Pulse EPR and ENDOR Study of Manganese Doped [(CH )2 NH ][Zn

Article Cite This: J. Phys. Chem. C 2017, 121, 27225−27232

pubs.acs.org/JPCC

Pulse EPR and ENDOR Study of Manganese Doped [(CH3)2NH2][Zn(HCOO)3] Hybrid Perovskite Framework

Mantas Šimeṅ as,*,† Lucyna Macalik,‡ Kȩstutis Aidas,† Vidmantas Kalendra,† Daniel Klose,§ † ∥ Gunnar Jeschke,§ Mirosław Ma̧czka,‡ Georg Völkel,∥ Juras ̅ Banys, and Andreas Pöppl †

Faculty of Physics, Vilnius University, Sauletekio av. 9, LT-10222 Vilnius, Lithuania Institute of Low Temperature and Structure Research, Polish Academy of Sciences, P.O. Box-1410, PL-50-950 Wroclaw 2, Poland § ETH Zürich, Department of Physical Chemistry, Vladimir-Prelog-Weg 2, 8093 Zürich, Switzerland ∥ Faculty of Physics and Earth Sciences, Universität Leipzig, Linnestrasse 5, D-04103 Leipzig, Germany ‡

S Supporting Information *

ABSTRACT: We present a pulse electron paramagnetic resonance (EPR) and electron−nuclear double resonance (ENDOR) study of a manganese-doped [(CH3)2NH2][Zn(HCOO)3] dense metal−organic framework which exhibits a structural phase transition at 163 K. The echo-detected field sweep Mn2+ EPR spectrum of the lowtemperature phase is in a perfect agreement with the previous continuous-wave EPR results, while the spectrum of the disordered phase reveals a significant EPR transition-dependent relaxation. The 1 H ENDOR pattern indicates several protons in the vicinity of the Mn2+ ion. The experimental ENDOR spectrum is successfully simulated using the proton hyperfine tensors calculated by the density functional theory. A multifrequency electron spin echo envelope modulation (ESEEM) spectroscopy shows a peculiar signal which is unaffected by the external magnetic field. The modulation depth of this signal starts to decrease above 40 K, coinciding with the temperature at which the methyl groups of the (CH3)2NH2+ cations start to rotate. We also relate the methyl group motion to the decrease of the phase memory time of the Mn2+ ions. The temperature dependence of the longitudinal relaxation time indicates a coupling between the Mn2+ electron spins and a hard optical phonon mode. This mode undergoes a damping at the phase transition point.

■

phase.14−16 In addition, some of these compounds containing paramagnetic transition metal ions show a long-range magnetic order making them potential single-phase multiferroics,11,13,17−20 a rare and highly attractive material property.21 The first reported and the most thoroughly studied members of the hybrid metal formate frameworks are [(CH3)2NH2][M(HCOO)3] (DMAM) compounds, where M is one of the following cations: Mn2+, Fe2+, Co2+, Ni2+, Cu2+, Zn2+, or Mg2+.10,11,22 These compounds exhibit a structural phase transition at the transition temperature Tc of about 160−180 K (except for the Mg2+ analogue for which the much higher Tc of 270 K was observed). The X-ray diffraction (XRD) data of the high temperature (HT) phase (T > Tc) indicates that these compounds crystallize into the perovskite topology.10,11 Here each dimethylammonium cation (CH3)2NH2+ (DMA+) occupies one of the three positions within a metal-formate cavity (see Figure 1a). Due to a significant degree of crystal twinning, the XRD methods encounter challenges while determining the low temperature (LT) structure (T < Tc) of these frameworks.

INTRODUCTION Metal−organic frameworks (MOFs) draw significant attention of the scientific community due to their broad potential applicability.1,2 These hybrid crystalline compounds contain metal ion nodes that are connected by organic linkers into various porous structures.3 Usually transition metal or lanthanide cations are used as metal centers due to their high coordination number and the ability to adapt different coordination geometries.4 Most of the exceptional MOF properties originate from their pronounced porosity which can be utilized for gas adsorption and similar applications.5−8 Recently, a new subclass of these materials called dense MOFs emerged.9 In these compounds, the pores are densely occupied with organic cations that cannot leave the framework. Representatives of dense MOFs are [A][M(HCOO)3] formate frameworks where A+ is a molecular cation [e.g., (CH3)2NH2+] and M2+ is usually a transition metal ion.10−14 In such compounds, the metal nodes are interconnected by the formate HCOO− linkers into semicuboid pores. In the center of each cuboid, a single A+ cation is situated and H-bonded with oxygen atoms of the MO6 octahedra. Many members of the formate frameworks demonstrate structural phase transitions, some of which are from the paraelectric to the ferroelectric © 2017 American Chemical Society

Received: October 9, 2017 Revised: November 15, 2017 Published: November 17, 2017 27225

