United Nations Economic Commission for Europe -
International Co-operative Program on Assessment and Monitoring of Air
Pollution Effects on Forests
______
Working Group on QA/QC in Laboratories
Minutes from the Meeting of the Heads of the Laboratories
7.-8. September 2017 in Verbania-Pallanza
Thursday 7. September 2017
(All presentations can be downloaded from the ICP Forests website: http://icp-forests.net/group/qualityinlaboratories/page/document-archive)
Nils König welcomed the participants of the meeting in Pallanza and thanked the team of the National Research Council – Institute of Ecosystem Study for the invitation.
Rosario Mosello as a former chairman of WG QA/QC in Labs recalled a few facts from the past activity of the group in ICP-Forests: from the first deposition WRT in 2002, 2005, 2009 and up until the recent ones. He mentioned the efforts that have been made over time to improve the quality of the chemical results: discussions with Expert Panel members, program of exchange knowledge between labs, and the publications. Mr Mosello pointed that some QA/QC problems have never been solved and therefore there is still a need for effort to make further improvement.
Topic: Report from the last meeting of the WG QA/QC in Labs in Zagreb (Croatia)
Nils König reported on the last meeting in March 2017. The main topics were: ring test organization, preparation of presentations for the meeting of heads of the labs on parameters that were problematic in last ring tests, organization of the next meeting of the heads of the labs, proposal of improved coding for C-CO3 and Corg determination on elemental analyzers, comparability of aqua regia digestion with reflux method, heavy metals in the ICP Forests monitoring program, and problems and possible changes in data submission.
Topic: Future ring tests – preparation, evaluation, financing, timetable, new features
Next 9th Atmospheric deposition and soil solution Working Ring Test is scheduled for 2018/2019
After the discussion the timetable was agreed:
Invitation - 17 September 2018
Registration deadline – 19 October 2018
Distribution of the samples and invoices – 12 (or 19) November 2018
Data submission and payment deadline – second half of January 2019
Requalification will continue through February 2019.
9th Soil Interlaboratory Test Programme is scheduled for 2018 with the timetable:
Invitation – February 2018
Registration till 1.5.2018
Distribution of test samples & invoice for registration fee - first week in May
Payment of registration fee (within 30 days)
Data submission first week in October 2018
First results October 2018
Re-qualification process starts November 2018
Discussion of the first results on ICP Forests expert panel meeting Spring 2019
20th Needle/Leaf Interlaboratory Comparison 2017/2018 is already in progress. Samples were sent to the participants in July 2017. Next steps are:
Data submission till January 1st, 2018
First results (January 2018)
Re-qualification process finished (September 1st, 2018)
Next, 21th Needle/Leaf Interlaboratory Comparison is scheduled for 2018/2019. The timetable is as follows:
Informing the participating labs via email (March 2018)
Registration date (Deadline July 9th, 2018)
Distribution of four test samples & invoice for the participation fee (July 2018)
Data submission till January 1st, 2019
First results (January 2019)
Re-qualification process finished (September 1st, 2019).
Topic: Presentation of the results of the 18th and 19th Foliar Interlaboratory Comparison Test
Mr. Fürst presented the results of the 18th and 19th Needle/Leaf Interlaboratory Comparison Tests. The numbers of participants were in both tests lower than before (18th:53 labs; 19th:45 labs). This decrease is also a consequence of the stop of funding of the monitoring program.
The samples in the 18th test were: beech leaves, pine branches and two spruce needles; in the 19th test: three samples spruce needles and one pine needle. In the 19th test there is a big group of laboratories with correct results but on the other side there is a big group with very bad results (one lab failed with nearly all results). Mr. Fürst tries to compare the results with the answers of the QA/QC questionnaire:
Participating in more than one ringtest per year has a positive influence on data quality. Laboratory accreditation and a higher number of analyzed samples per year have a small positive influence on data quality. The results are not so good if the lab has the opinion that the staff is good trained on the method and the matrix!
Microwave methods are the most used pretreatment method (target temperature 200°C over 20 minutes). For nitrogen and carbon element analyzers are the common determination method. Multi-element methods like ICP-AES and ICP-MS are the most used determination methods for all other parameters. It is important to trained the staff on these instruments and to service the instruments in regularly intervals!
Cadmium results are really good in both tests (only between 2.1-8.0% non tolerable results).
P, Mg, K, Cu and Pb are the most “problematic” parameters with more than 15% of non tolerable results in one or in both tests.
