Performance and Properties of a Ti-Al Composite Anodic Oxide Film on AC-Etched Al Foil

Abstract

AC-etched aluminum foils for an Al electrolytic capacitor were covered with a TiO₂ film by a sol–gel coating and then anodized to 25 V in an ammonium adipate solution. The structure, properties, and performance of the anodic oxide films were examined by transmission electron microscopy (TEM), scanning electron microscopy (SEM), electrochemical impedance measurements (EIS), a general digital LCR meter, a TV characteristic tester, and multicycle pulse charging–discharging. It was found that the anodizing of aluminum coated with TiO₂ films led to the formation of Al-Ti composite anodic oxide films, which consist of an outer Al-Ti composite oxide layer and an inner Al₂O₃ layer on the metal substrate. The capacitance (C_25V) of the anodic oxide films formed on specimens with a TiO₂ coating was about 10% larger than without a TiO₂ coating. The specific resistance (R_ox) of the Al-Ti composite film measured by EIS was lower than the blank one, accounting for a greater increase in the rise time (T_r) and a slight reduction in the withstand voltage (V_t). After hydration and a multicycle pulse charging–discharging destructive test, the Al-Ti composite anodic oxide film maintained the same good, comprehensive dielectric properties and performance as the blank one, thereby proving to be promising for acting as dielectric layers.

1. Introduction

Aluminum electrolytic capacitors are widely used in electric and electronic devices, such as mobile telephones, automobiles, notebook computers, inverter air conditioners, digital video cameras, and others. The integration and miniaturization trend of electronic components and products require the high capacitance of Al electrolytic capacitors, as well as a long service life. The properties of barrier-type anodic oxide films formed on etched Al foil play an important role as a dielectric film on the performance of the capacitor [1,2,3].

The electric capacitance, C, of anodic oxide films on Al foil can be expressed by the equation $C=\frac{{\epsilon }_0·{\epsilon }_r·A}{d}$, where ε₀ is the vacuum permittivity, ε_r is the relative dielectric constant of the anodic oxide film, A is the effective surface area of the etched Al foil electrode, and d is the film thickness. Increasing A is usually achieved by the electrochemical etching of Al foils with direct current (DC-etching) to form tunnel pits for high-voltage capacitors, or by the alternative-current etching (AC-etching) of Al foils for low-voltage ones before anodization. The value of d depends on the working voltage of the Al electrolytic capacitor. For a given working voltage, the value of d is essentially fixed [4,5,6,7].

Increasing A by the etching process is now encountering a bottleneck and limitations in the electrolytic capacitor manufacturing industry. The ε_r value of the Al₂O₃ anodic oxide film on the Al substrate is about 8~10, lower than other valve metal oxides, such as Ta₂O₅, ZrO₂, Nb₂O₅, and TiO₂, and far below ferroelectric materials, such as BaTiO₃, Bi₄Ti₃O₁₂, and (Ba₀.₅Sr₀.₅)TiO₃. Their ε_r values are summarized in

Table 1. ε_r value of some valve metal oxides and ferroelectric materials.

Oxide	Al2O3	ZrO2	Ta2O5	Nb2O5	TiO2	BaTiO3	SrTiO3
εr	9.8	22~25	27.6	41.4	90.0	500~6000	>200,000

. The formation of composite anodic oxide films with a high ε_r to replace the pure Al₂O₃ one is becoming a promising approach to improve the electric capacitance [8,9,10,11].

Over the past tens of years, sol–gel, hydrolysis precipitation, and electrodeposition have been developed to incorporate these high ε_r materials into the Al₂O₃ anodic oxide film on the Al substrate. For example, Yao et al. prepared a 100 V Ti-Al composite oxide film on flat Al foil by the sol–gel coating method and anodization [12]. Wang prepared a 200 V Al₂O₃-(Ba₀.₅Sr₀.₅)TiO₃ composite oxide film on tunnel-etched Al foil by the sol–gel coating method and anodization [13]. Lin et al. fabricated 20 V TiO₂-Al₂O₃ composite oxide films on tunnel-etched Al foil by electrodeposition and anodization [14]. Park et al. prepared 400 V ZrO₂-Al₂O₃ and Nb₂O₅-Al₂O₃ composite oxide films on flat Al foil by the sol–gel coating method and anodization [15]. Feng et al. formed 30 V Nb₂O₅-Al₂O₃ composite oxide films on low-voltage-etched Al foil by complexation precipitation and anodization [16]. Takahashi et al. fabricated 200, 400, 600, 800, and 1000 V Al-Si composite oxide films on DC-etched Al foil by the sol–gel coating method and anodization [17]. In these studies, the specific capacitance of the specimens with the composite oxide films had an increased tendency to some extent compared to those with the pure aluminum oxide ones.

