Matrix effects and the use of standard addition.
An attempt was made to quantify P in the soil extra
cts by
establishing a standard curve in MS using P solutio
ns in DI water. Peak areas for soil samples analyz
ed by MS
would have been compared to the standard curve to d
etermine P concentration, if there were no complica
ting matrix
effects. Differences in matrix effects can be gaug
ed by comparison of the slopes of the standard addi
tion curves to
the P standard curve. The slope of the P standard
curve was approximately 5 times greater than the sl
ope of the DI-
extracted PBC Inc standard addition curve and 50 ti
mes greater than the slope of the AmOx-extracted BI
O Inc
standard addition curve. This indicates a signific
ant matrix effect, likely due to components extract
ed from the soil
and the ammonium and oxalate ions. Electrospray pe
rformance is very dependent on the amount and type
of salts
that may be in the matrix. Due to these matrix eff
ects, two samples for each extractant were chosen f
or standard
addition analysis. In order to minimize analysis t
ime, the remainder of the samples were analyzed wit
hout standard
addition and compared to the reference (standard ad
dition) samples (as described above). For this to
be valid, the
matrix effects cannot significantly differ between
soil extract samples. The reference samples in the
DI-extracted
PBC Inc sample set had slopes of 0.0038 and 0.0045
(normalized peak area/ppm P) for SA6 and SA11, resp
ectively.
For the AmOx-extracted BIO Inc sample set, the slop
es of the reference samples were 0.0035 and 0.0051
(normalized peak area/ppm P) for SA3 and SA9, respe
ctively. This demonstrates a difference in the int
ensity of
signal response to phosphate due to matrix effects.
The resultant discrepancy between P values based
on the