Tunable high-performance microwave absorption for manganese dioxides by one-step Co doping modification

The frequencies of microwave absorption can be affected by the permanent electric dipole moment which could be adjusted by modifying the crystal symmetry of the microwave absorbing materials. Herein, we corroborate this strategy experimentally and computationally to the microwave absorption of manganese dioxides. Nanosized Co-doped cryptomelane (Co-Cryp) was successfully synthesized by a one-step reaction. The introduction of Co(III) induced a change of crystal symmetry from tetragonal to monlclinic, which could lead to an increase of its permanent electric dipole moment. As a result, the frequencies of maximum microwave absorption were regulated in the range of 7.4 to 13.9 GHz with a broadened bandwidths. The results suggested that microwave absorption of manganese dioxides can be tailored with Co doping to expand their potential uses for abatement of various microwave pollutions.

properties of materials could affect magnetic and dielectric loss when the materials absorb microwave 20 , it is worth noting that crystal structure and morphologies are important factors to microwave absorption. In the frequency range of 4-14 GHz more than 99% of EM wave energy was attenuated by a synthesized architecture of Fe 3 O 4 nanorod arrays and graphene sheets 8 . Moreover, MnO 2 nanorods prepared with a hydrothermal method reached a minimum reflection loss of 74.1% or − 5.9 dB at 7.57 GHz 21 .
Manganese dioxides (MOs) have excellent microwave absorption properties. They are made of corner-or edge-sharing [MnO 6 ] octahedra mostly forming tunnel-like or layered structures [22][23] . Cryptomelane (Cryp) is a type of MOs made of a 2 × 2 tunnel structure. Cryp is able to produce instantaneous polarization and strong activity centre under the action of microwave [24][25] , form reactive oxygen species, and have strong responses to microwaves 26 . Furthermore, due to the replacement of Mn(VI) by Mn(III) and Mn(II), charge balance is compensated by K + residing in the channels 27 . When different transitional metal ions enter the Cryp structure to substitute for Mn(III), a change of structure, morphology, or other physical properties could occur. Detailed changes could be linked to the types, amounts, and locations of the doped ions 28 . The structure of Co(II) doped Cryp is subject to change depending upon the Co(II) content 29 . With the change of crystal structure, the morphology and frequencies for microwave absorption are expected to change accordingly. In this study, the change of microwave absorption properties after Cryp was doped with different amounts of Co(III) was characterized. A one-step method to synthesis the Co(III) doped Cryp (Co-Cryp) that is able to regulate the microwave absorption frequencies and bandwidth was developed.

