Dear Johanna Ludwig ,

Once aromatic hydrocarbons absorb UV light, they emit fluorescent light at a longer

wavelength. By measuring the intensity of the UV fluorescence light, one can determine the amount of aromatic hydrocarbons, which can be related to the total amount of hydrocarbons providing that the ratio of aromatics to total hydrocarbons remains relatively constant.

The UV fluorescence technique is the second most widely used and accepted

after infrared. It is very sensitive and requires no use of CFC solvents, but it is important to remember that should the ratio of aromatic to aliphatic (or aromatic to total hydrocarbons) change due to, for example, the inclusion of new oil streams from different fields, then a new calibration should be established before a measurement is carried out.

Below info, may help for the calibration curve:

Calibration for Lab Methods

Two issues are specifically discussed here.

• Calibration oil

• Method for preparing calibration standards

For a reference method, usually calibration oil is defined. For example, in the

ISO9377-2 and OSPAR GC-FID methods, this calibration oil is a 50:50 (in weight)

mix of two types of mineral oils, Type A and Type B, both containing no additives.

An example of Type A is a diesel fuel; an example of Type B is a lubricant.

Despite the specification in the method for using a stabilized crude oil from the individual installations as a calibration standard, not everyone used crude oil as a calibration standard. The use of different calibration oils can have a marked effect on the final oil in water results. This is because instruments can respond differently to the different calibration oils.

Regards,

You might want to do what is called "procedural standardization"  Add a known amount (say 300 ppm) of your oil standard into water and extract with hexane. Use this as your 300 ppm standard calibration solution.

For quantitative analysis the other part of extracting water with organic solvent is you need to account for volumes.  It is best to extract sample 3 times, then concentrate to a known volume.  That way you are consistent and know initial volume of water extracted (say 500mL), and final volume (say 5 mL).  Reproduce this exactly with your calibration standard and it should be fine.

Take a look at the paper in this link.

 BASZANOWSKA, E. et al. Influence of oil-in-water emulsions on fluorescence properties as observed by excitation-emission spectra. Journal of the European Optical Society - Rapid publications, Europe, v. 8, oct. 2013. ISSN 1990-2573. Available at: . Date accessed: 10 Jul. 2015. doi:446.

Take a look at the paper in this link.

 BASZANOWSKA, E. et al. Influence of oil-in-water emulsions on fluorescence properties as observed by excitation-emission spectra. Journal of the European Optical Society - Rapid publications, Europe, v. 8, oct. 2013. ISSN 1990-2573. Available at: . Date accessed: 10 Jul. 2015. doi:446.

Hi, Johanna,  you have a very interesting question. I am not sure I can help you but let me also share some thoughts. I suppose that you are using the standard procedure and applying a suitable excitation/emission wavelength pair, having no any overlap with the elastic scatter peak. So, I was thinking about several points:

(1) Are you sure that after extraction, the hexane phase is "clean" of micro emulsion, i.e., micro-drops of water containing your oil? (I admit that the contribution of micro-drop phase to the overall concentration of oil components in a hexane phase should be low; nevertheless..; there is also an issue of a polar micro-environment effect on spectral shifts and intensities and differences in a micro-environment between a sample and standard)

(2) Another point is the question which people from service companies like to ask. Did you check how clean is your cell? Might it be that first, you have checked your standards, and then, the hexane extract was examined? You said you made a device calibration with the 5000 ppm oil solution in hexane, and, perhaps, then, the aqueous extract was checked. Small residuals from a sticking concentrated standard on cell walls may affect your readings.  Did you try to check your zero concentration standard in hexane in the end of your sample analysis? It should yet show zero. 

(3) I did not really understand the second calibration you made but my feeling is that you may deal not with a solution but with a micro-emulsion containing oil, water and hexane; there might be issues of scattering, too

(4) I do not think there is any chemical reaction with a distilled water but perhaps one should consider that oxygen is a recognized quencher of fluorescence, and its concentration  in water and hexane or oil will be very different. So, you may try to make measurements with argon or nitrogen-pre-purged samples.

