Chromatography Forum

Our lab analyses sVOC using EPA8270D and let me tell you...liquid/liquid extraction of wastewater is such a pain in the behind...

I've tried researching for alternatives, but found none.

Is there some sort of method I'm missing?
Alternatively, is there any reliable and simple automatisation available?
I doubt that our lab will invest a lot of money into costly automation, but knowing the options is better than not.

My experience - albeit limited - with these standard methods is that the developers are usually a committee. They generally have thought of and tried (during the interlaboratory collaboratives) everything that might work for the analysis. They probably found that the solvent-solvent extraction is the best way to go.

However, if they didn't. Have you given solid-phase extraction a shot? If your waste water is really high in solids, it probably won't work very well. Just a thought............... Good luck. I understand your frustration. Solvent extraction of beer doesn't work very well either.

We don't do liq-liq on waste water, but some of our surface waters aren't much better I think.

how we do it:
- 100ml sample in a glass centrifuge-tube with GL45 cap.
- ad 6ml of CH2Cl2
- put on mechanical shaker 5min (Edmund buhler)
- centrifuge
- take the CH2Cl2 out by using a glass serological pipette and a pipette balloon (is it called that in English??)
- transfer the solvent into a standard test tube
- add dry Na2SO4 and vortex
- centrifuge again
- take of the dry solvent and fill your vial.

In our experience we recover 3-4ml of the initial 6ml of solvent.

you can always add salt to lower solubility, and when you add just a little bit of CuSO4 it is a lot easier to see the difference between water and solvent.

It still is liq-liq, but it is a lot easier than working with seperation funnels etc. We do approx 50-60 samples (incl extracted calibration curve and QC) on a single 9-5 day by 1 analist.

My experience - albeit limited - with these standard methods is that the developers are usually a committee. They generally have thought of and tried (during the interlaboratory collaboratives) everything that might work for the analysis. They probably found that the solvent-solvent extraction is the best way to go.

However, if they didn't. Have you given solid-phase extraction a shot? If your waste water is really high in solids, it probably won't work very well. Just a thought............... Good luck. I understand your frustration. Solvent extraction of beer doesn't work very well either.

I have not tried SPE personally, but my boss did, and said that it worked well only with filtered leachates or water samples.

The problem is that 90% of our workload is rough n' dirty inorganic industrial wastewater of various sources and we're analysing them for remnants of organics.

Developing a separate method for 10% of samples seems kinda wasteful.

Are you a beer chemist?

We don't do liq-liq on waste water, but some of our surface waters aren't much better I think.

how we do it:
- 100ml sample in a glass centrifuge-tube with GL45 cap.
- ad 6ml of CH2Cl2
- put on mechanical shaker 5min (Edmund buhler)
- centrifuge
- take the CH2Cl2 out by using a glass serological pipette and a pipette balloon (is it called that in English??)
- transfer the solvent into a standard test tube
- add dry Na2SO4 and vortex
- centrifuge again
- take of the dry solvent and fill your vial.

In our experience we recover 3-4ml of the initial 6ml of solvent.

you can always add salt to lower solubility, and when you add just a little bit of CuSO4 it is a lot easier to see the difference between water and solvent.

It still is liq-liq, but it is a lot easier than working with seperation funnels etc. We do approx 50-60 samples (incl extracted calibration curve and QC) on a single 9-5 day by 1 analist.

Sounds quite a bit simpler, but is it a in-house method, or is there some kind of known method associated with it?

Do you extract just once or many times? Do you extract in 1 pH, or do you acidify/alkalinify? What kind of surrogate recoveries do you usually get?
What kind of calibration curve and for how many compounds?

I'm a bit restricted by the base method here...but I'm very interested to see what other people use.

Sounds quite a bit simpler, but is it a in-house method, or is there some kind of known method associated with it?
Yes it is an in-house method, but I don't really see a difference in shaking a sample in a separation funnel or in a centrifuge tube

Do you extract just once or many times?
Do you extract in 1 pH, or do you acidify/alkalinify?
just once at one pH

What kind of surrogate recoveries do you usually get?
We compare our surrogate recoveries to the mean of the recovery found in our calibration (which is made in drinking water). Alle recoveries must be within a 70-130% ratio (and most are within 80-110%)

What kind of calibration curve and for how many compounds?
Calibration is lineair, ignore zero and 1/x; R² 0.995 or better, used standards: 5-10-25-50-100-250-500-1000ng/l (standard conc is the concentration in the spiked drinking water that is extracted to produce the standard that is injected).

we use this method for PAH's, PCB's and organo chloro and posphor-pesticides, in total there are about 130 compounds that are reported.

injection on GC is large volume 50µl CH2Cl2

We don't do liq-liq on waste water, but some of our surface waters aren't much better I think.

how we do it:
- 100ml sample in a glass centrifuge-tube with GL45 cap.
- ad 6ml of CH2Cl2
- put on mechanical shaker 5min (Edmund buhler)
- centrifuge
- take the CH2Cl2 out by using a glass serological pipette and a pipette balloon (is it called that in English??)
- transfer the solvent into a standard test tube
- add dry Na2SO4 and vortex
- centrifuge again
- take of the dry solvent and fill your vial.

