Electrophoretic Deposition To Control Artificial Opal Growth

Langmuir 1999, 15, 4701-4704

4701

Electrophoretic Deposition To Control Artificial Opal Growth M. Holgado,†,‡ F. Garcı´a-Santamarı´a,†,‡ A. Blanco,†,‡ M. Ibisate,†,‡ A. Cintas,†,‡ H. Mı´guez,†,‡ C. J. Serna,† C. Molpeceres,§ J. Requena,† A. Mifsud,| F. Meseguer,†,‡ and C. Lo´pez*,†,‡ Instituto de Ciencia de Materiales de Madrid (CSIC), Cantoblanco, 28049 Madrid, Spain, Unidad Asociada UPV-CSIC at Departamento Fı´sica Aplicada, Camino de Vera s/n, 46022 Valencia, Spain, Instituto de Tecnologı´a Quı´mica UPV-CSIC, Avenida de los Naranjos s/n, 46022 Valencia, Spain, and Departamento Fı´sica Aplicada ETSII-UPM, Madrid, Spain Received February 16, 1999. In Final Form: June 6, 1999 In this work we propose and demonstrate a solution to the problems which arise when SiO2 monodisperse nanospheres of diameters under 300 nm or over 550 nm are used to obtain opal-based photonic crystals. If the nanospheres are too small, the sedimentation rate is very slow or even may not occur; if they are large enough, no significant order can be achieved because the velocity is too high. This method, based on the electrophoretic phenomenon, allows us to control the sedimentation velocity. Furthermore, other species of importance in this field, such as SiO2 spheres covered with a thick layer of TiO2, do profit from this method.

The photonic crystal technology has attracted an increasing interest in the past few years. The current focus of researches lies on the search for a three-dimensional full photonic band gap (PBG).1 This full PBG was first observed in the microwave regime,2 subsequent reductions in the wavelength were achieved,3 and very recently it has been developed in the 5-10 µm wavelength region by wafer fusion techniques4 and in the 1.35-1.95 µm wavelength with silicon processing techniques,5 but although it is promising, the huge amount of labor necessary to obtain them must be taken into account. The present goal is to achieve a drastic reduction on the operating wavelength range because of the enormous number of applications of these materials when operated in the near-infrared and visible ranges.6 One of the most important goals is the inhibition of the spontaneous emission of lucent materials embedded therein that can lead to a thresholdless laser. To get it, ordered arrays with micrometer and submicrometer parameters are needed. Various techniques based on the use of colloids have been developed to construct these solid arrays. The best known methods are based on solvent evaporation,7 template-directed crystallization,8 and natural sedimentation.9 New ways to obtain large enough domains are continuously developed.10 The used materials vary de* Corresponding author at CSIC. E-mail: [email protected]. † Instituto de Ciencia de Materiales de Madrid (CSIC). ‡ Unidad Asociada UPV-CSIC. § Departamento Fı´sica Aplicada ETSII-UPM. | Instituto de Tecnologı´a Quimica UPV-CSIC. (1) Yablonovitch, E. Phys. Rev. Lett. 1987, 58, 2059. John, S. Phys. Rev. Lett. 1987, 58, 2486. (2) Yablonovitch, E.; Gmitter, T. J.; Leung, K. M. Phys. Rev. Lett. 1991, 67, 2295. (3) O ¨ zbay, E. M.; Tuttle, G.; Biswas, R.; Ho, K. M. Opt Lett. 1994, 19, 1155. (4) Yamamoto, N.; Noda, S.; Chutinan, A. Jpn. J. Appl. Phys. 1998, 37, 1052. (5) Fleming, J. G.; Lin, S.-Y. Opt. Lett. 1999, 24, 49. (6) Joannopoulos, J. D.; Villeneuve, P. R.; Fan, S. Nature 1997, 386, 143. (7) Dimitrov, A. S.; Nagayama, K. Langmuir 1996, 12, 1303. Rakers, S.; Chi, L. F.; Fuchs, H. Langmuir 1997, 13, 7121. (8) van Blaaderen, A.; Ruel, R.; Wiltzius, P. Nature (London) 1997, 385, 321.

