That is a neat trick. I don't rightly know if this would be possible in Isodat.
Using a sizeable chunk of the cylinder gas "peak", e.g. a 12 s wide slice cut from the middle, flat-top part of a 20 s wide pulse, should/would provide a reasonable stable and uniform ratio value serving as a point of comparison to put actionable, decision-enabling sample deltas on the board.
Ultimately, it makes no difference which approach one takes to integrating the cylinder gas peak; a slice of e.g. 60% or the entire peak. As this thread shows once more, cylinder gas peaks are no substitute for reference materials (RMs) when it comes to generating true aka properly scale normalized deltas. On several counts, cylinder gases and cylinder gas peaks do not meet with the Identical Treatment principle (see the article penned by Arndt Schimmelmann and yours truly in Isotopes Health Environ Stud, 2019). "Delta-wise dialled in" cylinder gases are merely a convenient way to obtain raw measured sample deltas that are a half-way decent approximation of their "true" deltas so one can make an informed decision on the quality of the analysis.
As mentioned above, in SOP lingo the sole purpose of cylinder gas peaks as part of sample analysis is to obtain actionable, decision-enabling sample deltas. Hence, why I call this peaks "cylinder gas" peaks and not "reference gas" peaks. Actionable / decision-enabling means, one has a documented basis upon which one decides to continue with a sample batch run namely, if the deltas for the first, say three standards of a batch run (in-house #1 or international RM #Z), calculated by the proprietary software against the cylinder gas peaks, are within, say ± 2 ‰ of their accepted deltas. (The delta of the cylinder gas needs to be known reasonably well for this though dialling in the cylinder gas delta can be achieved by a few runs against a reference material).
"True" aka scale normalized sample deltas are only obtained by correction using a normalization (calibration) equation which one determines on the basis of 2 standards (reference materials; RMs) that were run alongside one's samples and whose measured or raw deltas were generated the same way as the samples', namely, by comparison with the cylinder gas delta. This way, raw (measured) sample delta and RM delta were generated under like-foe-like conditions (Identical Treatment principle).
For those who are still awake and with me, there are two distinctly different effects responsible for gas peaks to be isotopically anisotropic.
1. The chromatographic isotopic effect, also called the inverse isotope effect, because interaction between a given analyte (solute) and a given stationary phase (which is a two-phase partitioning process; phase A = carrier gas; phase B = stationary phase) results in the isotopically "heavier" fraction of the analyte to elute before the isotopically lighter fraction. Anybody with experience in quantitative GC/MS or LC/MS will know this. In some instances, the 2H labelled internal standard can even be baseline separated from the naturally abundant analyte; example: uniformly labelled 2H-benzene elutes before n.a. benzene by several seconds.
2. The mass transfer isotope effect which in case of a (cylinder) gas peak or pulse within a carrier gas envelope is driven by Fick's Law in which D (the diffusion coefficient) is proportional (amongst other parameters) to the size of particles according to the Stokes-Einstein relation. In this scenario the isotopically lighter isotopologue moves ahead of the isotopically heavier isotopologue. For cylinder gas pulses of the relatively heavy and viscous gas CO2 this can still be seen on the 46/44 ratio trace dipping down at the beginning of the cylinder gas peak. In the case of H2, one can see this already on the mass traces of m/z 2 and m/z 3. Mass traces of cylinder H2 gas peaks in TC/EA-IRMS clearly show m/z 2 arriving ahead of m/z 3. This is even true for sample H2 peaks because H2 is not retained by (does not interact with) the post TC/EA reactor MolSieve column.
Hope this explains rather than confuses matters. :-)
From: Stable Isotope Geochemistry [mailto:[log in to unmask]] On Behalf Of Hilary Stuart-Williams
Sent: 30 June 2020 23:30
To: [log in to unmask]
Subject: Re: [ISOGEOCHEM] [External] Re: [ISOGEOCHEM] Possible to take middle of square reference peak and not the highest point?
On the Isoprime we define which parts of the flat reference peak we ignore. So we ignore the start and end where valve opening and closing might influence things. So I typically use about 60 % of the reference peak width - losing the first 30% and last 10%.
