Improved Hydrogen De/absorption Kinetics of Magnesium (Mg) via Addition of Catalyst Y₂O₃

Introduction

Hydrogen gets attenuation in the whole world due to its dense energy content of 142 MJ/kg, compact, ecofriendly nature, and can easily be produced from water ¹. The main problem associates with its storage, which makes it not suitable for automobile as well as industrial applications. Its liquid and gas phase storage required cryogenic and high-pressure storage tanks. Both methods have a very high explosion risk and large energy lose due to boiling off ². The main challenge of hydrogen storage is closely linked to the need for the development of reliable and efficient hydrogen compression methods. Solid-state hydrogen storage provides a much easier, safer, effective, and economical way to store hydrogen ³. Throughout recent decades ammonia borane ^4,5 lithium compounds^6,7, lightweight metal hydrides (for example MgH₂, NaAlH₄, LiAlH₄, LiH, LaNi₅H₅, TiFeH₂, PdH₂) have been considered for their role as solid state materials used for storing hydrogen⁸.

In all above MgH₂ hasshown great potential as a solid-state hydrogen storage media because of its large hydrogen storage capacity (theoretical ~7.6 wt%), compact, most abounded element, low cost, and high reversibility^9,10. Due to high thermal stability and poor ab/desorption kinetics, it is not widely used in both automotive and industrial applications ¹¹. In later times, several attempts have been made to enhance sorption kinetics and heat dynamics of Mg/MgH₂, such as ball milling, catalysts, and alloys as dopants. For example, materials such as metals ^12-15, alloys ^16-21, oxide ^11,22-24, and halides ^25-28, have been doped with Mg/MgH₂.  Extensive research efforts have been dedicated to transitional metal oxides serving as efficient catalysts to store hydrogen in MgH₂. Oelerich et al. reported that the catalysing impact of TiO₂, V₂O₅, Cr₂O₃, Mn₂O₃, Fe₃O₄, and CuO on the hydrogenation properties of Mg are comparable²⁹. Fe₃O₄ demonstrated superior catalytic activity in the dehydrogenation reaction, which was succeeded by V₂O₅, Mn₂O₃, Cr₂O_3, and TiO₂. Remarkably, Barkhordarian et al. recorded the exceptional catalytic activity of Nb₂O₅ and the rapid hydrogen sorption kinetics of 0.2 mol.% Nb₂O₅-doped Mg with nanosized crystal lattice³⁰.

Xueping et al.³¹ examine the catalytic role of Y₂O₃ in promoting the dehydrogenation of NaAlH₄.  It was observed that hydrogen storage capacity rises with Y₂O₃ concentration. Where desorption rate firstly increases and then decreases with Y₂O₃ concentration. Furthermore, Liu et al.³² examined the hydrogenation properties of MgH₂-Y₂O₃/NiO hybrid. Their study revealed that the dehydrogenation temperature of MgH₂ reduces with an enhancement of Y₂O₃/NiO hybrid. They found that Y₂O₃ plays a superior role in enhancing hydrogen storage capacity, improving cycle life kinetics, and sorption kinetics of MgH₂.  Surface reactivity of Y₂O₃ powder toward hydrogen has been studied by Kogler et al. [33] and they found that due to the specific hydrophilicity of the oxide, sorption of water eventually takes place. The evidence supporting this surface-restricted mechanism is further strengthened by the absence of bulk structure and even at the highest reduction temperature (1173 K) morphological changes were observed upon reduction. To the authors’ knowledge, the catalytic effect of Y₂O₃ on MgH₂ has not been widely studied so far. In the present study, we conducted a comprehensive study of catalytic role of Y₂O₃ on the hydrogen storage properties and sorption kinetics of Mg/MgH₂.

Experimental Preparation

Nano-Composite Preparation

Mg and Y₂O₃ powder of purity 99.98% were used during the synthesis procedure. Mg-x wt% Y₂O₃ (x = 0, 5, 10) powder was loaded into a ball milling jar and then sealed with an O ring. Whole process is carried into a glove box filled with pure argon gas. Then sealed jar was transferred to ball milling machine (Model: Retsch PM 100 type).  Ball-Milling process was performed for 100 h at a speed of 150 rpm. To synthesis Mg-x wt %Y₂O₃ nanocomposites, ball to charge ratio of roughly 5:1 was used. Ball-Milling procedure was performed at room temperature under atmospheric condition.

