- Email: [email protected]
Neutron diffraction study of molybdenum dioxide
JOURNAL
OF SOLID
STATE
CHEMISTRY
19,36.5-368
(1976)
Neutron Diffraction Study of Molybdenum Dioxide JAYASREE HALLAM,*
GHOSE,* NORMAN N. DAVID A. READ*
GREENWOOD,?
GRAHAM
C.
AND
Department of Physics,* and Department of Inorganic and Structural CTniversity of Leeds, Leeds LS2 9JT, England
Chemistry,?
Received July 26, 1976 The magnetic properties of powdered, crystalline molybdenum dioxide have been studied by neutron diffraction at 293°K and 673°K. The results establish the absence of a permanent magnetic moment on the MO(W) ions in this compound and confirm that MoOz has a small temperatureindependent magnetic susceptibility. This latter is considered to arise from Pauli paramagnetism of the collective electrons.
Introduction
A) as shown in Fig. 1. This results in the formation of pairs of metal atoms and a There is some uncertainty in the literature consequent distortion of the Moo6 octahedra concerning the detailed magnetic properties which lowers the symmetry from tetragonal of molybdenum dioxide and we therefore for rutile to monoclinic for the MOO, strucundertook a neutron diffraction study on ture. powdered MOO, over the temperature range In an isolated Mo(IV) ion there are two 4d 293-673°K to seewhether the reported small electrons which could contribute to the paraparamagnetic susceptibility was indeed temp- magnetic moment, but if full double bonds erature independent. A further aim was to exist between the pairs of Mo(IV) ions, then determine the magnetic form factor of the no contribution to the paramagnetic moments Mo(IV) ion, if the molybdenum ions were by these d electrons would be expected. In found to have a permanent moment. Our 1915 Wedekind and Horst (4) reported that interest in this problem arosefrom recent work MOO, was weakly paramagnetic, the value of on the magnetically compensated inverse the moment being 4.2 pB. Later, Selwood (5) spine1FeZMoO in which we showed that the reported that MoOz had almost zero magMo(IV) ion had two unpaired electrons which netic susceptibility and no antiferromagnetic played a crucial role in the overall magnetic NCel point up to 1100°C. Tarama et al. (6), behaviour of the compound (I). in their study of the electric conductivity and Molybdenum dioxide has a deformed rutile- magnetic susceptibility of V,O, catalysts type structure (i?,3) in which eachmolybdenum containing small amounts of MOO, or LXis coordinated by six oxygen atoms to form A1,03 found that the addition of small MOO, octahedra. The octahedra are joined amounts of MOO, increased the effective by sharing edges to form strings which are magnetic moment (pu,rf), lowered the Neel mutually connected into a three-dimensional temperature, and decreased the activation structure by corner-sharing of the octahedra. energy for electric conductivity (E,,). Rast (7), In the ideal rutile structure, the metal atoms are working on alumina-supported MOO,, found equidistant within the strings but in the that crystalline MOO, showed a temperature MOO, structure the metal-metal distances are independent magnetic susceptibility of +0.33 alternately short (2.5106 A) and long (3.1118 x lop6 emu gg’ and that dilution of the MOO,
Copyright Q 1976 by Academic Press, Inc. All rights of reproduction in any form reserved. Printed in Great Britain
365
366
CHOSE ET AL.
