Stanton D. answered 02/04/22
Tutor to Pique Your Sciences Interest
Hi Camille A.,
I'll have to slightly revise your question! You are free to make a standard calibration curve over whatever range you want, but in order to confidently use that calibration curve your sample absorbance must fall between (or, coincide with) the plotted points on that curve. Usually, "curve" refers to a straight line, and data which obey a non-linear but still mathematically-well-defined function are transformed so that they will fit on the (linearized) equivalent calibration plot.
There are many major reasons why that must be; here are just two:
1) If the absorbance value is above the uppermost point of the calibration curve, it may be also above the linear limit of the reagents of the method, or above the linear working range of the spectrophotometer. Spectrophotometers might give you an absorbance reading of 4.0, but that datum is assuredly trash, and governed by stray light and dark-current of the detector. Most spectrophotometers advise reading only below 2.0 for that reason. The actual linear working range depends on the instrument, the wavelength, and sometimes the age of the lamp in use (deuterium lamps especially age out).
2) If the absorbance value is below the lowest point of the calibration curve, it is liable for interference by non-analyte components of the sample. You might, when you prepare your calibration solutions, wish to include those "placebo" components, but that is normally done as part of the method development and validation. But you should always be alert, since all sorts of contaminants may give interfering absorbances or equivalent, for instance diffractive scattering from suspended particulate material too fine to perceive by naked eye.
-- Cheers, --Mr. d.