Comprehensive Study of Structural, Optical and EPR Properties of Mn²⁺ Doped CdSe Nanocrystals

Highly undoped CdCl₂ powder, elemental Selenium (≥ 99 % Purity, Merck), Ethylene glycol and Hydrazine hydrate (≥ 99 % Purity, Rankem) were used without further purification in the typical synthesis of CdSe NCs by chemical route. MnCl₂ (≥ 99 % Purity, Sigma – Aldrich) was used as doping reagent for Mn²⁺ ions. In this synthesis process, CdCl₂ (16.0 g) was taken with deionised water in 200 ml capacity conical flask. The aqueous solution of CdCl₂ was divided into three equal parts. Remaining one out of these three parts, two parts were doped by weight 0.25% and 0.75% respectively of Mn²⁺ ions by adding drop wise aqueous solution of MnCl₂ with continuous stirring. Further, elemental selenium (8.0 g) was taken with deionised water, ethylene glycol and hydrazine hydrate in the volume ratio of 7:3:1 in a 200 ml capacity conical flask. This solution of elemental selenium was also divided into three equal parts. Now the three parts of the solution of elemental selenium were added separately with three parts of aqueous solution of CdCl₂ containing 0%, 0.25% and 0.75% Mn²⁺ ions. Then, these solutions were refluxed under vigorous stirring at 60^○ C for 6 hours. Three brown precipitates were collected and washed with anhydrous ethanol and hot distilled water several times. Afterwards, these were dried in vacuum at 50^○ C for 5 hrs to avoid agglomerations.

The exact mechanism for the formation of CdSe NCs is still unclear, but it is reasonably concluded that the appropriate ratio of volume of solvents may play the crucial role for the formation of CdSe NCs. In growth mechanism of the CdSe NCs Se source can be easily converted into Se^2- by hydrazine (N₂H₄), which results in a high monomer concentration. In the initial step, hydrazine hydrate (N₂H₄.H₂O) complexes with Cd²⁺ and forms the transparent soluble complex, which effectively decrease the concentration of Cd²⁺ and avoid the precipitation of CdSeO₃, and thus provides a more homogeneous solution environment for the reaction. The chemical reaction involved in the entire synthesis of CdSe NCs could be formulated as:

So, the application of N₂H₄ as the coordination agent is determinable for the obtained phase of the product. Thus, clearly the complexing ability of group containing atom N (such as NH₂ or NH₃) can effectively determine the final phase of the products. In comparison to the CdO deposit, it is easier for Cd(OH)₃^- to release Cd^2+, which can facilitate growth of NCs under nonequilibrium kinetic growth conditions with a high monomer concentration. A similar phenomenon was found during preparation of PbSe and Cu₂Te NCs using N₂H₄.H₂O as complexing agent and the exact mechanism was fully understood ^{22, 23}.

XRD was performed on Proto A-XRD diffractometer using CuKα1 radiation with wavelength 1.54Å in wide angle region and the spectra were used to determine the lattice parameter, crystallite size and phase identification. The structural and morphological analysis of Mn²⁺ doped CdSe NCs were done by HRTEM and SAED patterns using a Tecnai G2, F30 field emission gun - transmission electron microscope operating at 300 kV accelerating voltage with resolution, Point: 2.0 Angstrom Line: 1.0 Angstrom and magnification 58x to 1 million x. SEM photographs and EDS spectra were recorded on the JEOL JSM-7600F FEG-SEM with SEI Resolution: 1.0 nm at 15 kV and 1.5 nm at 1 kV, in GB mode, Magnification: Low: 25X to 10,000X, High: 100X to 1,000,000X at 4x5 photo size, Accelerating Voltage: 0.1 to 30 kV and Probe Current Range: 1 pA to ≥ 200 nA. UV/ Vis absorption spectra were recorded on a Unicam– 5625 spectrophotometer using sample in distilled water, which were used to determine the direct energy band gap. PL spectra of undoped and Mn²⁺ doped CdSe NCs with excitation wavelength 360 nm at room temperature were recorded on Perkin Elmer LS55 Luminescence Spectrometer. EPR spectra were recorded on JSE – FA200 ESR Spectrometer (ESR-JEOL, Japan) at the frequency 9.451878 GHz, field modulation 100 kHz, Sensitivity: 7×109 spins / 0.1 mT, Resolution: 2.35(micro)T or better at room temperature.