However, the other dielectric properties, such as withstanding voltage, the rise time, hydration resistance, and the charge–discharge resistance, apart from the specific capacitance, service performance, and microstructure of the composite oxide film on the AC-etched Al foil with holes and pits for low-voltage capacitors, have never been comprehensively evaluated. Meanwhile, in comparison with hydrolysis precipitation and electrodeposition methods, coating by the sol–gel process has distinct advantages in practical applications for its low cost, ability to coat substrates with a complicated surface, homogeneity, and controllability.

In this study, a TiO₂-Al₂O₃ composite oxide film was produced on an AC-etched Al foil substrate with numerous holes by the sol–gel coating method and anodization. Such samples have never been used before. Its composition and microstructure are thoroughly investigated by SEM, TEM, and EDS. Its properties and service performance are comprehensively examined by EIS, including its specific capacitance measurements and rise time, as well as the withstanding voltage test, water resistance test, and charge–discharge test.

2. Experimental Section

2.1. Specimen

Highly pure (99.99%) commercial AC-etched aluminum foils with numerous tiny holes for low-voltage Al electrolytic capacitors were cut into specimens to a size of 1 cm × 5 cm with a handle. The thickness of the Al foil is about 120 μm.

2.2. Sol–Gel Coating

The TiO₂-precursor sol was first prepared by mixing tetrabutyl titanate (AR, Tianjin Kermel Chemical Reagent Co., Ltd., Tianjin, China), deionized water, hydrochloric acid (65%; Laiyang Economic and Technological Development Zone Fine Chemical Plant, Yantai, China), and dehydrated ethanol (99.7%; Yuandong Fine Chemical Co., Ltd., Yantai, China) at the molar ratio of 1:5.8:1.1:46.8 and then stirring for 24 h at room temperature. The specimens were immersed in the sol for 10 min, followed by air-drying at room temperature. After drying, the specimens were heated at 500 °C for 5 min.

2.3. Anodizing

According to the specification of the EIAJ (Electronic Industries Association of Japan) code, coated specimens were first anodized to 25 V with a constant current of 25 mA/cm² in 13% (mass fraction) ammonium adipate solution at 85 °C. After the selected voltage was reached, the anodization continued for 10 min while the voltage was held constant and the current was allowed to decay. After first anodization, the specimens were removed from the electrolyte, rinsed with deionized water, blown dry and annealed at 500 °C for 2 min. After heat treatment, the specimens were given the second anodization in the 13% (mass fraction) ammonium adipate and 0.2% (mass fraction) ammonium dihydrogen phosphate solution at 85 °C, reducing holding time of 25 V to 2 min under other invariable conditions. The blank specimen without TiO₂ was anodized under the same process for comparison.

2.4. Micromorphology and Composition of Anodized Samples

The surface and cross-section of anodized specimens were examined by scanning electron microscopy (SEM, JEOL JSM-6300 F, JEOL, Tokyo, Japan), transmission electron microscopy (TEM; Hitachi H-700H, Hitachi, Tokyo, Japan) using an ultra-thin sectioning technique and analyzed with energy dispersed X-ray analyzer (EDX, JEOL JEM-2000ES, JEOL, Tokyo, Japan). For cross-section observation, the samples must be embedded in resin. The sample surface and cross-section is sprayed with carbon for electron microscope observation.

2.5. Property and Performance of Anodized Samples

According to the EIAJ code, the specific capacitance of the anodized samples was measured in 13% (mass fraction) ammonium adipate solution at 30 °C using a general digital LCR meter (YY2810A) and denoted as C_25V. A pure aluminum sheet with a very large area was used as the counter electrode. The testing frequency was set at 100 Hz.