Structural and Morphological Characterization

Catalytic role on crystal structure and crystalline dimension were examined by using cu-k_α radiation (λ=1.54 Å). XRD data was collected across the 20-80̊ range. The surface morphology of the synthesized samples was measured by Scanning electron microscope (SEM, Model Nova-Nano FE-SEM 450) at 15 kV. Chemical elemental composition was recorded by an EDX detector, which is inbuilt in SEM instrument.

Thermodynamics of Dehydrogenation

Thermodynamics of the hydrogen-depleting reaction of Mg and Mg-5 wt% Y₂O₃ nano-composites were measured via a Differential Scanning Calorimeter (DSC Model No. 7020). All DSC measurements were carried out at diverse heating rates of 5, 10, 15, and 20 ⁰C/min in scale of thermal conditions 30⁰C – 550⁰C. Roughly 5-6 mg sample was used for each DSC measurement. A constant nitrogen flow of 60 ml/min was maintained during the whole measurement.

Catalytic effect on dehydrogenation properties was carried out with Thermal Gravimetric Analyzer (TGA) measurement (STA Model Hitachi TGA-7300). Other Conditions are the same as of DSC measurement.

Pressure-Composition-Temperature (PCT) analysis

A sievert-type Homemade PCI/PCT setup was employed to measure the hydrogen storage attributes of MgH₂ – x wt% (x = 0, 5, 10). Roughly 1 gm sample was used for each PCT measurement. Before every measurement, a leakage test was performed for 24 h at 350 ⁰C. After that activation process was executed at 350⁰C and 2 Mpa H₂ pressure for 6 h followed by 6 cycles of activation. After each sample loading the same process was repeated. Then PCT   measurement was taken. Notably, Van’t Hoff equation was used to measure the enthalpy and entropy of re-hydrogenation reaction. Generally, enthalpy and entropy of most metal hydride are negative, except in certain instances.

Result and discussion

Figure 1 represents the XRD spectrum of Mg/MgH₂ and Mg- 5 wt% Y₂O₃ nanocomposites. XRD investigation reveals that the phase identity of Mg/MgH₂ and Mg- 5 wt% Y₂O₃ is maintained during ball milling. XRD spectrum indicates the existence of diverse phases within the nanocomposites, which possibly occur during ball-milling. The presence of these phases provides extra catalytic effects on the hydrogenation properties of Mg- 5 wt% Y₂O₃ nanocomposites. XRD data revealed diminished peak intensities and stretching of the Full width at half maxima (FWHM), value MgH₂ is 0.17, and for Mg+5 wt%Mg+Y₂O₃ is 0.2 signifying the presence of physical stress accumulation and reduction in particle dimension during ball milling. Ball-milling creates new active surfaces, resulting in enhanced hydrogenation properties of the nanocomposite. Analysing, the X-ray diffraction (XRD) spectra focus on the full width at half maxima (FWHM) and peak intensity, which are essential parameters for assessing crystallinity and grain size. A narrower FWHM signifies larger crystallite sizes and improved crystallinity, as broader peaks usually indicate smaller crystallites or greater structural disorder. Higher peak intensity reflects better crystallinity and larger grain sizes, indicating well-aligned and abundant crystalline planes. The lattice parameters for MgH₂ and Mg+5 wt%Mg+Y₂O₃ are as for MgH₂ a=b=4.49 Å &C=2.0 Å and for Mg+5 wt%Mg+Y₂O₃ are a = b = c= 10.606 Å as verified from JCPDS file no.- 01-079-6694 and JCPDS file no.- 01-089-5592

These values align well with previously reported data. The ratio c/2a lattice parameter for MgH₂ is 0.22 and for Mg+5 wt%Mg+Y₂O₃ is 0.5.

The crystallite size D was determined using Scherrer’s formula:

Where D is the average crystallite size, k is the shape factor, λ is the wavelength of Cu-Kα X-rays, β is the full width at half maxima (FWHM), θ is the Bragg angle. The value of D for MgH₂ is 79 nm and for Mg+5 wt%Mg+Y₂O₃ is 71.5 nm.

By calculating, the dimensions of dislocation lines per unit volume, the dislocation density(𝞭) is calculated using Williamson and Smallman’s formula is

The value for MgH₂ is 0.00016 nm^-2 and for Mg+5 wt%Mg+Y₂O₃ is 0.00019 nm^-2.

