Increasing the specificity of the forensic luminol test for blood.

It is shown that the presumptive luminol chemiluminescence test for the presence of traces of blood can be made more determinative by measuring the peak emission wavelength of the luminol chemiluminescence. When sprayed onto a surface containing traces of human haemoglobin, a 1 g/L solution of aqueous luminol containing 7 g/L sodium perborate gives an emission peak at 455 +/- 2 nm, whereas the same mixture gives an emission peak at 430 +/- 3 nm when sprayed onto a surface containing traces of sodium hypochlorite (household bleach). This spectral difference can readily be determined using spectroscopic equipment that either scans the spectrum before significant luminescence decay occurs or corrects the spectrum for the effects of any decay. It was found that bovine haemoglobin and human haemoglobin showed no significant spectral differences. Copyright 2001 John Wiley & Sons, Ltd.