Effect of [MnO6] Octahedra to the Coloring Mechanism of (Li1-x Na x )2MnO3.

(Li1-x Na x )2MnO3 (0 ≤ x ≤ 0.10) solid solutions were synthesized by a conventional solid-state reaction technique to investigate the relationship between the steric structure of the [MnO6] octahedra and coloration mechanisms. The color, optical properties, and crystal structure of the solid solutions were characterized. The (Li1-x Na x )2MnO3 (0 ≤ x ≤ 0.10) solid solutions absorbed the visible light at wavelengths shorter than 550 nm and around 680 nm. The former and latter optical absorption bands were attributed to the spin-allowed (4A2g → 4T1g, 4T2g) and spin-forbidden (4A2g → 2Eg, 2T1g) d-d transitions of tetravalent manganese ions, respectively. The absorption band assigned to the 4A2g → 4T2g transition shifted toward longer wavelengths with the enlargement of the average [Mn(2)O6] bond distance by doping Na+. In contrast, the latter absorption bands did not shift but the absorption intensities increased due to the distortion of the [Mn(2)O6] octahedra. Consequently, the red color purity of the sample gradually increased with the increase in the Na+ concentration. Among the (Li1-x Na x )2MnO3 (0 ≤ x ≤ 0.10) samples synthesized in this study, the highest red color purity was obtained in the (Li0.93Na0.07)2MnO3 (hue angle: h° = 39.1) sample. The results of this study provide important insights for the development of environment-friendly inorganic red pigments containing Mn4+ ions as a coloring source.

Introduction

Inorganic pigments consisting of metal oxides are applied in a wide range of fields, such as ceramics, paints, and plastics, because of their high hiding powers and fine coloring properties.[1] In particular, red inorganic pigments are much in demand due to their high visibility. Some red pigments such as lead oxide red (Pb3O4), mercuric sulfide red (HgS), and cadmium red (CdS·CdSe) have been popularly used as industrial inorganic color materials. However, these pigments contain elements (e.g., Pb, Hg, Cd) that are toxic to the environment and the human body. Accordingly, the use of the conventional pigments composed of these harmful elements is tended to be regulated or banned on a global scale. In particular, many nontoxic red pigments have been developed by several researchers to replace the harmful pigments with the environment-friendly ones.[2−11]

Because of this situation, we focused on tetravalent manganese (Mn4+) ion as a source of red color. Recently, Mn4+ has been investigated as an activator for the red-light-emitting phosphors due to low cost and nontoxicity.[12−20] K2XF6:Mn4+ (X = Si, Ge, Ti), Li2TiO3:Mn4+, and CaAl12O19:Mn4+ have been reported as examples of Mn4+-doped phosphors.[18−20] These phosphors absorb the visible light in the wavelength ranges from 350 to 550 nm, which is attributed to the electronic transition between two 3d orbitals (t2g and e2g) of Mn4+. The optical absorption band due to the d–d transition is influenced by the crystal field around the Mn4+ ions. The Mn4+ content is controlled approximately by 1 mol % to avoid concentration quenching in phosphors. However, it is considered that the visible light in the region from 350 to 550 nm is strongly absorbed by further increasing the Mn4+ concentration and coloring of the sample can be recognized. Since the visible light around 550 nm corresponds to green light, it is expected that a reddish color, which is a complementary color of green, will be obtained. For these reasons, Mn4+ is a promising coloring source to develop environment-friendly inorganic red pigments.

In this study, we focused on Li2MnO3 as a host material, and the components are nontoxic elements. It has been reported that the color of this material is brick-red or orange-red due to the optical absorption in the wavelength range between 350 and 550 nm, corresponding to the d–d transition of Mn4+ ions.[21,22] However, the red color purity of pure Li2MnO3 is not enough, and it is necessary to enhance the red color purity to make it a red pigment. Because the d–d transitions in transition metal cations are strongly influenced by the crystal field around them, the control of the optical absorption due to the d–d transition is possible by adjusting the crystal field energy. The d–d transition band shifts to the lower energy (i.e., longer wavelength) side, when the crystal field around the chromophore ions becomes weak due to the introduction of larger cations into the host material. Therefore, a more reddish color than that of pure Li2MnO3 will be obtained by doping larger cations. To enhance the red color purity of Li2MnO3, we selected Na+ (ionic radius: 0.116 nm)[23] as a dopant because this is a nontoxic element and has the same valence with Li+ (ionic radius: 0.090 nm).[23] Namely, Na+-doped Li2MnO3 samples, (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10), were synthesized by a solid-state reaction technique and the influences of differences in the geometric structure around the Mn4+ ions on the color properties were investigated.

