negative intercept in calibration curve

[kaka_munda]

Do you use an internal standard? And if so, the same internal standard for all compounds? And does an injection of solvents with no added internal standard or analyte give any peak - particularly in the location for the internal standard?

No, I did not use any internal standard. and yes the solvent blank was clean. It did not contain any peak.

[MSCHemist]

I studied this in detail.
LCGC ran an article titled "To b or not to b". The conclusion was if you find the standard error in the y intercept (excel linest array function can do this) and it is larger than your b then the b is not significant. Otherwise you should keep it.

Another issue I often saw was if your calibration range is large then errors at the top end can distort the low end. When you make a typical least squares line corrections to the Y value to get it on the line have the same weight at the top and bottom calibration points. It assumes homoscedacity in the errors (that you have the same abolute error at all calibration level rather than more often the errors are relative). So the equation is just as likely make a correction of the response equal to .5ppm at .5ppm as at 500ppm. Forcing zero often helps to overcome this effect. A more valid way to to create a weighted calibration curve (Chemstation can do this) and have it inversely weighted by ammount.

I recommend reading Statistics and Chemometrics for Analytical Chemistry.

[hoynj]

When making linear dilutions of analytes with a large dynamic range where your lowest point is far from 0, I frequently have issues with large y intercepts since small variation at the high end will result in huge changes near 0. I'd recommend continuing the dilution with more points at the low end to see if the plot is linear.