According to the EIAJ code, the rise time (T_r) and withstanding voltage (V_t) of the anodized samples were measured in 13% (mass fraction) ammonium adipate solution at 85 °C by an anode-foil TV characteristic tester (TV-3CH). During measurement, a 304 stainless-steel plate was used as the counter electrode. A constant current density of 10 A·cm⁻² was applied to the anodized aluminum specimen and the response voltage was recorded. As soon as the current was applied, the voltage increased immediately and then gradually reached a saturated value. The time taken to the formation voltage (25 V) is recorded as the rise time (T_r). After 3 min from Tr, the voltage was recorded as the retention voltage or withstanding voltage (V_t) of the anodized aluminum oxide film. T_r and V_t were considered as two parameters to evaluate the reliability of the oxide film.

In order to evaluate performance and service life of the anodized samples during process of application, the anodized samples undergo the following hydration test and charging and discharging test.

The anodized samples were given hydration treatment in boiling, deionized water. The hydration process time was set at 1 h and 24 h. After hydration, C_25V, T_r, and V_t of the samples was measured again and designated as C_25V-1h, T_r-1h, V_t-1h and C_25V-24h, T_r-24h, V_t-24h.

The anodized samples were strictly tested by multi-cycle pulse charging–discharging at 25 V. The pulse period was set to 1 s and the number of pulses was set to 10,000 and 23,000. After charging–discharging test, the C_25V and V_t of the samples was measured again and designated as C_25V-10,000, V_t-10,000 and C_25V-23,000, V_t-23,000.

Dielectric properties of the anodic oxide film were further examined by EIS measurement. In EIS measurement, specimens were immersed in 13% (mass fraction) ammonium adipate solution at 30 °C, and 100 mV of sinusoidal voltage in 10⁻²–10⁵ Hz range was applied. The EIS results were analyzed using the software ZSimpWin 3.20.

3. Results and Discussion

3.1. Surface SEM Images and EDS of Anodized Samples

The surface SEM image of the anodized specimen with blank oxide film is illustrated in Figure 1a–c and the Al-Ti composite one is shown in Figure 1d–f under different magnifications. Numerous tiny cavities, voids, pits, and holes induced by AC-etching can be found on the surface of two kinds of Al foil after anodization. The shape, size, and distribution of such etching structures are more clearly visible for the anodized specimen with Al-Ti composite oxide film than the blank one, contributing to the improvement in effective area of Al electrode foil and specific capacitance.

The area EDS analyses of two samples are listed in

Table 2. Area EDS analysis of two samples.

Area EDS of Sample with Al-Ti Composite Oxide Film			Area EDS of Sample with Blank Oxide Film
Element	wt%	wt% Sigma	Element	wt%	wt% Sigma
C	13.48	2.51	C	14.42	1.28
O	29.10	0.96	O	12.72	0.27
Al	54.55	1.62	Al	72.48	1.10
P	0.29	0.04	P	0.38	0.02
Ti	2.58	0.09	Ti	-	-
Total	100	-	Total	100	-

. Comparing EDS analysis of two foils shows that the surface of anodized specimen with Al-Ti composite oxide film contains small amounts of Ti element.

Numerous white particles or zones appear on two samples. The EDS analysis of such regions and its nearby zones on samples with Al-Ti composite oxide film are given in Figure 2 and Figure 3, respectively, further confirming the existence of Ti on the surface.

3.2. Cross-Section SEM Images and EDS of Anodized Samples

Figure 4a,b show cross-section SEM images of anodized samples with Al-Ti composite oxide film, and that with a pure aluminum oxide film, respectively. From both samples, it can be found that the anodized samples consist of two exterior layers of AC-etched structure and one interior layer of residual aluminum core. The AC-etched layer is about 40 μm thick and rich in tiny holes. Comparing the blank sample, more AC-etched holes are distinct and distributed on the anodized samples with Al-Ti composite oxide film, suggesting enhancement of effective area of Al electrode foil and specific capacitance, in consistent with observation in Figure 1

The anodic oxide film is formed on the surface of holes. At the same formation voltage, thick anodic oxide film is easy to clog the tiny holes and decrease the effective area. The anodic oxide film thickness, d, can be expressed by d = E_a·K, where E_a is the formation voltage, in this study E_a = 25 V, and K is the formation constant, defined as the ratio of film thickness against formation potential. It has been revealed that the K value of Al-Nb, Al-Zr, Al-Ti composite oxide film is lower than pure Al₂O₃. Change in etch hole morphology between two samples in Figure 1 and Figure 2 indicates that a thinner composite anodic oxide film has been produced after TiO₂-precursor sol immersion, anneal treatment, and anodization.