Unit Cell Volume (V)

The unit cell volume V for MgH₂ was calculated using the relation V = a² c and Mg+5 wt%Mg+Y₂O₃ is calculated by V=a³, where a and c are the lattice parameters values are 40.3202 ų and 1193.03 ų.

Microstrain (𝞊)

Stress is an important factor influencing the structural features of crystalline lattices when there is a geometric mismatch between their boundaries. The relationship illustrated below can be utilized to compute microstrain (𝞊).

The value for MgH₂ is 0.043855nm and for Mg+5 wt%Mg+Y₂O₃ is 0.04845 nm.

Then, it is summarised as MgH₂ has a slightly larger crystallite size, lower dislocation density, and larger unit cell volume than MgH₂– 5 wt% Y₂O₃. A slight peak shift for (110) was measured for MgH₂ with the addition of Y₂O₃. And two peaks of Y₂O₃ for the plane (222) and (431) was noted.

Figure 1: XRD spectrum of 100 h milled MgH2 and Mg-5 wt% Y2O3 nanocomposites. Click here to View Figure

SEM (Scanning electron Microscopy)

Surface morphology of all synthesis samples is shown in Figure 2.  Figure 2(a) represents the SEM micrograph of 100 h milled MgH₂ and 2(b) represents MgH₂– 5 wt%Y₂O₃ nanocomposites. The average crystalline size for MgH₂ was recorded at ~ 79 nm and for MgH₂– 5 wt% Y₂O₃ nanocomposites was in the range of ~71.5 nm. SEM imaging indicates a decrease in the particle size after ball milling, generating a refreshed surface area conducive to hydrogen absorption and release within the composites. Examination of SEM images shows that ball milling yields a nonuniform particle size distribution. Analysis of SEM images emphasize the dense aggregation of small particles in the vicinity of large particles and some clusters are present in the prepared samples. This property plays an important role in improving the kinetics of hydrogenation. Figure 2(c) and 2(d) represents the EDX mapping of 100 h milled MgH₂ and MgH₂– 5 wt%Y₂O₃ nanocomposites. The EDX analysis provides information on both elemental identification and quantitative composition. Hight of the peaks represents the chemical quantitive information of the nanocomposites.

Figure 2: FE-SEM for 100h milled (a) MgH2 (b) for Mg-5 wt%Y2O3 nanocomposites, EDX graphing for 100h milled (c) MgH2 (d) for Mg-5 wt%Y2O3 nanocomposites. Click here to View Figure

Figure 3 represent particle size distribution of MgH₂ and MgH₂– 5 wt% Y₂O₃ nanocomposites. For each plot about 200 data points have been recorded. The average particle size found for MgH₂ is 81.963 ± 32.045 nm and for MgH₂– 5 wt% Y₂O₃ is 81.021 ± 21.273 nm.

Figure 3: particle size distribution plots of 100 h milled (a) MgH2 and (b) MgH2– 5 wt% Y2O3 Click here to View Figure

Pressure-Composition-Temperature (PCT)

Pressure-composition-temperature (PCT) is carried out at diverse temperatures for 100 h milled Mg and Mg-5 wt%Y₂O₃. The maximum hydrogen absorption achieved for Mg is 5.2 wt% at 320 ̊ C as shown in Fig.4(a). The results indicate that as temperature increases, the plateau pressure shifts towards the higher end, correlating with a decrease in absorption capacity, and notably the plateau pressure does not persist unchanged with the temperature.

Fig4(b) represents the maximum absorption capacity achieved for Mg-5 wt% Y₂O₃ nanocomposites is 4.7 wt% at 280 ̊ C. The PCT plot recorded in the temperature range 280-340 ̊ C. PCT investigation indicate that with the increase of temperature absorption capacity reduces.  But the plateau pressure didn’t persist unchanged with the temperature.

Fig.4 (c) shows comparative study at temperature 280 ̊ C at different weight percentage. So, it shows that Mg- 5 wt%Y₂O₃ have higher hydrogen absorption capacity as compare to different weight percentage. From the PCT analysis it is clear that as we increase Y₂O₃ concentration absorption capacity of Mg/MgH₂ reduces.