pending on the method, with polystyrene11 and silica being the most usual. The sedimentation of silica spheres to form opal has been studied,12 and it was demonstrated that they present a face-centered cubic structure.13 The involved optical properties have been investigated as well.14 These structures can be used as a matrix to embed many different materials: graphite,15 CdSe,16 CdS,17 TiO2,18 dye molecules,11,19 and InP.20 This is probably one of the most tempting ways that will lead to the achievement of the full PBG because of the easy procedures involved. Nevertheless, there remain some problems. The first one is the time required to obtain an opal. If the silica spheres are too small (under 300 nm diameter), several weeks are needed for settling or they may not even settle at all because thermal agitation compensates gravitational forces. This problem has seemed as an inconvenience (9) Mayoral, R.; Requena, J.; Moya, J. S.; Lo´pez, C.; Cintas, A.; Mı´guez, H.; Meseguer, F.; Va´zquez, L.; Holgado, M.; Blanco, A. Adv. Mater. 1997, 9, 257. (10) Park, S. H.; Qin, D.; Xia, Y. Adv. Mater. 1998, 10, 1028. (11) Yamasaki, T.; Tsutsui, T. Appl. Phys. Lett. 1998, 72, 1957. (12) Salvarezza, R. C.; Va´zquez, L.; Mı´guez, H.; Mayoral, R.; Lo´pez, C.; Meseguer, F. Phys. Rev. Lett. 1996, 77, 4572. (13) Mı´guez, H.; Meseguer, F.; Lo´pez, C.; Mifsud, A.; Moya, J. S.; Va´zquez, L. Langmuir 1997, 13, 6009. (14) Bogomolov, V. N.; Gaponenko, S. V.; Germenenko, I. N.; Kapitonov, A. M.; Petrov, E. P.; Gaponenko, N. V.; Prokofiev, A. V.; Ponyavina, A. N.; Silvanovich, N. I.; Samoilovich, S. M. Phys. Rev. E 1997, 55, 7619. Mı´guez, H.; Lo´pez, C.; Meseguer, F.; Blanco, A.; Va´zquez, L.; Mayoral, R.; Ocan˜a, M.; Forne´s, V.; Mifsud, A. Appl. Phys. Lett. 1997, 71, 1148. Vlasov, Y. U.; Astratov, V. N.; Karimov, O. Z.; Kaplianskii, A. A.; Bogomolov, V. N.; Prokofiev, A. V. Phys. Rev. B 1997, 55, 357. (15) Zakhidov, A. A.; Baughman, R. H.; Iqbal, Z.; Cui, C.; Khayrullin, I.; Dantas, S. O.; Marti, J.; Ralchenko, V. G. Science 1998, 282, 897. (16) Romanov, S. G.; Fokin, A. V.; Tretijakov, V. V.; Butko, V. Y.; Alperovich, V. I.; Johnson, N. P.; Sotomayor Torres, C. M. J. Cryst. Growth 1996, 159, 857. Vlasov, Y. A.; Yao, N.; Norris, D. J. Adv. Mater. 1999, 11, 165. (17) Vlasov, Y. A.; Luterova, K.; Pelant, I.; Ho¨nerlage, B.; Astratov, V. N. Appl. Phys. Lett. 1997, 71, 1616. Blanco, A.; Lo´pez, C.; Mayoral, R.; Mı´guez, H.; Meseguer, F.; Mifsud, A.; Herrero, J. Appl. Phys. Lett. 1998, 73, 1781. (18) Wijnhoven, J. E. G. J.; Vos, W. L. Science 1998, 281, 802. (19) Petrov, E. P.; Bogomolov, V. N.; Kalosha, I. I.; Gaponenko, S. V. Phys. Rev. Lett. 1998, 81, 77. (20) Mı´guez, H.; Blanco, A.; Meseguer, F.; Lo´pez, C.; Yates, H. M.; Pemble, M. E.; Forne´s, V.; Misfud, A. Phys. Rev. B 1999, 59, 1563.

10.1021/la990161k CCC: $18.00 © 1999 American Chemical Society Published on Web 06/03/1999

4702 Langmuir, Vol. 15, No. 14, 1999

Letters

somehow difficult to solve for many researchers.10,21 The other difficulty is related to heavy spheres (over 550 nm diameter); their sedimentation velocity is such that it is quite hard to achieve an ordered array, and it becomes completely impossible if the diameter is further increased. It is known that when gravitational energy is much larger than thermal energy (kBT), the sedimentation occurs far from equilibrium and noncrystalline sediment is obtained. Both problems make it quite unpleasant to work out of the limits of this reduced diameter range (300-550 nm), which corresponds to sedimentation velocities from 0.2 to 0.7 mm/h according to Stokes law. In this point, the electrophoretic phenomena offer a good solution for these two problems. The forces between particles22 and the effects of electric fields over colloidal particles have been widely observed,23,24 e.g., modulation of lateral attraction between particles and particle clustering.25 Using the electric field to drive the sedimentation velocity and keep it around 0.4 mm/h would solve the difficulties mentioned before. The model of constant-velocity particle packing is based on the interaction of gravitational (Fg ) 1/6πFsgd3), Archimedes (FA ) 1/6πFwgd3), and frictional forces (Ff ) 3πηvd), where Fs and Fw are the sphere and water mass densities, g is the gravity acceleration, η is the viscosity of water, d is the sphere diameter, and v is their velocity. When all forces are balanced, Stokes law is obtained. Experimental observations fit to this expression in an excellent way.9 It is well-known that particles (SiO2 and TiO2) in a colloidal suspension have a surface charge density when they are away from the point of zero charge (PZC), in which case the electric charge is null. Taking into consideration the force produced by an electric field E parallel to all other forces, the following equation is obtained for the velocity:

v)

d2(Fs - Fw)g + uE 18η

(1)

where the first part of this equation is the classical Stokes law and the second part corresponds to the contribution of the electric field to the sedimentation velocity, related to the mobility of the spheres u. Now, the main problem is how to calculate the particle’s mobility. The application of the electrophoretic concept can solve it. Provided that Stokes velocity without electric field is calculated with great accuracy, the electrophoretic mobility can be obtained in a straightforward manner if Stokes velocity is subtracted from the experimental velocity of the sample under a known electric field. Once the mobility is determined, the electric field necessary to achieve a given velocity can be stated beforehand. Several experiments according to what has been explained were performed, and the mobility values obtained varied from -2 to -8 µm cm/V s in agreement with the range of values given by an independent study of the electrophoretic mobility in a Delfa Coulter 440. Because moving away from the PZC increases the particle’s charge, variations of the pH must involve changes in the mobility (21) Mei, D.; Liu, H.; Cheng, B.; Li, Z.; Zhang, D.; Dong, P. Phys. Rev. B 1998, 58, 35. (22) Atkins, D.; Ke´kicheff, P.; Spalla, O. J. Colloid Interface Sci. 1997, 188, 234. (23) Trau, M.; Saville, D. A.; Aksay, I. A. Science 1996, 272, 706. Trau, M.; Saville, D. A.; Aksay, I. A. Langmuir 1997, 13, 6375. (24) Sarkar, P.; Nicholson, P. S. J. Am. Ceram. Soc. 1996, 79, 1987. (25) Solomentsev, Y.; Bo¨hmer, M.; Anderson, J. L. Langmuir 1997, 13, 6058.

Figure 1. Experimental setup of the electrophoretic cell.

values. The colloidal suspensions of these experiments consisted of silica spheres in double-distilled water (with no added salt) in which the solid content was around 1.0 wt %. The cell where electrophoresis was performed (Figure 1) consisted of a cylindrical tube (2 cm of diameter) of methyl acrylate fixed to the basis where the opal should settle, obtained from a standard silicon wafer sputtered with titanium (with less than 0.1 nm of rugosity and thick enough to ensure a good conductivity). The problem we had to cope with in these experiments was the electrolysis phenomenon, which has been reported elsewhere.23 The solution adopted was the use of platinum for the upper electrodes because they have the highest redox potential. Then, both electrodes were connected to a dc source used to obtain an electrical field. With this method compacts with thickness ranging between a few monolayers and 1 mm (depending on the amount of silica spheres used) with surfaces of about 3.1 cm2 are produced. To measure the sedimentation velocity, the height descended by the colloid/ clear water interface (setting the initial height at 0 mm) was monitored with time. The velocity results from an experimental fit of height vs time. First, the response of SiO2 spheres was studied. An electric field was applied to colloidal suspensions of SiO2 spheres in which the original pH was varied by adding HCl to change the surface charge. The PZC of silica occurs at pH ) 2.5, so pHs of the suspensions were chosen to be different enough without being close to the PZC: pH ) 3.8 and the reference value (no acid added) of pH ) 8.4. The results of the sedimentation velocities for silica spheres of 500 nm of diameter are graphically compared with the theoretical Stokes fall of a sample without an electric field in the left panel of Figure 2. It can be clearly seen that, as we move away from the PZC, the mobility increases and so does uE. To study the effects of velocity variations on silica particle ordering, two more sedimentations were prepared from the same sample. One of them was left to fall without an electric field, whereas in the other one the electrodes were inverted to decrease the sedimentation velocity by opposing the electric field to gravity. Because the mobility can be extracted from the previous experiment (u ) -3.9 µm cm/V), as explained, the electric field needed to get the desired velocity (0.4 mm/h) was calculated to be 0.5 V/m. The experimental value (v ) 0.35 mm/h; see right graph in Figure 2) was close to it. In Table 1 the results from this experiment are numerically compared. Electronic and optical microscopy studies of all of these samples were made, and it was observed that the slowed sedimentation sample presented a better ordering than the one settled without field, while the accelerated samples

Letters

Langmuir, Vol. 15, No. 14, 1999 4703

Figure 2. SiO2 spheres of 500 nm diameter fall obeying the Stokes law. The sedimentation is faster (left) as the pH values move away from the PZC when an electric field is applied and slower (right) if it is inverted. The dotted lines are the data fits, and their slopes give the velocities.