Figure 5 shows a cross-section line EDS of anodized samples with Al-Ti composite oxide film, indicating the concentration depth profile and change in O, Al, Ti, P and C elements with the distance from interface between samples and resin. Additional elements of C and P are probably introduced by carbon spraying during sample preparation for SEM observation and electrolyte for anodization.

It can be found that the content of O in the exterior bilateral etching layers is basically stable and much higher than that in the interior residual Al core layer. The variety of Al concentration in the outer etching layers and middle residual Al core layer has an opposite trend to that of O concentration. More interestingly, the Ti content gradually decreases with the increase in distance from the interface between sample and resin, demonstrating the existence of air in the tiny AC-etching holes and such labyrinthic AC-etching micro-pore structure may prevent the precursor sol from getting inside the depth of the etching layers during the sol immersion pretreatment step.

3.3. Cross-Section TEM Images and EDS of Anodized Samples

The cross-section TEM images and EDS of anodized samples with Al-Ti composite oxide film are given by Figure 6. Figure 6a shows cross-section TEM images of anodized samples with Al-Ti composite oxide film. More cavities and voids of a large size unevenly appear on the cross-section, suggesting that tiny holes produced by AC-etching tend to scale out and merge during growth, very different from the tunnel growth induced by DC-etching for high voltage Al electrolytic capacitor. Figure 6b shows the TEM image of Al-Ti composite oxide film, located on the spot circled by the red line in Figure 6a. The composite oxide film is uniform, about 45 nm thick and consists of an outer layer adjoining the hole and an inner layer on the Al substrate. The two layers were identified and confirmed as the Al-Ti composite layer and Al₂O₃ layer by following EDS analyses. It can be found that when the specimen is continuously irradiated in the electron beam, dark spots in the oxide layer gradually appeared in the TEM image.

Figure 6c–f shows cross-section EDS line scanning of Al-Ti composite oxide film, indicating the concentration depth profile and change in O, Al, and Ti elements with distance from bilateral holes, as illustrated by Figure 6g. The analysis of these EDS line scanning further demonstrates that the out layer of composite oxide film is about 30 nm, rich in Ti and the inner layer of composite oxide film is about 15 nm, poor in Ti.

Figure 7 shows a schematic model of the formation of the anodic oxide film during anodizing after TiO₂ coating. It is assumed that the TiO₂ layer has a network structure of micropores and cracks which may have been formed by the evaporation of organic compounds during heating of the sol precursor film. The growth of the inner and outer layers is a result of the transport of O²⁻, TiO₄⁴⁻, and Al³⁺ ions [18,19]. During anodizing after TiO₂ coating, O²⁻ ions dissociated from water at the bottom of the TiO₂ layer transport inward across the anodic oxide film to form Al₂O₃ at the interface between the inner Al₂O₃ layer and metal substrate. Al³⁺ ions transport outward to form the composite oxide layer by filling the micropores with Al₂O₃ at the interface between the outer composite oxide layer and the TiO₂ layer. In addition to the formation of oxide at the two interfaces, a conversion of Al₂O₃ to composite oxide may occur at the interface between the outer and inner layers due to the inward transport of TiO₄⁴⁻ or TiO₃²⁻ ions under the electric field across the composite oxide layer. The growth rate of the outer layer would be much higher than that of the inner layer. This is because the formation of the outer layer consumes electric charge only for repairing cracks and voids in the TiO₂ film. Hence, the total growth rate of the outer composite layer is much higher than the growth rate of the inner layer.

3.4. Property and Performance of Anodized Samples

3.4.1. Effect of Hydration on T_r, V_t and C_25V

The T_r, V_t, and C_25V values of blank samples and ones with Al-Ti composite oxide film are listed in

Table 3. T_r, V_t, and C_25V values of blank samples and ones with Al-Ti composite oxide film.

