Study of monolayer dispersion of MoO₃ on muscovite powder and diffusing behavior of MoO₃ on muscovite wafer by SR-TXRF

Abstract

The dispersion capacity of MoO₃ on the surface of muscovite powder was studied by X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD) techniques. The results indicated that MoO₃ can disperse onto the surface of a support as a sub-monolayer. The dispersion capacity was 5.88 mg MoO₃ (g muscovite)⁻¹, or 3.24 Mo atoms nm⁻². The diffusing process of MoO₃ on muscovite wafer was investigated by means of Synchrotron Radiation excited Total-reflection X-Ray Fluorescence spectroscopy (SR-TXRF). A stripe of MoO₃ on the muscovite wafer was used as the diffusion source. After thermal treatment in air or dry N₂, MoO₃ diffuses onto the surface of muscovite and forms a sub-monolayer. Sublimation of MoO₃ was also detected during the diffusion process. The diffusion rate in two heating atmospheres differed significantly. The diffusion rate of MoO₃ on muscovite in dry N₂ was greater than that in air. A possible explanation of the phenomena is that surface diffusion onto the muscovite surface plays a more important role in the process than transportation via gas phase during the diffusion of MoO₃.

---

Introduction

Xie and Tang¹ suggested that many oxides and salts can disperse spontaneously onto the surface of supports to form a monolayer or submonolayer, at a temperature well below the melting point of the dispersates. For each system, there is an utmost dispersion capacity, namely, the threshold. Monolayer dispersion is a ubiquitous phenomenon, which has been applied extensively in the preparation of heterogeneous catalysts and adsorbents. Since surface diffusion plays an important role in preparation and application of monolayer-type catalysts, it is significant to study the mechanism of the diffusion of supported metal oxides.

The study of supported molybdena catalysts continues to be interesting because of their numerous applications in petroleum refining, chemical production, and pollution control industries. Traditionally, most of the research about monolayer dispersion focuses on the systems that consist of molybdenum trioxide dispersed on a support with a high specific surface area.^2–11 Since the surface structure of the support is usually too complicated to be adequately elucidated, it is hard to observe the diffusing process in a real powder system clearly.

Knözinger's group has already paid attention to the study of the diffusing process and the diffusing mechanism. At first, they investigated systems similar to real powder ones, which consisted of an alumina wafer in contact with a molybdenum trioxide wafer. Since they used the oxide powder as supports, this lead to uncertainty when studying the spreading phenomena.¹² In their later work, they used model systems such as alumina and titania thin films and observed the diffusion in the close vicinity of the homogeneously deposited MoO₃ crystal.^13,14

In our previous work, we investigated MoO₃ diffusion behavior on alumina, titania and silica thin films.^15–17 By vacuum evaporation, a stripe of MoO₃ on the oxide thin film was used as the diffusion source. The diffusion source had much stronger contact with the support than previously reported ones.^13,14 Because the supports are nano-scale flat (according to Atomic Force Microscopy (AFM)) they are less smooth than the surface of muscovite since muscovite has an atomic flat plane.

In this paper, we study monolayer dispersion of MoO₃ on the surface of muscovite powder by means of XPS and XRD. At the same time, muscovite wafer was also used as the support, whose atomic flat surface enables us to study the diffusion process of MoO₃ clearly. A small amount of MoO₃ was vaporized in vacuum and deposited onto the slot of a mask on the muscovite wafer. Thus the stripe of MoO₃ on the muscovite surface was used as the diffusion source. After thermal treatment, the signals of line scans across the diffusion source and the diffusion region were acquired by SR-TXRF.

SR-TXRF is a very surface sensitive technique for trace element analysis. Since synchrotron radiation has many merits such as very high brightness, natural collimation and polarization etc., SR-TXRF is a more sensitive analysis method than conventional XRF in elemental detection sensitivity and space resolution. In Beijing Synchrotron Radiation Facility (BSRF), the minimum detection limit obtained by SR-TXRF can be as low as ng g⁻¹, or ppb in the micro area, much lower than that of XPS and Secondary Ion Mass Spectrometry (SIMS). Therefore, this method was used to analyze the samples with a monolayer or sub-monolayer of MoO₃ on the surface.

Experimental

Preparation of the samples

Muscovite powder was used as support. The BET surface area of the support was 7.6 m² g⁻¹. Muscovite powder was mechanically mixed with the required amount of MoO₃ in an agate mortar. After complete mixing, all the samples were calcined at 673 K for 8 h.