DOI: 10.1021/acs.jpcc.7b09990 J. Phys. Chem. C 2017, 121, 27225−27232

Article

The Journal of Physical Chemistry C

Sánchez-Andújar et al. managed to solve the LT structure of DMAMn and assigned it to the monoclinic Cc space group, which is noncentrosymmetric as necessary for the ferroelectricity to occur.23,24 In this phase, each DMA+ cation occupies a single position within a cavity (see Figure 1b), and the system exhibits the long-range electric order. However, it was found that the monoclinic centrosymmetric C2/c group can also relatively well-describe the XRD data.23 The ordering in these compounds was investigated using various other experimental techniques such as heat capacity,10,11,25 heat flow,24,26 and pyrocurrent18,27,28 measurements, second-harmonic generation,29 dielectric,10,11,19,23,30,31 infrared, Raman,32 NMR,31,33−35 as well as electron paramagnetic resonance (EPR)36−38 spectroscopies. Most of these methods detected a first-order order−disorder phase transition at Tc. However, a proper ferroelectric hysteresis loop was obtained only for the deuterated DMACo compound.39 In addition, our recent single crystal continuous-wave (CW) EPR study with the applied external electric field raised serious concerns about the ferroelectric origin of the ordered phase in DMAZn.38 The uncertain nature of the LT phase in these compounds stimulates further attention to this problem. In our previous CW EPR studies of the manganese-doped DMAZn framework, we have successfully shown that the paramagnetic Mn2+ probes replace Zn2+ ions in the lattice and form MnO6 octahedra.37,38 The spin Hamiltonian parameters of the Mn2+ ions were found to be sensitive to the structural changes occurring at Tc. The temperature-dependence of the axial zero-field splitting (zfs) parameter revealed a strong firstorder phase transition.37 The single crystal EPR study confirmed the crystal twinning model proposed by the XRD study.38 The CW EPR methods can probe only the first coordination sphere of the paramagnetic center, and thus the much weaker hyperfine (hf) interactions of the distant nuclei are usually not resolved. Such interactions can be detected by the more powerful pulse EPR and pulse electron nuclear double resonance (ENDOR) spectroscopies.40,41 This allows investigating a much broader spatial region around the paramagnetic probe. In this study, we use these pulse techniques as well as the density functional theory (DFT) calculations to further investigate the LT structure, dynamics, and motion of the DMA+ cations in the manganese-doped DMAZn framework.

acid (98%, Fluka) were commercially available and used without further purification. DMAZn: 0.05 Mn2+ mol % powder sample (1) was obtained by a slow diffusion method. In a typical experiment, 16 mL methanol solution containing 12.8 mmol of (CH3)2NH and 12.8 mmol of formic acid was placed at the bottom of a glass tube (9 mm inner diameter). Sixteen milliliters methanol solution containing 1.5992 mmol of ZnCl2 and 0.0008 mmol of MnCl2 was gently added on this solution. The tube was sealed, and after 5 days colorless crystals were harvested. The previously reported powder XRD study37 of 1 confirmed that this compound corresponds to DMAZn (see Figure S1). EPR and ENDOR Spectroscopy. A Bruker E580 EPR spectrometer was used for pulse EPR and ENDOR measurements at X-band microwave (mw) frequency (∼9.5 GHz). Pulse EPR experiments at Q (∼34 GHz) and W-band (∼95 GHz) frequencies were performed with Bruker E600 and E680 spectrometers. The two pulse (2p) echo-detected field sweep, 2p (pulse sequence: π/2−τ−π−τ−echo) and three-pulse (3p) (π/2−τ−π/2−τ′−π/2−τ−echo) electron spin echo envelope modulation (ESEEM)42 experiments were recorded using nonselective mw pulses. The typical π/2 mw pulse length was tπ/2 = 16 ns. The baselines of the obtained ESEEM time domain (TD) patterns were corrected by subtracting exponential decays. The remaining signals were Fourier transformed to the frequency domain (FD) spectra. The longitudinal relaxation time T1 was determined using the inversion recovery pulse sequence.41 The temperature-dependent relaxation time measurements were performed on heating. For most of the ENDOR experiments, the Mims pulse sequence43 was used with the mw and radiofrequency pulse lengths of tπ/2 = 16 ns and tRF = 10 μs, respectively. To avoid the suppression effects (blind spots),41 the 3p ESEEM and Mims ENDOR experiments were performed using different values of the interpulse delay τ. The interpulse delay τ in 2p experiments was 100 ns. All simulations of the spectra were performed using EasySpin 5.1.5 simulation software.44 DFT Calculations. All computations of the proton hf coupling tensors were carried out using the unrestricted DFT framework as implemented in the Orca 3.0.3 program.45 The experimental LT crystal structure of DMAMn MOF23 was used for calculations. Figure S2 depicts the used geometry, which includes a single Mn2+ ion coordinated by six formate linkers and surrounded by eight nearest DMA+ cations. The total charge of this system is +4. The PBE0 exchange-correlation functional46 along with the EPR-II double-ζ basis set47 was selected for the organic moieties, whereas the aug-cc-pVDZ basis set48 was adopted for the metal atom. The resolution of the identity approach, the socalled RIJCOSX method49 as implemented in Orca, was utilized using the auxiliary aug-cc-pVDZ basis set in order to speed up the hybrid DFT calculations. The Lebedev 770 points grid was selected for the computation of the two-electron integrals. The tight criterion for the SCF procedure was chosen. The Fermi contact and spin dipolar contributions to the hf coupling tensors were computed, so that we did not account for the spin−orbit contribution.

EXPERIMENTAL AND COMPUTATIONAL METHODS Sample Preparation. ZnCl2 (99%, Fluka), MnCl2 (99%, Sigma-Aldrich), 2.0 M solution of (CH3)2NH in methanol (Sigma-Aldrich), methanol (99.8%, Sigma-Aldrich), and formic

RESULTS AND DISCUSSION Echo-Detected Field Sweep. The X-band echo-detected field sweep EPR spectra of 1 recorded at 170 and 6 K are presented in Figure 2. They correspond to the powder patterns of the Mn2+ ions with the 3d5 electronic configuration and the

Figure 1. (a) HT and (b) LT structures of DMAMn. In (a), the DMA+ cation is represented as a superposition of three orientations (hydrogen atoms are omitted for clarity). Structures taken from ref 23.