Samples		Tr (sec)		Vt (V)		C25V (μF/cm2)
Blank	1	49	Mean Value 46	25.7	Mean Value 25.7	109.01	Mean Value 109.19
	2	47		25.6		110.25
	3	42		25.8		108.31
Composite	1	67	Mean Value 66.7	25.5	Mean Value 25.5	120.21	Mean Value 120.2
	2	67		25.5		120.29
	3	66		25.5		120.11

and further expressed by Figure 8. The mean value of C_25V for Al-Ti composite oxide film is 120.2 μF/cm², about 10% higher than that of blank one. This can be attributed to the high relative dielectric constant and low formation constant of Al-Ti composite oxide film, comparing pure Al₂O₃ film, leading to an improvement in the effective surface area of the AC-etched Al foil electrode and its specific capacitance. The change in C_25V for both kinds of anodized specimens is agreed via SEM observation.

The mean value of V_t for Al-Ti composite oxide film is 25.5 V, almost the same as that for blank one. However, its T_r mean value is 66.7 s, 45% longer than blank one. The large prolongation of T_r from pure Al₂O₃ film to Al-Ti composite oxide film reflects the increase in internal defects in the anodic oxide film.

After 1h hydration, the T_r-1h, V_t-1h, and C_25V-1h values of blank samples and ones with Al-Ti composite oxide film are given in

Table 4. T_r-1h, V_t-1h, and C_25V-1h values after 1h hydration.

Samples		Tr-1h (s)		Vt-1h (V)		C25V-1h (μF/cm2)
Blank	1	26	Mean value 27.7	25.4	Mean value 25.4	109.32	Mean value 109.7
	2	30		25.3		110.87
	3	27		25.4		108.96
Composite	1	34	Mean value 35.3	25.2	Mean value 25.2	120.86	Mean value 120.7
	2	36		25.2		120.76
	3	36		25.2		120.71

and Figure 9. It can be found that for comparison of two samples, the respective change trend of T_r-1h, V_t-1h, and C_25V-1h is similar to data before hydration. Comparing with Table 3, the withstanding voltage and specific capacitance almost remain unchanged after one hour hydration, which reflects that both samples possess good water resistance. However, the rising time for both samples after one hour hydration is much shortened, respectively, which suggests that internal defects in the films have been repaired to some extent during hydration.

After 24 h hydration, the T_r-24h, V_t-24h, and C_25V-1h values of blank samples and ones with Al-Ti composite oxide film are given in

Table 5. T_r-24h, V_t-24h and C_25V-24h values after 24 h hydration.

Samples		Tr-24h (s)		Vt-24h (V)		C25V-24h (μF/cm2)
Blank	1	27	Mean Value 26	25.2	Mean Value 25.23	110.03	Mean Value 110.32
	2	26		25.2		111.99
	3	25		25.3		108.95
Composite	1	88	Mean Value 89.3	24.9	Mean Value 24.93	122.24	Mean Value 121.8
	2	90		24.9		121.95
	3	90		25.0		121.22

and Figure 10. It can be found that for comparison of two samples, the respective change trend of T_r-24h, V_t-24h, and C_25V-24h is similar to data in Table 3 and Table 4. Comparing Table 3 and Table 4, as far as one sample was concerned, the withstanding voltage and specific capacitance keep rather constant after 24 h hydration. However, the rise time for Al-Ti composite anodic film is greatly lengthened to 89.3 s and the rise time for blank sample is substantially shortened to 26 s, which suggests that more internal defects have been accumulated in the Al-Ti composite anodic film, its reliability has much degraded during long hydration, and the blank one shows superior performance maintenance against hydration.

3.4.2. Effect of Charging–Discharging on V_t and C_25V

Table 6. Withstanding voltage and specific capacitance after 1000 charging–discharging tests.

Samples		Vt-10,000 (V)		C25V-10,000 (μF/cm2)		Vt-23,000 (V)		C25V-23,000 (μF/cm2)
Blank	1	25.88	Mean Value 25.97	109.47	Mean Value 109.25	24.81	Mean Value 25.41	109.1	Mean Value 109.04
	2	25.91		109.55		24.76		109.19
	3	26.06		108.78		26.02		108.75
	4	26.01		109.21		26.05		109.13
Composite	1	24.61	Mean Value 24.84	120.86	Mean Value 121.19	24.54	Mean Value 24.81	121.82	Mean Value 121.17
	2	24.6		120.76		24.56		121.82
	3	25.33		120.71		25.32		119.89

gives the withstanding voltage and specific capacitance after 10,000 and 23,000 times charging–discharging test. It can be found that although mean value of specific capacitance for Al-Ti composite anodic film is higher than blank one after charging–discharging, the withstanding voltage for the former is always below the latter. A lower withstanding voltage usually indicates that more internal defects have been produced in the dielectric film, probably leading to a serious degeneration in its reliability and performance.

