From my point of view an appropiate extraction in urban dust doesn't assure that the same extraction is going to work in a complete different sample matrix such as marine sediment. Nevertheless, taking intto account that low availability of SRM it can a complementary test for the method validation.

My personal view is that it's almost impossible to get an SRM/CRM for every matrix in existence, especially one that's certified for the substance for which you're monitoring. The comparability of the physicochemical properties of the two matrices will determine whether the results are comparable. To Neus's point, numerous soil-based matrices exist, and depending on the analysis and extraction method, and the nature of the chemical you're analyzing, the results between two can be considered comparable.

However, there appear to be a couple of SRM's and CRM's for both inorganic and organic components in marine sediment (NIST SRM's 1941b, 1944, 2702, NIES CRM 12, etc). While it would probably be best to use matrices that are closest to your matrix, you could possibly make an argument for the urban dust depending on what you're measuring. Which compound are you looking at?

Thank you very much for your enlightening ideas. Im using total digestion technique (HF+HClO4+HNO3) in 45 minute microwave oven. I plan to measure inorganic elements such as major and trace metals (mostly heavy metals).

As stated above, I would obtain a SRM that matches your matrix as much as possible. For many organic compounds, there are not a lot of choices, but for inorganic elemental analysis, there are several. We typically use the CRMs from the National Research Council Canada, since they are more reasonably priced than NIST standards and are excellent.

Here is a link to their inorganic CRM site:

http://www.nrc-cnrc.gc.ca/eng/solutions/advisory/crm/inorganic/list_products.html

Better to find SRM with comparable matrix. For sediments there are now any problems. But if you use using digestion as sample preparation method you can prove your measuring technique, but not sample preparation method. Urban dust have more black carbon compare to marine sediments, but anyway you can use recovery test for your sediment samples to check sample preparation method.

Marine sediment can be one of the nastiest matrices, and can vary widely from location to location. You will never get a true representative SRM for your area especially for sediment. Urban dust is better than nothing though. I would try for a true marine sediment SRM - it will be loaded with sulfur and other nasty interferences.

No.

First of all I think that you should try to find a SRM/CRM which is as close to your sample matrix as possible. But as already mentioned it is almost impossible to find a SRM/CRM which matches perfectly your samples if they are marine sediments. I would suggest that you use a combination of reference materials. For marine sediments I use MESS-2 (now MESS-3) and GBW 07313. GBW 07313 is a Chinese reference material which also reports reference values for lanthanides. I have also used NIST 8704, which is a Buffalo River sediment reference material, but it has shown out to work well with Baltic sea sediments. Of course it is a question of money, but I feel more confident with two or normally three CRM:s in every sample batch.

When an appropriate SRM is not available, I have found that inter-laboratory comparisons can be useful, especially if totally different analytical approaches are employed by the different labs involved. If the agreement is "reasonably" good, then you can have confidence that your results are reasonably accurate and that your recoveries are good. But, if the agreement among the labs is poor, that also is important to know and worthwhile following up.

no problem if your sample matrix is not exactly matching with SRM  provided if  you have good quality  calibration standards and you are using uniform sample preparation procedure for sample,srm and blank .we have rich experience of analysis on ICP-OES ,it is our observation and we have experimently done this.

Thank you very much for your contributions. Limited funding for research is a common problem in developing countries. Therefore sometimes we need to make wise decision without compromising the data quality. I am convinced with most of your answers and it helps me a lot. Thank you very much!

You can use closest matrix as possible. Marine and dust samples can't be too similar, but useful in case if you can find second independent method. In scientific research second method is good possibility for limited finances

Dear it is true that mostly it is not possible to have SRM close match to your sample but if you have srm having some matching it  works , practice which we adopt in our lab we make set of standard mixtures of various matrices from commercially available standard solutions and verify  with available srm  validate our  analytical  results/standard mixtures  by analysing the same sample by other compilable  technique and then apply these to real sample analysis ,