constants unchanged. The pure compounds were weakly paramagnetic (xlcr< 100 x 10b6) emu g-l, but doping raised the susceptibility markedly to -2500 x 10e6.Resistancestudies showed that these materials were metallic conductors, the room temperature specific resistivity being of the order of (10-4-10-3) ohm cm, decreasingby one order of magnitude at liquid nitrogen temperature. Doping also substantially lowered the conductivity. The present experiments were designed to obtain a more intimate insight into the magnetic properties of the MO(W) atoms in pure molybdenum dioxide. Experimental
Measurements were carried out on MOO, powder of 99.999% purity. Magnetization was measured from liquid helium temperature to 400°K in Foner vibrating magnetometer. Neutron diffraction was done at 293°K and 673°K on ‘Curran’ at AERE Harwell. Results and Discussion FIG. 1. Structure of molybdenum dioxide showing
The neutron diffraction pattern at 293°K is shown in Fig. 2, and Table I gives the values of calculated and experimental integrated on the alumina increased the susceptibility intensities at 293°K and 673°K. In addition of the MOO, considerably. For the highest to the reflections indexed from X-ray data, dilution with an atomic ratio of MO/AI = three weak peaks were observed at low angles l/99, the MOO, susceptibility was +27 x in the neutron diffraction data both at 293°K 10m6at 78”K, +9.8 x lop6 at 195°K and+6.5 x and 673°K. The peaks at 28 N 22.9” and 21.4” lO+j at 295°K. His results showed that plots are probably the 1/2 reflections for (31i, of l/x versus T obeyed the Curie Law for the 220) and (201, 22i), respectively. However, most highly diluted MoOz, whereas plots for the reflection at 20 = 20.20” could not be the preparations richer in MOO, showed a identified. The data at 673°K indicate a very curvature towards the temperature axis. small thermal expansion of the lattice without He suggestedthat the increase in paramagnet- actually causing any deformation. The small, positive, temperature-indeism on dilution was probably caused by an increasing number of unpaired electrons due pendent magnetic susceptibility of 0.33 x 10m6 to a decrease in the number of MO-MO emu g-’ reported by Rest (7) is consistent with bonds present in the crystal lattice of MOO,. our own measurements, and demonstrates Recently Ben-Dor and Shimony (8) reported the absenceof any permanent moment on the on the crystal structure, magnetic suscepti- molybdenum atoms in MOO,. On the other bility, and electrical conductivity of pure, and hand, the moment per molybdenum atom in NiO-doped molybdenum(IV) oxide and tung- MOO, diluted with A&O3 (as measured by sten(W) oxide single crystals. X-ray diffrac- Rest) extrapolates at infinite dilution to tion showed the crystal to be monoclinic and p,rr = 2.76 pB indicating a spin of S = 0.98 for doping up to 5% left the crystallographic the isolated MO atom, on the assumption of
MO-MO pairsand consequent distortions from theideal rutile structure.
NEUTRON
DIFFRACTION
OF MOLYBDENUM
367
DIOXIDE
TABLE
I
POWDER NEUTRON DIFFRACTION INTENSITIES OF MoOz AT 293°K AND 673°K I ohs
i
Ir75 Counter Angle (2 8)
FIG. 2. Neutron
diffraction pattern at 293°K.
complete orbital quenching, consistent with two aligned 4d-electrons. In MOO, the two molybdenum 4d electrons are in the tz9 orbitals. These overlap along the c-axis, i.e. d,, and d,,, forming metal-metal double bonds. The energy levels of these bonds fall in the same range as the span of the MO-O TI* band and the overlap of these bonds with the rc* band gives rise to metallic conductivity, as first proposed by Goodenough (9). Rest and others indicated that by diluting MOO, with a magnetically inert matrix these bonds are broken and the two 4d electrons of MO become free to contribute to the magnetic susceptibility. However, the distorted rutile structure of MOO, is due to the presence of these double bonds, so when they are broken the structure should change. Rest has given no information regarding the change in structure, although he has mentioned that molydenum compounds were not detected by X-rays in the mixtures with small amounts of MOO,. Experiments by Ben-Dor and Shimony (8) are not very helpful in this context as they have used nickel, a paramagnetic ion, as a dopant. It would, however, be interesting to study the influence of a nonmagnetic dopant on the crystallographic and magnetic properties of Mn02.