I also hope that this gets around the isotopic chromatographic separation issue: after a while the isotopologue ratios move to what in the plant world we call steady state: what comes out is what goes in?
On 1/07/2020 3:36 am, Wolfram Meier-Augenstein (pals) wrote:
> Hi Peter
> Isodat (like virtually any "quantifying" chromatography peak detection software) determines peak start, peak apex and peak end, and uses area under the curve = peak area as measure of quantity for amount of e.g. 13CO2 and 12CO2. Depending on conditions, e.g. temperature, stationary phase and carrier gas flow rate, the shape of a peak and thus its peak width and peak amplitude can vary. However, the area under the curve, the peak area will remain constant. This is why in the early days of GC/FID analysis, chromatograms were recorded on special paper, peaks were cut out, the paper cut-out was weighed and the weight translated into compound amount based on a calibration curve.
> You can easily verify this by making stark changes to the conditions for peak start and peak end in the method (open a sample data file in Workspace, click on the tab "Edit Method", click on the tab "Peak Detection" and change settings for Start Slope and End Slope). As the peak area changes in line with the changes you made, so will calculated isotope ratios and resulting deltas.
> There is a reason why this is particularly important in CF-IRMS and that reason can be seen by close inspection of the intensity traces of m/z 45 and m/z 44, and even better in the ratio trace. Compound peaks as well as the gas peaks resulting from compound conversion are subject to the chromatographic isotope effect. A detailed discussion of this effect and what causes it can be found in chapter II.4.4.3 of the 2nd edition of the book "Stable Isotope Forensics" (URL below). Its practical consequences are this. As to whether it’s a compound peak in GC or the CO2 peak from a compound after combustion in CF-IRMS, the leading part of the peak is richer in 13CO2 whereas the tailing part is richer in 12CO2. To put it another way, the isotopic distribution within a peak is anisotropic.
> This is why baseline peak separation and acquisition of the entire sample gas peak is so important to obtain ultimately accurate results, i.e. deltas that accurately reflect the isotopic composition of the sample compound. Partial interrogation of a sample or compound gas peak such as taking a thin slice at peak maximum would not achieve this because this slice would not truly represent the isotopic composition of the sample or compound.
> This effect may not be as pronounced in a square-ish cylinder gas peak but it is still there and can be seen on the trace of the 46/44 ratio.
> The option Square Pulse Recognition tells the software how to deal with a cylinder gas pulse which typically has steep flanks and a flat top (kind of like a square) and therefore do not look like a typical sample (gas) peak.
> Prof. Dr W Meier-Augenstein, CChem, FRSC Stable Isotope Forensics &
> Analytical Sciences
> School of Pharmacy & Life Sciences
> Robert Gordon University
> The Sir Ian Wood Building
> Garthdee Road
> AB10 7GJ
> E-mail: [log in to unmask]
> Stable Isotope Forensics book, 2nd edition:
> -----Original Message-----
> From: Stable Isotope Geochemistry [mailto:[log in to unmask]] On
> Behalf Of Peter R
> Sent: 30 June 2020 17:29
> To: [log in to unmask]
> Subject: [External] Re: [ISOGEOCHEM] Possible to take middle of square reference peak and not the highest point?
> CAUTION: External sender. Do not click on links or attachments unless you expect them from the sender and know it's safe!
> If in doubt please contact the Helpdesk: [log in to unmask] +44 (0) 1224 262777
> Thank you for the insight. I assumed ISODAT took the tip of the reference peak and used that instantaneous point to calculate ratios and final isotope per mil values rather than an integration of the square peak and using the areas to determine ratios/ per mil. Looks like its back to the drawing board for getting better reference standard deviations.
> Also, according to the Gas bench II manual it seems the "square peak detection" does not allow ISODAT to timeshift correct for isotopically substituted species that would normally be needed when using a column due to slight retention time differences. It detects the square peak using factor "f" = Area of peak / (peak height * peak width) (section 3-35 Gasbench II operating manual).
