Lecture given at the Spring Conference of the DPG 1993 in Greifswald
A. Tusche1, R. Lebert1, D. Rothweiler1, M. Krumrey2, W. Neff3
(1Chair of Laser Technology, RWTH Aachen; 2PTB, BESSY, Berlin; 3Fraunhofer Institute for Laser Technology, Aachen)
For use in X-ray diagnostics on pinch plasmas in the wavelength range of the water window (2.2 - 4.4 nm) and for image acquisition on an X-ray microscope, there is a need for a detector that offers higher sensitivity, the prospect of more stable calibration and faster availability of results (online) than photographic film. By coating a front-illuminated slow-scan CCD with an X-ray phosphor, its sensitivity was adapted to the spectral range of soft X-rays and hardening against radiation damage was achieved. To enable absolute measurements, the CCD was characterised and calibrated at BESSY. It is more than an order of magnitude more sensitive than Kodak 101 film. The properties of the system in quantitative X-ray diagnostics in conjunction with transmission grating spectrographs and zone plate spectrometers are discussed.
Slide 1: Outline
- Laboratory X-ray Microscope
- Wavelength range
- Pinhole grating spectrograph
- Film as detector
- CCD as detector
- Phosphor coating
- Response Quantum Efficiency
- Comparison Film - CCD
- Pinhole grating structure
- Zone plate structure
Slide 2: Laboratory X-ray Microscope
A laboratory X-ray microscope is being developed as part of a co-operation with the University of Göttingen and the Institute for Laser Technology in Aachen. We are working on the development and optimisation of the X-ray source, which in this case consists of a pinch plasma system.
Slide 3: Wavelength range
To observe biological samples, radiation is used that lies within the so-called water window. This is limited by the absorption edges of oxygen and carbon, the two main components of biological tissue, and provides a natural contrast mechanism.
The diagram below shows a spectrum of our source. We use the line radiation of hydrogen-like nitrogen which lies within the water window. This is already a matched spectrum whose harder components have been reduced by using oxygen in the beam tube.
Slide 4: Requirements
X-ray microscopy places certain demands on the source. For example, the X-rays should be as monochromatic as possible. To avoid radiation damage in biological samples, the pulse duration should also typically not exceed 10 ns. For us, source optimisation means increasing the time-integrated brilliance.
Achieving this goal places certain demands on the detector. For example, a simultaneous spectral and spatial resolution is required to measure the time-integrated brilliance. The detector must be long-term stable for comparative measurements. Furthermore, we need a high sensitivity. If possible, the detector should also enable absolute measurements.
Slide 5: Pinhole grating spectrograph
A typical setup for determining the time-integrated brilliance is the pinhole grating spectrograph. It is used to spectrally decompose a pinhole camera image of the source profile. This makes it possible to determine the wavelength interval from which the radiation is emitted and the size of the source in one wavelength simultaneously. The setup itself specifies the solid angle range dΩ used, which means that only the measurement of the emitted energy is required to determine the time-integrated brilliance.
Slide 6: Film as a detector
Up to now, we have mainly used X-ray film for our measurements. Our experience shows that the film is subject to production-related fluctuations. This is also confirmed by a publication by Schwanda and Eidmann. The blackening curves of this film from different production days are shown, as measured by the authors, as well as a comparison with the value of another author. According to Schwanda and Eidmann, the fluctuations from film to film can be a factor of two to three.
Should small individual measures at the source result in small improvements by a factor of, say, 1.5 or 2, this cannot be proven with certainty to be significant with film, unless one had the possibility of calibrating each film individually.
Slide 7: CCD as a detector
In the search for an alternative detector, we investigate whether a CCD would be suitable. A commercially available CCD which is illuminated from the front suffers radiation damage in the area of soft X-rays. It is therefore necessary to protect the chip from the destructive effect of the radiation. In addition to other conceivable concepts, such as luminescent screens or coated fibre optics, we have opted for direct coating of the chip with an X-ray phosphor. The phosphor must absorb the radiation as completely as possible and convert it into light that can be detected by the CCD with a high degree of efficiency; it must be sufficiently thick to protect against radiation, but also thin enough to minimise the spread from pixel to pixel.