Figure 4: shows PCT profile of (a) 100 h milled Mg/MgH2, (b)Mg-5wt%Y2O3 nanocomposites, and (c) Comparative study at different wt%. Click here to View  Figure

Enthalpy and entropy for hydrogenation reaction were determined applying the Van’t Hoff equation to PCT measurements and analyzing by Van’t Hoff plots. Fig. 5 represents the Van’t Hoff plot of 100 h milled (a) Mg/MgH₂ (b) Mg- 5 wt% Y₂O₃, and the calculated enthalpy and entropy for MgH₂ is 52.502 kJ/mol and 104.25 J/mol-k, for Mg-5 wt% Y₂O₃ is 76.995 kJ/mol and 149.236 J/mol-K, and for Mg-10 wt%Y₂O₃ is 55.156 kJ/mol and 108.976 J/mol-K respectively. So, from the results we conclude that the lowest enthalpy was obtained for Mg-5 wt% Y₂O₃ nanocomposites.  Entropy calculations are conducted to understand the transition from gaseous molecular hydrogen to dissolved atomic hydrogen. Enthalpy governs the heat released or absorbed during hydrogen uptake or release reaction.

Figure 5: Van’t Hoff Plot of 100 h milled (a) Mg/MgH2, (b) Mg- 5 wt%Y2O3 nanocomposites, Click here to View Figure

Thermodynamics and Kinetics of the dehydrogenation process

TGA and DSC analysis of MgH₂ and Mg-5 wt% Y₂O₃ nanocomposites at different heating rates are shown in fig. 6 (a, b). Thermogravimetric analysis (TGA) of MgH₂ shows that the maximum weight loss is observed at 15 ̊ C/min and loss is reached up to 3.1 % at 460 ̊ C and Mg-5 wt% Y₂O₃ shows that the maximum weight loss is 1.6% up to 477 ̊ C at the heating scale of 20 ̊ C/min

Thermodynamics analysis of Mg and Mg-5 wt%Y₂O₃ nanocomposites were examined through Differential Scanning Calorimetry (DSC) as shown in fig 6 (c, d). The decomposition of metal hydride often leads to an exothermic reaction. The exothermic reaction represents the hydrogen is liberated through the decomposition of MgH₂ at 449 ⁰C at the heating scale of 15 ⁰C. As, the heating rate decreases the peak shifted towards lower side. Heat flow profile obtained by the Mg-5 wt%Y₂O₃ show decomposition at 447 ⁰C at the heating scale of 20 ⁰C.  With decreasing heating rate, the peak tends to move towards lower end. DSC and TGA analysis indicate that on set temperature of Mg/MgH₂ shifted toward lower end with increasing Y₂O₃ concentration.

Figure 6: (a, b) shows TGA for 100 h milled MgH2 and Mg-5 wt% Y2O3 nanocomposites (c) shows comparative study at the rate of 10 ̊ C/min Click here to View Figure

Figure 6: (c, d) shows DSC for 100 h milled MgH2 and Mg-5 wt% Y2O3 nanocomposites. Click here to View Figure

The activation energy (E_a) for dehydrogenation reaction of the synthesized nanocomposites has been derived. By performing TGA measurements at varying heating rates under the correspondent experimental setup, the activation energy can be determined using Kissinger equation

Where k= β/ T_p² , β represents the heating rate, T_P represents the peak temperature, R represents the gas constant and E_a represents the activation energy. When plotting a graph of ln(k) contrary to 1/T, slope corresponds to -E_a/R. The activation energy value is accumulated from the slope.  Fig.7 shows activation energy for (a) MgH₂ and (b) Mg-5 wt% Y₂O₃. The calculated activation energy for MgH₂ is 218.57 kJ/mol and for Mg-5 wt% Y₂O₃ is 203.03 kJ/mol. The activation energy acts as threshold that must be surpassed to initiate the liberation of hydrogen and its nanocomposites. Figure 7 illustrates decrease of activation energy with catalyst concentration. The behavior affected by factors such as catalytic impact, particle dimension, surface area, and catalytic properties affects the desorption rates. Sample with higher activation energy are indicative of heightened thermal stability.

