Thin Solid Films, I73 (1989) 299-308 GENERAL
ELECTROCOLOURING EFFECT OF POROUS G. PASTORE,
OF ANODIZED ALUMINIUM WITH COPPER: AND BARRIER OXIDE FILM THICKNESSES M. PAEZ
AND J. H. ZAGAL
de Quimica, Facultadde Ciencia, Universidadde
(Received June 3,1988 revised January
Santiago de Chile, Casilla 5659, Santiago ?
In this work, we have investigated the effect of the thickness of both porous and compact films of aluminium anodized in 15% H,S04 on the electrocolouring process using copper sulphate and magnesium sulphate solutions. With thin layers (less than 10000 A) the colours obtained depend on the porous oxide layer thickness and they are controlled mainly by an interference phenomenon. With thicker films (6000&320000 A) the colour becomes independent of the amount of copper incorporated into the pores and an interference phenomenon is not observed.
In recent years, the applications of coloured anodized aluminium for architectural purposes have increased as a consequence of the development of new techniques for colouring. At the beginning of the century the methods used for colouring included immersion of the anodized aluminium specimens in solutions containing organic dyes. However, because of the limited stability of the colours so obtained, these methods have been replaced by processes known as “integral colouring” and “electrolytic colouring”‘-5. In the electrolytic colouring process, metal atoms are incorporated into the porous matrix of aluminium oxide obtained by anodization 4v5. The metal to be employed as the colouring agent is present in the electrolyte as a cation and the colour obtained depends on the type of electrolyte and concentration of the metal ion in the solutio#. The colour also depends on the conditions under which the aluminium is anodized, such as the type of electrolyte employed for anodization, since this will determine the properties of the porous oxide matrix obtained’. Although electrolytic and intergral colouring techniques are widely used now, the range of colours achieved is rather narrow, almost always shades from bronze to black. Those colours are produced by depositing metals such as nickel, cobalt, silver, tin, iron and copper. The resulting bronze and black shades are all very similar 6*8,9. Deposition of copper, however, also produces pink to maroon shades”. There are several patents in the literature related to the electrolytic colouring of aluminium using copper salts”-“. However, the papers that have addressed the fundamental aspects of the process are rather scarce18-22. 0040-6090/89/$3.50
0 Elsevier Sequoia/Printed
in The Netherlands
The thicknesses of both the porous and the barrier films of aluminium oxide are important in electrocolouring, depending on the practical applications that one expects for the coloured aluminium. Nobuyoshi and Issell* have anodized aluminium in oxalic acid, obtaining thicknesses of less than 10000 A, and they coloured these oxides with copper sulphate. These researchers obtained a wide range of colours in the visible spectrum by varying the thickness of the porous layer. In this work we study the effect of both porous and barrier film thicknesses in determining the colour of aluminium anodized in sulphuric acid and electrocoloured in copper sulphate solutions and we characterize the colours using the remission spectra by constructing chromatic diagrams as described by the Commission Internationale de 1’Eclaire (CIE)23. 2.
6063 T-5 aluminium alloy was used for all experiments. The aluminium samples (33mm x 30mm x 1.5 mm) were degreased in acetone, etched and brightened in acid. When thin oxide films were obtained, the samples were mechanically polished with 1 and 0.1 urn alumina before the etching procedure. The aluminium specimens were anodized at constant current with a Kepko model AT-lOO-10M power source monitoring the current with a Fluke digital multimeter. Different film thicknesses (34000, 60000, 79000, 102000, 120000, 144000, 170000, 195000, 216000, 231000, and 277000 A) were obtained using different anodizing times (5,10,15,20,25,30,35, 40,45,50 and 60 min) with a current density of 1.5 A dmP2 in 15% H,SO, at 25 “C. Different barrier film thicknesses were obtained using current densities of 1.0, 2.5, 4.0, 5.5 and 7.0A dme2 and anodizing times were chosen according to the total thickness desired, i.e. 12OOOOA (thick films) and 5OOOA (thin films). Thick film thicknesses were measured with a Dermitron model D-8 using the Faucault currents principle. The thin films’ thicknesses were estimated using interference spectrometry modified for reflection24. This was performed with a Carl Zeiss model DMR-22 spectrometer. The electrocolouring process was carried out in an electrolyte containing 10 g CuSO, 1-l and 1 g MgSO, 1-r (pH about 2) using an electronic programmer designed for this purpose 25. This device provided a wide variety of electrical programmes which could be reproduced accurately. The different colours obtained were characterized using the three-stimulus technique described by Billmeyer and Saltzman23 and was adapted to this work by using a diffuse reflectance device coupled to the Carl Zeiss spectrometer. This method for characterizing colours is based on the fact that each colour can be determined by tristimulus values x, y and z which represent the three primary colours, i.e. red (650 nm), green (530 nm) and blue (425 nm), which when combined give the sensation of colour. The tristimulus values of the equal-energy spectrum of colours (Fig. 1) in the X, Yand 2 system define the CIE standard which is generally used in colour technology23. For example, a monochromatic colour with 2 = 450 nm has X = 0.336, j = 0.038 and Z = 1.772. In order to represent a colour graphically, a tridimensional representation would be necessary. In fact, only a two-dimensional graph is used and the values of X, Yand Z are determined from Fig. 1 as follows:
Al WITH CU
Y, 1.0 ?I 2 0.6 F L” E 0.2
Fig. 1. Tristimulus standards.