■

■

27226


Article


parameters. The spectrum obtained in such a way is in agreement with the experiment (Figure 2a). The EPR transition-dependent relaxation time was also recently observed in frozen solutions of Gd3+.50 This behavior was assigned to the fluctuating (transient) zfs. The energy levels of the outer fs transitions are much more susceptible to these fluctuations, resulting in much shorter relaxation compared with the central transition. In our case, such fluctuations occur due to the DMA+ cation hopping which affects the MnO6 octahedra. The hf interactions between the Mn2+ ions and distant nuclei are too weak to be resolved in the field sweep spectra. Therefore, we performed ENDOR and ESEEM experiments which can detect these small interactions allowing us to further study the LT phase of 1. ENDOR Spectroscopy. We used Mims ENDOR spectroscopy to investigate the local environment of the Mn2+ probes. Figure 3d presents the 1H ENDOR spectrum of 1 recorded at 325 mT, τ = 144 ns, and 6 K. It consists of well-resolved hf splittings at the perpendicular edge singularities of the powder pattern, indicating several proton species in the close vicinity of the Mn2+ ions. The splittings corresponding to the parallel components of the 1H hf coupling tensors are poorly resolved and thus are not explicitly presented. We also performed the Davies ENDOR measurements of 1, but the recorded 1H spectrum shows much worse signal-to-noise ratio (see Figure S3). The obtained ENDOR pattern was simulated using the 1H hf coupling tensors obtained by the DFT calculations (see Computational Methods). The calculated tensors are listed in Table S1. In the simulations, we also used spin Hamiltonian parameters of the Mn2+ centers previously determined in the CW EPR studies.37,38 First of all, the simulation was performed by using only the protons from the formate HCOO− linkers. The corresponding ENDOR pattern is presented in Figure 3a, indicating that the strongest experimentally observed hf coupling originates from these linkers. Note that the same hf splitting was observed in our previous studies of manganese doped [NH3(CH2)4NH3][Zn(HCOO)3]2 and [CH3NH2NH2][Zn(HCOO)3] dense MOFs which also contain MnO6 octahedra connected via the formate anions.51,52 This demonstrates that such hf coupling could be used as a fingerprint to identify whether the Mn2+ centers successfully replaced the diamagnetic ions in similar frameworks. The simulation was further improved by including protons from the DMA+ cations surrounding the Mn2+ probe. The simulated pattern is presented in Figure 3b, revealing a satisfactory agreement with the experiment. An even better match was obtained after we slightly adjusted the calculated hf couplings of protons from the methyl groups of the DMA+ cations (see Figure 3c and the Supporting Information for more details). This modification corresponds to a decrease of the Mn−H distance by about 0.08 Å. The small adjustment of the 1 H hf couplings could be justified, since the crystal structure of DMAMn23 MOF was determined at 100 K while the ENDOR pattern was recorded at 6 K. Another possible origin of this discrepancy is that our DFT calculations are based on the LT structure of DMAMn compound, while the experiments are performed for the manganese-doped DMAZn framework. We also simulated the ENDOR spectrum using a spin system with an effective S = 1/2 electron spin to account for only the 1 1 ms = − 2 ↔ 2 transition. The obtained results are presented in

Figure 2. X-band two-pulse echo-detected field sweep spectra of 1 recorded at (a) 170 K and (b) 6 K. Simulations indicated in red were obtained by taking into account all fs transitions, while blue simulation corresponds to the central fs transitions. The arrow indicates the field position at which most of the pulse EPR and ENDOR experiments were performed. See the Supporting Information for simulation parameters.

high-spin 6S5/2 ground state (electron and nuclear spins are S = 5/2 and IMn = 5/2). The 6 K spectrum is in a perfect agreement with our previous CW EPR studies and corresponds to the LT phases of 1. We simulated this spectrum using the previously determined electron Zeeman, hf, and fine structure (fs) parameters of the Mn2+ center in DMAZn (see Figure 2b).37,38 Simulation parameters are presented in the Supporting Information. We also tried to simulate the HT spectrum measured at 170 K using the parameters obtained from the CW EPR experiment.37 A clear mismatch between the spectra can be seen in Figure 2a, since the experimental pattern does not contain broad spectral wings outside the central EPR transition 1 1 (ms = − 2 ↔ 2 , here mS denotes the magnetic electron spin quantum number). As discussed in our previous study, these 5 3 3 1 wings correspond to the fs (ms = ± 2 ↔ ± 2 and ± 2 ↔ ± 2 ), which can be described using highly distributed zfs parameters.37 The hopping motion of the DMA+ cations in the HT phase causes distinct positions of these cations in the vicinity of the MnO6 groups. This leads to differently distorted octahedra (distribution of zfs parameters) and unresolved spectral wings in the CW EPR spectrum above Tc. The outer fs transitions are absent in the experimental echo-detected spectrum, while it simultaneously shows the forbidden hf lines indicating a substantial zfs of the Mn2+ centers. At first glance, these two observations contradict each other. A plausible explanation of this problem is that the phase memory time is substantially shorter for the outer fs transitions. Therefore, these transitions cannot be detected in the echo-detected field sweep experiment. This assumption is supported by simulation of only the central fs transition using the same set of the Mn2+ ion 27227


Article


325 mT and 10 K are presented in Figure 4 (panels a and c). The TD pattern shows superimposed short and long period

Figure 4. X-band 2p and 3p ESEEM (a and b) TD and (c and d) FD signals of 1 recorded at 325 mT and 10 K. 3p ESEEM experiments were performed at τ = 108 and 144 ns. The proton signals are indicated in the FD spectra.