Figure 1(A) presents the XRD pattern of CdSe NCs with different Mn²⁺ ion concentrations at room temperature. The four characteristic broad peaks of undoped and Mn²⁺ doped CdSe NCs corresponding to the (111), (220), (311) and (331) lattice planes are almost located at 2θ values 25.36^ο , 42.28^○ , 49.60^ο and 67.18^o as for standard cubic CdSe (Joint Committee on Powder Diffraction Standards (JCPDS) card No. 19-0191) (a = 6.077 Å and space group:𝐹43𝑚).The diffraction peaks for undoped and Mn²⁺ doped CdSe NCs are at same values of 2θ. This suggests that the Mn²⁺ ions are introduced in the lattice of CdSe. Here, Mn²⁺ ions are present as dopant, not as alloy. Because Mn²⁺ doped CdSe NCs were synthesised at 50^○C with low concentration of Mn²⁺ ions (weight percentages 0.25% and 0.75%), the alloying effect is absent. The temperature 270^○C has been defined as the ‘‘alloying point’’ analogous to melting or boiling points for the CdSe/ZnSe and CdSe/ZnS systems ²⁴. Therefore, Mn²⁺ doped CdSe NCs have the same lattice parameters as core shell ²⁵. As expected, the XRD peaks of CdSe were considerably broadened as compared to bulk material due to small size of the NCs. The crystallite size of the CdSe NCs was estimated to be about 3 - 6 nm using Scherer’s formula ²⁶:

Figure 1.(A) XRD spectra of 0%, 0.25%, and 0.75% Mn2+ doped CdSe NCs at room temperature

where D is the crystallite size, β_hkl is the full - width at half - maximum (FWHM) of the diffraction peak measured in radian, λ (1.54 Å) is the wavelength of X-ray CuKα1 radiation, and θ is the angle of diffraction.

Lattice constant of all cubic CdSe NC samples were calculated by using formula ^{27, 28}:

where, 𝑎 is lattice constant, h, k, l are Miller indices and d is the interplanar spacing which was calculated using Bragg’s formula:

where, θ is the angle between the incident beam and the reflection lattice planes, n = 1 for first order. Corresponding to the planes (111), (220), (311) and (331), the calculated values of d- spacing, lattice constant 𝑎 and crystallite size for cubic phase 0%, 0.25% and 0.75% Mn²⁺ doped CdSe NCs are given in Table 1

Furthermore, with powder, the peak broadening is due to a linear combination of the contribution from nanocrystalline nature and local strain in the crystal structure from defects ^{29, 30}. The strain ε induced in powders due to crystal imperfection and distortion can be calculated using the formula:

The strain ε and crystallite size D of the Mn²⁺ doped CdSe NCs were estimated by Williamson – Hall (W – H) method ³⁰. The observed FWHM of diffraction peaks is simply the sum of the Eqs.(6) & (9) which results in the Williamson – Hall Eq. (10).

In Figure 1(B) the values of 𝛽_hkl𝑐𝑜𝑠𝜃 on y-axis were plotted as a function of 4sin𝜃 on x-axis. From the linear fit of data, the crystallite size D was estimated from the y-intercept, and the strain ε, from the slope of the linear fit. In this study, the crystallite size D obtained at zero strain were about 3.21, 4.91 and 5.09 nm with different Mn²⁺ concentrations (0 – 0.75%), respectively. These crystallite sizes are smaller than the exciton Bohr radius of CdSe 𝛼_B~5.6𝑛𝑚 ³¹ and increase as the concentration of Mn²⁺ ions increases in CdSe NCs. The result indicates that the quantum-size effect appears in all the samples. The strain values ε were found to be -0.01538, -0.02417 and - 0.01538 for 0%, 0.25% and 0.75% Mn²⁺ doped CdSe NCs, respectively. The negative sign means that the Mn²⁺CdSe NCs have a compressive strain in all the samples ²⁹. The appearance of the compressive strain is due to the substitution of Cd²⁺ (0.92 Å) ions with Mn²⁺ (0.67 Å) ions of the smaller radius in the CdSe host. In CdSe host the Mn-Se bond length (2.364 Å) is slightly smaller than the Cd-Se bond length (2.597Å) which leads to a lattice disorder and the formation of strain in the CdSe lattice ³².

