| Summary: | In the preparation of microdroplets of biological fluids for X-ray microanalysis, we have found that incorporation of a macromolecular solute, dextran, to a final concentration of 1.5-2.5% retards crystal formation and produces sufficiently uniform deposits on thin films to allow droplets to be analysed without prior freeze-drying. Analyses have been carried out at 20 kV in a scanning electron microscope, using energy-dispersive spectrometry. Absorption of Na X-rays by the added solute can be significant but its effect is minimized by preparing droplets as thin as possible, and by using standards of similar composition. The minimum detectable concentrations are increased because of the extra background contribution, and for a single determination are about 6 mM for Na and 2 mM for Cl and K. These concentrations can be further reduced by measuring replicates. The reproducibility of analysis is significantly improved (to less than 5% for Na and K) over the use of calibration curves by calculating the element concentrations from a known element in the sample, chlorine. Under our analytical conditions loss of Cl did not occur. This method requires that the Cl is measured separately by microcoulometry, but eliminates the need for a range of standard droplets on the grid, and determination of the unknowns is then independent of droplet volume, beam current, counting time and magnification. We have compared, with biological samples, the results from using Cl as an internal standard with those obtained using an added standard element, cobalt. The reproducibility using Cl was approximately two-times better than that obtained with Co, probably because of unavoidable volumetric errors when the Co is pipetted separately.
|
|---|