Improved Energy Stability in the NIST Microcalorimeter X-ray Detector

Terrence Jach, John A. Small, Dale E. Newbury

Purpose: Microcalorimeter x-ray detectors make up a new technology that combines some
of the positive features of wavelength dispersive (high resolution) and energy dispersive
(parallel detection over a wide energy range) detectors that have gained broad acceptance
in the microanalysis community. Microcalorimeters which use a transition edge sensor
(TES) have proven to be effective over energy ranges of 10 keV or more in application
such as x-ray fluorescence analysis with electron microscopes. TES microcalorimeter x-ray
detectors have successfully demonstrated energy resolutions better than 5 eV. However,
serious drifts in energy scale over extended counting times have set limits on both the
long-term resolution and the calibration of these detectors. This is because the operating
point of the detectors is the middle of the superconducting-normal phase transition. The
successful operation of a microcalorimeter as an x-ray detector puts considerable
constraints on the stability of all the electrical and thermal inputs to the instrument. We
investigated the sources of energy scale drift in the microcalorimeter x-ray detectors
developed by NIST (Boulder) and have addressed the most critical elements. Previously
observed drifts of >10 eV/h have been reduced to 1-2 eV/h. This improved stability, shown
in Figure 1, has resulted in the observation of x-ray fluorescence linewidths of 12-15 eV
over a 6 h time period.

Major Accomplishments: The detector is cooled to a substrate temperature of only 70
mK and maintained at its operating point by a complex feedback control system connected
to a large superconducting magnet. We carried out a careful analysis of the performance of
all the elements in the control system including the response function of the magnet under
typical conditions of operation. We determined that the desired stability and performance
of the detector required control of its substrate temperature to a precision of 23 µK, that is,
a variation of less than 5 parts in 10,000. By careful modification of the control circuitry,
we have been able to realize this degree temperature stability of the substrate. The energy
scale of spectra is now observed to be stable to within about =1 eV/h over a period of
hours, under operating conditions in which a linewidth of 12-15 eV is readily obtained. Xray
spectra acquired over long durations under these conditions of substrate temperature
stabilization show vastly improved stability and resolution.

Impact: The drift in energy scale of the microcalorimeter x-ray detector has been a major
limitation to the commercial development and marketing of this type of detector. There are
currently a number of potential applications of this detector, particularly for microanalysis,
which are eagerly awaited in the semiconductor industry. Semiconductor manufacturers
have already indicated an interest in using this type of detector if the problems can be
solved.

Future Plans: Additional measures to stabilize the operating temperature are possible.
Once realized, we can carry out a demonstration of the microcalorimeter detector with an
electron microscope for quantitative microanalysis. We also anticipate replacing the actual
detector element with higher resolution (< 4 eV) versions which have been developed at
NIST, Boulder. Combined with the achieved stability of the energy scale, we can start to
look at characterizing chemical state of some elements by the energy of their fluorescence
lines.

Web Contact micro@nist.gov