A good ringtest result means that the % Recovery is close to 100% and the variation between the replicates is small. A requalification is mandatory for all laboratories, if they plan to submit monitoring results to PCC! Use your labcode, if you submit the requalification papers per email. Use only the Excel sheet to summarize your requalification results and don’t forget to submit additional printouts (as pdf) about calibration, dilution factors.… It should be possible to recalculate your results! Submit your ideas about your opinion for missing the qualification and how you avoid the same error in future.
FFCC offers reference samples (http://bfw.ac.at/rz/bfwcms2.web?dok=5146) and a new spruce sample with known content of heavy metals. The registration deadline for the 20th Needle/Leaf Interlaboratory Comparison Test is exceeded, the samples are submitted to the labs.
Because of the use of multi-element determination methods (ICP-AES and ICP-MS) a lot of labs already analyze more and more elements without extra costs. A special evaluation for heavy metals (As, Co, Cr, Hg, Mo, Ni, Tl, V) was made for the 18th and 19th test. A repetitious accuracy was reached in both tests for these elements and the reached results are comparable to them for Cu, Cd and Pb. Therefore As, Co, Cr, Hg and Ni are including in the normal ringtest evaluation as optional parameters.
Foliage ringtest material is needed; if someone plans to collect such sample should contact Mr. Fürst first.
Topic: Presentations and discussion of analytical problems, new methods, method comparisons etc. proposed by the participants - plant analysis
Carmen Iacoban presented the ideas about comparability of dry ashing and wet digestion for total phosphorus analysis from needles/leaves. 20 samples from foliar ringtest were in parallel wet digested and dry ashed for P and the results were compared to the mean value for the ringtest. Dry ashing seems to cause the underestimation of P while wet digestion causes the overestimation compared to the ringtest mean value. In the discussion some possible reasons were supposed: element losses due to the dry digestion or the interference of silicates after wet digestion in molybdenum method. Possibly choosing of the other method of determination and/or the other instrument would be helpful in identifying the reason.
The issue of drying of foliage and litterfall has been raised at the last Expert Panel meeting for Foliage and Litterfall. The influence of drying procedure on the content of the elements in foliar and litterfall material was presented by Alfred Fürst. According to the Manual part XII dry mass of 100 leaves and 1000 needles has to be reported for 105°C drying temperature. Chemical analyses for C, N, Hg has to be done in the material dried at <70°C and results have to be corrected for the moisture. Moisture determination is made for the subsample not chemically analysed and dried at 105°C.
In May 2016 Manual Part XII was updated, equivocal sentence “Only the total concentration of elements in needles or leaves must be given by reference to 105°C-dried material (Table 2)” was replaced with “The total concentration of elements in needles or leaves and dry mass of 100 leaves or 1000 needles must be given by reference to 105°C-dried material (Table 2).”
However, the correction factor might be inaccurate if the subsample of ground needle is stored for a long time before drying at 105ºC as the weight might change. Moisture content should be then determined within a short time or right before the chemical analysis.
In litterfall samples (Manual Part XIII) dry weight of biomass and dry mass of 100 leaves or 1000 needles is determined after drying at max. 70°C but results of chemical analyses are corrected to dry mass at 105°C. This way the input of elements in kg m-2a-1 cannot be calculated from the results stored in the database. Correction factor for dry 105°C weight of biomass is necessary. To solve the problem heads of the labs will propose to the Expert Panel of Foliage and Litterfall to add a new sentence to the Manual Part XIII: “The total concentration of elements in needles or leaves and dry mass of 100 leaves or 1000 needles must be given by reference to 105°C-dried material.” so that now dry weight at 70°C and at 105°C will be submitted, both mandatory.
At the last EPM Foliage/Litterfall Mathias Neumann raised an issue of the potential loss of the volatile compounds like phenoles, terpenoides or alcohols in samples dried at 105°C or even at 70°C.
To avoid the losses samples should be freeze-dried but in case of volatile compounds in air dried samples losses are expected; this problem cannot be solved at the moment.
Correct sulphur determination of maple leaves with higher calcium content. Alfred Fürst presented the experience with analysing S in maple leaves with high content of Ca on LECO analyser.
The peak of inorganic S is not symmetric and appears later than in most of the typical leaf samples. As a result, the endpoint of the peak is too early recognized by the software and the S content is underestimated. To improve the integration of peak it is necessary to increase the analysis time from 140 to 180 sec and to reduce the sample weight from 0.20 to 0.15 g.
Topic: 8th Deposition and Soil Solution Interlaboratory Comparison Test.
Anna Kowalska summed up the results of the last deposition and soil solution ring test 2016/2017.