oscillations. The corresponding FD spectrum reveals three peaks at about 2.1, 13.8, and 27 MHz. The line at 13.8 MHz corresponds to protons, while the high-frequency bump at roughly 2νH is the proton combination peak.41 The 2p ESEEM TD signal decays with a relatively short phase memory time, resulting in broad and poorly resolved FD lines. In contrast, the TD signal in a 3p ESEEM experiment is limited by the much longer transverse nuclear or longitudinal electron relaxations providing narrower FD lines.41 The 3p ESEEM TD traces and FD spectra obtained at 325 mT and 10 K are presented in Figure 4 (panels b and d). Measurements were performed using the interpulse delays τ = 108 and 144 ns. The obtained TD decay for τ = 108 ns also exhibits superimposed modulations of short and long period oscillations. The corresponding FD spectrum consists of much narrower lines, although the internal structure of the proton line is still hardly resolved. The width of the proton peak is about 3 MHz, which is in agreement with the 1H hf couplings determined by the ENDOR spectroscopy. The lowfrequency signal consists of several lines below 3 MHz with the most intense peak at 1.85 MHz. The TD decay recorded for τ = 144 ns reveals only the long period modulation, since the proton signal is absent due to the suppression effect at the 1H Larmor frequency.41 The resulting FD spectrum consists of the low-frequency signal which coincides with the τ = 108 ns case. A possible origin of the low-frequency signal could be nitrogen 14N nuclei (IN = 1) of the DMA+ cations. To investigate this, we performed 3p ESEEM measurements at X, Q, and W-band mw frequencies. The obtained spectra are presented in Figure 5. They reveal that the observed lines are independent of the external magnetic field. This surprising result indicates that this signal cannot originate from the nitrogen nuclei, since one would expect a significant nuclear Zeeman effect for a system with nuclear spin I > 0. Several NMR studies revealed that the motion of the DMA+ cations persists in the ordered phase of DMAZn.33−35 Thus, we performed ESEEM experiments at different temperatures to

Figure 3. (a−c) Simulated and (d) measured (6 K) X-band 1H Mims ENDOR pattern of 1 at 325 mT and τ = 144 ns. The simulation in (a) was performed by taking into account only protons from the HCOO− linkers, while in (b), protons from the (CH3)2NH2+ cations were also included. The 1H hf couplings used in (a) and (b) simulations were obtained from the DFT calculations. The simulation in (c) was performed using slightly adjusted 1H hf couplings of the methyl groups (lines indicated by the arrows). The bottom spectrum in (c) was 1 1 obtained taking into account only the ms = − 2 ↔ 2 transition and neglecting the zfs.

Figure 3c, revealing an absence of the lines at |ν − νH| ≳ 3 MHz (here νH is the Larmor frequency of protons). This indicates that the origin of these lines is not the strongly coupled 1H species but the nuclear transitions in the |mS| > 1/2 spin manifolds. Slight changes of the central part of the spectrum are also observed, although the agreement with the experimental pattern remains acceptable. The comparison between the pulse ENDOR experiment and the DFT-based simulations confirms the successful incorporation of the Mn2+ probes at the Zn2+ lattice sites in DMAZn framework. More importantly, our ENDOR data verifies the proton positions determined by the XRD. This confirms the refinement of the LT space group of these frameworks as the monoclinic noncentrosymmetric Cc.23 ESEEM Spectroscopy. The local environment of the Mn2+ ions in 1 was further probed using 2p and 3p ESEEM spectroscopy. The 2p ESEEM TD and FD signals recorded at 27228


Article


Figure 7. Temperature dependence of the modulation depth k of the low-frequency X-band 3p ESEEM signal of 1. The inset explains the definition of the parameter k.

Figure 5. Low-frequency 3p ESEEM signal recorded at 325, 1195, 3350 mT and 10 K.

NMR, the characteristic time of this motion is about 10−7 s at 70 K.33,34 The highest peak of the low-frequency ESEEM signal is roughly at 2 MHz, which is in a good agreement with the time scale of the methyl group rotation. Thus, we speculate that the decrease of the modulation depth of the low-frequency signal may be related to the averaging of some anisotropic magnetic interactions due to the onset of the methyl group rotations. The precise origin of this unusual low-frequency ESEEM signal remains unknown and will be thoroughly investigated in a separate study using samples with different isotopes. Relaxation Time Measurements. We also performed measurements of the phase memory time Tm and the longitudinal relaxation time T1 of the Mn2+ centers to further investigate the low-temperature phase of 1. Figure 8a reveals a monotonous decrease of Tm with increasing temperature. The relaxation time starts to diminish at about the same temperature as the modulation depth of the low-frequency 3p ESEEM signal. Thus, we relate the decrease of Tm to the methyl group rotations that cause fluctuations of the Mn2+ ion magnetic environment. The phase memory time at the phase transition point exhibits anomalous behavior, which is related to the critical dynamics of the order parameter fluctuations or enhanced motion of the DMA+ cations above Tc (see Figure S5). The anomaly of Tm was previously observed in trissarcosine calcium chloride (TSCC),53 KH3(SeO3)2,54 and similar materials. Note that we were unable to reliably measure Tm above 180 K due to too fast relaxation. The phase memory relaxation rate exhibits the Arrheniustype behavior and can be described using the following equation:

check whether we can observe the influence of these motional effects on the echo modulation. The temperature-dependent 2p and 3p ESEEM spectra are presented in Figure 6 (see the

Figure 6. Temperature-dependent X-band (a) 2p and (b) 3p ESEEM FD spectra of 1 recorded at 325 mT. 3p ESEEM experiments performed at τ = 108 ns. Spectra are normalized to the proton line.

Supporting Information for the TD traces), indicating that the low-frequency signal starts to diminish above 40 K and completely disappears at about 70 K. Figure 7 reveals the temperature-dependence of the modulation depth parameter k of this signal, which was determined from the 3p ESEEM TD traces at 400 ns (see inset for the definition of k). A sudden decrease of the modulation depth is observed above 40 K, while k = 0 above 70 K. Note that Besara et al.33 observed a glassy behavior in DMAZn below 40 K, while Asaji et al.34 reported a small anomaly in the spin−lattice relaxation time of 1H at 79 K. The hopping motion of the DMA+ cation should be frozen on the ESEEM time scale below Tc as demonstrated by dielectric30 and NMR34,35 spectroscopies. However, the reorientation of the methyl groups around the C3 symmetry axes is still expected in the ordered phase.34 As determined by