Figure 1.(B) W-H plots of CdSe: Mn (0%, 0.25% and 0.75%) NCs

Figure 2 shows transmission electron microscopy (TEM) images, high-resolution transmission electron microscopy (HRTEM) micrographs and selected area electron diffraction (SAED) patterns of 0%, 0.25% and 0.75% Mn²⁺ doped CdSe NCs. HRTEM was used for structural and morphological analysis of undoped and Mn²⁺ doped CdSe NCs. TEM images of undoped and Mn²⁺ doped CdSe NCs are shown in Figure 2(A, B &C) which indicates that the average size of undoped and doped CdSe NCs lies between 5 - 15 nm and increases with increase in doping concentration of Mn²⁺ ions in CdSe NCs. The difference between XRD and TEM data may be due to:

(i) there may be more than one crystallite in single grain;

(ii)there may be some agglomeration of the particles ³³.

Figure 2.TEM images (A-C), HRTEM micrographs (D-F) and SAED patterns (G-I) of 0%, 0.25%,and 0.75% Mn2+ doped CdSe NCs, respectively

HRTEM micrographs of undoped and Mn²⁺ doped CdSe NCs are given in Figure 2(D, E & F) which shows that particles are well defined. The lattice fringes visible in the HRTEM micrographs of a single CdSe NC are indicative of the crystalline nature of the particles. The lattice spacing ~3.5 Å corresponds to (111) planes and ~2.1 Å corresponds to (220) planes of cubic CdSe ³⁴.

Figure 2(G, H & I) shows broad diffuse SAED patterns of undoped and Mn²⁺ doped CdSe NCs. These patterns show a system of spots which form concentric rings corresponding to diffraction from various atomic planes. Each spot corresponds to reflection from a single – crystal particle of specific orientation. The presence of single crystals of various orientations leads to the formation of rings of spots ³⁵. SAED pattern shows principally three rings ascribed to Miller indices (111), (220) and (311), respectively of the cubic CdSe phase. This indicates that as the concentration of Mn²⁺ ions increases, crystallinity of the sample slightly changes. Table 2 shows the values d- spacing for 0%, 0.25%, and 0.75% Mn²⁺ doped CdSe NCs corresponding to Miller indices (111), (220) and (311), respectively obtained from SAED pattern which are in good agreement with the cubic structure of CdSe (JCPDS card No. 19-0191).

Scanning electron microscopy (SEM) is a versatile technique to study morphology of nanomaterialsFigure 3(A, B & C) shows the FEG-SEM photographs of 0% (undoped), 0.25% and 0.75% Mn²⁺ doped CdSe NCs. The average diameter (crystallite size) of almost spherical undoped and doped CdSe NCs lies between 5 nm to 20 nm and increases with increase in doping concentration of Mn²⁺ ions in CdSe NCs ³⁶, which is in good agreement with HRTEM result discussed above.

Figure 3.SEM photographs (A-C) and EDS spectra (D-F) of 0%, 0.25%, and 0.75% Mn2+ doped CdSe NCs, respectively

The Energy Dispersive Spectrometer (EDS) analysis can be used to determine the composition of a specimen as a whole as well as the composition of individual componentsFigure 3 (D, E & F) shows the EDS spectra of 0%, 0.25% and 0.75% Mn²⁺ doped CdSe NCs, respectively, which confirm the presence of the Cd, Se and Mn elements. The data reveal the formation of undoped and Mn²⁺ doped CdSe with appropriate ratio of Cd : Se : Mn in the prepared NCs. The elemental composition of the synthesized 0%, 0.25%, and 0.75% Mn²⁺ doped CdSe NCs are given in Table 3 It is very important to note that no additional peaks corresponding to impurities or contaminants are observed, which confirms the purity of the NCs prepared ^{37, 38}.

