[browniee]

calibration is important prior to using GC analysis and i am researching on what are the ways it are done. and i came across this calibration curve (http://www.environics.com/environics-post/bid/59842/Gas-Chromatography-and-Calibration-Standards ) as i am slightly unclear.

what are doing the calibration curve for ? this graph is done before the "actual analysis" is done. So once i got my "final result" . how does this compare to the calibration curve ? or in other words,what must i do to this final graph with respect to the calibration curve to get an accurate reading?

[Corribus]

To understand why you need a calibration curve, you have to understand how the instrument works. The MS part of the GC/MS is what actually does the detecting. The typical mode of action is that your molecules (separated by the GC) are ionized before flying into the mass spectrometer. Once in the MS, they are subjected to an electromagnetic field. Because the ions are charged, they are subjected to a force and travel with or against the field. The length of distance they travel in the MS is proportional to their mass. The ions eventually hit the detector, which generates a current at the point where the ions strike. The magnitude of the current, which is recorded by the instrument, is proportional to the number of ions that strike the detector.

An instrument is not a human, though. It doesn't really understand what you're doing. All it "knows" is that a certain current was generated at a certain location. By itself, this carries no useful chemical information. In order to turn the current/spatial information into useful mass/concentration information, you need to tell the instrument two things: (1) what is the relationship between location on the detector and mass, and (2) what is the relationship between the measured charge and the number of ions striking the detector in a given time.

Because the field strength is typically the same for each measurement, and the geometry of the instrument never changes, the first piece of information is usually set in the instrument software and you don't have to worry about it. Occasionally, maybe during a yearly maintenance, the instrument will be calibrated to ensure that the relationship between mass and distance hasn't changed. This may be done with a set of certified mass standards (e.g., you inject a series of compounds with known molecular weights, and the software records where they hit and stores this information for later use).

The relationship between measured charge and number of ions striking the detector, however, needs to be defined for every experiment, because every molecule and sample matrix is different. For quantitative analysis, it is standard procedure to obtain a "pure" sample of your analyte (or, if unavailable, something as chemically similar as possible) and inject at least three different concentrations that span a range that will encompass the concentration of your unknown. In this way, you tell the instrument, "if you measure so-and-so charge, it corresponds to so-and-so concentration". Building a curve allows the instrument software to interpolate (or extrapolate, although this isn't recommended) the concentration of an unknown based on a pre-determined mathematical relationship between measured signal and concentration, with a certain degree of tolerable error.

This is, more or less, standard procedure for ANY quantitative analytical technique. Very rarely are physical parameters measured directly. Most instruments measure some kind of voltage change, which is assumed to be proportional to the physical quantity of interest. Verifying the proportionality, and measuring the proportionality constant using a standard curve, is the first step in practically any analytical measurement.