Results and Discussion

X-ray Powder Diffraction (XRD) and Field Emission–Scanning Electron Microscopy (FE-SEM) Image

The XRD patterns of the synthesized (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples are shown in Figure . The XRD pattern of Li2MnO3 (No. 01-073-0152) from the inorganic crystal structure database (ICSD) is also depicted as a reference on the bottom. A single-phase Li2MnO3 structure was observed for all samples, and no diffraction peaks of other phases or impurities were detected in the patterns. The diffraction patterns of the samples corresponded to that of the monoclinic Li2MnO3 from ICSD. The lattice volumes of the (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples were calculated from the diffraction angles in the XRD patterns. The compositional dependence of the lattice volume for the (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) solid solutions is shown in Figure . The cell volume increased with the increasing Na+ concentration, which indicated that some Li+ (ionic radius: 0.090 nm)[23] ions in the Li2MnO3 structure were substituted with larger Na+ (ionic radius: 0.116 nm)[23] ones. These results indicate that the solid solutions of the monoclinic Li2MnO3 phase were successfully formed.

Figure 1

XRD patterns of the (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples.

Figure 2

Composition dependence of the cell volume for the (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples.

The crystal structure of Li2MnO3 has been investigated using the Rietveld method by many researchers.[24−29] Li2MnO3 has a layered rock-salt-type structure. It adopts a monoclinic unit cell with the C2/m (No. 12) space group. This layered structure is composed of Li+ (2c and 4h sites) and Li+/Mn4+ (2b and 4g sites) layers. These layers were alternately stacked via oxide anions. It has been reported by Boulineau et al. that Li+ and Mn4+ ions in the Li+/Mn4+ layer mainly occupy the 2b (Li-rich) and 4g (Mn-rich) sites, respectively, but the partial exchange between Li and Mn is possible on the 2b and 4g sites.[24]

The Rietveld structure refinement of the (Li1–Na)2MnO3 (x = 0 and 0.07) samples was carried out to investigate the geometric structure around the Mn4+ ions. Figure shows the Rietveld refinement profiles of the Li2MnO3 and (Li0.93Na0.07)MnO3 samples. The detailed crystallographic and structural parameters are tabulated in Tables and 2, respectively. Figure displays the crystal structure of Li2MnO3 depicted by the VESTA program on the basis of the crystallographic parameters obtained from the Rietveld refinement.[30] As shown in Figure , the broadening of the peaks consisting of reflections was observed in the range from 2θ = 20 to 25°, corresponding to the [√3ahex × √3ahex] superstructure caused by a Li(Na)/Mn ordering arrangement.[24] Therefore, the refinement was conducted in the 2θ ranges except in this region. The low R-factors were obtained for both (Li1–Na)2MnO3 (x = 0 and 0.07) samples. As shown in Table , almost all of the Na+ ions doped into the host lattice were located at the Li(1) site, and this result was in good agreement with the structure of (Li0.95Na0.05)2MnO3 reported by Dong et al.[31]

Figure 3

Observed intensities (black crossed symbol) and calculated patterns (solid red line) for the Rietveld structural refinement from the XRD data of Li2MnO3 (a) and (Li0.93Na0.07)2MnO3 (b). The green vertical bars represent the Bragg reflection peak positions. The bottom blue line indicates the difference curve between the observed and the calculated patterns.

Table 1

Crystallographic Parameters of (Li1–Na)2MnO3 (x = 0 and 0.07) Obtained by Rietveld Structural Refinement Analysisa

	x = 0	x = 0.07
Lattice Parameter
a (nm)	0.49286(3)	0.493658(12)
b (nm)	0.85311(3)	0.85404(2)
c (nm)	0.50222(2)	0.503443(11)
β (deg)	109.236(4)	109.4020(14)
V (nm3)	0.19938(2)	0.200200(8)
R-Factor
Rwp	1.275	1.683
Rp	0.873	1.061
Re	0.694	0.958
S	1.837	1.757
RF	5.265	5.668

Crystal symmetry: monoclinic, space group: C2/m (No. 12), number of formula units per unit cell: Z = 4.

