Bright Yellowish-Red Pigment Based on Hematite/Alumina Composites with a Unique Porous Disk-like Structure.

Inspired by a bacteriogenic, iron-based oxide material and a traditional Japanese red pigment, a bright yellowish-red pigment was prepared by heating an Al-containing iron oxyhydroxide precursor. The obtained red pigment had a unique porous disk-like structure, comprising Al-substituted hematite particles and crystalline alumina nanoparticles. Although these disk-like structures loosely gathered to form an aggregate in powder, they can be easily dispersed into a single, disk-like structure by simple ultrasonic irradiation. The powder exhibited a bright yellowish-red color and high thermostability, making it attractive as a coloring material for various industrial products needing a bright-red color, high weather resistance, and durability. Quantitative color measurements revealed extremely high L*, a*, and b* values that are much greater than those of commercially available hematite. The thermostability test showed that even after exposure to high temperatures, the pigment retained the red color, indicating its high thermostability. The unique microstructure should be strongly related to the bright yellowish-red color and the high thermostability of the developed red pigment.

Introduction

In modern society, almost all the daily necessities, such as clothes, sundries, toys, cosmetics, electrical appliances, vehicles, and buildings, are colored. Among various colors, red has a special meaning for humans. A pigment workshop was discovered in an old cave that is around 100,000 years old, and red ochres were detected.[1,2] Red pigments were used to decorate a 73,000-year-old abstract drawing, which was excavated from the Blombos cave in South Africa; they were used in prehistoric cave arts found in the caves of Chauvet in France (30,000 years old), Altamira in Spain, and Lascaux and Niaux in France (12,000–17,000 years ago).[3−5] In addition, the red color has been regarded as a symbol of transformation and was used in burials and for decorating the Venus figurines in the Paleolithic.[6] The red pigments used by humans are mainly composed of a simple iron oxide, α-Fe2O3 (hematite), and humans used mined hematite ores or heat-treated iron oxyhydroxides.[7]

Organic pigments generally show vivid colors and have various color variations as well as excellent tinting strength. Although organic pigments are often used to give industrial products an intense and vivid color, there are some problems: color fading; repainting is needed when using them for painting applications; and they cannot be used in high-temperature applications, such as ceramic and asphalt, because of their chemical instability. On the other hand, inorganic pigments often do not show vivid colors, although they exhibit extremely high weather resistance. Hematite—a simple iron oxide that shows a red color and does not contain toxic elements—has been excavated from many remains of the Paleolithic.[1−4,6] Excavated hematite retained its red color even after exposure to the natural environment for several tens of thousands of years—or over 100,000 years—, which proves that oxide pigments are ultimate weather-resistant pigments that do not rust anymore. However, the red color of hematite is not vivid and intense; it is actually close to brown. If a bright-red inorganic pigment can be realized on the basis of hematite, such a pigment could be used for coloring various industrial products as a cost-effective, safe, and environmentally benign weather-resistant pigment.

It is known that the main absorption in the visible region of hematite is attributed to the charge transfer transition from the O 2p to Fe 3d levels,[8,9] resulting in a red color. In addition, the color tone of the hematite powder strongly depends on its particle size and dispersibility: the color can range from red for small and/or well-dispersed particles to dark gray for large and/or strongly aggregated particles.[10] Therefore, it is quite important, for obtaining a bright-red color, to precisely control the particle size and dispersibility. Previously, various red pigments based on hematite have been developed. Contemporary research on hematite red pigments has focused on controlling the particle size, distribution, dispersibility, and nanostructure of hematite, yielding novel red pigments with bright color and high chemical and/or thermal stabilities.[11−37]

We have developed novel red pigments based on a unique concept[38,39] that differs from conventional approaches. A porous microtubular structure (∼1 μm in diameter), which is composed of hematite/silicate composites, can be prepared by heating a bacteriogenic, iron-based oxide material produced by iron-oxidizing bacteria.[38] The hematite particles are covered with amorphous silicate, and many pores exist between them. The tubular structures do not sinter strongly, and the hematite particles are highly dispersed in air, which results in a bright yellowish-red powder. This red pigment does not show color fading even after exposure to high temperatures of 800 °C. In addition, Al-substituted hematite was synthesized, inspired by a traditional Japanese red pigment comprising hematite fine particles incorporating 1 mol % Al, the first artificial red pigment produced in Japan, showing that Al substitution in the hematite structure intrinsically enhances lightness and chroma in hematite color.[39] Based on these results, nanocomposite materials comprising Al-substituted hematite particles and ultrafine, Fe-substituted, corundum nanoparticles were synthesized, resulting in a bright yellowish-red color and a high thermostability in color.[39]