−1 −Ea / kBT Tm−1 = Tm, ∞e

(1)

where Ea denotes the activation energy of the motional process and T−1 m,∞ is the phase memory relaxation rate at infinite temperature. In the LT phase, we identified two processes with the activation energies Ea = 17(1) meV and Ea = 5.8(4) meV. The inset in Figure 8a presents the corresponding fits. Note that the same analysis provided Ea of 15 meV in the related Mn2+ doped [CH3NH2NH2][Zn(HCOO)3] perovskite.52 We also observed a third process in the HT phase with an activation energy of about 78 meV (see Figure S5b). The obtained activation energy in the HT phase is similar to the values obtained by NMR for the methyl group rotations in 27229


Article


Figure 8. Temperature dependence of (a) the phase memory time and (b) the most probable longitudinal relaxation rate of the Mn2+ centers in 1. The inset in (a) shows the Arrhenius-type behavior of T−1 m below Tc. The curves in (b) are the total (black) and separate (blue and red) contributions to the longitudinal relaxation rate. The inset in (b) emphasizes the behavior of the longitudinal relaxation rate about Tc. If not indicated, the error bars are smaller than data points. Measurements performed at 325 mT and X-band mw frequency.

DMAZn framework.33−35 However, the energies of the LT phase are significantly smaller. Such activation energies are frequently observed for quantum tunnelling motion of the methyl groups.55 This may indicate a substantial contribution of such process to the observed low-temperature phase memory relaxation in 1. This effect will be investigated in more detail in a separate study using deuterated compounds. The longitudinal relaxation time T1 was determined by the inversion recovery pulse sequence. The measured magnetization kinetics were fit using a recovery function for the continuously distributed relaxation times (see refs 52 and 56 for more details). The obtained temperature dependence of the most probable longitudinal relaxation rate T−1 1 is presented in Figure 8b. The experimental data was approximated using the following equation:54 T1−1 = AT + Bcsch2(hνopt /2kBT )

the phase transition56 or due to the structural changes occurring at Tc.

■

SUMMARY AND CONCLUSIONS We used pulse EPR and ENDOR spectroscopies to study the Mn2+ doped [(CH3)2NH2][Zn(HCOO)3] perovskite framework which exhibits a structural order−disorder phase transition. The pulse methods allowed us to probe a much wider environment of the Mn2+ ions compared to the CW EPR spectroscopy. The echo-detected field sweep EPR spectra of Mn2+ centers were recorded for the HT and LT phases of the framework. The LT spectrum was simulated using the spin Hamiltonian parameters taken from previous CW EPR studies. Analysis of the HT phase spectrum revealed much faster phase memory relaxation of the outer fs transitions. This behavior was assigned to the fluctuating zfs of the Mn2+ ions in the HT phase of the framework. These fluctuations are caused by the hopping motion of the DMA+ cations, confirming a tight interaction between the two subsystems of the material. The X-band 1H Mims ENDOR spectrum displayed several proton species in the vicinity of the Mn2+ probes. We simulated the experimental spectrum using the proton hf coupling tensors obtained by the DFT calculations of the DMAMn framework. A good agreement between the experimental and simulated ENDOR patterns verifies the structural framework model and the proton positions that are often difficult to determine accurately by the XRD methods. The ESEEM spectroscopy revealed a signal of weakly coupled protons in agreement with the ENDOR results. We also observed a remarkable low-frequency ESEEM signal which is unaffected by the external magnetic field. The modulation depth of this signal starts to decrease above 40 K and reaches zero at 70 K. This temperature range and the frequency of the signal suggest that the modulation disappears due to the rotations of the methyl groups of the DMA+ cations. However, the precise origin of this peculiar ESEEM signal remains obscured and demands further attention. The relaxation time measurements of the Mn2+ ions revealed further information about the motion of the DMA+ cations and the dynamics of the metal-formate lattice. We observed a

(2)

where the first term describes a direct (one-phonon) relaxation process with the acoustic lattice phonons. The second term takes into account a Raman (two-phonon) process which involves the optical phonon branch of frequency νopt. A and B are constants. The best fit provided νopt = 3.5(3) × 1012 Hz which corresponds to 118(10) cm−1. Note that IR and Raman study reports a similar frequency of the lattice phonons that involve metal−oxygen octahedra in the DMAZn and DMAFe frameworks.32 The behavior of the longitudinal relaxation rate around the phase transition point is presented in the inset of Figure 8b. A small dip of T−1 1 is observed at Tc, which we verified by several measurements. Such a minimum of the longitudinal relaxation rate at the phase transition was measured for many other ferroelectric or related compounds such as rochelle salt,56 KH3(SeO3)2,54 betaine phosphite,57 TSCC,58 triglycine sulfate (TGS),59 and lead titanate.60 Note that in our previous study of the Mn2+ doped [CH3NH2NH2][Zn(HCOO)3] framework, we also observed indications of the same behavior.52 A minimum of T−1 1 occurs due to the enhanced damping of the optical phonon mode νopt, which governs the longitudinal relaxation. The observed increase of the mode damping may originate from the coupling with the relaxational mode responsible for 27230


