Wonderful question...to which there exists a variety of wonderful answers. This is a problem with which all environmental chemists have likely grappled. We have extracted literally thousands of animal and plant tissue, soil and sediment etc. samples for PAHs and diverse (and continually evolving) suites of organohalogen POPs over the years using accelerated solvent extraction (and supercritical fluid extraction) with subsequent GC-MS analysis (EI and CI/NCI etc.) characterization and we have achieved remarkably consistent (both accurate and precise) recoveries of our standard compounds. Our recovery standards are usually heterogeneous mixtures of PAHs, PCBs, hetero-PAHs and BFRs etc. at concentrations determined by our experiences with such sample matrices. After some initial method development, we found that spiking 1 mL of our recovery standard mixtures directly onto the surface of the sample (in the extraction vessels) works great. Our sample sizes typically range from 1-10 g for freeze-dried animal/plant tissues and 20-30 grams for freeze-dried soils/sediments. For my SFE work, however, as I am typically constrained by sample sizes of only about 1-5 grams (depending upon matrix), I have achieved nicely consistent results by reducing the spiking volume to 0.1 mL (100 uL). With such smaller sample sizes, during initial method development work, I found that 1 mL prematurely eluted my recovery and target analytes. I have experimented with spiking and homogenization techniques as well, but, presumably on account of excessive sample manipulation, have achieved significantly lower recoveries. I hope that our experiences are of some value to you. Good luck!!

Wonderful question...to which there exists a variety of wonderful answers. This is a problem with which all environmental chemists have likely grappled. We have extracted literally thousands of animal and plant tissue, soil and sediment etc. samples for PAHs and diverse (and continually evolving) suites of organohalogen POPs over the years using accelerated solvent extraction (and supercritical fluid extraction) with subsequent GC-MS analysis (EI and CI/NCI etc.) characterization and we have achieved remarkably consistent (both accurate and precise) recoveries of our standard compounds. Our recovery standards are usually heterogeneous mixtures of PAHs, PCBs, hetero-PAHs and BFRs etc. at concentrations determined by our experiences with such sample matrices. After some initial method development, we found that spiking 1 mL of our recovery standard mixtures directly onto the surface of the sample (in the extraction vessels) works great. Our sample sizes typically range from 1-10 g for freeze-dried animal/plant tissues and 20-30 grams for freeze-dried soils/sediments. For my SFE work, however, as I am typically constrained by sample sizes of only about 1-5 grams (depending upon matrix), I have achieved nicely consistent results by reducing the spiking volume to 0.1 mL (100 uL). With such smaller sample sizes, during initial method development work, I found that 1 mL prematurely eluted my recovery and target analytes. I have experimented with spiking and homogenization techniques as well, but, presumably on account of excessive sample manipulation, have achieved significantly lower recoveries. I hope that our experiences are of some value to you. Good luck!!

We analyzed PAH from aerosols in our laboratory and we use a internal standard (IS) mix of deuterated-PAH and some deuterated surrogates according to EPA-TO13A. IS is applied directly on samples (PUF and Quartz filters) is possible a good % recovery between 80-105%. Extraction is via Soxhlet around 18 h with 10% diethyl ether/ hexane. The spiking withsurrogates for soxhlet is 20 uL at 50 ug/mL, the same for both field and blank samples.

I recently performed this analysis with 2g soil samples. I added the recovery standards then shook on a mechanical shaker for approximately an hour. I have found that mechanical shakers are best used for spiking a soil sample with recovery standards, and also for ensuring the extractant solution is thoroughly distributed throughout the sample.

the best to my knowledge is to add the standards in several drops in the soil surface (usually we only sieve the soil - best after freeze-drying - grinding may not be an advantate), and let the solvent evaporate. It depends a bit on the chemicals' polarity what you add. Nice is acetone, because it evaporates rapidly, but it is hard to control standard solution concentrations (it also evaporates from the flask).. Methanol we only add when we extract with sovlent -water mixtures. So finally our recommendation is to solve the standards in ethylacetate - it is stable - evalporates as well though a bit more slowly. ANnd then just distribute it at several drops...see, e.g., attachment done for pestides, but we did it also for a range of pharmaceuticals.