In this study, inspired by both the bacteriogenic, iron-based material and the Japanese traditional red pigment, an Al-added low-crystalline iron oxyhydroxide was synthesized as a precursor, and it was then heated to obtain a novel bright-red pigment with high thermostability in color. Especially, we investigated in detail the effect of the structure (e.g., crystallographic structure and microstructure) on the color tone and the thermostability of the pigment, comparing it with the Al-free sample and commercially available hematite.

Experimental Section

Samples containing Al molar ratios of 0–0.5 [x = Al/(Fe + Al)] in 0.1 increments were prepared as follows: Fe(NO3)3·9H2O (Wako Pure Chemical Industries, 99.9%) and Al(NO3)3·9H2O (Wako Pure Chemical Industries, 99.9%) were used to prepare 0.1 dm3 of 0.5 mol dm–3 solution in total metal amounts. Then, 0.6 mol NH4HCO3 (Wako Pure Chemical Industries) was added to the solution, stirred for 15 min using a propeller agitator (IKA, Eurostar 20) at 200 rpm, and left to stand for 1 h. The resultant precipitate was collected by vacuum filtration and washed with ∼3 dm3 deionized water and ∼0.05 dm3 ethanol. The wet pastes were dried under vacuum at room temperature. The obtained precursor powders were heat-treated by elevating the temperature at a rate of 10 °C min–1 to 700 °C, kept for 2 h, and cooled in a furnace. To clarify the effects of heat on the color tone, the heat-treated samples were reheated at 900–1200 °C for 1 h. A commercially available hematite (Nakalai Tesque, ≥95.0%) sample was used as the reference.

The crystallographic structures, reflectance spectra and colors, and microstructures of the obtained powder samples were characterized by X-ray diffractometry (XRD, MiniFlex600, Rigaku, Japan), spectrophotometry using the CIE standard illuminant D65 (CM-5, Konica Minolta Sensing, Japan), scanning electron microscopy (SEM, JSM-6701F, JEOL), and transmission electron microscopy (TEM, JEM-2100F, JEOL, Japan) coupled with secondary electron detection and energy-dispersive X-ray spectrometry (EDS). High-resolution TEM images were acquired in the TEM mode, while the secondary electron and EDS elemental mapping images were acquired in the scanning TEM (STEM) mode. The particle size distribution of the sample was determined using a laser particle size analyzer (LA-950, HORIBA, Japan). The powder sample was added to a 0.1 wt % sodium hexametaphosphate aqueous solution, and the measurement was performed during ultrasonic irradiation for 0–50 min. To determine the pore size distribution and the specific surface area, the samples were measured by the nitrogen adsorption method at −196 °C using a BELSORP-mini-II system (MicrotracBEL, Japan). Prior to the measurement, all samples were degassed under vacuum for 5 h at 300 °C. Data were analyzed by the Barrett–Joyner–Halenda method to obtain the pore size distribution and the Brunauer–Emmett–Teller (BET) method to estimate the specific surface area.

Results and Discussion

Figure shows the XRD patterns of the precursor and heat-treated samples. The XRD patterns of all the precursors were similar. Therefore, the pattern of the Al-free sample is displayed in the bottom of Figure . Two broad peaks with 2θ values of around 35 and 60° were detected for the precursor samples, indicating the formation of a low-crystalline iron oxyhydroxide of 2-line ferrihydrite (2Fh).[40] Sharp diffraction peaks were detected as well. Although these peaks have not been assigned, they could be due to a kind of salt related to Fe3+, nitrate, carbonate, and/or ammonium ions, which was decomposed by heat treatment.

Figure 1

XRD patterns of the precursor sample with x = 0 (bottom) and heat-treated samples with x = 0–0.5 (upper). Red dotted vertical lines were placed at the diffraction peaks of (214) and (300) planes for the sample with x = 0 to visually know the peak shift depending on the Al molar ratio.