Article


(9) Cheetham, A. K.; Rao, C. N. R. There’s Room in the Middle. Science 2007, 318, 58−59. (10) Jain, P.; Dalal, N. S.; Toby, B. H.; Kroto, H. W.; Cheetham, A. K. Order-Disorder Antiferroelectric Phase Transition in a Hybrid Inorganic-Organic Framework with the Perovskite Architecture. J. Am. Chem. Soc. 2008, 130, 10450−10451. (11) Jain, P.; Ramachandran, V.; Clark, R. J.; Zhou, H. D.; Toby, B. H.; Dalal, N. S.; Kroto, H. W.; Cheetham, A. K. Multiferroic Behavior Associated with an Order-Disorder Hydrogen Bonding Transition in Metal-Organic Frameworks (MOFs) with the Perovskite ABX3 Architecture. J. Am. Chem. Soc. 2009, 131, 13625−13627. (12) Li, W.; Wang, Z.; Deschler, F.; Gao, S.; Friend, R. H.; Cheetham, A. K. Chemically Diverse and Multifunctional Hybrid Organic-Inorganic Perovskites. Nat. Rev. Mater. 2017, 2, 16099. (13) Ma̧czka, M.; Gagor, A.; Ptak, M.; Paraguassu, W.; da Silva, T. A.; Sieradzki, A.; Pikul, A. Phase Transitions and Coexistence of Magnetic and Electric Orders in the Methylhydrazinium Metal Formate Frameworks. Chem. Mater. 2017, 29, 2264−2275. (14) Xu, G.-C.; Ma, X.-M.; Zhang, L.; Wang, Z.-M.; Gao, S. DisorderOrder Ferroelectric Transition in the Metal Formate Framework of [NH4][Zn(HCOO)3]. J. Am. Chem. Soc. 2010, 132, 9588−9590. (15) Zhang, W.; Xiong, R.-G. Ferroelectric Metal-Organic Frameworks. Chem. Rev. 2012, 112, 1163−1195. (16) Asadi, K.; van der Veen, M. A. Ferroelectricity in Metal-Organic Frameworks: Characterization and Mechanisms. Eur. J. Inorg. Chem. 2016, 2016, 4332. (17) Wang, X.-Y.; Gan, L.; Zhang, S.-W.; Gao, S. Perovskite-like Metal Formates with Weak Ferromagnetism and as Precursors to Amorphous Materials. Inorg. Chem. 2004, 43, 4615−4625. (18) Wang, W.; Yan, L.-Q.; Cong, J.-Z.; Zhao, Y.-L.; Wang, F.; Shen, S.-P.; Zou, T.; Zhang, D.; Wang, S.-G.; Han, X.-F.; et al. Magnetoelectric Coupling in the Paramagnetic State of a Metal-Organic Framework. Sci. Rep. 2013, 3, 2024. (19) Ma̧czka, M.; Ga̧gor, A.; Macalik, B.; Pikul, A.; Ptak, M.; Hanuza, J. Order-Disorder Transition and Weak Ferromagnetism in the Perovskite Metal Formate Frameworks of [(CH3 ) 2 NH 2 ][M(HCOO)3] and [(CH3)2ND2][M(HCOO)3] (M = Ni, Mn). Inorg. Chem. 2014, 53, 457−467. (20) Clune, A. J.; Hughey, K. D.; Lee, C.; Abhyankar, N.; Ding, X.; Dalal, N. S.; Whangbo, M.-H.; Singleton, J.; Musfeldt, J. L. Magnetic Field-Temperature Phase Diagram of Multiferroic [(CH3)2NH2][Mn(HCOO)3]. Phys. Rev. B: Condens. Matter Mater. Phys. 2017, 96, 104424. (21) Eerenstein, W.; Mathur, N. D.; Scott, J. F. Multiferroic and Magnetoelectric Materials. Nature 2006, 442, 759−765. (22) Pato-Doldan, B.; Sanchez-Andujar, M.; Gomez-Aguirre, L. C.; Yanez-Vilar, S.; Lopez-Beceiro, J.; Gracia-Fernandez, C.; Haghighirad, A. A.; Ritter, F.; Castro-Garcia, S.; Senaris-Rodriguez, M. A. Near Room Temperature Dielectric Transition in the Perovskite Formate Framework [(CH3)2NH2][Mg(HCOO)3]. Phys. Chem. Chem. Phys. 2012, 14, 8498−8501. (23) Sánchez-Andújar, M.; Presedo, S.; Yánez-Vilar, S.; CastroGarcía, S.; Shamir, J.; Senarís-Rodríguez, M. A. Characterization of the Order-Disorder Dielectric Transition in the Hybrid Organic-Inorganic Perovskite-Like Formate Mn(HCOO)3[(CH3)2NH2]. Inorg. Chem. 2010, 49, 1510−1516. (24) Sánchez-Andújar, M.; Gomez-Aguirre, L. C.; Pato Doldan, B.; Yanez-Vilar, S.; Artiaga, R.; Llamas-Saiz, A. L.; Manna, R. S.; Schnelle, F.; Lang, M.; Ritter, F.; et al. First-Order Structural Transition in the Multiferroic Perovskite-Like Formate [(CH3)2NH2][Mn(HCOO)3]. CrystEngComm 2014, 16, 3558−3566. (25) Ma̧czka, M.; Sieradzki, A.; Bondzior, B.; Deren, P.; Hanuza, J.; Hermanowicz, K. Effect of Aliovalent Doping on the Properties of Perovskite-Like Multiferroic Formates. J. Mater. Chem. C 2015, 3, 9337−9345. (26) Lopez-Beceiro, J.; Gracia-Fernandez, C.; Gomez-Barreiro, S.; Castro-Garcia, S.; Sanchez-Andujar, M.; Artiaga, R. Kinetic Study of the Low Temperature Transformation of Co(HCOO)3[(CH3)2NH2]. J. Phys. Chem. C 2012, 116, 1219−1224.

monotonous decrease with increasing temperature of the phase memory time in the ordered phase of the framework. This behavior was assigned to the methyl group motion which causes fluctuations of the local Mn2+ ion environment. We identified two motional mechanisms in the LT phase with the activation energies of 17(1) meV and 5.8(4) meV. Indication of a third process with much higher activation energy was observed in the HT phase. We found that the longitudinal relaxation of the Mn2+ ions is dominated by the direct process with the acoustic lattice phonons and the two-phonon Raman process with the optical lattice mode of 118(10) cm−1 frequency. We also observed a small anomalous dip of the longitudinal relaxation rate at Tc, which is caused by the increased damping of the hard optical mode governing the relaxation.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcc.7b09990. Additional XRD, DFT, EPR, and ENDOR data (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: mantas.simenas@ff.vu.lt. Tel: +370 5 2234537. Fax: +370 5 2234537. ORCID

Mantas Šimėnas: 0000-0002-2733-2270 Lucyna Macalik: 0000-0002-7067-9691 Mirosław Ma̧czka: 0000-0003-2978-1093 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the COST Action MP1308-TOBE (STSM grant) and by the Research Council of Lithuania (Project TAP LLT-4/2017). The authors thank High Performance Computing Center “HPC Sauletekis” of Vilnius University Faculty of Physics for the computational resources.