hkl
I CllC
100
1
011 110 102 202 211 020 002 200 111 212 021 210 302 122 121 215 222 113 311 022 220 222 300 313 13T 03 1 01 I 322
loo 1 7.1 90.5 46.8 7.6 8.9 0.11 89.9 28.8 93.4 8.3 21.8 0.4 36.7 30.5 6.2 4.2 119.8 35.1 7.4 1.1 10.4 3.0 10.0 8.7 1.9
293°K
673°K
0
0
100
100
0
0
38
36.4
103 117
94.8 111
18.4
16.7
291
267
0
0
35
32
8.45
7.2
-
From Rest’s results it might have been supposed that the two 4d electrons of molybdenum do not contribute wholly towards the CJ+ II bond formation and a very small contribution is responsible for the small moment. If this were so it should show in the neutron diffraction studies. However, comparison of the neutron diffraction peak intensities at 293°K and 673°K show no magnetic contributions in the reflections at smaller angles. Also, the absence of a strongly sloping background probably indicates the absence of paramagnetic diffuse scattering. This confirms the temperature-independent nature of the magnetic susceptibility of MOO,.
368
GHOSE ET AL.
We conclude that the very small magnetic moment of Mo(IV) in MOO, is due to Pauli paramagnetism arising from the collective electrons in the overlapping MO-O z* band which itself is similar in energy to the MO-MO o and n bonds. We thank the Neutron Diffraction Group at Harwell for the use of their facilities and the SRC for financial support. References 1. J. GHOSE, N. N. GREENWOOD, G. C. HALLAM, AND D. A. READ, J. Solid State Chem. 11, 239 (1974).
2. A. MAGNELI AND G. ANDERSSON,Acta Chim. &and. 9, 1378 (1955). 3. B. G. BRANDT AND A. C. SKAPSKI, Acta Chim. Stand. 21,661 (1967). 4. E. WEDEKIND AND C. HORST,Ber. 48, 105 (1915).
5. P. W. SELWOOD,“Magnetochemistry,” p. 336, Interscience, New York (1956).
2nd ed.,
6. K. TARAMA, S. TERANISHI, K. HATTORI, AND A. YASUI, Mem. Fat. Eng., Kyoto Univ. 31, 611 (1969). 7. E. RUST, J. Amer. Chem. Sot. 81,3843 (1959). 8. L. BEN-D• R AND Y. SHIMONY, Mat. Rex Bull. 9, 837 (1974). 9. J. B. GOODENOUGH, Bull. Sot. Chim. France 4,
1200 (1965).
OF SOLID
STATE
CHEMISTRY
19,36.5-368
(1976)
Neutron Diffraction Study of Molybdenum Dioxide JAYASREE HALLAM,*
GHOSE,* NORMAN N. DAVID A. READ*
GREENWOOD,?
GRAHAM
C.
AND
Department of Physics,* and Department of Inorganic and Structural CTniversity of Leeds, Leeds LS2 9JT, England
Chemistry,?
Received July 26, 1976 The magnetic properties of powdered, crystalline molybdenum dioxide have been studied by neutron diffraction at 293°K and 673°K. The results establish the absence of a permanent magnetic moment on the MO(W) ions in this compound and confirm that MoOz has a small temperatureindependent magnetic susceptibility. This latter is considered to arise from Pauli paramagnetism of the collective electrons.