Table 2

Structural Parameters of the (Li1–Na)2MnO3 (x = 0 and 0.07) Samples Refined by the Rietveld Method for the XRD Patterns Obtained at Room Temperaturea

atom	site	occupancy (g)	multiplicity × g	x	y	z	Biso (Å2)
Li2MnO3 (x = 0)
Li1	2b	0.618b	1.236	0	1/2	0	1.0
Mn1	2b	0.382(5)	0.764(10)	0	1/2	0	1.0
Li2	2c	1	2	0	0	1/2	1.0
Li3	4h	1	4	0	0.680(2)	1/2	1.0
Mn2	4g	0.809b	3.236	0	0.1642(4)	0	0.5
Li4	4g	0.191b	0.764	0	=y(Mn2)	0	0.5
O1	4i	1	4	0.215(2)	0	0.2247(13)	0.8
O2	8j	1	8	0.2471(14)	0.3197(6)	0.2188(7)	0.8

The values of the isotropic atomic displacement parameter (Biso) of the lithium, manganese, and oxygen sites were fixed at 1.0, 0.5, and 0.8 Å2, respectively, with reference to the literature.[24,25,29]

The occupancies (g) of 2b and 4g sites in Li2MnO3 were linearly constrained to be the stoichiometric composition; g(Li1) = 1 – g(Mn1), g(Li4) = 0.5 × g(Mn1), and g(Mn2) = 1 – g(Li4).

The occupancy factors of 2b, 4h, and 4g sites in (Li0.93Na0.07)2MnO3 were also linearly constrained: g(Li1) = 0.72 – g(Mn1) + 2 × g(Na2), g(Li3) = 1 – g(Na2), g(Li4) = 0.5 × g(Mn1), g(Mn2) = 1 – g(Li4), and g(Na1) = 0.28 – 2 × g(Na2).

Figure 4

Crystal structure of Li2MnO3 (Li: yellow, Mn: violet, O: red) refined by the Rietveld method.

The average Mn(2)–O bond distance and distortion index (D) of the [Mn(2)O6] octahedra in both (Li1–Na)2MnO3 (x = 0 and 0.07) samples are tabulated in Table . The numbers in parentheses indicate the standard deviation. A distortion index, D, based on bond lengths was calculated with the formula[32]where l is the distance between the central atom and the ith coordinating atom and lav indicates the average bond length. The average bond distance between Mn(2) and O increased as the Na+ content increased. The increase in the D value indicates that the distortion of the octahedron has become larger. For the [Mn(2)O6]octahedra in the (Li1–Na)2MnO3 (x = 0 and 0.07) samples, the D value increased twofold by Na+ doping, as shown in Table . These results indicate that the [Mn(2)O6] octahedra were distorted by the introduction of Na+ into the lattice.

Table 3

Average Bond Length of Mn(2)–O and Distortion Index (D) of the [Mn(2)O6] Octahedra in the Refined (Li1–Na)2MnO3 (x = 0 and 0.07) Structure

x	average Mn(2)–O bond length (nm)	D
0	0.1900(2)	0.0068
0.07	0.1911(2)	0.0154

Figure shows the FE-SEM photographs of the Li2MnO3 and (Li0.93Na0.07)2MnO3 samples. In the undoped Li2MnO3 (x = 0) sample, the primary particles aggregated and thermally fused to form the spherical secondary particles with lumpy surfaces. In contrast, the faceted primary particles, which had wide and flat surfaces, were observed for the (Li0.93Na0.07)2MnO3 sample. In addition, the primary particle size for the sample with x = 0.07 was about 12 μm and almost equal to the secondary particle size of the undoped one.

Figure 5

FE-SEM images of Li2MnO3 (a) and (Li0.93Na0.07)2MnO3 (b).

