L.; Karim, A.; Douglas, J. F.; Han, C. C. Phys. ReV. Lett. 1996, 76, 4368.
Figure 5. Series of AFM images of PS films (in height mode) which are treated in different time interval under saturated 10% NaOH-water solution vapor at 35 ( 1 °C in left column: (a) 6 h, (b) 18 h, and (c) 75 h; and under saturated 30% NaOH-water solution vapor at 35 ( 1 °C in right column: (d) 59 h, (e) 80 h, and (f) 180 h.
the PS film/substrate, the water/air, and the water/substrate, respectively. According to the Young equation, γp/s + γp/g cos θp/s ) γs/g, and γw/s + γw/g cos θw/s ) γs/g. In our experiments, we used γp/g ) 27.75 mN/m, γw/g ) 72.8 mN/m, θp/s1 ) 40°,34 θp/s2 ) 20°,34 θw/s1 ) 7°, and θw/s2 ) 80° (s1 is the silicon wafers with the oxide layer and s2 is the silicon wafers without the oxide layer). We calculate the spread coefficient of the replacement process S1 ) 5.984 mN/m > 0 on the silicon wafers with the oxide layer and S2 ) -58.468 mN/m < 0 on the silicon wafers without the oxide layer. Our calculation results indicate that water can replace the PS on the silicon wafers with the native oxide layer and cannot replace the PS on the silicon wafers without the oxide layer. The thickness of the native oxide layer (34) We obtain the contact angles from AFM image by the formulas tg(θ/2) ) H/R, (where H is the height of droplet and R is the radius of droplet) on Karim et al. Langmuir 2002, 18, 7041. The thin PS film (h ) 7nm) on the Si wafer without oxides layer is treated at 140°C after 10 hours, and we estimate θp/s2 from the AFM image of the treated PS film, and θp/s1 from Figure 1f.
Dewetting of Thin Polymer Films
Figure 6. Series of AFM images of PS films (in height mode, h ) 7 nm, h is the thickness of film) which are treated in different time interval under saturated water vapor at 20 ( 1 °C: (a) 26 h, (b) 48 h, (c) 75 h, (d) 125 h, (e) the PS film (h ) 7 nm) was treated 123 h under saturated water vapor at 16 ( 1 °C, and (f) a magnification of the boxed region in panel e.
greatly influences the wetting/dewetting properties of PS.18,35 In order to further verify the results of the calculation, we change the substrate into Si wafer without the oxide layer by removing the oxide layer with hydrofluoric acid and repeat the above experiments at 35 °C. It is found that the holes are about the depth of 3 nm on the PS film after 45 h, indicating that the replacement process does not take place. The phenomenon further proves that the above results of calculation are true (see Figure 3). Finally, we place the thin PS film in the dry conditions, and no nucleation holes are found at 35 and 38 °C after 12 h; however, there are some obvious holes on the thin PS film at 41 °C after 12 h (see Figure 4). These phenomena show that the above nonsolvent-induced dewetting process is not thermal induced, and the glass transition temperature of the thin PS film (Mw ) 3600 g/mol, 7 nm) is higher than 35 °C. (35) Seemann, R.; Herminghaus, S.; Jacobs, K. J. Phys.:Condens. Matter 2001, 13, 4925.
Langmuir, Vol. 23, No. 18, 2007 9285
Figure 7. Series of AFM images of PS films (in height mode) which are treated in different time interval under saturated water vapor (left column; a-c) and saturated ethanol vapor (right column; d-f). (a) h ) 7 nm, t ) 63 h, T ) 20 ( 1 °C; (b) h ) 7 nm, t ) 60 min, T ) 35 ( 1 °C; (c) h ) 20 nm, t ) 50 h, T ) 35 ( 1 °C; (d) h ) 7 nm, t ) 20 s, T ) 20 ( 1 °C; (e) h ) 7 nm, t ) 30 s, T ) 35 ( 1 °C; (f) h ) 20 nm, t ) 60 min, T ) 35 ( 1 °C.
In this context, we prove our hypothetical mechanism of the nonsolvent-induced dewetting by the results of experiments, thermodynamic assumption, and dynamic analysis. Our experimental results also show that the PS is not in a viscous state during the process of nonsolvent-induced dewetting, and the present process differs much from the usual thermal dewetting process and solvent-induced dewetting process. Besides, we investigated the influence of the variation of humidity on the dewetting behavior by adjusting the concentration of the NaOH solution at 35 °C. It is found that the morphology on the film at low humidity is analogous to that at high humidity, and the decreasing of humidity only decelerates the dewetting process in our experimental humidity range (see Figure 5). On the other hand, we investigated the influence of temperature on the dewetting process under saturated water vapor. The morphology on the film at 20 °C is different from that at 35 °C (see Figures 1 and 6). The process of penetration and coalescent become more difficult, and water may directly spread under PS film at low temperature. Therefore, more irregular morphology of film and obvious cracks appear on PS film (see Figure 6). In Figure
9286 Langmuir, Vol. 23, No. 18, 2007
6e,f, we find annular corrugations in the two AFM images. Water probably spreads under PS film at some places because the process of holes growth and coalescent is very slow at low temperature and the substrate is hydrophilic. The formation of the annular corrugations may be due to the evaporation of water under the PS film. The difference of morphology on the thin PS film may be due to the mobility of PS chains in our experiments. In order to study the factor of the mobility of PS chains, we used ethanol as nonsolvent to induce the PS film. It is observed that the velocity of dewetting under saturated ethanol vapor is faster than that under saturated water vapor, and the morphology of the film in the dewetting process under saturated ethanol vapor is different from that under saturated water vapor (see Figure 7).
4. Conclusion We use water as a nonsolvent to study the process of nonsolvent-induced dewetting of thin polystyrene (PS) films on hydrophilic surfaces at room temperature. The dewetting process is much different from other previous dewetting processes. The present dewetting process first undergoes water penetration followed by water replacement of the PS, which we described
Xu et al.
by a penetration-replacement mechanism. The analysis from thermodynamics and dynamics also support the hypothetical mechanism. Except for the difference of the dewetting process, the PS is not in a viscous state during the dewetting process, which makes the morphology much different from previous one. In addition, we find that the decreasing of humidity only decelerates the dewetting process in our experimental humidity range, and the mobility of PS chains is an important factor to influence the morphology of the PS film in the nonsolventinduced dewetting process. Acknowledgment. Thanks are due to Dr. P. K. Dutta of MNNIT, Allahabad, India for a proof-reading of the manuscript. This work is supported by the National Natural Science Foundation of China (20334010, 50503022, and 20620120105) Programs and the Fund for Creative Research Groups (50621302), and subsidized by the Special Funds for National Basic Research Program of China (2003CB615600). L.J.A. thanks the supports from the National Natural Science Foundation of China (20674086 and 50390090) Programs. LA700805F