In order to study the diffusing behavior of MoO₃ on the muscovite surface, freshly cleaved muscovite wafer was prepared as the support. The wafer was cut to the size of 20 mm (X direction) × 10 mm (Y direction). Then, by using a mask with a slit 2 mm wide, MoO₃ was evaporated onto the muscovite surface in vacuum (JEE-4C Vacuum Evaporator) forming a stripe which was the diffusion source. After being heated in an air atmosphere (78 vol.% nitrogen, 21 vol.% oxygen, ca. 65% relative humidity) or in a dry N₂ atmosphere (99.995 vol.%, little water) at 673 K in a quartz tube furnace for different numbers of hours, the samples were prepared for further characterization.

Characterization of the samples

XRD was carried out by means of a Rigaku D/max 2000 X-ray powder diffractometer, employing Cu Kα (Ni filtered) radiation of wavelength 1.542 Å. The accelerating voltage was 50 kV and the anode current was 120 mA.

XPS spectra were acquired with a VG-ESCA LAB 5 spectrometer with Mg Kα X-ray source operated at 9 kV and 18.5 mA. The pressure in the chamber was less than 1.0 × 10⁻⁷ Torr. Electron binding energies were calibrated against the C1s emission at E_b = 284.8 eV. The XPS peak intensity ratio between Mo 3d and Si 2p was taken as a measurement of the relative content of MoO₃ on the surface of the muscovite.

The Synchrotron Radiation excited Total-reflection X-Ray Fluorescence (SR-TXRF) data were acquired by Beijing Synchrotron Radiation Facility. White light was used as the excitation source. After it passed through a 40 µm (X) × 80 µm (Y) slit, it contacted the sample with grazing incidence. The fluorescence spectra were recorded by the Si (Li) detector. Then the sample stage was moved perpendicular to the diffusion source (along the X direction) and in the plane parallel to the surface. Thus, a line scan was made. The origin of the scan is the center of the diffusion source determined by microscopy.

Results and discussion

Dispersion capacity of MoO₃ on muscovite powder

Since XPS is a surface sensitive technique, it can be used to study the surface state of MoO₃ supported on muscovite. The intensity ratio of the XPS peak I_Mo3d/I_Si2p as a function of MoO₃ content is illustrated in Fig. 1. The figure consists of two straight lines with different slopes; and at the intersection point the MoO₃ content is 5.22 mg g⁻¹ muscovite. This value is called the utmost dispersion capacity of MoO₃ on muscovite. When the MoO₃ content is below the value, the line slope is rather steep, which indicates that MoO₃ disperses on the muscovite surface as a highly dispersed monolayer.¹ With an increase in highly dispersed MoO₃ loading, I_Mo3d/I_Si2p increases quickly. However, beyond the value, the slope of the line becomes smaller, which reveals that the dispersity of MoO₃ goes down. Crystalline MoO₃ appears and might exist on the support surface of dispersed MoO₃ or mixed with it mechanically. Although the loading is increasing, the surface area of MoO₃ exposed to X-rays does not increase to the same degree. Therefore the increasing rate of I_Mo3d/I_Si2p slows.

Fig. 1 XPS peak intensity ratio I_Mo3d/I_Si2pversus the content of MoO₃.

The dispersion capacity of MoO₃ on muscovite is also obtained by XRD measurement. XRD quantitative phase analysis to determine the dispersion capacity of the metal oxides on the support has been discussed elsewhere.¹ The intensity ratio of the XRD peak I_{MoO3(2θ = 12.768)}/I_{muscovite (2θ = 17.738)} as a function of MoO₃ content is illustrated in Fig. 2.

Fig. 2 XRD peak intensity ratio I_MoO3/I_muscoviteversus the content of MoO₃.

In Fig. 2, for lower loading, no crystalline MoO₃ is detected. This indicates that the period structure of crystalline MoO₃ is destroyed, i.e. Mo species are highly dispersed onto the surface of muscovite powder as mentioned above. When the loading is increasing, crystalline MoO₃ begins to appear in an orthorhombic crystal form. The intensity ratio of the XRD peak I_MoO3/I_muscovite above the intercept point represents the relative amount of residual crystalline MoO₃ in the samples. The intercept of the straight line with the X-axis corresponds to the utmost dispersion capacity of MoO₃ on the surface of the muscovite. Below the threshold the amount of residual crystalline MoO₃ is zero. The intercept is 5.88 mg MoO₃ g⁻¹ muscovite, close to the value determined by the XPS technique. We determined the dispersion capacity of MoO₃ to be 5.88 mg g⁻¹ muscovite, equal to 3.24 Mo atoms nm⁻², since the specific surface area of muscovite is 7.6 m² g⁻¹. This calculation is based on the results of XRD because the error of the intensity ratio is a bit larger for XPS analysis of low loading samples.

For 1.4 mg MoO₃ g⁻¹ muscovite sample, no diffraction peaks of MoO₃ were seen in the XRD pattern. When physically mixing the system containing the same content of MoO₃ and muscovite powder without the following calcination, the characteristic diffraction peaks of MoO₃ in its XRD pattern could be seen. This means that the disappearance of diffraction peaks of MoO₃ in low loading samples was due to the dispersion of MoO₃ onto the surface of muscovite, not because of the detection limit of the X-ray diffractometer.

