I would just like to describe my experience measuring 18O in tree ring cellulose and state my complete agreement with Pier's comments below and Willi Brand's reply also posted to ISOGEOCHEM.

We keep all cellulose samples dry in desiccators for a least several weeks before analysis to dry them out.  Them we do the analysis using a sealed sampler in Helium to keep the samples dry during analysis, and delay analysis for three hours after transferring the samples and standards from the desiccator to the sampler.  So I regard all comments on the problems of water build up in the cellulose as being completely accurate.

In addition there are numerous other problems:

We use a low temperature furnace (1130 degree C) with a nickel carbon catalyst as used for %O analysis for organic compounds, and have had to add a) doping of helium carrier gas with chloropentante to improve results and reduce memory effects, b) use tin capsules to avoid problems of shiver build up, c) use of a Porapak QS column before a molecular sieve column to separate organic contaminants from the CO, and then d) back flushing of these columns to get rid of the contaminants, e) different sized standards to correct for the non-linearity of CO in the source of the mass spectrometer, and f) a very complex post analysis spreadsheet that corrects for non-linearity and drift of the high and low calibration standards drifting at different rates.  An example of this spreadsheet "Generic Template Berkeley 1.00" can be downloaded at .

We have had poor results trying to use standards other than cellulose and now use three standards:

high standard)  IAEA C3 which we assign a value the 32.2 VSMOW to, based on the average value in the literature, we have not yet calibrated successfully with IAEA 601 and 602, (see problem described below).
low standard) A commercially bought Sigma cellulose that we analyzed against IAEA C3 intensively and have assigned a value of 26.4 delta 18O VSMOW.  We have not yet calibrated successfully with IAEA 601 and 602.
quality control) IAEA V9 our quality control standard which has an average value of 28.57 VSMOW plus or minus 0.21 for all of these run in the last two years (long term external quality control).

We are confident that these standards are internally reproducible, but because of the problems that Pier and Willi describe are not quite sure of an exact number on the VSMOW scale. This is certainly a problem and I completely agree with Pier that a certified standard, preferably two, would be desirable.  However, Willi's comment that this is difficult for the international standard is certainly true.  However, if we are trying to analyze cellulose or other organic matter, it would appear that we have to work out reproducible methods that will remove the effect of water contamination from organic samples, otherwise the analysis results may not be useful.

Our standards only cover a small range but 90% of our tree ring standards fall in this range.

We have found that the benzoic acid seems to analyze differently from cellulose in our system.  For example, the IAEA benzoic acid seem to have a memory effect we do not see in cellulose.  In order to get a stable number from benzoic acid, it appears we have to analyze at least 3 and preferably 4 standards in a row.   Strangely, there even seem to be a memory effect from the benzoic acid into the cellulose V9 quality control.  It appears the analyses including benzoic acid do not run well and we have been forced to analyze just with the cellulose standards as I described.  Since the structure of benzoic acid is quite different, we have speculated that it decomposes quite differently to form CO compared with cellulose.

Do any other labs have 18O numbers for IAEA C3 and IAEA V9?

I would have to say that this analysis is not for the faint of heart!  However, good for those who enjoy a challenge. 

I can send you a power point describing our method off list if you like. 

Hope this helps,

Paul Brooks.

At 01:18 PM 3/24/2008, you wrote:
Hi Simon,

The reality is that there is no standard. The only way one can go is to
'calibrate' a 'local working standard (cellulose)' against the international
scale. The problem is here the comparison of two different materials with
different matrices, or in other words comparing apples with pears.

There clearly is a need for certified reference material for d18O in
cellulose. Storing cellulose is not straightforward, since it is highly
hygroscopic (which basically does not effect the d13C, but does the d18O),
one of the reasons we have no cellulose standard available for d18O...

What people do is comparing with each other (e.g. Robin-round), but that is
no calibration. It creates a 'sub-scale' and the data obtained that way in
fact cannot be used for comparison with anybody not included in the
robin-round. Also, it introdices more uncertainty in measurements if this
'local-scale' is calibrated against the international scale (more steps in
traceability chain increase the combined uncertainty of a measurement). If
applying thermometry with such data, the result is very uncertain (actually
useless, because all thermometers are calibrated to the international scale,
and not to a local scale), and no small temperature differences can be
determined in any way.

Of course, if only used for comparison of measurements made by the same
method in the same laboratory, only repeatability is sufficient. But no
temperature determination can be done than, only relative difference can be
determined on the 'local' measurements.
I will explain matters like this in the short course that will be presented
by Bob Vocke and myself at the EGU2008 (13-18 April, in Vienna).

We really need a better situation here....

Best wishes,
Delta Isotopes Consultancy

Dr. Pier A. de Groot
Pastoor Moorkensstraat 16
2400 Mol - Achterbos
Tel. +32 (0)14 326 205
e-mail: [log in to unmask] or [log in to unmask]

Associate editor for stable isotopes of eEarth on-line magazine

Head of Isotopes in Geosciences of the European Geoscience Union (EGU).
EGU Home web-site:
Isotope program for the EGU2008 General Assemblee can be found at:

Organization Committee member of BASIS (Benelux Association of Stable
Isotope Scientists).

Visit my WEB-site about my ³Handbook of Stable Isotope Analytical
Techniques², with a link to the Elsevier web site on the handbook (marked:
ŒOrder Now¹):
last update: August 15, 2005
Volume I is available. Volume II is in press and will be available second
half of 2008.


> From: Simon Poulson <[log in to unmask]>
> Reply-To: Stable Isotope Geochemistry <[log in to unmask]>
> Date: Mon, 24 Mar 2008 11:56:57 -0700
> To: <[log in to unmask]>
> Subject: [ISOGEOCHEM] cellulose-d18O standards
> Dear All,
> We were planning to run some cellulose-d18O analyses in our lab, and
> I had a question about d18O standards.  A number of papers that have
> run similar analyses state that they calibrate a working cellulose
> standard (e.g. from Aldrich or Sigma) for d18O and use this when
> running samples, but the papers don't say what standard they use to
> calibrate this working standard.
> Cellulose standard IAEA-CH-3 is available from IAEA, but is certified
> for d13C only, not d18O.  Is there a consensus d18O value that people
> use for this standard?  Or do people use other certified standards,
> such as benzoic acid IAEA-601 (+23.3 per mil), or barium sulfate
> NBS-127? (+9.3 per mil)
> Thanks in advance for any comments and suggestions.
> Cheers,
> Simon Poulson
> --
> *****
> Simon Poulson
> Research Associate Professor
> Dept. of Geological Sciences & Engineering MS-172
> University of Nevada - Reno
> 1664 N. Virginia St.
> Reno, NV 89557-0138. USA
> (775) 784-1104
> Fax: (775) 784-1833
> [log in to unmask]
> *****

Center for Stable Isotope Biogeochemistry
Valley Life Science Building Room 3060
Integrative Biology - MC3140
Berkeley CA-94720
Phone: (510)-643-1748
Fax: (510)-643-1749
[log in to unmask]