values of equal-energy
in the f, j and P system according
It is not necessary to determine 2 because X + Y+Z = 1. The values of X and Y obtained are represented in a typical graph that is illustrated in Fig. 223. The point where the lines meet is white. Once the X and Yvalues of a colour are determined, they are represented by a dot on the graph and the predominant wavelength is obtained by joining the point that represents white to the dot and extrapolating until the line intersects the horseshoe-shaped curve. In our case the method consists essentially of recording the diffuse reflectance spectra and comparing them with a
Fig. 2. Typical chromatic
for all possible colours,
to CIE standards.
white standard of barium sulphate. The remission of light data for different wavelengths is processed by a computer to calculate the tristimulus values which are then used to construct the chromatic diagrams. Details ofthe technique are given by Billmeyer and Saltzmanz3 and the computational method will be described in a separate publicationz6. 3. RESULTS AND DISCUSSION 3.1 Thick anodicfilms studies Figure 3 shows the dependence of the percentage of remission at a wavelength of 630nm as a function of the thickness of the porous layer. The percentage remission is essentially independent of the thickness of the porous layer. Figure 4
THdkSS Fig. 3. Percentage thickness.
18 22 / “r-n.163 of anodized
26 at 1 = 630nm
as a function
of the barrier
Fig. 4. Chromatic diagram as a function of porous layer thickness of anodized aluminium: line 1,340OO and 60OOOA; line 2, 79000, 102000 and 120OOOA; line 3, 144000 and 170OOOA; line 4, 195000 and 216OOOA;line 5,231OOO and 277OOOA.
Al WITH CU
shows the chromatic diagram for 12 different thicknesses of the porous layer. For these 12 samples the dominant wavelength is approximately constant and lies around 600 nm. Thus the colour and its intensity are not a function of the thickness of the porous film. This suggests that the copper electrodeposition process is not limited by the diffusion of cupric ions through the channels that constitute the porous film. These results agree with those obtained by Lichtenberger-Bajza et ~1.” who claim that the colour intensity is a function of the amount of copper deposited inside the pores and that this quantity is independent of the thickness of the porous layer for a given electrical programme. If the colour is independent of the thickness of the porous film, it is possible that it could depend on the thickness of the barrier film. In order to check this, samples anodized at different current densities giving barrier film thicknesses from 160 to 320 A were electrocoloured. It is well known that, when an anodizing process is carried out at a certain current density, the voltage increases with time, goes over a maximum and then decreases slightly, reaching a constant value7. This constant voltage is reached when the formation of the barrier film is completed; the porous film then starts to form. This constant voltage value can be used to estimate the thickness of the barrier film and when H,SO, is used as the electrolyte the thickness corresponds to 10 A V-r (ref. 7). The total thickness was kept constant by adjusting the anodization times. Figure 5 shows the variation in the percentage remission at 630 nm as a function of the thickness of the barrier film for samples coloured under the same conditions. We observe differences in the percentage of remission for thicknesses between 160 and 320 A. At first sight we observe four colours as a function of the thickness of the barrier film, namely brown (thickness, 160 A), reddish purple (thickness, 248 A), and purple and blue in the range of thicknesses between 280 and 320 A. However, the spectra
Fig. 5. Percentage
at 1 = 630 nm as a function
of barrier film thickness.