3.4.3. EIS of the Anodic Oxide Films

The electrochemical impedance spectroscopy (EIS) of blank Al anodic oxide film and Al-Ti composite one is given in Figure 11. The Nyquist plots in Figure 10a suggest that a single time-constant capacitance behavior was dominant over the frequency range in this investigation [20].

In Bode plots in Figure 11b, the |Z| vs. f curves decrease linearly with a slope of −∞ in the very low frequencies region and immediately turn to a horizontal line with f increasing. The phase shift angle vs. f shows the similar changing trend like |Z| vs. f curves. These behaviors are typical of dielectric materials. The two specimens exhibit the similar behaviors of impedance and phase, indicating that they possess similar dielectric properties. However, the |Z| value in the Bode plot and the arc radius in the Nyquist plot for blank Al anodic oxide film are larger than that for composite one.

As per the inserted diagram in Figure 11a, the EIS can be well-simulated and interpreted by an equivalent electric circuit, consisting of a parallel combination of C_ox and R_ox, connected in series to R_s, where C_ox and R_ox represent the specific capacitance and resistance of the film and R_s is the solution resistance. Measured data and calculated data show good coincidence. According to the equivalent circuit and spectra data, the C_ox, R_ox, and R_s values of the film can be calculated and are listed in

Table 7. Electrical properties of specimens with pure Al anodic oxide film and Al-Ti composite one.

Specimen	Rox (MΩ·cm−2)	Cox (μF·cm−2)	Rs (Ω·cm−2)	C25V (µF·cm−2)	Vt (V)	Tr (s)
Blank Film	1.847	34.68	1.512	109.19	25.7	46
Al-Ti Composite Film	1.115	52.08	1.166	120.2	25.5	66.7

, which also includes the film’s other electrical properties, such as C_25V, V_t, and T_r.

As shown by Table 7, the C_ox and C_25V values for the Al-Ti composite anodic film are higher than those for the blank one. According to the equation C = ε_rε₀S/d = ε_rε₀S/(E_a·K), the C value depends on ε_r/K under the same formation voltage, E_a. It has been reported that the ε_r/K values of TiO₂ and Al₂O₃ are 30.0 and 7.7, respectively. Therefore, the C_ox and C values of specimen with Al-Ti composite oxide film was higher than blank Al₂O₃ film in this study.

However, the R_ox of the Al-Ti composite anodic film is lower than that for blank one. The decrease in R_ox is possibly due to more defects in the Al-Ti composite oxide film caused by doping, which accounts for the reduction in V_t and prolonging in T_r.

4. Conclusions

AC-etched Aluminum foil was anodized to 25 V in an ammonium adipate solution after TiO₂ coating by sol–gel processing to examine the dielectric properties and performance of the anodic oxide films. The following conclusions may be drawn.

1. Anodizing of aluminum coated with TiO₂ films by sol–gel dip-coating leads to the formation of oxide film which consists of an outer Al-Ti composite oxide layer and an inner Al₂O₃ layer on the metal substrate.

2. The capacitance (C_25V) of anodic oxide films formed on specimens with TiO₂ coating is about 10% larger than that without TiO₂ layer. This is due to the formation of the Al-Ti composite oxide layer, which can sustain a higher electric field than Al₂O₃. The rise time (T_r) for Al-Ti composite film is 45% longer than blank film, although their withstanding voltage (V_t) is almost equivalent, because the resistance (R_ox) of the Al-Ti composite film is 65.6% lower than the blank film.

3. After 1 h and 24 h hydration, C_25V-1h, T_r-1h, V_t-1h and C_25V-24h, T_r-24h, V_t-24h show the same pattern of change as C_25V, T_r, and V_t.

4. After 10,000 and 23,000 multi-cycle pulse charging–discharging, the Al-Ti composite film maintains the same considerable stability in the specific capacitance and withstanding voltage as blank film.