Hello Greg:

I had a similar inquiry a while ago and both Patrick McLoughlin and Robert Shneider gratefully ran to my "rescue" in understanding the underlying algorithm of Isodat. I am including below both of their contributions. Hope this will help address your question:

From Patrick McLoughlin:

1)      You put in the del C13/C12 and del 18O/16O for your reference gas.

2)      The software used standard del notation stuff to convert those to R(13C/12C) and R(18O/16O).

3)      If you have configured it to use SSH it uses the equations from that paper (I have the reference around somewhere, write if you need it) to calculate R45 and R46 for your reference gas. I will refer to these as R45_knw & R46_ knw.

4)      The raw numbers are used to construct the raw ratios rR45 (= rArea45/rArea44) and rR46 (=rArea46/rArea44). These raw numbers are already background corrected. I don't know too much about time shifts, they may be corrected for that as well. Note: this is also done for the rR45 and rR46 of the peak (or peaks) you have designated as reference peaks.  Also note: by convention, lower case r stands for "raw" – this is an integer number that is in mVs, typically 5 integers long for my system. Do NOT use theVS numbers that contain decimals. I only use one reference peak for each run, so I'm not sure about how to use more than one. From here on out I assume 1. Anyway, for that one peak I will refer to rR45_mrf & rR46_mrf, where the mrf stands for measured reference )

5)      In the spirit of internal standardization, calibration factors are created:

CF45 = R45_knw/rR45_mrf
CF46 = R46_knw/rR46_mrf

6)      (This is why your dels change when your reference changes.) All peaks are normalized by those calibration factors to produce:
R45 = rR45 * CF45

R46 = rR46*CF46

7)      The SSH equations are then used to find R(13C/12C), R(18O/16O), R(17O/16O). (Note: if you use SSH, the equations are not analytical. I did them in MS Excel and used "Solver", you should probably do whatever you are most comfortable with – or just take it on faith.)

8)      Standard delta equations are used …

This is what the reference peak does. It also is insensitive to minor variations in the multiplier feedback resistors, etc. Basically, this technique allows it to be a lot more precise. I think it is kind of cool, and I can't figure out why nobody wanted to tell me about it.

 

 

Patrick McLoughlin, Ph.D.

Technical Director

Microseeps, Inc.

==================================================

From Robert Schneider:

I thought about your question concerning the change of the R values when
changing the delta values of the reference gas. Also, I had the look to
the isodat software and reproduced your observation.

If you look on the 45R and 46R values, the same effect will take place,
but these values should be equal to one for the peak, that isodat uses
as reference peak (possibly peak one). For this peak, the 45R which is

45R = (Area_45/Area_44)/(Area_45_

ref_peak/Area_44_ref_peak)

should be one because for the reference peak Areas are equal to the peak
Areas. But they are not equal to one. They are equal to

 R_RM_45 = R_13_VPDBCO2*(dC_13_RM_VPDBCO2
+1.0)+2*R_17_VPDBCO2*(dO_18_RM_VPDBCO2+1.0)^a

(a = 0.516                            # condition between R17 and R18:
R17 = K*R18^a
R_13_VPDBCO2 = 0.0112372             # VPDBCO2 standard of 13C
R_17_VPDBCO2 = 0.0004023261          # VPDBCO2 standard of 17C
R_18_VPDBCO2 = 0.0020052             # VPDBCO2 standard of 18C)

where R_13_VPDBCO2 is the ratio as measured for the VPDB scale and
similar for R_17_VPDBCO2. dC_13_RM_VPDBCO2 and dO_18_RM_VPDBCO2 are just
the delta values which you changed and 'a' is the exponent of 0.516
using Santrock 17O correction methods.
And this correction is the crucial point. To get the 13R and 18R values
from the measured 45R and 46R values, one needs to solve a system of
three equations with three unknown variables (after the relationship
between 17R and 18R was assumed to be R17 = K*R18^a). I attached a pdf
file concerning this 17O correction method which I got from

http://www.cstl.nist.gov/div837/837.01/outputs/standards/algorithm/algorithm.html#Calc
(have a look to ' Notation and Strategy ')

Thus, before calculating 18R and afterwards using this to get 13R, which
we are interested in, 45R and 46R are converted to VPDB standard. So, if
you change the delta values of the reference gas, the 45R and 46R values
as well as the 18R and 13R values will change too because they are
already in VPDB standard.

I hope this could explain the puzzling isodat procedure.
If there is still something uncertain or unclear, or if I got your point
wrong, do not hesitate to contact me.

respect,

Robert