> On Mon, 29 Jun 2020 12:24:45 -0700, mwolhowe . <[log in to unmask]> wrote:
>> You may be thinking of how ISODAT determines what value to give for
>> peak amplitude (which would be the highest point). My understanding
>> is, as Jack said, that it integrates areas under each trace to give a
>> ratio, regardless of shape. THAT SAID, I am actually somewhat unclear
>> on what the 'square peak detection' option does - the Occam's Razor
>> explanation is it handles peak start/end detection differently so it
>> doesn't decide the top of the peak is a new baseline, but I can
>> imagine other possibilities. Can anyone confirm?
>> Dr. Matthew Wolhowe
>> Research Scientist/Engineer III
>> Lab Manager, Oceanographic and Geochemical Stable Isotope Facility
>> School of Oceanography College of the Environment University of
>> On Mon, Jun 29, 2020 at 11:43 AM Jack Hutchings <[log in to unmask]> wrote:
>>> Hi Peter,
>>> My understanding of Isodat is that the entire peak's ratios are
>>> integrated to yield a final, weighted-average of the peak's ratio.
>>> That is, no single 'point' on the peak defines the entire ratio. I
>>> could be wrong about this, but I would be surprised (and disturbed).
>>> I haven't rigorously tested the 'weighted-average' part of this
>>> claim (i.e., I'm not 100% sure how the averaging occurs), but you
>>> can sort of test the general averaging behavior yourself: use the 'Def Peak'
>>> option to define new peaks along a reference pulse. You'll find that
>>> while some of these peak chunks approximate the Isodat-integrated
>>> peak, you only really get close to the automatically integrated
>>> value when you manually integrate basically the entire peak (including the rising/falling ends).
>>> You can have it perform tighter (narrower) integrations by
>>> increasing the values of the Start Slope and End Slope, and I find
>>> that higher-than-default performs a bit better for our setup. But,
>>> you can't really crank them up so far as to completely exclude the
>>> rising/falling sides of the peak... since the system has to detect
>>> those in the first place! I could see value in what you're
>>> describing... i.e., detect a peak, and then chop off the first/final
>>> X% of the width. However, to my understanding, this isn't possible.
>>> It also probably wouldn't be desirable for any continuous flow
>>> application, since alternate flow paths/chromatography would
>>> probably yield differently structured (relative to ref. pulses)
>>> intra-peak isotope heterogeneity (i.e., peaks are not isotopically
>>> homogeneous and you need to integrate the whole peak to get the 'right' value).
>>> Jack Hutchings, Ph.D.
>>> Manager, Konecky Lab
>>> Department of Earth and Planetary Sciences Washington University in
>>> St. Louis
>>> On Mon, Jun 29, 2020 at 1:07 PM Peter R <[log in to unmask]> wrote:
>>>> I'm looking to find a way to tell Isodat to take the middle of a
>>>> square reference peak instead of the highest point on an IRMS
>>>> system automatically. Is this possible?
>>>> In the isogeochem archives it seems someone confirmed what I found
>>>> is that Isodat treats references like any other peak in which it
>>>> takes the highest point of the peak and uses the ratios at that
>>>> time point. There doesn't seem to be an easy way to tell Isodat to
>>>> that the ratio at the midway point from the start and end of the detected peak.
>>>> Peter R
> [Robert Gordon University]
> Robert Gordon University has been awarded a TEF Gold award for the quality of its undergraduate teaching and learning, placing it in the top 20% of Universities in the UK.
> Robert Gordon University, a Scottish charity registered under charity number SC 013781.
> This e-mail and any attachment is for authorised use by the intended recipient(s) only. It may contain proprietary material, confidential information and/or be subject to legal privilege. It should not be copied, disclosed to, retained or used by, any other party. If you are not an intended recipient then please promptly delete this e-mail and any attachment and all copies and inform the sender. Please note that any views or opinions presented in this email are solely those of the author and do not necessarily represent those of Robert Gordon University.
> Thank you.
Note: ANU is making a staged reöpening after COVID-19. We're getting back to normal(ish)!
Dr Hilary Stuart-Williams, Research Manager - Stable Isotopes, Farquhar Lab, Research School of Biology, R.N. Robertson Building (46), The Australian National University, Canberra ACT 2601
Work: 61 02 6125 2099