The diffusion of MoO₃ on muscovite wafer

To determine the species of the diffusion source, XPS and XRD measurements were used. The binding energies of Mo 3d_3/2 and Mo 3d_5/2 were 235.79 and 232.60 eV respectively, which confirms that the diffusion source is MoO₃ and it belongs to orthorhombic MoO₃ as shown by the results of the XRD measurements. Within the detection limits of SR-TRXF, no Mo species were found outside the stripe of the diffusion source without heating.

The SR-TXRF line scans for MoO₃/muscovite samples are illustrated in Fig. 3–6.

Fig. 3 SR-TRXF line scans in the X direction for MoO₃/muscovite samples heated for 4 h in two different atmospheres.

Fig. 4 SR-TRXF line scans in the X direction for MoO₃/muscovite samples heated for 12 h in two different atmospheres.

Fig. 5 SR-TRXF line scans in the X direction for MoO₃/muscovite samples heated for 18 h in two different atmospheres.

Fig. 6 SR-TRXF line scans in the X direction for MoO₃/muscovite samples heated for 24 h in two different atmospheres.

The common feature of the four figures is that Mo species were detected in the diffusion region. That means, after thermal treatment, MoO₃ can spontaneously diffuse onto the surface of muscovite. From all the figures, it can be clearly seen that the integral area of Mo Kα peak (about 17.444 keV) decreases dramatically outside the diffusion source. In the diffusion source, the concentration of MoO₃ of samples prepared in air is higher than in dry N₂. On the contrary, outside the diffusion source, the concentration of MoO₃ of samples prepared in dry N₂ is higher than in air. The second data point of the sample heated for 4 h in dry N₂ and the third data point of the sample heated for 12 h in air are significant. Perhaps for the places where the two data points lie, the roughness of the sample is a bit greater and lead to the deviation of the total reflection. Under the two heating atmospheres, increasing the heating time, the diffusion regions become more and more distant. With the same heating time, the diffusion distance in dry N₂ is longer than that in air. That means, the presence of water in the gas phase cannot enhance the spreading velocity.

From all of the above results, it can be seen that after being heated at 673 K for some time, MoO₃ spontaneously disperses onto the surface of muscovite.

XPS was used to analyze the whole diffusion region, which had been prepared by cutting away the diffusion source from the heated sample. For MoO₃/muscovite powder system prepared by mechanical mixing, when the content of MoO₃ is up to the threshold, the intensity ratio (I_Mo3d/I_Si2p) is 0.358. For the MoO₃/muscovite surface system, although the sample is heated for 24 h in dry N₂, in the diffusion region the intensity ratio (I_Mo3d/I_Si2p) is 0.132, much less than that of the threshold.

As with MoO₃ supported on Al₂O₃, SiO₂ and TiO₂ thin film systems,^16,17 besides the surface diffusion, sublimation of MoO₃ was also detected during thermal treatment under both atmospheres. Using SR-TXRF, Mo species were detected on the fresh muscovite surface placed beside the sample during the heating process. However, XPS cannot detect the occurrence of sublimation under both atmospheres due to the detection sensitivity. S. Günther¹⁴ found that the presence of water in the gas phase could enhance the spreading of MoO₃ on a flat Al₂O₃ support. Yan¹⁵ also suggested that moisture could accelerate the diffusion rate of MoO₃ on the surface of Al₂O₃. In this study the spreading rate of MoO₃ on muscovite in dry N₂ atmosphere is greater than that in air. This discrepancy may be related to the different surface properties of the supports or to different experimental conditions.

Although the transport mechanism for the diffusion of MoO₃ on a molecular level is still not fully understood, Knözinger and Taglauer have suggested three possible explanations for the phenomenon:¹⁸ (a) unrolling carpet mechanism: Mo oxide species, diffusing on top of crystalline MoO₃ particles, are trapped at the edge of the crystal before reaching the support surface; (b) surface diffusion on the support: Mo oxide species, detached from the MoO₃ crystals, diffuse and nucleate on the surface of the support foil and eventually recystallize at certain spots; (c) transport via the gas phase: Mo oxide species from the MoO₃ crystals sublimate and spread on the support via subsequent recondensation and finally recrystallize at certain spots.