Figure 7: Kissinger plot of 100 h milled (a) MgH2, and (b) for Mg-5 wt% Y2O3 nanocomposites. Click here to View Figure

Conclusion

The study focused on examination how catalysts Y₂O₃ influence structural, morphological, and ab/desorption kinetics of Mg/MgH₂ by utilizing XRD, FE-SEM with EDX and simultaneous TGA-DSC techniques. XRD and SEM investigation shows that ball-milling provides non-uniform mixing and creates new active fresh surface, which leads to increase hydrogen ab/desorption kinetics as well as capacity of Mg/MgH₂. PCT analysis revels that the highest absorption achieved for Mg/MgH₂ is at 5.2 wt% at 320 ̊ C and for Mg-5 wt% Y₂O₃ nanocomposites is 4.04 wt% at 280 ⁰C. From PCT analysis it is clear that that the hydrogen absorption capacity of Mg/MgH₂ reduces with increasing Y₂O₃ concentration. The DSC analysis indicates that that the decomposition of metal hydride often leads to an exothermic reaction. Through the decomposition of MgH₂ hydrogen is liberated at 449 ⁰C and Mg-5 wt%Y₂O₃ shows decomposition at 447 ⁰C. DSC analysis revels that the onset hydrogenation temperature of Mg/MgH₂ shifted towards lower end with addition of Y₂O₃ catalyst. TGA analysis shows that activation energy for MgH₂ is 218.57 kJ/mol and for Mg-5 wt% Y₂O₃ is 203.03 kJ/mol. So, by the addition of catalysts activation energy decreases.

Acknowledgment

The authors would like to acknowledge the Department of Physics, University of Rajasthan, Jaipur for their assistance with TGA and DSC facilities respectively. Also thankful to MNIT Jaipur for Characterization XRD and FE-SEM with EDX facility.

Conflict of Interest

The authors have no relevant financial or non-financial interest to disclose.

Funding Sources

There is no funding Sources

Ethical Approval

All the ethical Standard maintain during the research. On the behalf of all authors Corresponding Authors (Dr. Chhagan Lal) demonstrate that we have followed the accepted ethical standard of genuine research study in the research article.

Data Availability Statement

On the behalf all authors Dr. Chhagan Lal declare that all authors have freely (Voluntarily) entered the research with full information and that consent has been taken from all authors before entering the research.

Authors Contribution

Madhu Yadav, Prof. S.N. Dolia involved in formal Investigation, Methodology and writing, Deepak Kumar Yadav helped in formal analysis of measurements and the entire work was done in the supervision of Dr. Chhagan Lal. 

Future Directions

Future directions may involve further optimization of nanocomposite synthesis parameters, exploration of alternative additives, and investigation of the performance of MgH₂ nanocomposites under realistic operating condition. Additionally, studies focusing on scale-up and commercial viability of the developed materials warranted to facilitate their practical application in hydrogen storage system.