Compared to the previous ring test, time for analyses and for requalification was shortened to 4 months starting from the registration due to the risk of instability of the samples. 40 labs (25 countries) registered and 39 of them submitted results. Five of the samples were natural waters: stemflow (beech), throughfall (spruce), bulk open field, two soil solutions. Labs were requested to analyse in these samples all the mandatory parameters and P-PO4, Fe, Al, and Mn. Sixth sample was syntetic for alkalinity measurements only.
P-PO4 in samples 2, 3, and 4 and N-NH4 in sample 4 were excluded from the evaluation due to the low concentration – more than 33% of results was <LOQ. DOC in one sample was excluded from evaluation because it did not pass the stability test. Alkalinity in soil solutions (sample 4 and 5) was not evaluated because pH was <5.
For mandatory parameters, 13% of Ntot, 11% of DOC and 7% of alkalinity results were missing. Nine labs did not submit all the mandatory parameters.
Percentage of acceptable results was above 90% for 15 labs, between 80% and 90% for next 14 labs. Three labs reported less than 40% of acceptable results. Requalification was necessary for 25 labs but most of them failed for only 1 or 2 parameters; 20 labs successfully requalified.
Too high limit of quantification was reported by 1 - 3 labs for Cl, DOC, Mg and N-NO3 and these results did not pass the test.
The most problematic parameters were: Ca, pH, alkalinity, and N-NH4; less than 80% of acceptable results was recorded for those parameters and more than 5 labs had to requalify for them. As much as 10 labs failed for Ca.
Labs reported in requalification the following reasons for failure:
- technical problems (instruments, sensors, electrodes failure/ageing, poor compensation of temperature): Ca, alkalinity, Ntot, pH, N-NH4, DOC, conductivity;
- Quality control material or standards of poor quality: Ca, Alkalinity, Ntot, pH, DOC;
- Calibration (failed for low concentration samples): Ca, Ntot, DOC;
- Contamination of samples, spectral interferences: Ca
The common problem was reporting wrong units or lack of recalculation for alkalinity, S-SO4, N-NH4, N-NO3, P-PO4. It happened also to the labs participating regularly in WRT’s. Some of the common errors could be avoided if labs would read the instruction delivered with samples and if the quality tests (conductivity, ionic balance, N check) would be performed as a routine. Excel file for such tests is downloadable from ICP-Forests web-page (Working Group Quality in Laboratories-> Quality Assurance and Control in Laboratories -> new excel file for analytical data validation (with DOC)). If the lab uses LIMS system, please contact Aldo Marchetto for help with automation of the transfer of data between system and excel file.
Percentage of non tolerable results although remarkable lower than in ring tests organized between 2002 and 2009, stays at the level comparable with two previous rounds of ring tests (2013 and 2015). For most of the parameters, except for Ca, less than 20% results falls outside the tolerable limits. Obviously, concentration of the analytes in samples can influence the score of the labs. Higher % of unacceptable results was recorded in samples with low concentration.
In future ring test submission of all requested parameters in all the samples will be mandatory for labs; the list of requested parameter per sample will be delivered with samples by ring test provider. 9th Deposition and Soil Solution ring test is planned to start in autumn 2018; deposition and soil solution samples are needed for the ringtest, unfiltered in the amount of 60l each. Please, contact Anna Kowalska in case you plan to deliver such samples.
Topic: Tolerable limits and LOQ´s for Al, Fe, Mn for deposition and soil solution ring tests – Anna Kowalska, Nils König
Three parameters that are obligatory in soil solutions if pH<5: Al, Fe, and Mn were not measured in 6ht and 7th WRT but were requested in 8th WRT. Al, Fe, and Mn are going to be included in the next ring tests. The number of labs that submitted results in 8th WRT (i.e. 31-32 labs) was high enough to make the statistical evaluation of the results. Maximum tolerable quantification limits (LOQ) and tolerable limits for future ring tests were calculated. Tolerable limit for Al is proposed to be 30% for concentration ≤0.1 mg/L and 15% for concentration >0.1 mg/L. For Fe in whole concentration range, proposed tolerable limit is 30%. For Mn proposed tolerable limit is 15% for concentration ≤0.025 mg/L and 10% for concentration >0.025 mg/L. A questionnaire sent by Nils König to the labs before the meeting has revealed that in different labs using the same type of instrument (ICP-OES) LOQ’s may differ by a factor of 100 for Al (from 0.001 to 0.1 mg/L), 1000 for Mn (from 0.0001 to 0.1 mg/L) and even 3000 for Fe (from 0.0003 to 0.1 mg/L). After the discussion, maximum LOQ for future ringtests evaluation was decided to be 0.05 mg/L for Aluminium, 0.02 mg/L for Iron, and 0.01 mg/L for Manganese.