Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 77, 1948-49
This text is also available in PDF
(2 MB) Opens in new window
– 41 –

X-Ray Diffraction Of Ironsand.

New Zealand's ironsands, particularly those on the west coast of the North Island, have long been regarded as a potentially rich source of iron. Now their relatively high titanium content has aroused considerable interest, and their smaller vanadium content is also considered worth attention.

The sands consist of grains of many different minerals. The proportions of these vary from place to place, even from one part of a beach to another; and the South Island ironsands differ very considerably from those of the North. In the North Island sands, the mineral present in quantity from which iron should most conveniently be extracted is magnetite (Fe3O4), and as this is strongly magnetic it can readily be separated almost completely from the other materials. The grains of magnetite are found to contain one-tenth as much titanium as iron, and a small quantity of vanadium.1 This unusually large proportion of titanium is the main reason for the failure of past attempts to work the sands.2 In a blast furnace titanium forms heavy infusible slags which accumulate at the bottom and soon put the furnace out of action. No method of reducing the titanium content before sending the magnetite to the furnace has yet been devised. Monro and Beavis1 concluded from chemical evidence that the titanium atoms are actually in the magnetite lattice, i.e., some of the lattice points normally occupied by iron atoms are occupied by titanium atoms. This arrangement is known to occur in some minerals in other parts of the world. Chemical evidence alone, however, cannot settle this matter, and the only way of obtaining the necessary further information is to apply the methods of X-ray crystallography.

The apparatus used to do this is that designed by Williamson and constructed at the Dominion Physical Laboratory.3 It consists of an X-ray tube with interchangeable anticathodes and the necessary pumps and power supplies, and a Debye-Hull powder camera. The method requires substantially monochromatic X-rays. These are obtained by strongly exciting the characteristic K radiation of a suitable metal used as the anticathode or target, and removing the unwanted Kβ line with a suitable filter. The material to be studied is ground up to a fine powder, bound together with a gum containing only light atoms, and formed into a very thin rod which is slowly rotated at the centre of the camera (with its axis vertical) while a narrow (horizontal) beam of monochromatic X-rays

– 42 –

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

is shone upon it. The purpose of this treatment is to present as many orientations of crystal planes as possible to the incident beam so that the diffracted radiation will form continuous cones about the incident beam as axis. In order to record the positions of these cones a strip of photographic film is held in the shape of a cylindrical surface coaxial with the specimen so that upon development a set of dark lines mark the intersections of the cones of radiation with the cylinder of film. From the measured positions of these lines the angles of diffraction can immediately be deduced and the corresponding spacings of crystal planes calculated from the well-known Bragg equation: nλ = 2nd sin θ

where n = the order of diffraction

λ = the wavelength of the X-rays

d = the spacing of the crystal planes

θ = the angle between the incident or diffracted ray and the crystal plane.

From the positions and relative densities of the lines it would be possible with much labour to find the position of every atom in the unit cell of the crystal studied. However, this has not been necessary in the present work as the materials so far encountered have their diffraction patterns listed in the A.S.T.M. X-ray Data Cards.4, 5 These cards are so arranged that any listed element, chemical compound or mineral can be identified from its diffraction pattern.

Plate 1 (d), a contact print from the original, shows the diffraction pattern of silver, using copper Kα radiation. This was tried first to check the reliability of the apparatus. The lattice constant found agreed to one part in a thousand with the published value. This pattern is interesting in that it shows plainly the rapid increase in resolution as the ends of the pattern are approached. The doubling of the last line is due to the fact that the Kα line is really a doublet.