The color of the precursors was basically brown, but changed to a light yellowish-brown upon adding Al and to red after heat treatment. Upon heating these precursors at 700 °C, a single phase of hematite, which was assigned by using the powder diffraction file of the Inorganic Crystal Structure Database (ICSD) 81248, was confirmed for samples with x ≤ 0.3. In addition to hematite, weak peaks of AlFeO3 (ICSD 203202) were detected in the samples with x = 0.4 and 0.5. Diffraction peaks slightly shifted toward higher angles with increasing Al molar ratios, but the shift was mostly saturated ≥0.2 (see the high angle around 60–65° using dotted red vertical lines as guidelines), suggesting lattice parameter change. The lattice parameters a and c of heat-treated samples with monophasic hematite (x = 0–0.3) were determined by XRD measurements (Figure ). The parameters a = 5.038 nm and c = 13.752 nm were calculated for the heat-treated Al-free sample, which are nearly consistent with the reported values of hematite. On the other hand, for the heat-treated Al-containing samples, the parameters were found to be shortened. The ionic radii of six-coordinated Fe3+ and Al3+ are 0.0645 and 0.0535 nm, respectively, and the radius of Al3+ is shorter than that of Fe3+. Hence, the shortening of lattice parameters has originated from the differences in ionic radii. Lattice parameters decreased with increasing x up to 0.2 and showed nearly constant values. The amount of Al substitution for the heat-treated sample with x = 0.2 and 0.3, as calculated using Vegard’s law, was ∼0.08, which is much lower than the starting composition. This will be discussed later. The Al substitution value (∼0.08 at 700 °C) is similar to the value reported previously in the Fe2O3–Al2O3 phase diagram.[39,41]

Figure 2

Relationship between lattice parameters (a and c) of hematite and Al molar ratio x.

Figure shows the results of reflectance curves (upper) and quantitative color measurements (lower) on heat-treated samples with x = 0–0.5. The results obtained for a reference sample and photographs of the powders are also shown. All the reflectance spectra showed the typical profile of hematite; the reflectance value greater than 550 nm enhanced with the increasing Al molar ratio, and the color systematically changed depending on the molar ratio of Al (x). The L* value indicating lightness was found to monotonically increase with increasing x. On the other hand, a* and b* values, whose positive values show reddish and yellowish colors, respectively, improved with increasing x, thereby showing the brightest yellowish-red color at x = 0.3 and gradually fading at x > 0.3. The results obtained for a reference sample and photographs of the powders are also shown. For the heat-treated sample with x = 0 (Al-free sample), the lightness (L* value) and chroma (a*, and b* values) were extremely low, with values of L* = 32.9, a* = 11.2, and b* = 4.2. On the other hand, upon adding Al, for example, with x = 0.3, these values increased drastically to L* = 48.3, a* = 37.2, and b* = 48.8. A comparison with a commercially available hematite sample exhibiting a relatively bright-red color (L* = 38.6, a* = 30.4, and b* = 21.0) shows an obvious difference, with the L*, a*, and b* values of the Al-added sample being 1.25, 1.22, and 2.32 times greater, respectively, than those obtained for the commercially available sample. Especially, the Al-added sample showed a high b* value, indicating the bright yellowish-red color. To the best of our knowledge, such a bright yellowish-red color has not been reported so far in hematite-based pigments. In addition, these values are higher than those reported in previous literature for red pigments based on bismuth oxide[42] or attapulgite[43] and those observed for model materials: the red pigment prepared by us previously using a bacteriogenic iron oxide as a precursor[38] and yellowish-red Al-substituted hematite powders prepared by us previously, inspired by Fukiya Bengala, the first artificial hematite-based red pigment produced in Japan.[39] From the absorbance spectra calculated using the Kubelka–Munk function, f(R) = (1 – R)2/2R, where R is the reflectance, band gap energies of the commercially available hematite and heat-treated samples with x ≥ 0.2 were roughly estimated to be ∼2.1 and ∼2.15 eV, respectively. The band gap increase, that is, the absorption edge shifting to lower wavelengths, by Al substitution supports the enhancement of b* (yellowish) value in Al-added samples.