■

REFERENCES

(1) Kuppler, R. J.; Timmons, D. J.; Fang, Q.-R.; Li, J.-R.; Makal, T. A.; Young, M. D.; Yuan, D.; Zhao, D.; Zhuang, W.; Zhou, H.-C. Potential Applications of Metal-Organic Frameworks. Coord. Chem. Rev. 2009, 253, 3042−3066. (2) Meek, S. T.; Greathouse, J. A.; Allendorf, M. D. Metal-Organic Frameworks: A Rapidly Growing Class of Versatile Nanoporous Materials. Adv. Mater. 2011, 23, 249−267. (3) Rosseinsky, M. Recent Developments in Metal-Organic Framework Chemistry: Design, Discovery, Permanent Porosity and Flexibility. Microporous Mesoporous Mater. 2004, 73, 15−30. (4) Kitagawa, S.; Kitaura, R.; Noro, S. Functional Porous Coordination Polymers. Angew. Chem., Int. Ed. 2004, 43, 2334−2375. (5) Murray, L. J.; Dinca, M.; Long, J. R. Hydrogen Storage in MetalOrganic Frameworks. Chem. Soc. Rev. 2009, 38, 1294−1314. (6) Wu, H.; Gong, Q.; Olson, D. H.; Li, J. Commensurate Adsorption of Hydrocarbons and Alcohols in Microporous Metal Organic Frameworks. Chem. Rev. 2012, 112, 836−868. (7) Li, J.-R.; Sculley, J.; Zhou, H.-C. Metal-Organic Frameworks for Separations. Chem. Rev. 2012, 112, 869−932. (8) Kreno, L. E.; Leong, K.; Farha, O. K.; Allendorf, M.; Van Duyne, R. P.; Hupp, J. T. Metal-Organic Framework Materials as Chemical Sensors. Chem. Rev. 2012, 112, 1105−1125. 27231


Article

The Journal of Physical Chemistry C (27) Tian, Y.; Stroppa, A.; Chai, Y.; Yan, L.; Wang, S.; Barone, P.; Picozzi, S.; Sun, Y. Cross Coupling between Electric and Magnetic Orders in a Multiferroic Metal-Organic Framework. Sci. Rep. 2015, 4, 6062. (28) Šimėnas, M.; Balciunas, S.; Ma̧czka, M.; Banys, J.; Tornau, E. E. Structural Phase Transition in Perovskite Metal-Formate Frameworks: a Potts-Type Model with Dipolar Interactions. Phys. Chem. Chem. Phys. 2016, 18, 18528−18535. (29) Jain, P.; Stroppa, A.; Nabok, D.; Marino, A.; Rubano, A.; Paparo, D.; Matsubara, M.; Nakotte, H.; Fiebig, M.; Picozzi, S.; et al. Switchable electric polarization and ferroelectric domains in a metalorganic-framework. Npj Quantum Materials 2016, 1, 16012. (30) Yadav, R.; Swain, D.; Bhat, H. L.; Elizabeth, S. Order-Disorder Phase Transition and Multiferroic Behaviour in a Metal Organic Framework Compound (CH3)2NH2Co(HCOO)3. J. Appl. Phys. 2016, 119, 064103. (31) Abhyankar, N.; Kweon, J. J.; Orio, M.; Bertaina, S.; Lee, M.; Choi, E. S.; Fu, R.; Dalal, N. S. Understanding Ferroelectricity in the Pb-Free Perovskite-Like Metal-Organic Framework [(CH3)2NH2]Zn(HCOO)3: Dielectric, 2D NMR, and Theoretical Studies. J. Phys. Chem. C 2017, 121, 6314−6322. (32) Ma̧czka, M.; Ptak, M.; Macalik, L. Infrared and Raman Studies of Phase Transitions in Metal-Organic Frameworks of [(CH3)2NH2][M(HCOO)3] with M = Zn, Fe. Vib. Spectrosc. 2014, 71, 98−104. (33) Besara, T.; Jain, P.; Dalal, N. S.; Kuhns, P. L.; Reyes, A. P.; Kroto, H. W.; Cheetham, A. K. Mechanism of the Order-Disorder Phase Transition, and Glassy Behavior in the Metal-Organic Framework [(CH3)2NH2]Zn(HCOO)3. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 6828−6832. (34) Asaji, T.; Ashitomi, K. Phase Transition and Cationic Motion in a Metal-Organic Perovskite, Dimethylammonium Zinc Formate [(CH3)2NH2][Zn(HCOO)3]. J. Phys. Chem. C 2013, 117, 10185− 10190. (35) Asaji, T.; Yoshitake, S.; Ito, Y.; Fujimori, H. Phase Transition and Cationic Motion in the Perovskite Formate Framework [(CH3)2NH2][Mg(HCOO)3]. J. Mol. Struct. 2014, 1076, 719−723. (36) Abhyankar, N.; Bertaina, S.; Dalal, N. S. On Mn2+ EPR Probing of the Ferroelectric Transition and Absence of Magnetoelectric Coupling in Dimethylammonium Manganes e Formate (CH3)2NH2Mn(HCOO)3, a Metal-Organic Complex with the PbFree Perovskite Framework. J. Phys. Chem. C 2015, 119, 28143− 28147. (37) Šimėnas, M.; Ciupa, A.; Ma̧czka, M.; Pöppl, A.; Banys, J. EPR Study of Structural Phase Transition in Manganese-Doped [(CH3)2NH2][Zn(HCOO)3] Metal-Organic Framework. J. Phys. Chem. C 2015, 119, 24522−24528. (38) Šimėnas, M.; Kultaeva, A.; Balciunas, S.; Trzebiatowska, M.; Klose, D.; Jeschke, G.; Maczka, M.; Banys, J.; Pöppl, A. Single Crystal Electron Paramagnetic Resonance of Dimethylammonium and Ammonium Hybrid Formate Frameworks: Influence of External Electric Field. J. Phys. Chem. C 2017, 121, 16533−16540. (39) Fu, D.-W.; Zhang, W.; Cai, H.-L.; Zhang, Y.; Ge, J.-Z.; Xiong, R.-G.; Huang, S. D.; Nakamura, T. A Multiferroic Perdeutero MetalOrganic Framework. Angew. Chem., Int. Ed. 2011, 50, 11947−11951. (40) Schweiger, A.; Jeschke, G. Principles of Pulse Electron Paramagnetic Resonance; Oxford University Press, 2001. (41) Schweiger, A. Pulsed Electron Spin Resonance Spectroscopy: Basic Principles, Techniques, and Examples of Applications. Angew. Chem., Int. Ed. Engl. 1991, 30, 265−292. (42) Rowan, L. G.; Hahn, E. L.; Mims, W. B. Electron-Spin-Echo Envelope Modulation. Phys. Rev. 1965, 137, A61−A71. (43) Mims, W. B. Pulsed Endor Experiments. Proc. R. Soc. London, Ser. A 1965, 283, 452−457. (44) Stoll, S.; Schweiger, A. EasySpin, a Comprehensive Software Package for Spectral Simulation and Analysis in EPR. J. Magn. Reson. 2006, 178, 42−55. (45) Neese, F. The ORCA Program System. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2012, 2, 73−78.