(5) the last point (which may reflect my lack of understanding of your system): fluorescence spectroscopy is working perfectly to detect traces of aromatic compounds in water. For example, PAHs (present in oil, too) may be easily detected in water as is. So, why do you need the extraction at all? Why not analyze water sample as is? is it  because you have no true solution but an micro-emulsion, from a very beginning? or it is just the issue of a standard curve? In that case you could prepare the oil stock solution in methanol or acetone, dilute it in ddwater and (if very needed), even to purge out acetone?

eventually you water or apparatus got contaminanted?

Thank you for your answers! I checked, that the cells and the apparatus are clean. I don't know about the deionized water, a contamination might be possible... Also, I did not have a milky solvent phase, it was perfectly clear. 

I cannot analyze the oily water without hexane, because the oil is not distributed perfectly in the sample. It is no stable emulsion. So I am not able to catch a representative amount of oil in a 1ml cuvette. 

Mikhail, your point about the oxygen quenching sounds interesting. Do you know if there usually can be big impact on the fluorescence? It is true, that in my oily water the oxygen concentration might be elevated. 

Hi Johanna -

Can you tell us what instrument you are using?

Also running the hexane through a sodium sulfate column will remove any water in your hexane extract.

I am using a TD500d by Turner Instruments. Do you use the sodium sulfate method regularly? No oil will be removed?

Oil will not be removed using sodium sulfate, we use it all the time.  You can actually just add it to the extract shake and let it sit for a couple minutes.  You may want to check out this app note by Turner and think about using the suggested compound they use.  There is also lots of good info by an author named Robyn Comny.

http://www.turnerdesigns.com/t2/doc/appnotes/S-0113.pdf 

Oil will not be removed using sodium sulfate, we use it all the time.  You can actually just add it to the extract shake and let it sit for a couple minutes.  You may want to check out this app note by Turner and think about using the suggested compound they use.  There is also lots of good info by an author named Robyn Conmy.

http://www.turnerdesigns.com/t2/doc/appnotes/S-0113.pdf 

Hi, Johanna,

the issue of oxygen as a quencher is known. Whether it is essential in your specific case, I do not know. 

You may want to look at the paper published by Chin,Y.P., Buterbaugh,J.S., Gustafson,T.L., Traina,S.J., Danielsen,K.M., 1996. Solubility Enhancement and Fluorescence Quenching of Pyrene by Humic Substance: The Effects of Dissolved Oxygen on Quenching Processes, published in Environmental Science and Technology.

It seems to me that one idea there was that"solubility" of oxygen in organic pseudophase is increased, as compared with aqueous oxygen solubility, and it should (may?) different impact on oxygen quenching effect. You may wish also to look at the comments and rebuttle:

Comment:  Green,S.A., Blough,N.V., 1996. Solubility enhancement and fluorescence quenching of pyrene by humic substances: The effect of dissolved oxygen on quenching processes - Comment, Environmental Science & Technology 30, pp. 1407-1408.

Response: Chin,Y.P., Buterbaugh,J.S., Gustafson,T.L., Traina,S.J., Danielsen,K.M., 1996. Response to Comment on "Solubility Enhancement and Fluorescence Quenching of Pyrene by Humic Substance: The Effects of Dissolved Oxygen on Quenching Processes", ENVIRON. SCI. TECHNOL. 30, pp. 1409-1410.

and also on another old (but, to my taste, very good) paper: 

 Vaughan,W.M., Weber,G., 1970. Oxygen Quenching of Pyrenebutyric Acid Fluorescence in Water. A Dynamic Probe of the Microenvironment, Biochemistry 9, pp. 464-473.

Yes, in your case of oily phase, oxygen solubility may be different as compared with aqueous solubility; log L (Ostwald coefficient; solution/vapor concentration) is -1.51 in water and -0.723 in hexadecane (the data from Abraham et al., J. Chem. Soc. Perkin Trans. I 1994). Thus, oxygen concentration in an oily phase is greater almost by order of magnitude, as compared with the aqueous phase.