Figure 3

Reflectance curves (top); L*, a*, and b* values (bottom); and photographs (insets in the lower graph) of heat-treated samples with x = 0–0.5. The results obtained for commercially available hematite (CA) are also shown.

We performed microscopic observations to clarify the mechanism responsible for the high lightness and chroma of the Al-added samples in terms of microstructure. Figure shows the SEM images of heat-treated samples with x = 0–0.5. Primary particles with diameters of 70–160 nm were strongly sintered and aggregated in Al-free sample (x = 0, Figure a). Upon increasing the Al molar ratio up to 0.1 (Figure b), fine particles were loosely aggregated to form a crumbling disk-like secondary structure. The disk-like, secondary structure matured at x ≥ 0.2, and its shape and size were not dependent on the Al molar ratio (Figure c–f). The drastic improvement of a* and b* values in heat-treated samples with x = 0.1 and 0.2, as shown in Figure , should be related to the formation of the disk-like structure. Figure shows the detailed microscopic observations of the precursors and heat-treated samples with x = 0 and 0.3. The Al-free precursor consisted of irregularly shaped agglomerates (Figure a), whereas disk-like particles with diameters and thicknesses of 0.4–0.9 and 0.2–0.3 μm, respectively, were observed in the case of the Al-added precursor (Figure b). The shapes of the precursors were preserved after the heat treatment (Figure c,d), indicating that the disk-like structures were formed in the stage of precursor formation.

Figure 4

SEM images of heat-treated samples with x = 0–0.5 (a–f) in 0.1 increments.

Figure 5

Electron micrographs of samples with x = 0 and 0.3. SEM images of precursors with x = 0 (a) and 0.3 (b) and heat-treated samples with x = 0 (c,e) and 0.3 (d). STEM images of the heat-treated sample with x = 0.3 (f–h). The inset of (g) shows a high-resolution TEM image of the surface fine particles. White lines show lattice planes with a d-spacing value of ∼2.78 Å. Elemental mapping images of (i) Fe K and (j) Al K edges for the heat-treated sample with x = 0.3. (k) Merged image of (i,j).

The particle size distributions of the Al-added and subsequently heat-treated sample (x = 0.3) were measured before and after ultrasonic irradiation for various periods of time to determine the dispersibility of the sample (Figure a). Although the initial state [before ultrasonic irradiation: black line (0) in Figure a] shows a broad distribution containing large particle sizes greater than 100 μm, meaning the aggregation of disk-like particles, this distribution changed to a single distribution, with a peak maximum of ∼0.7 μm, which is nearly consistent with the diameter of disk-like particles, and the distribution sharpened with increasing ultrasonic irradiation time. This result indicates that the initial aggregated disk-like particles can easily be dispersed into single particles by a simple treatment. This easily dispersive character can be a great advantage for industrial pigment applications.

Figure 6

(a) Particle size distribution of the heat-treated sample with x = 0.3, measured by ultrasonic irradiation. The numbers in the figure show ultrasonic irradiation times (min). (b) Pore size distributions of heat-treated samples with x = 0 (open circles) and 0.3 (solid circles).

The sizes of primary particles composed of irregularly shaped agglomerates and of disk-like particles estimated from SEM images were 70–160 nm (Figure e) and 20–40 nm (Figure f) for the Al-free and Al-added samples, respectively, after heat treatment. These and the above-mentioned XRD results indicate that the primary particles are hematite. The Al-added sample was observed in more detail using the secondary electron mode in STEM measurements (Figure g) and high-resolution TEM measurements (inset to Figure g). Interestingly, the disk-like particles were not just an assembly of hematite particles, and extremely small particles with diameters of ∼5 nm were dispersed on the surface of the hematite particles. As noted above, the amount of Al substitution, calculated from the lattice parameters, was lower than the initial composition. This result implies that the amount of Al exceeding the solubility limit of Al in hematite exists as fine particles and/or in the amorphous state. Actually, the nanoparticles (indicated by an arrow in Figure g) were confirmed by STEM measurements. Hence, these particles could be an Al-containing oxide. For high-resolution TEM measurements of nanoparticles, as representatively shown in the inset to Figure g, lattice fringes with a d-spacing value of 2.78 Å, which most likely correspond to the (112) plane of γ-Al2O3 (ICSD 99836), were observed. Additional characterization experiments of STEM/EDS analysis (Figure h–k) revealed the spatial distribution of Fe and Al elements. The signal of Fe was detected only from hematite particles with diameters of 20–40 nm (Figure i). On the other hand, the signal of Al was detected from hematite particles as well as from surface nanoparticles with a diameter of ∼5 nm (Figure j). The merged elemental mapping image (Figure k, red: Fe and green: Al) clearly shows the distribution. On the basis of these microscopic experiments, we confirmed that the nanoparticles observed using STEM (Figure g) were crystalline transition alumina such as the γ-phase. Moreover, Al-substitution in the hematite structure was directly evidenced by elemental analysis, in addition to the shrinkage of lattice parameters (Figure ).