obtained by visible diffuse remission give curves that are parallel to each other, exhibiting variations only in colour intensity. These differences are due to different amounts of copper deposited for the different samples. The thickness of the barrier film can affect the amount of copper deposited since this thickness affects the potential drop inside the film. The thicker the film, the higher the potential drop. This would then affect the electric field strength at the barrier-porous oxide interface, thereby affecting the rate of copper deposition. Figure 6 shows a chromatic diagram for samples with different barrier film thicknesses. The wavelength that predominates for a barrier film thickness of 160 A is 598 nm and this value increases to 700 nm for one sample with a film thickness of 248 A. The wavelength that predominates for a barrier film thickness of 280 8, is 500 nm. For samples with barrier film thicknesses from 300 to 320 A the wavelength that predominates is 470 nm. These three last wavelengths correspond to the range of purple colours described by Billmeyer and Saltzman23. The dependence of the colour on the barrier film thickness is probably not due to the thickness itself. It is known that anodizing at different current densities produces oxide films with different degrees of porosity. This will affect the amount of copper that is deposited into the pores and will then determine the colour obtained, as it is well known that the colour is a function of the amount of copper present in the pores19*20.In order to clarify this point two aluminium samples with the same total oxide coating thickness (120000 A) and with the same barrier film thickness (160 A) were prepared. One of them was electrocoloured according to the electrical programme in which a linear ramp of voltage was applied until a final voltage of 13 V was achieved in 15 s. Immediately after the final voltage was reached the applied voltage was shut off. A
Fig. 6. Chromatic diagram for anodized aluminium 160 A; x ,248 A; A, 280 A; l , 300 and 320 A.
samples with different barrier
second sample was electrocoloured using the same conditions but the final voltage of 13 V was held for 5 s. The main difference between the two treatments is that the second sample was polarized at the final voltage for a longer period of time, which gives the copper ions more time to diffuse through the pores. The results are summarized in the form of a chromatic diagram in Fig. 7 for the samples electrocoloured using these two electrical programmes. When there is less copper present in the pores, the wavelength for remission that predominates is close to the IR (7OOnm) region whereas, as the amount of copper increases, the wavelength moves to 600 nm.
Fig. 7. Chromatic diagram voltage of 13 V for 5 s.
for two samples electrocoloured
with (0) and without
3.2 Studies with thin$lms Table I presents the colours obtained with samples which have the same barrier film thickness but different thicknesses for the porous layer. The total thickness of the oxide for these samples is then different. The conditions under which the electrocolouring process was carried out were kept constant. It can be noticed from TABLE I C~L~IJR~~BT.~I~DONTHIN THICKNESS,
Total oxide thickness (A)
Purple Blue-green Green-yellow Pink Reddish
2000 2500 5000 7500 10000
the results shown in Table I that with thin oxide films it is possible to obtain colours in the whole visible spectrum and that the colours obtained depend on the thickness of the porous layer. The colours obtained can be attributed to an interference phenomenon as illustrated by the visible spectra shown in Fig. 8. The colour is influenced by the refractive index n, by the total anodic film thickness Tand by the wavelength”. These three variables will determine the colour according to the following expressions: nTcos r = (m/2)1
nT cos r = ((2m + 1)/4}A
(4) where r is the angle of incidence of the light beam and m is an interference coefficient that can have integer values equal to or greater than untiy. We also investigated the influence of the barrier film thickness in thin films on the colour. Table II summarizes the experimental results obtained with aluminium
Fig. 8. Interference and electrocoloured
spectra for anodized aluminium with copper: l, 2000 A; 0,250O
specimens having dimerent barrier film thicknesses A; qi,5000 A; n , 75001(; A, 10000 A.
