Eshwar Udho

CHE 331
Dr. Rahni
May 6, 2003
Lab#4
Title: Atomic Absorption Spectroscopy
Purpose: To optimize parameters of instrument and use setting to determine copper
content of unknowns.
Introduction: Atomic absorption spectroscopy (AAS) is similar to UV-VIS
spectroscopy (see Lab1) in that they both measure the absorption of energy due to
electronic transitions in the molecules of interest. In UV-VIS spectroscopy, the samples
naturally absorb energy when in solution, making direct absorption measurements from
light through a solution feasible. Metals however only absorb light (UV only) and
undergo those electronic transitions when heated. Therefore, to determine the amount of
metals in solutions, the solutions is first aspirated into a flame when the metals are then
vaporized. They are then able to absorb light and hence absorption measurements can be
taken. It is this process just described that is referred to as atomic absorption
spectroscopy. The name makes sense since the vaporized metals are now in their atomic
states and we are performing absorption spectroscopy on those atoms.
In this lab, we will be using an air-acetylene flame which has an average
temperature of about 2300 degrees Celsius. We will determine the best height and fuel to
air ratio for the instrument. We will also construct a calibration curve from known
samples (the absorbance follows the Beer-Lambert Law as in Lab 1) and use it to
determine the amount of analyte in unknown samples. The calibration curve is better

than a direct application of the A=Ecl Berr-Lambert formula because of variations in

vaporizing efficiency leading to variations concentrations of gas-phase atoms.
The AA spectrophotometer consists of a light source (usually a holoow cathode
lamp) emitting a single set wavelength of light, an atomizing apparatus and a detector
linked to an output device.

The atomizing apparatus usually involves an aspirator to draw the solution into the gasflow path which then enters a nebulizing/mixing chamber to form small droplets before
entering the flame. The flame itself is a slot flame to ensure maximal vaporization and
hence maximal absorption by the sample. The detector is usually a photomultiplier tube
which can amplify the very weak incoming signal to something many orders of
magnitude greater making it a very sensitive detector.

Discussion: The standard curve came out with a rather height R-squared value indicating
a very good fit for the data points. This indicates that the parameters were very well
optimized and that there should be high precision of measurements. This did in fact turn
out to be the case seen in the replicate measurements: the percent coefficient of variation
had a maximum value of 2.5. I think therefore that the calculated concentrations of the
unknowns are very accurate. It seems that the alcohol itself absorbs no light at that
wavelength. The chloride and nitrate salts of the cupric ion both had lower absorbances
compared with the sulfate salt. This may be due to vaporization efficiency of the
compounds, and if it is the case, would imply that the sulfate salt is more readily
dissociated in the flame than the other salts compared. The alcohol solutions are
inconclusive as to the effects of alcohol in the solvent. Either it is that we made an error
and switched the data for the 5ppm copper in the 20% and 50% alcohol solutions or the
graphs for copper content versus absorbance cross each other for varying the alcohol
content of the solution. If the first case is true, we could then conclude that the greater
the alcohol content of the solvent, the less the absorbance. Here we would expect similar
shaped graphs for absorbance versus ppm copper but the higher alcohol content for
graphs would be below the lower alcohol content graphs (See Mixed Up Data Graph). In
the other case, it would seem to suggest that in a mixed solvent system, the absorbance is
a function of both the solvent system composition and the solute ion concentration, hence
the crossed graphs if the data is taken as is. The regression line added is linear although
the relationship may be more exponential.