The STEM images shown in Figure f,g indicate that the hematite particles were not densely sintered in disk-like particles—and there are many fine spaces between the hematite particles—suggesting a specific porous structure for the disk-like particles. The porous structure of the heat-treated sample with x = 0.3 was analyzed by the nitrogen adsorption method (Figure b), which demonstrated that the sample had a large specific surface area of ∼60 m2 g–1 and numerous mesopores with diameters ranging from 2 to 190 nm. In addition, characteristic mesopores with local peak maxima of ∼20, ∼45, and ∼90 nm were observed, which are presumably related to the interspaces generated by the primary hematite particles inside the disk-like particles and the interspaces generated by the assembly of disk-like particles.

These results show that the bright yellowish-red pigment synthesized in this study is composed of disk-like structures with diameters of 0.4–0.9 μm and thicknesses of 0.2–0.3 μm, which contain Al-substituted hematite particles with diameters of 20–40 nm and crystalline alumina nanoparticles with diameters of ∼5 nm; there are also many pores with diameters of 2–190 nm in the disk-like structures. The following four points could be the possible reasons for realizing high L*, a*, and b* values: (i) the particle size of hematite is small; (ii) the Fe3+ species in hematite are partly substituted by Al3+; (iii) crystalline alumina nanoparticles are tightly attached to hematite particles; and (iv) nanocomposites of hematite/alumina are loosely connected to form disk-like structures. The formation mechanism of the disk-like particles in the precursor has not been clarified yet, but the structure is observed only in the Al-added precursor. Hence, Al addition into the starting solution is a key factor for the formation of disk-like particles.

To study the thermostability, the heat-treated sample with x = 0.3 was reheated at 900–1200 °C for 1 h, and the color change was assessed. For comparison, commercially available hematite was heated in the same manner as a reference. The variations in a* and b* values are shown in Figure and SEM images of the samples after thermostability test are shown in Figure . The a* and b* values decreased as the heating temperature increased for both samples. The main reason for color fading should be the particle growth of hematite, as shown in Figure . For the reference sample, a* and b* values decreased drastically and reached extremely low values of a* = 22.8 and b* = 11.2 at 900 °C, and the color changed to nearly black at 1100 °C. The particle size of the reference sample grew from 60–300 to 120–440 nm at 900 °C and to 160–920 nm at 1100 °C (Figure d–f). On the other hand, for the Al-added sample, although the b* value slightly decreased upon heating at 900 °C, they still remained high (a* = 39.8 and b* = 45.7). These values are much greater than those observed for the as-received reference sample. The disk-like particles were preserved even at 900 °C, and the primary particle size grew from 20–40 nm before reheating to 50–150 nm at 900 °C (Figure a,b); this size is even smaller than that of the as-received reference sample. At high temperatures of 1100 °C, the reference sample showed a nearly black color, while the Al-added sample showed much higher a* and b* values (35.7 and 33.5) to those of the as-received reference sample. After reheating at 1100 °C, disk-like particles disappeared and the particle size grew to 100–400 nm (Figure c), which is larger than that of the as-received reference sample. At 1200 °C, although the a* and b* values were slightly lower than those observed for the as-received reference sample, the Al-added sample even retained the red color, regardless of its large particle size of 190–500 nm, suggesting that additional Al substitution enhanced the red color. To confirm this suggestion, XRD measurements were carried out. Figure a shows the XRD patterns and Figure b shows the variation of lattice parameters for the samples after thermostability test. Diffraction peaks of α-Al2O3 (corundum) were detected in addition to hematite, and the XRD peaks of hematite sharpened concomitantly with the shifting of the hematite peak toward higher angles. These results indicate that the transition-phase of crystalline alumina nanoparticles observed on the hematite surface transformed into the thermodynamically stable α-phase, the crystallite size of hematite grew, and the crystalline alumina nanoparticles reacted with hematite to enhance Al-substitution in hematite during the thermostability test. The lattice parameters, a and c, were both shortened and the value decreased with increasing temperature (Figure b). The additional Al-substitution to hematite should suppress the reduction in color values at high temperatures even though it was accompanied by strong particle growth.