Results and discussion
Characterization of Cryp/Co-Cryp structure. The XRD pattern of raw Cryp matched well with the JCPDS 42-1348 (Joint Committee on Powder Diffraction Standards), indicating a tetragonal system with a space group of I4/m (Fig. 1a). The tetragonal structure is made of double chains of edge-sharing [MnO 6 ] octahedra parallel to the c axis, and takes four such double chains with the same vertex angle forming a 2 × 2 tunnel framework 23,27 . As the amount of Co(III) doping increased, the XRD peaks of Co-Cryp became broader (Fig. 1b), and the structural symmetry changed to monoclinic with its XRD patterns matching well with JCPDS 44-1386 ( Fig. 1c) when the amount of Co(III) doping was 3% or more.
Occupancy of Co(III) the Mn sites was confirmed for both tetragonal and monoclinic structures of Co-Cryp in a structure refinement (Fig. 2). Modeling the atomic positions and site occupancies of the two Co-Cryp structures by the Rietveld method resulted in good discrepancy indices of R exp = 8.14, R wp = 11.35, R p = 8.04, GOF = 1.39 (Table 1) and R exp = 8.68, R wp = 11.93, R p = 7.59, GOF = 1.37 (Table 2), suggesting a tetragonal cell with 15% (atom) of Co and a monoclinic cell with 31% (atom) of Co, respectively. In fact, element analysis showed the molar ratios of Co(III) doped into Cryp. As the amount of Co(III) doping increased, more Co(III) was incorporated into the Co-Cryp. For 5-Co-Cryp and 4-Co-Cryp, The amounts of Co(III) were 2.5% and 2.1%, respectively ( Table 3). As the amount of CoCl 3 used in the synthesis of 5-Co-Cryp was two times that for 4-Co-Cryp, the minute difference in Co(III) contents between 5-Co-Cryp and 4-Co-Cryp suggested gradual saturation of Co(III) in 5-Co-Cryp.
The morphology of raw Cryp was nano-fibrous and the particle size was uniform, with a length of tens of micrometers and a diameter about 50 nm (Fig. 3a). As the doping amount of Co(III) increased, aggregates of granular particles micrometers in diameter became dominant (Fig. 3). The diameter of Co-Cryps changed from 50 nm to 860 nm as the amount of Co(III) increased (Fig. 4). In previous studies, the crystal morphology of Cryp changed from nanofibers to short rods or even granules and the degree of partial disorder of the Cryp structure increased 30 , as the doping of transition metal ions increased [31][32] .
The specific surface area (SSA) of Cryp was determined using fully automatic nitrogen adsorption BET. As the amount of Co(III) doped increased, the SSA became smaller with the SSA of Cryp being 62.0 m 2 /g and that of 5-Co-Cryp being 14.9 m 2 /g (Fig. 5).
The Mn2p 3/2 peak was deconvoluated by a multi-component spectral fitting to determine the distribution of various valence states of Mn in Co-Cryp (Fig. 6a-c). With the increase of initial Co input the amount of Mn(II)    Table 3. The elements of Co-Cryp by the method of XRF.
and Mn(IV) increased while that of Mn(III) decreased ( Table 4). The Co(2p) XPS spectra of Co-Cryp showed that the binding energies of Co(2p 1/2 ) and Co(2p 3/2 ) were ca. 795.5 and 780.5 eV for all of the Co-Cryp samples (Fig. 6d). These values matched well with those of Co(III) 33 . In addition, the split of Co(2p 1/2 ) and Co(2p 3/2 ) was 15 eV which was nearly identical to that of Co(III) but 1 eV smaller than that of Co(II) 34 . On the contrary, Co(II)     Co-Cryp/paraffin composites had a higher resistivity. When a fixed amount of EM wave was applied on a material, a small amount of reflected EM wave would indicate an effective attenuation of EM wave energy and the material potentially has a good microwave absorbing property. A smaller coefficient (R) would indicate a larger absorption of the EM wave. The performance of the MAMs can be evaluated by a reflection R defined in the Supplementary Information section. The reflection losses of Cryp-and Co-Cryp/paraffin composites were calculated from equation 3 (SI) and Cryp/Co-Cryp accounted for 20% and paraffin for 80% of the total losses in Cryp-or Co-Cryp-paraffin composites (Fig. 7). Thus, the maximal attenuation of microwaves by the composites occurred as a function of both specimen thickness and microwave frequency (Fig. 7). With increasing specimen thickness, the maximal attenuation occurred at lower frequency (smaller f m value) (SI: equation 3). In the thickness range considered, the width of frequency in which maximal microwave attenuation could be achieved by Co-Cryp was much greater than that by Cryp and it shifted towards higher frequency with increasing Co(III) doping. On the other hand, with increasing specimen thickness, the microwave absorption properties also increased.
Changes of EM absorption properties may stem from the phase transition of the MAMs 37-39 . The maximal reflection losses by the Co-Cryp composites mostly had less negative decibel values than − 10 dB (i.e., 90% of the incoming power was absorbed), in comparison to − 35.4 dB (> 99% power absorbed) for the Cryp composite at 3 mm composite thickness (Fig. 7a). An increase of Co(III) doping resulted in a phase transition from tetragonal to monoclinic structure and a decrease (less negative decibel value) of reflection loss and energy absorption (Fig. 8a). However, the frequency and bandwidth for EM attenuation became larger as the Co(III) doping increased (Fig. 8b). This feature suggests potentially broader uses of the functional materials for EMP attenuation 40 .
Microwave absorption was attributed to magnetic loss and dielectric loss 41 . Permittivity mainly originates from polarizations of electrons, ions, and intrinsic electric dipoles. Magnetic properties could be affected by crystal structure, special geometrical morphology and size [42][43] . Complex permittivity and complex magnetic permeability of Co-Cryp were analyzed (Fig. 9). The real (ε ′ ) and imaginary (ε ″ ) complex permittivity, and the real (μ ′ ) and imaginary (μ ″ ) permeability are displayed in Fig. 9a-d. As the amounts of Co(III)-doped increased, the ε ′ became smaller, suggesting gradual weakening of the dipole polarization and electric polarization of Co-Cryp. Because ε ′ represents material characterization of polarization. It could clearly be seen the values of ε ″ exhibited very complex nonlinear behaviour 44 , and also decrease with the increase of Co(III) doping in Cryp. As shown in Fig. 7b, there were relaxations of Cryp. Relaxation I, was located at the frequencies of 7.1, 9.4, and 13.5 GHz for raw Cryp, 1-Co-Cryp, and 2-Co-Cryp, respectively. The relaxation of Cryp was mainly from interfacial polarization and defect dipole polarization 39,45,46 . As the Co(III) doping increased, the frequency of relaxation I shifted to higher frequencies. Compared with the complex permittivity, the changes in complex magnetic permeability were minute and close to zero. Thus, the complex permittivity was the key factor affecting microwave absorption of Cryp.      (Table 5). Accordingly, the anisotropicity increased 20 and the morphology changed from nano-fibers to equidimentsional granules of micrometers in size [49][50][51] (Figs 3 and 4). Nanosize materials tend to have better microwave absorption properties [52][53][54][55] . As evidenced in this work, the nanosize Cryp fibers showed greater EM absorption than the micrometer-sized Co-Cryp. Moreover, the permanent electric dipole moment increased from 0 to 3.3792 × 10 −29 C·m per [MnO 6 ] octahedron with the increase of anisotropicity ( Table 6). The permanent electric dipole moment is closely related to the composite magnetic permeability and complex dielectric constant of materials (Equations 2-4 in SI), and could cause different f m values and absorption bandwidths [56][57][58][59] as is the case in this study. With an increase in the amount of Co(III) doping, a right shift of f m occurred and the bandwidth of maximal microwave attenuation became significantly larger, albeit with a slight decrease of absorption performance. These features make it optimistic to use Cryp or Co-Cryp as MAMs in a wide range of environments for potential microwave abatement [60][61][62] .