Introduction
A) as shown in Fig. 1. This results in the formation of pairs of metal atoms and a There is some uncertainty in the literature consequent distortion of the Moo6 octahedra concerning the detailed magnetic properties which lowers the symmetry from tetragonal of molybdenum dioxide and we therefore for rutile to monoclinic for the MOO, strucundertook a neutron diffraction study on ture. powdered MOO, over the temperature range In an isolated Mo(IV) ion there are two 4d 293-673°K to seewhether the reported small electrons which could contribute to the paraparamagnetic susceptibility was indeed temp- magnetic moment, but if full double bonds erature independent. A further aim was to exist between the pairs of Mo(IV) ions, then determine the magnetic form factor of the no contribution to the paramagnetic moments Mo(IV) ion, if the molybdenum ions were by these d electrons would be expected. In found to have a permanent moment. Our 1915 Wedekind and Horst (4) reported that interest in this problem arosefrom recent work MOO, was weakly paramagnetic, the value of on the magnetically compensated inverse the moment being 4.2 pB. Later, Selwood (5) spine1FeZMoO in which we showed that the reported that MoOz had almost zero magMo(IV) ion had two unpaired electrons which netic susceptibility and no antiferromagnetic played a crucial role in the overall magnetic NCel point up to 1100°C. Tarama et al. (6), behaviour of the compound (I). in their study of the electric conductivity and Molybdenum dioxide has a deformed rutile- magnetic susceptibility of V,O, catalysts type structure (i?,3) in which eachmolybdenum containing small amounts of MOO, or LXis coordinated by six oxygen atoms to form A1,03 found that the addition of small MOO, octahedra. The octahedra are joined amounts of MOO, increased the effective by sharing edges to form strings which are magnetic moment (pu,rf), lowered the Neel mutually connected into a three-dimensional temperature, and decreased the activation structure by corner-sharing of the octahedra. energy for electric conductivity (E,,). Rast (7), In the ideal rutile structure, the metal atoms are working on alumina-supported MOO,, found equidistant within the strings but in the that crystalline MOO, showed a temperature MOO, structure the metal-metal distances are independent magnetic susceptibility of +0.33 alternately short (2.5106 A) and long (3.1118 x lop6 emu gg’ and that dilution of the MOO,
Copyright Q 1976 by Academic Press, Inc. All rights of reproduction in any form reserved. Printed in Great Britain
365
366
CHOSE ET AL.
constants unchanged. The pure compounds were weakly paramagnetic (xlcr< 100 x 10b6) emu g-l, but doping raised the susceptibility markedly to -2500 x 10e6.Resistancestudies showed that these materials were metallic conductors, the room temperature specific resistivity being of the order of (10-4-10-3) ohm cm, decreasingby one order of magnitude at liquid nitrogen temperature. Doping also substantially lowered the conductivity. The present experiments were designed to obtain a more intimate insight into the magnetic properties of the MO(W) atoms in pure molybdenum dioxide. Experimental
Measurements were carried out on MOO, powder of 99.999% purity. Magnetization was measured from liquid helium temperature to 400°K in Foner vibrating magnetometer. Neutron diffraction was done at 293°K and 673°K on ‘Curran’ at AERE Harwell. Results and Discussion FIG. 1. Structure of molybdenum dioxide showing
The neutron diffraction pattern at 293°K is shown in Fig. 2, and Table I gives the values of calculated and experimental integrated on the alumina increased the susceptibility intensities at 293°K and 673°K. In addition of the MOO, considerably. For the highest to the reflections indexed from X-ray data, dilution with an atomic ratio of MO/AI = three weak peaks were observed at low angles l/99, the MOO, susceptibility was +27 x in the neutron diffraction data both at 293°K 10m6at 78”K, +9.8 x lop6 at 195°K and+6.5 x and 673°K. The peaks at 28 N 22.9” and 21.4” lO+j at 295°K. His results showed that plots are probably the 1/2 reflections for (31i, of l/x versus T obeyed the Curie Law for the 220) and (201, 22i), respectively. However, most highly diluted MoOz, whereas plots for the reflection at 20 = 20.20” could not be the preparations richer in MOO, showed a identified. The data at 673°K indicate a very curvature towards the temperature axis. small thermal expansion of the lattice without He suggestedthat the increase in paramagnet- actually causing any deformation. The small, positive, temperature-indeism on dilution was probably caused by an increasing number of unpaired electrons due pendent magnetic susceptibility of 0.33 x 10m6 to a decrease in the number of MO-MO emu g-’ reported by Rest (7) is consistent with bonds present in the crystal lattice of MOO,. our own measurements, and demonstrates Recently Ben-Dor and Shimony (8) reported the absenceof any permanent moment on the on the crystal structure, magnetic suscepti- molybdenum atoms in MOO,. On the other bility, and electrical conductivity of pure, and hand, the moment per molybdenum atom in NiO-doped molybdenum(IV) oxide and tung- MOO, diluted with A&O3 (as measured by sten(W) oxide single crystals. X-ray diffrac- Rest) extrapolates at infinite dilution to tion showed the crystal to be monoclinic and p,rr = 2.76 pB indicating a spin of S = 0.98 for doping up to 5% left the crystallographic the isolated MO atom, on the assumption of
MO-MO pairsand consequent distortions from theideal rutile structure.