Figure 7

Color variations of the heat-treated sample with x = 0.3 and commercially available hematite (CA) after reheating at 900–1200 °C for 1 h.

Figure 8

SEM images of (a) heat-treated sample with x = 0.3 after reheating at (b) 900 and (c) 1100 °C and (d) commercially available hematite after reheating at (e) 900 and (f) 1100 °C.

Figure 9

(a) XRD patterns and (b) lattice parameters of the heat-treated sample with x = 0.3 after reheating at 1000 °C and 1200 °C. Red dotted vertical lines in (a) were placed at the diffraction peaks of (214) and (300) planes for the sample with x = 0.3 before reheating to visually know the peak shift depending on the reheating temperature.

The red pigment synthesized in this study shows extremely high thermostability in color. This excellent thermostability is because: (i) the particle size of hematite after reheating is much smaller than that of typical hematite, when heat-treated at the same temperature, as the starting hematite particles are quite small (20–40 nm); (ii) excessive particle growth could be suppressed by crystalline alumina nanoparticles on the hematite surface, which act as a sintering inhibitor; (iii) excessive color fading in hematite resulting from particle growth could be partly canceled by the effect of additional Al substitution, which enhances the lightness and chroma in hematite, during reheating. In addition, the pigment showed reasonably good chemical stability in base, salt, organic solvent, and alcoholic solutions, whereas it is slightly unstable in strong acid solutions (see Table S1 in the Supporting Information).

The bright yellowish-red pigment synthesized in this study can be used not only for ceramic applications involving high temperatures, but also for the coloring of cosmetics, glasses, plastics, cars, smartphones, and so on. If this pigment can replace a part of unstable organic pigments, extremely stable red coloring will be realized in terms of weather resistance and durability. In addition, using the numerous pores formed in the disk-like structure as active and/or storage sites could enable application of this material in new application fields—besides coloring—in which hematite has not been used yet.

Conclusions

To obtain a red pigment based on hematite exhibiting high lightness and chroma, an Al-added iron oxyhydroxide was prepared through a wet process; then, the obtained precursor was heat-treated at 700 °C to obtain a hematite-based red powder. The powder sample was evaluated by spectrophotometry, XRD, electron microscopy, and nitrogen adsorption method. The Al-added precursor was basically a low-crystalline iron oxyhydroxide, 2-line ferrihydrite, with some iron-, nitrate-, carbonate-, and/or ammonium-ion-related salts. The Al-added and subsequently heat-treated sample with x = 0.3, which was mainly composed of hematite, showed an extremely bright yellowish-red color with L* = 48.3, a* = 37.2, and b* = 48.8; these values are 1.25, 1.22, and 2.32 times greater than those of commercially available hematite. Microscopic observations combined with the nitrogen adsorption method revealed that the microstructure of the Al-added and subsequently heat-treated sample is composed of micrometer-sized, disk-like structures containing Al-substituted hematite particles and crystalline alumina nanoparticles—and there are many fine pores in these disk-like structures. Such a unique microstructure is strongly related to the bright yellowish-red color. Additionally, Al-added and subsequently heat-treated sample showed excellent thermostability in color, so that the sample retained relatively high values of a* = 35.7 and b* = 33.5, even after exposure to high temperatures of 1100 °C. The obtained bright yellowish-red pigment, which exhibits excellent thermostability, can be used in diverse coloring applications, especially in those needing thermostability, weather resistance, and durability.