Figure 4. SEM image of a cleaved edge of 870 nm diameter SiO2 spheres opal settled under electric field and its Fourier transform showing the presence of periodicity.

Table 1. Mobilities and Velocities from SiO2 Spheres of 500 nm Diameter at Different pH and Electric Fields pH

E (V/m)

u (µm cm/V s)

v (mm/h)

3.8 8.4 8.4

-33 -33 0.5

-2.0 -3.9 -3.9

2.9 5.2 0.35

Figure 5. Bragg diffraction from (a) sintered 870 nm diameter SiO2 spheres opal whose sedimentation was slowed and (b) as-grown 205 nm diameter SiO2 spheres opal settled under acceleration.

Figure 3. SEM image of a cleaved edge of 870 nm diameter SiO2 spheres opal settled without electric field and its Fourier transform showing the absence of order.

from the previous experiment presented no order at all. Bragg diffraction was performed as well, showing that the slowed opal presented well-defined Bragg peaks. To prove how useful this method could be, SiO2 spheres with a diameter of 870 nm were used. The purpose was to obtain a well-ordered array by decreasing the natural velocity of this colloid (no electric field applied). Figure 3 shows a scanning electron microscopy (SEM) of a cleaved edge of the naturally settled opal. A high velocity (1.54 mm/h) is obtained for these large spheres, and no longrange order is achieved as by the Fourier transformed image shown in the inset of Figure 3. An equal colloid of the same spheres was prepared and settled under a slowing electric field, in which the velocity was kept close to 0.35 mm/h. Figures 3 and 4 show that only very small domains appear when the electrophoretic technique is not applied while large domains are obtained when sedimentation is performed under an appropriate electric field. To

check this, Fourier transforms from both images were performed; the opal settled under an electric field presents a clear pattern that is not present in the natural settled opal. A Bragg diffraction study from the slowed opal was performed after sintering,26 and very clear peaks were measured as shown in Figure 5a while the other sample did not present any kind of peak as a result of the lack of large enough ordered domains. In addition, a small percentage of small spheres was present in this sample. They were observed in SEM images of the naturally settled sample, but they were not present in the other one because electric force compensated for the gravity force. This suggests that the electrophoretic concept could be used to control the presence of small spheres in sedimentation when monodispersity is not granted. A sample of quite small (205 nm of diameter) SiO2 spheres, which would take 2 months to settle, was prepared for sedimentation. It was accelerated from 0.09 (natural velocity) to 0.35 mm/h in order to complete the sedimentation in less than 2 weeks without decreasing the optical quality. Again, the diffraction study from the as-grown (26) Mı´guez, H.; Meseguer, F.; Lo´pez, C.; Blanco, A.; Moya, J. S.; Requena, J.; Mifsud, A.; Forne´s, V. Adv. Mater. 1998, 10, 480.

4704 Langmuir, Vol. 15, No. 14, 1999

Figure 6. Height of the interface of sedimentation of TiO2/ SiO2 spheres plotted versus time. The dotted lines are the data fits, and their slopes give the velocities.

opal presented Bragg peaks as shown in Figure 5b, which denoted the presence of order within the opal. To check the effects of electric fields on other species, spheres of TiO2 with a nucleus of SiO2 (silica core of 250 nm diameter covered with a 180 nm layer of TiO2) having a different PZC (pH ) 6.6) from silica were prepared (details will be published elsewhere). In this case, the average density of these heterogeneous spheres was used; once its value was introduced into the first part of eq 1,

Letters

the natural sedimentation velocity was deduced. Taking these data into account, the first experiment was repeated with different pH values (1.5 and 5.1). The results shown in Figure 6 confirmed what was expected: moving away from the PZC involves a change in mobility and the ability of these spheres to be driven by an applied electric field. In summary, the importance of applying the electrophoretic concept to opal sedimentation has been demonstrated. With this method we have been able to assemble opals presenting an ordered array of spheres with diameters out of the range of 300-550 nm. This technique has proved to be an efficient way to control the sedimentation velocity of SiO2 and heterogeneous SiO2-TiO2 spheres, so it allows us to save time when small spheres are used and solves as well the existing problems for too heavy spheres. This fact opens new fields of research for photonic crystals beyond the visible and near-infrared ranges. Acknowledgment. We acknowledge M. Planes for his help during SEM characterization and P. Morales for her advice about mobility measurements. This work has been partially financed by the Fundacio´n Ramo´n Areces, the European Project PHOBOS, and the Spanish CYCyT Project MAT97-0698-C04 and Generalitat Valenciana Project GV-D-CN-08-129-96. LA990161K