Diffuse Reflectance Spectra

The ultraviolet–visible (UV–vis) reflectance spectra for the (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples are shown in Figure a. An enlarged view from 400 to 550 nm is also depicted in Figure b. Optical absorption at the wavelength shorter than 550 nm was observed for all samples. The absorption band around 400 nm was attributed to the d–d transition of Mn4+, which was assigned to the spin-allowed 4A2g → 4T1g transition, according to the Tanabe–Sugano diagram.[33] The absorption in the wavelength region between 400 and 550 nm corresponded to the spin-allowed 4A2g → 4T2g transition.[21] In addition to these optical absorptions, the absorption bands corresponding to the spin-forbidden 4A2g → 2Eg, 2T1g transitions were observed in the range of 650–710 nm.[21]

Figure 6

UV–vis reflectance spectra (a) and enlarged spectra (b) of the (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples.

In the visible-light region from 400 to 700 nm, the light scattering capacity of a material tends to increase as the particle size decreases.[34] Accordingly, Li2MnO3 is expected to scatter more visible light than (Li0.93Na0.07)2MnO3 because the primary particle size of the former is smaller than that of the latter. Although such a result was obtained at wavelengths longer than 550 nm in the reflectance spectra, there was no significant change in the wavelength range from 400 to 550 nm, as shown in Figure b. Therefore, the effect of the particle morphology on the optical properties is quite small, and the change in the reflectance is due to the difference in the coordination environment around the Mn4+ ion caused by the substitution of Li+ with Na+ in the lattice.

As seen in Table , in both of Li2MnO3 and (Li0.93Na0.07)2MnO3, the amount (= multiplicity × g) of Mn(2) ions in the unit cells was significantly larger than that of Mn(1) ions. Therefore, the absorption wavelength of the spin-allowed 4A2g → 4T2g transition predominantly depended on the coordination environment at the Mn(2) site. The absorption band attributed to the 4A2g → 4T2g transition shifted to the longer wavelength side with Na+ doping because the crystal field around the Mn4+ ion was weakened by the introduction of Na+, which was larger than Li+. In fact, the average Mn(2)–O bond distance for the Na+-doped sample at x = 0.07 was longer than that for the undoped sample (x = 0), as shown in Table .

In contrast, the optical absorption bands corresponding to the spin-forbidden 4A2g → 2Eg, 2T1g transitions appeared more intensely with the increasing Na+ content, but no peak shift was observed. As already explained for the results in Table , the [Mn(2)O6] octahedra in the lattice were more distorted when some Li+ ions were substituted with Na+ ions. Thus, the 4A2g → 2Eg, 2T1g transitions were partially allowed because of the symmetry reduction caused by the enlargement of distortion. According to the Tanabe–Sugano diagram for d3 electron configuration, the energy levels of 2Eg and 2T1g are hardly affected by the crystal field strength. Therefore, the absorption intensity of the 4A2g → 2Eg, 2T1g transitions of Mn4+ increased due to the loss of symmetry, but the absorption wavelength hardly depended on the crystal field energy around the Mn4+ ion.

Color Properties

The L*a*b*h° chromatic parameters for the (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples are summarized in Table . The sample photographs are also shown in Figure . The values of the brightness (L*), redness (a*), and yellowness (b*) decreased as the concentration of Na+ increased. These relationships were attributed to the differences in the absorption wavelengths and the intensities of the d–d transitions. As already discussed with respect to the results in Figure , the yellow–orange light (580–605 nm) was absorbed because the optical absorption band around 550 nm was shifted to the longer wavelength side by Na+ doping. Additionally, the red-light reflection in the range from 605 to 700 nm was also reduced. As a result, all values of L*, a*, and b* decreased. However, the hue angle (h°) became smaller by Na+ doping and the color of the samples changed from orange to deep red. Among the (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples obtained in this work, (Li0.93Na0.07)2MnO3 exhibited the lowest h° value, and the sample was, namely, reddish in color.

Table 4

L*a*b*h° Chromatic Parameters for the (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) Samples

x	L*	a*	b*	h°
0	51.5	+30.8	+44.0	55.0
0.04	43.1	+30.7	+31.3	45.6
0.06	40.7	+25.3	+24.3	43.8
0.07	38.3	+24.8	+20.2	39.1
0.08	36.5	+23.3	+19.8	40.3
0.10	33.0	+22.1	+19.0	40.7

Figure 7

Photographs of the (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples.