(46) Adamo, C.; Barone, V. Toward Reliable Density Functional Methods Without Adjustable Parameters: The PBE0Model. J. Chem. Phys. 1999, 110, 6158. (47) Chong, D. Recent Advances in Computational Chemistry; World Scientific Publishing, 1995; Vol. 1. (48) Kendall, R. A.; Dunning, T. H.; Harrison, R. J. Electron Affinities of the First-Row Atoms Revisited. Systematic Basis Sets and Wave Functions. J. Chem. Phys. 1992, 96, 6796−6806. (49) Neese, F.; Wennmohs, F.; Hansen, A.; Becker, U. Efficient, Approximate and Parallel Hartree-Fock and Hybrid DFT Calculations. A ’Chain-of-Spheres’ Algorithm for the Hartree-Fock Exchange. Chem. Phys. 2009, 356, 98−109. (50) Raitsimring, A.; Dalaloyan, A.; Collauto, A.; Feintuch, A.; Meade, T.; Goldfarb, D. Zero Field Splitting Fluctuations Induced Phase Relaxation of Gd3+ in Frozen Solutions at Cryogenic Temperatures. J. Magn. Reson. 2014, 248, 71−80. (51) Šimėnas, M.; Ciupa, A.; Ma̧czka, M.; Völkel, G.; Pöppl, A.; Banys, J. EPR of Structural Phase Transition in Manganese- and Copper-Doped Formate Framework of [NH3(CH2)4NH3][Zn(HCOO)3]2. J. Phys. Chem. C 2016, 120, 19751−19758. (52) Šimėnas, M.; Balciunas, S.; Trzebiatowska, M.; Ptak, M.; Maczka, M.; Völkel, G.; Pöppl, A.; Banys, J. Electron Paramagnetic Resonance and Electric Characterization of a [CH3NH2NH2][Zn(HCOO)3] Perovskite Metal Formate Framework. J. Mater. Chem. C 2017, 5, 4526−4536. (53) Völkel, G.; Brunner, W.; Windsch, W. Critical Fluctuations in Trissarcosine Calcium Chloride (TSCC) Observed by the Electron Spin Echo (ESE) Method. Solid State Commun. 1975, 17, 345−348. (54) Müller, H.-E.; Völkel, G.; Brunner, W.; Cevc, P.; Kurkin, I. N. Electron Spin Relaxation in Ferroelastic KH3(SeO3)2. Phys. Status Solidi B 1987, 141, 343−349. (55) Langen, H.; Montjoie, A. S.; Muller-Warmuth, W.; Stiller, H. NMR T1-analysis of Methyl Tunnelling in Molecular Crystals at Intermediate Barriers. Z. Naturforsch., A: Phys. Sci. 1987, 42, 1266− 1274. (56) Völkel, G.; Müller, H.-E.; Flohrer, W. Critical Anomaly of the Electron Spin-Lattice Relaxation in X-Irradiated Rochelle Salt. Phys. Status Solidi B 1981, 108, 501−506. (57) Pöppl, A.; Völkel, G.; Hoentsch, J.; Orlinski, S.; Klopperpieper, A. Electron Spin Relaxation of the PO32− Radical in Ferroelectric Betaine Phosphite and in the Proton Glass Betaine Phosphate/Betaine Phosphite. Chem. Phys. Lett. 1994, 224, 233−237. (58) Brunner, W.; Völkel, G.; Windsch, W.; Kurkin, I.; Shlenkin, V. Electron Spin-Lattice Relaxation in Mn2+-Doped Ferroelectric TSCC. Solid State Commun. 1978, 26, 853−855. (59) Völkel, G.; Brunner, W.; Windsch, W.; Steudel, D. Electron Spin-Echo (ESE) Investigations of Critical Paramagnetic Relaxation Anomalies in X-Irradiated Triglycine Sulphate (TGS:X). Phys. Status Solidi B 1979, 95, 99−105. (60) Böttcher, R.; Brunner, W.; Milsch, B.; Völkel, G.; Windsch, W.; Kirillov, S. EPR, ENDOR and ESE Investigations on Chromium Modified PbTiO3 Ceramics. Chem. Phys. Lett. 1986, 129, 546−549.

27232


Pulse EPR and ENDOR Study of Manganese Doped [(CH )2 NH ][Zn

Recommend Documents