Newsletter contents... UP

First Experience with Novel Microsource X-Ray Tube

A C Bloomer & U W Arndt
MRC Laboratory of Molecular Biology
Hills Road, Cambridge CB2 2QH, UK

Last month we took delivery of the first commercially-produced MicroSource X-ray generator from Bede Scientific Instruments Ltd (Durham, UK). This is a sealed tube, as distinct from the continuously pumped laboratory prototypes upon which development work has been done over the past several years. We present here the initial performance indications of this novel generator and some prospects for its future development. A more detailed report is being prepared for submission to J. Applied Crystallography.


Macromolecular crystallographers seek X-ray beams where the highest number of photons illuminate the crystals whose diffraction they need to measure. Over the years, the most challenging crystals being studied tend to have smaller size and/or larger unit cells. A desirable X-ray source has high brightness and low cross-fire, combined with great stability and reliability. For most biological crystallographers, the early sealed tubes for laboratory production of X rays have long been overtaken by rotating anode generators, and more recently by centralised facilities for synchrotron radiation.

However, a new design of sealed X-ray tube for use in the laboratory specifically aims to meet the requirements in this field of crystallography for good collimation and utilisation of smaller crystals. Design features used in electron microscopes and similar devices have been used in the development of this low-power copper-target X-ray tube [1]. This has been optimised for use with focusing mirrors, chosen to maximise the solid angle of collection of the emitted X rays, thus achieving a high intensity at the sample. The additional requirement of small cross-fire in the beam imposes a geometrical constraint that the X-ray source should be very small in relation to the size of the sample. The present instrument, which has a typical source size of 20micron, is well-matched to samples of up to 300micron in size, positioned at 60cm from a specularly reflecting ellipsoidal mirror [1].

Comparison of different types of focusing mirrors for use with a microfocus X-ray tube [2] shows the advantages of ellipsoidal mirrors of very small diameter (typically less than 1mm). These can be manufactured by an electroforming technique. Data presented elsewhere [2] introduce the method of determining the performance of mirrors by their "insertion gain". This is the ratio of the intensity from an X-ray source delivered into a defined small area (comparable to the size of a specimen crystal) at the focus of the mirror when it is in position, to the intensity when the collimating mirror is removed. The expected insertion gain for a perfect system can be calculated for comparison with that achieved in practice [2].

Results from early prototype designs have led to the arrival on the market this year of a commercially-produced system.

Experimental Arrangement

The MicroSource tube which is now at LMB (Cambridge) has been operating at 25watts (50kV 0.5mA) and is mounted to enable its alignment to an early prototype of the MAR (18cm) diffractometer positioned so that the sample is at the optimum distance (60cm) from this source. A vacuum pipe covers much of this distance. Good diffraction data have been collected on this system from crystals of both lysozyme and also an antibody Fab fragment.

We have not yet completed a detailed comparative assessment of data obtained using this new installation with that from our standard laboratory rotating anode installation. However, we can already compare the characteristics of the incident X-ray beam at the position of a crystal sample. The rotating anode installation is a Nonius GX13 (big wheel) operating at 2,400watts (40kV 60mA) with a fine focusing cap (100micron nominal). This is used with a standard MAR (30cm) diffractometer equipped with double Franks mirrors and two pairs of slits to control the size and cross-fire of the beam at the crystal.

X-ray intensity is measured by an ionisation chamber positioned at or just behind the sample position. In some cases a small aperture (300micron) is carefully aligned on a goniometer head so that it is at the sample position. Measurements of the flux passing through this aperture are representative of the X-rays incident upon a crystal of size 300micron or less.

We take as our local standard of comparison the flux obtained with the MAR slits all set to 0.3mm and which passes through our 300micron aperture. Table I shows that the flux obtained from the GX13 tube running at 100* the power of the BEDE tube is less than 3 times greater. For use with larger crystals, the flux measurement (without the 300micron aperture) from the GX13 flux is 3.7 times greater, and if larger cross-fire can be tolerated (opening up all the MAR slits to 0.4mm) on bigger crystals the flux is 4.6 times greater from a GX13 than from the present BEDE tube operating at 1/100th of the power.

The closest comparison comes in the case where a reduced cross-fire is needed in the X-ray beam. When the MAR slits are all reduced to 0.2mm, the GX13 output exceeds that now obtainable from the BEDE tube by a factor of less than 2.

The expected life-time of this novel tube is several months, just as for other sealed X-ray tubes. Early indications of the present tubes are that the filaments survive for several weeks at least, but have not yet been tested for a complete life-time. The other question frequently raised about long-term usage concerns the rate of radiation damage to mirrors. Our calculations, which will be detailed elsewhere, indicate that the flux per unit area falling onto the small ellipsoidal mirror used with the BEDE tube is much less than that incident upon the first Franks mirror used with our rotating anode tube: the difference is a factor of about 5, within the range from 3 to 7 depending upon the exact geometry chosen for the ellipsoidal mirror. Experience with the Franks mirrors on our GX13 tube is that the first mirror usually needs replacing after about six months; by analogy one would expect the ellipsoidal mirrors to last for several times longer than this.

Future Prospects

Earlier ideas by one of us (UWA) of a rotating anode version of this novel generator, which would have increased the flux by a factor of about 5, have since been superseded by two other potential developments which both retain the great simplicity of the present MicroSource tube in comparison with other high brightness tubes.

  1. Theoretical calculations indicate an insertion gain of 250-3000 from use of optimum ellipsoidal mirrors [2]. The gain presently achieved is only about 50, such that improvements in both the profile accuracy and the surface smoothness of mirrors can be expected soon to enable higher flux by a factor of 5-30. As the mirrors are readily inter-changeable, this can be a simple upgrade for any existing installations.

  2. The maximum operating power of the BEDE tube is determined both by the size of the focal spot, and the efficiency of cooling of the target. We have operated the BEDE tube at 25 watts, taking a cautious view of the safe operating limits. Power loading of up to 100 watts should be readily achievable, with a consequent gain in flux by a factor of 4. This will depend both upon alternative methods of cooling the copper target within the tube [1] and upon further study of the effect of increased power loading upon the size of the source.

One major advantage of the MicroSource tube is its adaptability, with simple inter-change between different mirrors to match the beam to the needs of varying crystals. The ultimate limit to the flux at the sample depends upon which configuration is needed. However, the present system already delivers onto a small sample a well collimated beam of intensity comparable to that obtained from some existing synchrotron stations using bending magnets.


Arndt, U. W., Long, J. V. P., and Duncumb, P. (1998). A micro-focus X-ray tube with focusing collimators. J. Appl. Cryst.. In press.
Arndt, U. W., Duncumb, P., Long, J. V. P., Pina, L., and Inneman, A. (1998). Focusing mirrors for use with microfocus X-ray tubes. J. Appl. Cryst.. In press.

Table I

Intensity measurements, using an ionisation chamber placed at the sample position, for two X-ray tubes, with and without a limiting aperture of 300micron. For details, see text.

Data: Normalised Raw figures
(arbitrary units)
300micron aperture in position: NO YES NO YES
BEDE tube, mirror ME-II 1.0 1.0 0.44 0.44
operating power of generator 25w .
GX-13, slits set to 0.2mm 1.9 1.9 0.83 0.84
GX-13, slits set to 0.3mm 3.7 2.7 1.62 1.21
GX-13, slits set to 0.4mm 4.6 2.9 2.02 1.28
operating power of generator 2400w .

Newsletter contents ... UP