TABLE II COLDURS OBTAINED
SAMPLES COATED WITH THIN OXIDE FILMS AND WITH DIFFERENT
Barrier film thickness (A)
Bronze Yellow Light yellow Purple Purple-blue Light green
96 152 200 220 260 278
samples of constant total oxide film thickness and electrocoloured under the same conditions. The results obtained with anodized aluminium coated with thin oxide films are in agreement with a model reported in the literature that proposes that light interference is produced between the planes defined by the base metal (aluminium) and the bottom of the pores27*28 where most of the copper is deposited. In our case, for the conditions of Table I, the optical path is altered by changes in porous layer thickness and, for the conditions of Table II, the optical path is affected by changes in the barrier layer thickness. 4.
Our experimental results and those presented in the literature clearly indicate that, when anodized aluminium specimens coated with thick oxide layers are electrocoloured with copper, the colour obtained depends on the amount of copper incorporated into the pores in the aluminium oxide film, this amount being a function of barrier film thickness but independent of porous film thickness. When anodized aluminium specimens coated with thin oxide layers (less than 7000 A total thickness) are electrocoloured with copper the colour obtained not only is a function of the amount of copper present in the pores of oxide but also depends on the thicknesses of both the barrier and the porous films as a result of the interference of light that occurs in the films. With thin films it is then possible to obtain a greater variety of colours. ACKNOWLEDGMENTS
This work has been supported by DICYT and by FONDECYT Grant 0203/87.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
K. Wada, T. Hatano and K. Uchida, J. Appl. Electrochem., 9 (1979) 445. G. A. Dorsey, Plating, 59 (1972) 948. J. Misuki, N. Baba and T. Yoshino, J. Met. Finish. Sot. Jpn., 30 (1979) 294. P. G. Sheasby and W. E. Cooke, Trans. hst. Mef. Finish., 52(1974) 103. J. Patrie, Trans. Inst. Met. Finish., 55 (1977) 57. A. W. Brace and P. G. Sheasby, The Technology of Anodizing Aluminium, Technicopy 1979, Chap. 14. J. W. Diggle, T. C. Downie and C. W. Goulding, Chem. Rev., 69 (1969) 365. N. Akira, S. Yasumori, R. Mineo and S. Yoshikazu, Jpn. Patent 798947, 1979. T. S. Kogyo, Jpn. Patenr8206519,1982. N. Akira, S. Yasumori and S. Yoshikazu, Jpn. Patent 7989948, 1977. M. Masashi, S. Norio and K. Kenj, Jpn. Putenf 7992533,1977. Nippon Light Metal Research Laboratory, Jpn. Pafent 8232 119, 1982. Ronain S.A., Spanish Patent 482,203,1980. S. Katsuo, S. Isao, I. Daiji and I. Yasuji, Jpn. Pafenf 79116351, 1978. Y. Kazuo, Jpn. Patent 7985139,1979. D. Rodriguez-Martinez, Spanish Pafenr 498,578, 1981. E. Hoffmann and W. Eitle, Ger. Patent 2556146, 1977.
18 19 20 21 22 23 24 25 26 27 28
G. PASTORE et al.
B. Nobuyoshi and M. Issel, Aruminyumu Kenkyu Kaishi, 137 (1979) 34. V. V. Koleda, G. L. Shchulkin, V. V. Sviridov and A. L. Belanovich, Zashch. Met., 20 (1984) 795. E. Lichtenberger-Bajza, F. Dolmolkki and I. Imre-Baan, Mer. Finish., 71(1973) 50. N. Baba and T. Sato, Kinzoku Hyomen Gijutsu, 27 (1976) 491. A. S. Doughty, G. E. Thompson, J. A. Richardson and G. C. Wood, Trans. Inst. Met. Finish., 53 (1975) 33. F. W. Billmeyer and M. Saltzman, Principles of Colour Technology, Wiley, Chichester, 2nd edn., 1981. F. Muiioz and G. Pastore, Bol. Sot. Chil. Quim., 27(1982) 112. D. PCrez. G. Pastore and J. M. Orpinas. Proc. 5th Brazilian Symp. on Electrochemical Corrosion, SL?O Paula, University of Sfo Paulo, Sio Paulo, Brazil, March 1986, p. 350. G. Pastore, A. Gevert and M. Paez, to be published in Bol. Sot. Chil. Quim. P. G. Sheasby, Trans. Insl. Met. Finish., 58 (1980) 41. P. G. Sheasby and F. P. Short, Met. Finish., 83 (1985) 47.