The choice of wavelength of radiation to be used for best results in the magnetite problem depends on several factors. Molybdenum radiation is a good general-purpose one because it is sufficiently penetrating for absorption in the specimen usually to give no trouble. However, its short wavelength results in a diffraction pattern that extends only a short distance on each side of the undeviated beam, so that it is of little use when we seek weak lines which will be obscured in this case by the crowding of stronger ones. Copper radiation is the most commonly used when a greater spread is required in the pattern. But copper radiation is a short distance on the wrong side of the K-absorption edge of iron, which means the radiation consists of photons energetic enough to eject electrons from the K shells of iron atoms and is thus strongly absorbed in any specimen containing iron (see Text Fig. 1). This results in a considerable reduction in the intensity of the diffracted beams and an increase in the background intensity due to re-radiation in random directions by the iron. Under such circumstances only the strongest lines in the diffraction pattern are detectable. If a still softer radiation is used to avoid this difficulty with iron we run into the same difficulty with vanadium and titanium which are also present. If, however, we go so far as to use vanadium radiation this will be on the right side of the K-absorption edges of all these elements; but there are two difficulties. The first is that solid or sheet vanadium metal for use as a target is hard to obtain. The second is that vanadium radiation is so soft that it will be fairly strongly absorbed in the specimen, with the result that at least those lines near the undeviated beam will be rather weak.

A chromium target was available and this was tried first, as the radiation is at least on the right side of the K-absorption edges of both iron and vanadium, but the patterns obtained were not as clear as could be desired. However, every line of the magnetite pattern could be detected and one line besides. Extra lines in a pattern almost invariably mean additional chemical compounds present, but this one line was much too far along the pattern to be the strongest line of any known compound, and was not likely to be the second or even third strongest. Thus the most likely explanation seemed to be that there was just one lattice present (magnetite) and that replacement of some of the iron atoms by titanium atoms (as suggested by Monro and Beavis) permitted a line to appear that would normally be cancelled out. (This explains the abstract appearing in the Congress programme, as this had to be written at this stage in the investigation.)

As the patterns obtained with chromium radiation could not be regarded as conclusive and as no vanadium target was available, an effective absorption

– 43 –
Picture icon

Text Fig. 1.

curve for the specimen as a whole was drawn assuming the proportions of iron to vanadium to titanium found by Monro and Beavis. This showed plainly that cobalt radiation is likely to be the best for the purpose (see Text Fig. 2). (It is as far away as possible from the titanium- and vanadium-absorption edges without exciting iron.) The patterns obtained with cobalt radiation give one definite piece of evidence. Besides the magnetite lattice there is also present the lattice of ilmenite (FeTiO3).

The reason why the strongest line of this did not show up in the earlier patterns is presumably that it should appear in that part of the pattern fairly near the undeviated beam and would consequently be strongly absorbed in the specimen, chromium radiation being very soft. Another strong line of ilmenite is irresolvable from the strongest line of the magnetite pattern, which completely conceals its presence.

– 44 –
Picture icon

Text Fig. 2.

A rough quantitative estimate (as good as the information in the X-ray Data Cards will allow) of the relative proportions of ilmenite and magnetite present indicates that there is probably enough ilmenite to contain most or all of the titanium. The quantity of vanadium present is so small that even the most refined of X-ray-diffraction techniques would be unlikely to reveal its state of combination. Plate 1 shows the diffraction patterns obtained with molybdenum, cobalt and unfiltered iron radiation. In (e) the doubling of the number of lines due to failure to remove the Kß radiation can be seen. Most specimens were made up from sand from Patea (west of Wanganui), but one made up from sand from Muriwai (north-west of Auckland) gave precisely the same diffraction pattern.

It should be pointed out that this is an account of only the beginning of X-ray investigation of New Zealand ironsands. There are many more minerals to be investigated and many refinements of technique to be tried, especially those leading to more accurate quantitative results.

The author is pleased to acknowledge the very helpful interest of Dr. E. R. Cooper, Director of the Dominion Physical Laboratory, in this work.

References.

1. Monro and Beavis (1946). N.Z. J. Sci. and Tech., 27B, 237.

2. Wylie (1937). N.Z. J. Sci. and Tech., 19, 227.

3. Williamson, K. I. (1946). N.Z. J. Sci. and Tech., 27B, 393–397.

4. (1946.) A.S.T.M. Standards, Part 1-B, 865–871.

5. Smith and Barrett (1947). J. Applied Physics, 18, 177.