The UV/V is absorption spectra for both undoped and Mn²⁺ doped CdSe NCs are shown in Figure 4(A). Direct optical band gap energies were obtained as 2.54 eV, 2.27 eV and 2.24 eV for 0%, 0.25% and 0.75% Mn2+ doped CdSe NCs with the help of Tauc plot as shown in Figure 4(B) using the classical relationship of near edge optical absorption of semiconductors ^{39, 40}:

where A is a constant, E_g is the energy band gap of the materials, h is the Planck’s constant, ν is the frequency and exponent n depends on the type of transition. For direct allowed transition n = 1/2, for indirect allowed transition n = 2, for direct forbidden transition n = 3/2 and for forbidden indirect transition n = 3 ⁴¹. To determine the possible transitions, (αhν)² vs. hν curves are plotted and corresponding band gap energies are obtained from extrapolating the straight line of the curves for zero absorption coefficient on hν axis. The optical band gap energy decreases from 2.54 eV to 2.24 eV as the concentration of Mn²⁺ ions increases from 0% to 0.75% in CdSe NCs. This small change in band gap energy may be due to direct transfer of energy between semiconductor – excited states and 3d state of Mn²⁺ ions ^{40, 41}.

The UV/V is absorption spectra for both undoped and Mn²⁺doped CdSe NCs show blue shift of ~0.6 eV as compared to the absorption edge (1.74 eV) for bulk CdSe at 300 K ²⁹ as shown in Figure 4(A). The shift to shorter wavelength in absorption peak was obviously caused by quantum confinement effect due to decrease in particle size.

Figure 4.(A) UV – Visible absorption spectra for undoped and Mn2+ doped CdSe NCs and (B) Tauc plot of (αhν)2 versus hν for undoped and Mn2+ doped CdSe NCs

Furthermore, for strong confinement, in which NC diameter D is smaller than the exciton Bohr radius

for bulk CdSe), the quantized energy levels of electrons and holes can be found from the solution of Schrödinger equation of Hamiltonian in Eq. (13) for a particle in a spherical box ⁴²

where the first two terms are the kinetic energy of an electron and a hole, the third and fourth terms represent the confinement potential energies and finally, the last term represents the Coulomb electron-hole interaction. The Eigen values of Hamiltonian in Eq. (13) yield the following expression for band gap energy of NC (the 1s - 1s excited state energy) ⁴²:

where E^nc_g and E^b_g are the band gap energies for CdSe NCs and bulk CdSe semiconductor, respectively h=h/2π with h representing Planck’s constant, D is the diameter of the NCs, 𝜀_∞ is the optical dilectric constant of the bulk CdSe semiconductor and µ is the reduced mass of the exciton given as µ=

is the electron mass at rest in vacuum 𝜀_∞=8.5𝜀_𝑜⁴³). Second and third terms represent the quantum confinement and the Coulomb interaction, respectively. Finally, last term represents the Rydberg correlation energy (polarisation term). The diameter of the NC can be calculated by converting Eq. (14) into a quadratic equation for the D as unknown:𝐴𝐷²+𝐵𝐷+𝐶=0, where A, B and C are material dependent parameters. From the calculated band gap energies as described above and by using Eq. (14) (neglecting polarization term), the average size of 0%, 0.25% and 0.75% and Mn²⁺ doped CdSe NCs has been calculated as 1.69 nm, 1.77 nm and 1.78 nm. The calculated values of average size of CdSe NCs from optical absorption study also increase with increase in doping concentration of Mn²⁺ ions and matches well with those obtained from XRD data described above Table 1.