Slide 8: Efficiencies
First of all, we need to be clear about the type of X-ray phosphor. A CCD chip, such as the one we use, has a sensitivity curve which has its maximum at around 700 nm, i.e. in the red range. Phosphors that have their emission maximum close to this range promise good efficiency, provided that they themselves have sufficient conversion capacity. The diagram below is obtained by multiplying the conversion capacity of the phosphor by the sensitivity of the CCD in the wavelength range in question.
We decided in favour of terbium-doped gadolinium oxosulphide based on these considerations and on work carried out by the Göttingen group.
Slide 9: Phosphor coating
The size of the phosphor grains used is less than 3 µm. In order to achieve the most homogeneous coverage of the chip, at least two layers must be applied. Our coating thickness is approx. 6 µm. When using soft X-rays with photon energies of around 500 eV, the intensity drops to 1/e of the incident intensity every 0.2 µm. With a coating thickness of 6 µm, the CCD is therefore sufficiently protected.
The crosstalk behaviour from pixel to pixel can be studied by looking at the MTF. An MTF of 0.5 certainly promises good images. Due to the pixel structure of the CCD, the MTF naturally drops rapidly at high spatial frequencies. The same applies to the phosphor layer due to its granularity. The solid curve shows a view of the overall system.
Slide 10: Response Quantum Efficiency
In order to be able to carry out absolute measurements with the CCD, it is necessary to know the response behaviour of the coated CCD. Thanks to the friendly support of the PTB, it was possible to characterise the chip at BESSY. The signal height of the CCD, measured in counts, is plotted against the photon energy of the incoming radiation. The visible edge structure may be due to the absorption behaviour of the phosphor and the sedimentation agent used. However, this requires further measurements. In the energy range of 500 eV that interests us, the response behaviour shows a maximum of approx. 1 counts/photon.
Slide 11: Comparison film - CCD
By varying the ring current of the BESSY and the exposure time, it was possible to determine the sensitivity of the CCD. While Kodak 101 film typically requires around 0.3 erg/cm² to obtain a usable signal, the CCD requires about an order of magnitude less energy per detector area in normal operation. Furthermore, the coated CCD is linear over slightly more than two orders of magnitude. By combining several pixels, it is possible to further increase the sensitivity of the CCD. The price to be paid for this is a reduction in the spatial resolution. It would be conceivable, for example, to combine all the pixels of a CCD line. Using the PHG spectrograph, for example, it would be possible to record a spectrum with undiminished spectral resolution while losing the spatial information.
Slide 12: Pinhole grating structure
We use the described CCD system for quantitative plasma diagnostics. A spectrum of our source is shown here, which was taken with a PHG spectrograph with an aperture diameter of 50 µm and a grating constant of 10,000 lines/mm. To protect the CCD, the area of the harder X-rays was blanked out with a bar. It would now be possible to derive plasma parameters from such a spectrum. As such spectra can be obtained with the CCD system in one pulse, it is now possible to make statements about the reproducibility of the system.
Slide 13: Zone plate structure
A zone plate spectrometer is better suited for viewing the beam profile alone, as it has a better spatial resolution than a pinhole camera. This image shows the beam profile of our system at a wavelength of 2.5 nm in two-dimensional resolution. It can be seen that the source profile is quite symmetrical.
Slide 14: Summary
To summarise, it can be said that a coated CCD offers sufficient resolution for the projects mentioned, is more sensitive than Kodak 101 film and offers the prospect of calibration capability. Another advantage of the CCD that has not yet been mentioned is the rapid availability of the data.
It would now be interesting to use this concept to adapt the CCD to other wavelength ranges or to increase the overall sensitivity of the system by using newer phosphors that are now coming onto the market. For further measurements in the field of X-rays, backside thinned chips promise a further increase in sensitivity.