3.2.7.2. Fluorescence based aluminum assay
J. Ruzicka, M. Davis, M. Hatta and Ch. I. Measures © February 2024
The key advantage of CCD measurement is that it provides, in contrast to PMT measurement spectral information in real time. For routine applications, this feature supports trouble shooting, for method development serves as valuable tool since intensity and wavelength of emitted light depends on composition of chromogenic reagents. This section documents this feature by comparing CCD with performance of PMT based method.
Determination of aluminum, based on its reaction with lumigallion was carried out at pH 5.5, and at 55 °C by using excitation wavelength 500 nm while monitoring emission at 595 nm (S.H. Sutheimer and S. E. Cabaniss Anal. Chim. Acta 303, (1995) 211-221. They automated the assay in continuous Flow Injection format, using Shimadzu RF535 fluorescence monitor equipped with appropriate filters and PMT detector. Their calibration had linear response and limit of detection LOD=3.7 nM Al.
We modified their protocol into programmable Flow Injection format (A) by including two stop flow periods, of which the first one (in HC 2) is used to allow reactions to reach equilibrium at 50 °C, while the second one, when sample is in flow cell, allows use of long CCD integration time and positioning of data collection (WIN) when response has stabilized.
The corresponding software protocol (B) comprises control of microfluidics (A) and data collection and processing (autocalibration, spectrum). The autocalibration generates a calibration graph by autodilution of a single standard solution in five steps, each step performed in triplicate, thus yielding a fifteen-point calibration graph. Detailed explanation of this Single Standard Calibration (SSC) technique are described in Section 3.2.5.
CCD spectrometer is controlled by two panels that select the method: fluorescence or spectrophotometry. For fluorescence measurement set wavelength F595 and integration time (0.32 sec) on panel (C). With both light sources switched OFF, set dark spectrum on panel (D). With blue LED ON, get reference spectrum and adjust its light output to Irradiance between 10000 to 20000. (Blue LED has max. @450nm).
Set up and prime instrument as follows: carrier (P1 and P2) DI water. Reagent: 25uM lumigallion in 0.1M ammonium acetate buffer pH 5.5. Calibration standard 100.0 nM Al in 0.005 M HCl. (Should be prepared daily by dilution with DI of more concentrated Al in 0.1 M HCl see Sutheimer and Cabanis).
The calibration run will yield spectrum of each dilution step (E) as well as series of fluorescence (counts/sec)/ time response curves (F).
Maximum fluorescence, recorded during stop flow period within data collection window (WIN) and zero baseline (BS) plotted versus aluminum concentration yields calibration graph.
Shown are two calibration ranges obtained with 100nM Al single standard (G) and 30nM Al single standard (H) solutions. Because the purpose of this study is to show that instruments furnished with PMT or CCD are equally sensitive, limit of detection , %r.s.d and slope of calibration lines are relevant figures of merit. Sutheimer and Cabaniss report for PMT instrument LOD= 35 nM Al with sample volume 20 uL and LOD= 3.7 nM Al with 100 uL sample volume while r.s.d was 0.8% (slopes were in uV/uM). Therefore data obtained with CCD (G and H) are comparable (LOD = 14 nM and 1.4 nM Al). However, it is interesting that LOD values of 3.2 nM Al and 0.06nM Al are obtained when outliers in (G and H) are excluded. Obviously LOD data can only serve as guide since single data point (out of fifteen) considerably affects LOD value.
To identify the source of variable defining LOD in absence of reagent blank, spectra were recorded when flow cell was filled with DI (Fig. F at BS window) is shown in (I). Assuming r.s.d of 36 % and slope of 43counts/sec/nM Al (Fig. H) the amplitude (noise) of CCD and LED corresponds to LOD= 1.0 nM Al. Hopefully, the new generation of OI CCD spectrometers offers better performance than first generation USB 4000-F used in this work.