Chemical Stability and Humidity Resistance Tests

The chemical stability and humidity resistance of (Li0.93Na0.07)2MnO3 were also tested using a powder sample. To assess the chemical stability, the sample was soaked into 4% CH3COOH and 4% NH4HCO3 aqueous solutions. After leaving them at room temperature for 3 h, the samples were washed with deionized water and ethanol. Finally, the samples were dried at room temperature. The humidity resistance of the (Li0.93Na0.07)2MnO3 sample was tested in a thermohygrostat at 90% relative humidity (RH) and 80 °C for 24 h. The sample color after the leaching and humidity resistance tests was evaluated with a calorimeter. Unfortunately, a slight color deterioration was observed after the leaching and humidity resistance tests, as summarized in Table . To suppress the color degradation, therefore, it is necessary to cover the surface using an inert substance such as silica.

Table 5

Color Coordinates of (Li0.93Na0.07)2MnO3 Before and After the Chemical Stability and the Humidity Resistance Tests

treatment	L*	a*	b*	h°
none	38.3	+24.8	+20.2	39.1
4% CH3COOH	35.2	+21.8	+16.5	37.1
4% NH4HCO3	38.1	+23.2	+17.7	37.3
80 °C, 90%RH	35.2	+23.4	+18.7	38.6

Conclusions

(Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples were synthesized by a solid-state reaction process. The Rietveld analysis of the (Li1–Na)2MnO3 (x = 0 and 0.07) samples indicated that the average Mn(2)–O bond distance was increased, and the [Mn(2)O6] octahedra were distorted by doping Na+ into the Li+ site. The spin-allowed 4A2g → 4T2g transition observed at the wavelength shorter than 550 nm shifted to the longer wavelength side due to the decrease of the crystal field strength around Mn4+. No shift was observed in the spin-forbidden 4A2g → 2Eg, 2T1g transitions around 680 nm, but the absorption intensities increased with the increase in the distortion of the [Mn(2)O6] octahedra. Consequently, the sample color gradually turned from orange to deep red with increasing Na+ concentration. This study proposes that the coordination environment around Mn4+ is one of the important factors to develop environmentally friendly inorganic red pigments containing Mn4+.

Experimental Section

Materials and Methods

The (Li1–Na)2MnO3 (0 ≤ x ≤ 0.10) samples were synthesized by a conventional solid-state reaction technique. Stoichiometric amounts of Li2CO3, MnO2, and Na2CO3 were mixed with 5 cm3 of ethanol in an agate mortar to obtain 1 g of the final product. The mixtures were heated in an alumina crucible at 1050 °C for 6 h in the atmosphere. Before characterization, the sample was ground in an agate mortar.

Characterization

The crystal structure of the samples was identified by X-ray powder diffraction (XRD; Rigaku, Ultima IV) with Cu Kα radiation (40 kV and 40 mA). The data were collected by scanning over in the 2θ range of 20–80°, and the sampling width was 0.02°. The lattice volumes of the samples were calculated from the diffraction angles, which were refined using α-alumina as a standard and CellCalc Ver. 2.20 software. Rietveld refinement of the resulting XRD patterns in the 2θ range of 10–120° was performed by the RIETAN-FP software package to determine the precise crystal structure and investigate the coordination environment around Mn4+ ions for the (Li1–Na)2MnO3 (x = 0 and 0.07) samples.[35] From the Rietveld refinement, the following final R-factors were obtained: weighted pattern R-factor (Rwp), pattern R-factor (Rp), R-expected factor (Re), R-structure factor (RF), and goodness-of-fit indicator (S).

The particle size and morphology of the Li2MnO3 and (Li0.93Na0.07)2MnO3 samples were observed with a field emission–scanning electron microscope (FE-SEM; JEOL, JSM-6701F). The optical reflectance spectra were measured using an ultraviolet–visible–near-infrared (UV–vis–NIR) spectrometer (JASCO, V-770) with a standard white plate as a reference. The chromatic properties of the samples were estimated in terms of the Commission Internationale de l’Éclairage (CIE) L*a*b*h° system using a colorimeter (Konica-Minolta, CR-300). The L* parameter expresses the brightness or darkness of a color based on neutral grayscale, and the a* (the red–green axis) and b* (the yellow–blue axis) parameters indicate the color quantitatively. The value of h° (hue angle) ranges from 0 to 360° and is calculated with the formula, h° = tan–1(b*/a*). When h° is in the region of 0 ≤ h° ≤ 35, the color of the sample is red.