The SPEX FLUOROLOG spectrofluorometer is capable of detecting sub-picomolar ( < 1012 M) fluorescein according to the conventions established by the ASTM Committee E-13 on Molecular Spectroscopy (ANSI/ASTM E579-76). Other commercial manufacturers of spectrofluorometers specify their Minimum Detectable Concentration (MDC) somewhat differently. This technical note demonstrates how we achieve such a low MDC.
The measurements were taken on a ing mode FLUOROLOG spectrofluorometer with a 450-W (521 nm) xenon lamp, and a cooled R928 photomultiplier operated at 900 V in the photon-counting mode. The bandpass was set to 4.0 nm on both excitation and emission spectrometers. Integration time was 11020 s, with a single scan and no smoothing. Excita-490 542.5 595 tion of the sample was at 480 nm. The acquisition Wavelength (nm)mode was S1/R1, that is, the emission signal was Figure 2. Raman spectrum of water. compensated by a reference quantum counter with voltage adjusted for a reading of 1.0 A. The scans At the highest wavelengths, at the extreme right were taken under ambient room conditions.
Results and Discussion
Figure 1 shows spectra of 400 fM (0.4 pM or 4 1013 M) fluorescein in 0.01-N NaOH, a blank of solely 0.01-N NaOH, and the subtracted spectrumof fluorescein without the solvent. The large peak near 520 nm clearly shows the presence of fluorescein at a 0.4-pM concentration.
At the highest wavelengths, at the extreme right of Figure 1, you can see the edge of the water Raman O-H stretch. A clearer graph of this region is shown in Figure 2, under 480-nm excitation. The Raman peak is inherently broad, centered at 575 nm, along with a weaker H-O-H bending mode near 521 nm. The bending mode is also visible in the solvent-containing spectra in Figure 1. This weak bending mode demonstrates the excellent sensitivity of the FLUOROLOG. In Figures 1 and 3, spectr al subtraction completely eliminates the O-H and H-O-H modes.
The Standard Test Method according to the
ASTM was intended to establish the minimum detectable
fluorescence of quinine sulfate. We applied
this method to fluorescein because a number of customers
specifically requested detection limits for
that compound. The MDC is determined by the
limiting signal-to-noise ratio according to the equation:
where C is the concentration of the test solution, S is a signal, and the Noisep-p is the total rms noise from all sources. For the subtracted spectrum in Figure 1, we calculate an MDC of ~10 fM (~1014 M) fluorescein.
Measurements at the femtomolar level are plagued with technical problems originating primarily in the sample. Because of contamination or improper cleaning, detection of fluorescein at zero concentration is not uncommon. This is one reason why ASTM resorted to the above method of extrapolation. By running scans at higher concentrations, freedom from background contamination can be better assured. Our spectrum of fluorescein results from a subtraction of a blank, guaranteeing that our signal is greater than the background. As further insurance against impurities and contamination, the cuvettes were soaked in alcoholic KOH, rinsed soaked in chromic acid, rinsed, and then soaked in nitric acid. Aliquots of all solutions were scanned before and after the test solution to eliminate any effects from leaching of contaminants from the glassware.
We therefore conclude from these results that the ASTM method reliably predicted the actual MDC (scanned) of the FLUOROLOG system.