Figure 5 presents PL spectra of undoped and Mn²⁺ doped CdSe NCs with excitation wavelength 360 nm at room temperature. The PL spectra show distinctive dual colour emissions. The narrow band emission centered at 380 nm with FWHM ~11 nm is assigned to the CdSe host and the broad emission nearly at 460 nm with FWHM ~49 nm is assigned to Mn²⁺ ions doped in the CdSe NCs, which can be attributed to a transition between ⁴T₁ and ⁶A₁ energy levels of the Mn²⁺ 3d states ⁴⁴. The intense Mn²⁺ luminescence indicates that the doped Mn²⁺ ions could serve as an efficient acceptor of the energy from the excited CdSe host ⁴⁵. The PL spectrum sufficiently confirms the existence of Mn²⁺ ions in the synthesized CdSe NCs. The luminescence intensity and FWHM of CdSe NCs change with the change in concentration of Mn²⁺ ions ⁴⁶. 0.25 % Mn²⁺ doped CdSe NCs show maximum luminescence intensity while 0.75 % Mn²⁺ doped CdSe NCs show less luminescence intensity but greater than undoped CdSe NCs. This means that for larger concentration of Mn²⁺ ions the luminescence intensity becomes smaller. In highly doped sample (CdSe: Mn, 0.75 %) Mn ions are placed near the surface of the core and that the Mn – Mn pair formation is a possible reason for the reduction of intensity of luminescence. The magnetization of Mn reduces due to antiferromagnetic coupling of Mn spins and therefore luminescence intensity reduces. In lightly doped sample (CdSe: Mn, 0.25 %) the Mn related splitting energy decreases thereby relatively high luminescence intensity is observed ⁴⁷.

Figure 5.PL spectra of 0%, 0.25%, and 0.75% Mn2+ doped CdSe NCs at room temperature.

0.25% and 0.75% Mn²⁺ doped CdSe NCs show well-resolved X-band EPR spectra Figure 6, at room temperature. Between the six major peaks that arise from the electron-nuclear hyperfine interaction at the Mn²⁺ (S=5/2, I=5/2, L=0), several smaller features are observed that contain information about the zero-field splitting of the Mn²⁺⁶A₁ ground state ^{8, 48}. The spectrum can be explained by the following spin Hamiltonian:

where 𝜇_B is the Bohr magneton, B is the applied magnetic field, S and I are the electron and nuclear spin

operators ⁸. The first term gives the Zeeman interaction with isotropic_g. The second term is the zero-field splitting term and the third term is the cubic-field splitting. The fourth term represents the hyperfine interaction with the ⁵⁵Mn nucleus. In strong applied magnetic field for Mn, the first term (𝜇_B 𝑩⋅𝒈⋅𝑺)≫𝑎. Therefore, the third term is neglected in Eq. (15) ^{49, 50}. The above hyperfine features are reproduced in spectral simulations (Figure 6, bottom) using the same spin Hamiltonian (Eq. 15) and spectroscopic splitting factor 𝑔, hyperfine parameter A and second- rank axial zero-field splitting D parameters deduced from EPR study of Mn²⁺ doped CdSe NCs. This demonstrates Mn²⁺ substitution at the axial cation site of the cubic CdSe lattice in these NCs. For Mn²⁺ there are only six allowed transitions corresponding to ΔM_s = ±1 and ΔM_I = 0. In CdSe lattice containing substitutional Mn²⁺, hyperfine lines are possible due to ±5/2↔±3/2, ±3/2↔±1/2 and +1/2↔-1/2 transitions. But the anisotropic contributions from ±5/2↔±3/2 and ±3/2↔±1/2 transitions are cancelled due to random orientations. Thus only +1/2↔-1/2 transition gives six lines spectrum ¹⁰.

The Mn²⁺ doped CdSe NCs contain various types of disorders and differences in surface passivation of each crystal. The EPR detects average of all the above properties. The values of 𝑔, A and D parameters for Mn²⁺ doped CdSe NCs obtained from EPR spectra are given in Table 4.The 𝑔 values show that the doping of Mn²⁺ ions in CdSe NCs is present. The values of A for Mn²⁺ doped CdSe NCs are presenting the characteristics of Mn²⁺ ions in an octahedral crystal field. The smaller values of D for Mn²⁺ doped CdSe NCs show little distortion. In EPR

Figure 6.Experimental and Simulated EPR spectra of (A) 0.25% Mn2+ doped CdSe NCs and (B) 0.75% Mn2+ doped CdSe NCs at room temperature (Frequency = 9.45 GHz)

spectra all the samples provide sextet splitting. The peak-to-peak line widths of EPR lines are about 66 G for 0.25% Mn²⁺ and 76 G for 0.75% Mn²⁺ from which we conclude that the Mn²⁺ ions are incorporated within the CdSe and not on the surface of CdSe NCs ¹⁰. As Mn concentration is increased, Mn – Mn interaction increases due to which the intensity, line width and distortion of EPR spectra slightly increase ³⁹.