Best

Wulf

Michael, having done so much PAH analysis, what is your feeling about the best solvent system for extraction from soil & sediment? The two matrices are very different, but I'd prefer to use a single solvent system if possible. Our current plan is to use 1:1 acetone:hexane. Also, do you have a sense of the utility of including a silica (florisil) cleanup step with this solvent system? I see it in several methods, including EPA's, but I have a hard time believing that the cleanup works all that well with acetone in the mix (pretty strong eluent in adsorption chromatography) unless perhaps a solvent exchange to neat hexane were done first.

Charles, I have observed that acetone as solvent, extracts several polar compounds that would be interfered with your target analytes. In case of aerosols or emissions from wood-stoves collected in quartz filters, there are some compounds that posses same m/z ions than PAH and it's possible that cuantification is overestimated. for example Syringol (2,6 dimethoxy phenol) has m/z=154 same for Acenaphthene, even both have different retention times in column (RTX-5MS phase) and SIM mode in MS is used. Sometimes it's important the choice of cleanup columns, in my experience Silica has given good results.

Charles,

Thanks for your note. Another great question. Our current analytical approaches (as it usually does for all of us doing this type of work) required quite a bit of developmental effort to validate (especially given the magnitude and diversity of our sample stream). In the early going, we experimented with a variety of solvent regimes (including acetone:hexane mixtures in fact), but, as we have historically chased such a diverse suite of analytes in such a diverse suite of sample matrices, we ultimately found that 100% DCM gave us the best and most consistent recoveries for the largest suite of analytes across the largest spectrum of sample matrices. We call this our “brute force approach”…and it has been pretty good to us for many years now. We have a couple of the Dionex ASE 200 systems with extraction conditions set to the following for virtually all extractions: 2 extraction cycles using 100% DCM (~30 mL total per sample) at P=1000 psi and T =100°C; 5-min heat cycle (to ensure thermal equilibration of vessel/sample), followed by two 5-min static extraction cycles, followed by a vessel flush and 3-min solvent/sample purge.

I presume you are referring to use of silica directly in the extraction vessel at the exit side of your samples to achieve simultaneous cleanup? If so, funny you should ask, as I did a fair amount of development work on this very thing. Though the story is long and tedious (as most method development sagas usually are), the take-home result is that we found that filling the remaining vessel void volume with silica did not provide sufficiently compelling results to justify reducing our sample sizes (especially for biota samples (often aquatic organisms), as we try to maximize their masses in the vessel to maximize detection on the back end).

That said, in support of my basic research efforts where I am forced to deal with much smaller sample sizes (e.g. 0.5-2 g dry weights for my work with POP uptake by soil invertebrates), I have actually realized tangible cleanup benefits using silica directly in the extraction vessel and extracting with 100% DCM using the aforementioned conditions. My best results to date have been achieved with crickets and other soil arthropods (with which I must grapple with small, e.g. ~1-2 g dry weight sample sizes). For such samples, I have been able to go directly to a final cleanup on 2 g silica columns and then straight to GC-MS after solvent concentration. As you might imagine, this makes analytical life quite a bit easier ☺. But, again, there is always a trade-off on sample sizes.

In regards to your proposed acetone:hexane mixture, my experience would have to agree with yours…acetone is likely to elute a fair amount of the polar (especially biogenic) compounds that we would typically seek to retain. Now, in my SFE work (both basic and contractual high sample throughput work), I am able to achieve appreciable sample cleanup efficiency using e.g. 7-10 g silica added to the extraction vessel. However, as mentioned before, because of the vessel designs on my instruments, I am usually constrained to ~1-2 g dry weight sample sizes. This typically leaves about half the vessel volume to be filled with silica. With this approach, I use cryogenic analyte trapping (on a few grams of modified silica actually) and then elute analytes directly to GC vials using about 1.5 mL of solvent (usually isooctane or some combination of DCM/hexane depending upon targeted analytes). I can then reduce the solvent volume directly in the vial, spike in my quantitation standard(s) and go directly to GC-MS. Now, that said, in achieving this, I must also swap out my injector liners a bit more frequently, as they tend to get fouled up a bit more frequently (as you might imagine). I must also clip off a few centimeters of the GC column head a bit more frequently than some of my colleagues as well ☺

I hope our experiences help move your own work forward Charles. If I can assist further, don’t hesitate to drop me a message.

Good luck!

Michael

At start the question was about spiking. Somehow this turned onto the question of recovery from environmental samples, if I noticed correctly. Regarding the latter, nobody seems to care about bound residues. For this two step extraction including saponification seems necessary.