In our study of diffusion in the MoO₃/muscovite system, surface diffusion and sublimation are both observed. The unrolling carpet mechanism is ruled out, because after a long time of heating the Mo oxide species on the diffusion region are still less than a monolayer. The mechanism that diffusion of MoO₃ completely depends on transport via the gas phase is also ruled out. The reason is as follows: MoO₃ is known to sublime at around 700 K, which was observed in our experiment. In the presence of water, the oxyhydroxide MoO₂(OH)₂ can be formed,¹² a compound with a higher vapor pressure than MoO₃. Therefore moisture can promote the sublimation; that is to say, the sublimation in air is more obvious. If transportation via the gas phase plays a critical role in the diffusion, then the spreading rate of MoO₃ on muscovite should be greater in air than in dry N₂. Although in air sublimation is more obvious than in dry N₂, MoO₃ in the gas phase can condense onto any surrounding place. If only considering the sublimation the amount of MoO₃ in the diffusion source in dry N₂ is greater than in air, which means the initial concentration of spread phase is greater in dry N₂ atmosphere. Therefore we propose that the surface diffusion is of a critical role on the diffusion of MoO₃ on muscovite.

From the results of XPS analysis for the diffusion region and sublimation, it can be concluded that sublimation has only a little effect upon the diffusion of MoO₃ on the muscovite wafer.

Conclusions

In the present report, we firstly obtained that MoO₃ can be highly dispersed as a sub-monolayer on the surface of muscovite powder by conventional mechanical mixing and heating. The dispersion capacity of MoO₃ on muscovite powder is 3.24 Mo atoms nm⁻² determined by XPS and XRD techniques. Meanwhile we investigated the diffusion behavior of MoO₃ on the flat surface of a muscovite wafer. It is proved, by SR-TXRF line scans, that MoO₃ can spontaneously disperse onto the surface of the muscovite, and form a sub-monolayer. The diffusion rate of MoO₃ on muscovite in dry N₂ atmosphere is greater than that in air. Sublimation of MoO₃ is observed during thermal treatment under two heating atmospheres. A possible explanation is that surface diffusion on the support plays a more important role compared with transportation via gas phase during the diffusion process of MoO₃.

Acknowledgements

This program is financially supported by National Natural Foundation of China (No. 29733080) and the Major State Basic Research Development Program (Grant No. G2000077503). We are grateful to Meijuan Zhao for cooperation during the preparing of samples.

References

1. Y. C. Xie and Y. Q. Tang, Adv. Catal., 1990, 37, 1.
2. R. Radhakrishnan, C. Reed and S. T. Oyama, J. Phys. Chem. B, 2001, 105, 8519.
3. M. Zdrazil, Catal. Today, 2001, 65, 301.
4. T. Olorunyolemi and R. A. Kydd, Catal. Lett., 2000, 65, 185.
5. F. Xu, Y. H. Hu, L. Dong and Y. Chen, Chin. Sci. Bull., 2000, 45, 214.
6. M. E. Harlin, A. O. I. Krause, B. Heinrich, C. Pharm-Huu and M. J. Ledoux, Appl. Catal. A-Gen., 1999, 185, 311.
7. D. Klissurski, D. Petridis, N. Abadzhieva and K. Hadjiivanov, Appl. Clay Sci., 1996, 10, 451.
8. T. Suzuki, H. I. Iwanami, T. Yoshizawa, H. Yamazaki and Y. Yoshida, Int. J. Hydrogen Energy, 1995, 20, 823.
9. R.W. Borry, Y. H. Kim, A. Huffsmith, J. A. Reimer and E. Iglesia, J. Phys. Chem. B, 1999, 103, 5787.
10. L. Dong and Y. Chen, J. Chem. Soc. Faraday Trans., 1996, 92, 4589.
11. L. J. Lakshmi and E. C. Alyea, Catal. Lett., 1999, 59, 73.
12. J. Leyrer, D. Mey and H. Knözinger, J. Catal., 1990, 124, 349.
13. S. Günther, M. Marsi, A. Kolmakov, M. Kiskinova, M. Noeske, E. Tagauer, G. Mestl, U. A. Schubert and H. Knözinger, J. Phys. Chem. B, 1997, 101, 10004.
14. S. Günther, L. Gregoratti, M. Kiskinova, E. Taglauer, P. Grotz, U. A. Schubert and H. Knözinger, J. Chem. Phys., 2000, 112, 5440.
15. J. F. Yan, N. Z. Wu, H. X. Zhang, Y. C. Xie, Y. Q. Tang, Y. F. Zhu and W. Q. Yao, Acta Phys-Chim. Sin., 1999, 15, 684.
16. W. M. Xu, J. F. Yan, N. Z. Wu, H. X. Zhang, Y. C. Xie, Y. Q. Tang and Y. F. Zhu, Surf. Sci., 2000, 470, 121.
17. W. M. Xu, J. Q. Xu, N. Z. Wu, J. F. Yan, Y. F. Zhu, Y. Y. Huang, W. He and Y. C. Xie, Surf. Interface Anal., 2001, 32, 301.
18. H. Knözinger and E. Taglauer, in Catalysis, The Royal Society of Chemistry, Cambridge, 1993, vol. 10, p. 1.

---