NEUTRON
DIFFRACTION
OF MOLYBDENUM
367
DIOXIDE
TABLE
I
POWDER NEUTRON DIFFRACTION INTENSITIES OF MoOz AT 293°K AND 673°K I ohs
i
Ir75 Counter Angle (2 8)
FIG. 2. Neutron
diffraction pattern at 293°K.
complete orbital quenching, consistent with two aligned 4d-electrons. In MOO, the two molybdenum 4d electrons are in the tz9 orbitals. These overlap along the c-axis, i.e. d,, and d,,, forming metal-metal double bonds. The energy levels of these bonds fall in the same range as the span of the MO-O TI* band and the overlap of these bonds with the rc* band gives rise to metallic conductivity, as first proposed by Goodenough (9). Rest and others indicated that by diluting MOO, with a magnetically inert matrix these bonds are broken and the two 4d electrons of MO become free to contribute to the magnetic susceptibility. However, the distorted rutile structure of MOO, is due to the presence of these double bonds, so when they are broken the structure should change. Rest has given no information regarding the change in structure, although he has mentioned that molydenum compounds were not detected by X-rays in the mixtures with small amounts of MOO,. Experiments by Ben-Dor and Shimony (8) are not very helpful in this context as they have used nickel, a paramagnetic ion, as a dopant. It would, however, be interesting to study the influence of a nonmagnetic dopant on the crystallographic and magnetic properties of Mn02.
hkl
I CllC
100
1
011 110 102 202 211 020 002 200 111 212 021 210 302 122 121 215 222 113 311 022 220 222 300 313 13T 03 1 01 I 322
loo 1 7.1 90.5 46.8 7.6 8.9 0.11 89.9 28.8 93.4 8.3 21.8 0.4 36.7 30.5 6.2 4.2 119.8 35.1 7.4 1.1 10.4 3.0 10.0 8.7 1.9
293°K
673°K
0
0
100
100
0
0
38
36.4
103 117
94.8 111
18.4
16.7
291
267
0
0
35
32
8.45
7.2
-
From Rest’s results it might have been supposed that the two 4d electrons of molybdenum do not contribute wholly towards the CJ+ II bond formation and a very small contribution is responsible for the small moment. If this were so it should show in the neutron diffraction studies. However, comparison of the neutron diffraction peak intensities at 293°K and 673°K show no magnetic contributions in the reflections at smaller angles. Also, the absence of a strongly sloping background probably indicates the absence of paramagnetic diffuse scattering. This confirms the temperature-independent nature of the magnetic susceptibility of MOO,.
368
GHOSE ET AL.
We conclude that the very small magnetic moment of Mo(IV) in MOO, is due to Pauli paramagnetism arising from the collective electrons in the overlapping MO-O z* band which itself is similar in energy to the MO-MO o and n bonds. We thank the Neutron Diffraction Group at Harwell for the use of their facilities and the SRC for financial support. References 1. J. GHOSE, N. N. GREENWOOD, G. C. HALLAM, AND D. A. READ, J. Solid State Chem. 11, 239 (1974).
2. A. MAGNELI AND G. ANDERSSON,Acta Chim. &and. 9, 1378 (1955). 3. B. G. BRANDT AND A. C. SKAPSKI, Acta Chim. Stand. 21,661 (1967). 4. E. WEDEKIND AND C. HORST,Ber. 48, 105 (1915).
5. P. W. SELWOOD,“Magnetochemistry,” p. 336, Interscience, New York (1956).
2nd ed.,
6. K. TARAMA, S. TERANISHI, K. HATTORI, AND A. YASUI, Mem. Fat. Eng., Kyoto Univ. 31, 611 (1969). 7. E. RUST, J. Amer. Chem. Sot. 81,3843 (1959). 8. L. BEN-D• R AND Y. SHIMONY, Mat. Rex Bull. 9, 837 (1974). 9. J. B. GOODENOUGH, Bull. Sot. Chim. France 4,
1200 (1965).