The non catalytic data can be corrected by interpolating the baseline under the peak from the charging current on each side of the faradaic signal. Interpolation uses a cubic spline procedure.
At least 3 markers are required to define the baseline. These markers are input clicking with the mouse.
o, the marker switches between Smooth, eXact or Off data. (Definitions of markers here).
ato Add a marker at the position of the cursor.
rto Remove the nearest marker,
qto Quit and subtract base line,
mto get the following additional commands:
uto qUit replacing the data with the baseline.
vto quit replacing the data with data/baseline.
dturn on and off the diplay of the derivative of the baseline. If you quit (
u) while the derivative is displayed, you replace data with derivative.
bturn on and off the display of the baseline full scale (without the data).
d, the baseline can still be modified by adding/removing cursors.
2to add 10 markers to the left or right of the cursor's position, respectively (this is to encourage Lars' laziness).
0to add 20 markers evenly spaced all over the data.
-to delete all markers.
tto turn on or off the display of the ratio lines.
ito zoom out or in along the Y axis (this is useful when the limits of the plot don't allow the baseline to fit in).
pto turn on/off a one-peak search in the residuals. If it is on, the peak current and potential are written in the graphic window, above the area/last area values.
I set the maximum number of marker to 50 because putting hundreds of markers makes no sense : a few are enough if the data are smooth enough for the analysis to be worth doing. When you put markers on each side of a peak, you assume that you can interpolate a SMOOTH polynomial under the peak...if the CV is so rough that you need many markers on each side, the base line is unlikely to be smooth under the peak...So there is hardly any way of working it out.
Using the spline interpolation to smooth or differentiate :
Smoothing: An efficient way to analyze noisy data is to put 20 Smooth
markers all over the data (typing
0), then qUit
u) replacing data with the baseline (the data are now
smoothed), and then subtract the Base line again with eXact markers
and Quit replacing data with data-baseline (
Differentiation: Once the baseline is interpolated using the
spline routine, it is literally free to compute its derivative. This
can be obtained by typing d when more than 3 markers are set on the
data. The derivative of the baseline (which may follow the data if the
markers are put all over the potential range) is plotted in the main
window instead of the raw data + baseline. Type
d again to get
the raw data back. If you quit while the derivative is displayed, the
current buffer is updated with the derivative. Note that it is not
strictly speaking the derivative of the data, but the derivative of
the spline function, the latter containing only intervals of cubic
polynomials ; its derivative is therefore made of intervals of second
This command uses the public domain spline/seval procedures from the NETLIB numerical libraries (http://www.netlib.org/sfmm/spline.f).
Christophe Leger 2009-02-24