In our experience we recover 3-4ml of the initial 6ml of solvent.

you can always add salt to lower solubility, and when you add just a little bit of CuSO4 it is a lot easier to see the difference between water and solvent.

It still is liq-liq, but it is a lot easier than working with seperation funnels etc. We do approx 50-60 samples (incl extracted calibration curve and QC) on a single 9-5 day by 1 analist.

Sounds quite a bit simpler, but is it a in-house method, or is there some kind of known method associated with it?

Do you extract just once or many times? Do you extract in 1 pH, or do you acidify/alkalinify? What kind of surrogate recoveries do you usually get?
What kind of calibration curve and for how many compounds?

I'm a bit restricted by the base method here...but I'm very interested to see what other people use.

We are looking into the same thing, and for EPA8270 you can use 3510C as the extraction method with a sample size from 100-1000ml. The only difference would be that you need to do three 6ml extractions at both pH levels to comply with the method. If you do EPA625.1 for waste waters you can use the same procedure.

http://epa.ohio.gov/Portals/30/brownfie ... alysis.pdf

Ohio EPA has this presentation on the subject, which is one of the things that got me going on this.

Has anyone looked into validating EPA 3511 to use instead of 3510?
I know that many EPA methods take the "prove it and use it" approach, just wanted to know if anyone succeeded in adapting it for wide 8270 usage. The in house method mentioned above is close enough, but we're an ISO 17025 accredited laboratory and I was wondering how feasible it is.

For me (here in Belgium) the EPA numbers don't say a lot since we don't use them.
But we are a drinking water lab and this method is under accreditation.
As you say we had to prove it works by proficiency tests and by validation results.

Method 3511 is a nice alternative if you can achieve adequate sensitivity.
Extraction pH is dependent on the target analytes, e.g. there is no reason to do pH adjustment for PAHs, no reason to do a basic extraction for organophosphate pesticides.

UCT has a SPE method that works great for 8270. You should be able to find their how to video on YouTube.

Don

I've just recieved the large VOA vials needed for 3511, wish me luck

I have run a Demonstration of Capability study using 100ml boston round bottles as the separatory funnel and shaking on a horizontal shaker table. Using the solvents listed in the Bottle Liquid Liquid Extraction procedure http://blle.org/method-approval-process/ which has been approved for several methods so far. The shaker table is a modification that I did to speed up the process. Shaking for 10 minutes each time then 10 minutes in a centrafuge(definitely helps separate the layers and for actual samples almost necessary for emulsion breaking) I use a pipette to remove the organic layer. Three extractions in the acid, then add salt and three extractions in the base, just as the blle procedure. I just managed to average almost 70% recoveries on EPA625 analytes from DI water and 60% on actual waste water samples. It is actually better recoveries than we were getting on continuous liquid liquid extraction. I then use 5-10ul injections to improve sensitivity.

I have sent an email to the MICE site to ask if the sample bottle can be considered equivalent to a separatory funnel so we can claim EPA Method 3510C as the extraction method, or if we have to claim EPA Method 3511 microscale extraction method with modification to larger vial size. If a separatory funnel can be defined as not having to include a stopcock at the bottom then either method should be applicable.

- 100ml sample in a glass centrifuge-tube with GL45 cap.

BMU_VMW, where did you find these centrifuge tubes? Those I think would be even better than doing the extraction in the bottle, just for purposes of centrifuging since you have to limit the centrifuge speed on the bottles to prevent breaking them.

Hey James,

May I contact you personally? The only option on your profile is Yahoo messenger and that does not seem to work.

BMU_VMW, where did you find these centrifuge tubes?

we use these :

https://shop.omnilab.de/info1841_Centri ... centrifuge

Our centrifuge is a Hettich Rotana 460R with 4 buckets. The adaptors for these centrifuge tubes were custom made.


edit extra info:
For our 'low matrix' samples we don't use the centrifuge tubes but standard (amber) 150ml GL45 bottles. MAKE SURE YOU HAVE A TEFLON SEAL IN THE CAPS !!
When using 100ml sample + 6ml CH2Cl2 there is still enough air left to shake well. Extra advantage is that these bottles are smaller than the centrifuge tubes and we can place up to 25 on the shaker at once.

Our shaker is a Edmund Bühler and we did order a model with a longer stroke than standard in order to get the solvent mixed better.