References

1. Chawla, K; Yadav, DK; Sharda, P; Lal, N; Sharma, S; Lal, C. International Journal of Hydrogen Energy 2020; 45(44):23971-23976.
   CrossRef
2. Chawla, K; Yadav, DK; Bajpai, A; Kumar, S; Lal, C. Environmental Science and Pollution Research 2021; 28:3872-3879.
   CrossRef
3. Sakintuna,B; Lamari-Darkrim,F; Hirscher,M. International journal of hydrogen energy 2007; 32(9):1121-1140.
   CrossRef
4. Hu, MG; Geanangel, RA; Wendlandt, WW. Acta, 1997,23,249-255.
   CrossRef
5. Stephens, FH; Pons, V; Baker, RT; Dalton Trans.2007, doi:10.1039/B7030532C.
6. Chen, P; Xiong, Z; Luo, J; Lin, J;Tan ,KL. Nature,2002,420: 302–304.
   CrossRef
7. Chen, P; Xiong, Z; Luo, J; Lin, J; Tan, KL.J. Phys. Chem. B,2003,107,10967–10970.
   CrossRef
8. Jain, I; Jain, P; Jain, A. J Alloy Compd 2010,503(2),303-39.
   CrossRef
9. Schlapbach, L; Zuttel, A. Nature,2001, 414:353-358.
   CrossRef
10. Jain, IP; Lal, C; Jain, A. Int. J. Hydrogen Energy, 2010,35,5133-5144
    CrossRef
11. Yadav, DK; Chawla, K; Jain, IP; Lal, C. Environmental Science and Pollution Research 2021,28,3866-3871.
    CrossRef
12. Ouyang, LZ; Cao, ZJ; Wang, H; Liu, JW; Sun, DL; Zhang, QA; Zhu, M. Int. J. Hydrogen Energy 2013,38,8881-8887.
    CrossRef
13. Hanad, N.; Ichikawa, T.; Fujii, H., J. Phys. Chem. B,2005,109,7188-7194.
    CrossRef
14. Dai, J.H.; Song, Y.; Yang, R. J. Phys. Chem. C,2010,114,11328-11334.
    CrossRef
15. Mahmoudi, N; Kaflou A.; A. Simchi A. J. Power Sources,2011, 196,4604-4608.
    CrossRef
16. Zhong, H.C.; Wang, H.; Ouyang, L.Z. Int. J. Hydrogen Energy,2014,39,3320-3326.
    CrossRef
17. Yu, X.B.; Yang, Z.X.; Liu, H.K.; Grant, D.M.; Walker, G.S. Int. J. Hydrogen Energy,2010,35,6338-6344.
    CrossRef
18. Janot, R.; Darok, X.; Rougier, A., Aymard, L., Nazri, G.A.; Tarascon, J.M., J. Alloys Compd.,2005,404-406 ,293-296.
    CrossRef
19. Chawla, K., Sharma, G.; Bajpai, A.; Roy, P.K.; Jain, I.P.; Lal C. International Journal of Hydrogen Energy,2024, 57,1253-1262.
    CrossRef
20. Chawla, K., Yadav, D.K., Bajpai, A., Jain, I.P. and Lal, C. Sustainable Energy Technologies and Assessments,2022,51,101981.
    CrossRef
21. Friedrichs, O.; Aguey -Zinsou F.; Ares Fernandez J.R.; Sanchez-Lopez J.C.; Justo A.; Klassen T.; Bormann R.; Fernandez A. Acta Materialia,2006,54 ,105-110.
    CrossRef
22. Révész, A.; Fátay, D.; Zander, D.; Spassov, T., Journal of Metastable & Nanocrystalline Materials,2015,24-25,447-450.
    CrossRef
23. Cabo, M.; Garroni, S.; Pellicer, E.; Milanese, C.; Giella, A.; Marini, A.; Rossinyol, E.; Surinach, S.; Baro, M.D., Int. J. Hydrogen Energy;2011;36; 5400-5410.
    CrossRef
24. Yadav,D.K.;Chawla,K.;Lal,N.;Choudhary,B.L.;Lal,C. MaterialsToday:Proceedings, 2021;46;2326-2329.
    CrossRef
25. Yavari, A.R.; Le Moulec, A.; Castro F.R.D.; Deledda, S.; Friedrichs, O.; Botta, W.J.; Vaughan, G.; Klassen, T.; Fernandez, A.; Kvick, A. Scripta Materialia,2005,52,719-724.
    CrossRef
26. Jin, S.A.; Shim, J.H.; Cho, Y.W.; Yi, K.W.; J. Power Sources,2007, 172, 859-862.
    CrossRef
27. S. Kumar, A. Jain, S. Yamaguchi, H. Miyaoka, T. Ichikawa, A. Mukherjee, G.K. Dey, Y. Kojima, Int. J. Hydrogen Energy,2017,42,6152-6159.
    CrossRef
28. Ismail, M.; Energy,2015, 79,177-182.
    CrossRef
29. Oelerich, W.; Klassen, T.; Bormann, R. J Alloys Compd, 2001,315,37-42.
    CrossRef
30. Barkhordarian, G.; Klassen, T.; Bormann, R. Scr Mater,2003,49,213-217.
    CrossRef
31. Xueping, Z., Guo, X., Shenglin, L., Xin, F. and Jiaojiao, Z., 2013. Effect of CeO2 and Y2O3 as Catalysts on Hydrogen Desorption Properties of NaAlH4. Rare Metal Materials and Engineering, 42(7),1321-1324.
    CrossRef
32. Liu, Y.; Wang, S.; Li, Z.; Gao, M.; Liu, Y.; Sun, W.; Pan, H.Processes, 2021, 9(5), 892.
    CrossRef
33. Kogler, M.; Köck, E.M.; Bielz, T.; Pfaller, K.; Klötzer, B.; Schmidmair, D.; Perfler L.; Penner S.; The Journal of Physical Chemistry C, 2014,118(16), 8435-8444.
    CrossRef

---

This work is licensed under a Creative Commons Attribution 4.0 International License.