Conclusions
Cryp was demonstrated to be a very good microwave absorption material. The micowave reflection loss of Cryp reached to − 35.4 dB. As the amount of Co(III) doped increased, the bandwidth of high attenuation notably expanded and the frequencies shifted significantly to 13.9 GHz. In a nutshell, optimal microwave absorption frequency and bandwidth of Cryp materials can be tuned with a control of crystal structure via selective Co doping.

Experimental Section
Experimental materials. KMnO   Preparation of Cryp/Co-Cryp-paraffin composites. The composites for microwave absorption measurement were fabricated by mixing Cryp or Co-Cryp (20%) with paraffin matrix (80%). The mixtures were pressed into a cylindrical shape with Φ out = 7.00 mm and Φ in = 3.04 mm, respectively. The reflection losses of the composites should be attributed to that of the Cryp or Co-Cryp as the reflection loss of paraffin matrix was similar to that of the air.

Methods of analyses.
The powder X-ray diffraction (XRD) patterns of Cryp and Co-Cryp were recorded with a CuKα radiation at 40 kV and 100 mA, a scanning speed of 8° 2θ /min with 0.02° per step. To further investigate the changes in crystal structure of Cryp, Rietveld refinement was performed using the Topaz 3.0 program with the structural parameters of Cryp used as initial parameters.
The purities of the final products were carefully checked by XRD analysis from 5° to 90°.
Step scan performed in structural analysis at a rate of 2.35 s/step and a step size of 0.02. TOPAS package was deployed to calculate the structural details in the Rietveld refinement.
Elemental composition of Cryp was determined by X-ray Fluorescence spectrometry (XRF). Micro textures of the Cryp and Co-Cryp precipitates were observed on an FEI Quanta 250 field-emission environmental scanning electron microscope (ESEM) with a voltage of 15 kV. Samples were prepared by drying thick sample suspension on a silicon slide before the ESEM observation. Detailed microstructures and crystallite sizes were analyzed by high-resolution transmission electron microscopy (Model: JEOL JEM-2010F).   An X-ray photoelectron spectroscopy (XPS, Thermo Scientific Co., Ltd.) was used to determine the surface valence of Cryp after being doped with different quantities of Co(III) under a monochromatic CuKα source at 150 W and a base pressure of 6.5 × 10 −10 mbar in the measuring chamer.
The permittivity and permeability of the composites were measured by a coaxial wire method in the 2-18 GHz range using a phasor network analyzer PNA N5244A (Agilent).
The formula for the calculation of the geometric centre follows that for the calculation of the permanent electric dipole moment 64