7..4  Analytical Experiments with an External Reference Standard Calibration

7.4.1  Caffeine Concentrations in Human Urine by Nathan Conroy

As the use of drugs has become more commonplace, so has the concern and apprehension toward the misuse and abuse of drugs.  For example, professional athletes and Olympians are subject to random testing for performance enhancing drugs.  Drug testing has even become routine as part of many job applications.  Often drug analyses have to be designed to test for metabolites of the drug of interest, rather than the drug itself.  Cocaine drug analyses involve not only the quantification of cocaine, by also benzoylecgonine (a metabolite of cocaine formed in the liver) and ecgonine methyl ester (both a metabolite and precursor of cocaine) [1].  Typically, urinary drug analysis procedures require the use of an internal standard that accounts for most losses during a liquid-liquid extraction to remove analyte from the protein and compound aqueous bio-layer [1-2].  The isolated solution of analyte and internal standard is then analyzed on an instrument such as a gas chromatography-mass spectrometer (GC-MS) or high performance liquid chromatography-mass spectrometer (HPLC-MS).   Laboratory analysis of many drugs requires a costly license for possession of the drug, thus most academic laboratories do not teach these extraction procedures.  But similar extraction using street legal compounds can be used as surrogates.  This caffeine lab procedure is designed to introduce students to the techniques and procedures used in drug analyses where caffeine is used as a surrogate for many drugs.  Students will determine the concentration of caffeine in their urine after having consumed caffeinated beverages, and see how this concentration changes as a function of time; as well as across different caffeine consumption habits.

Multi-step sample preparation techniques do not quantitatively transfer an analyte from starting material to the final extraction solution; small percentage losses can occur at each step in a procedure.  Therefore, when trying to determine an unknown concentration of an analyte, the analyst must account for the sum of these experimental losses using an internal standard (ISTD).  A chosen internal standard should behave similarly to the analyte under reaction conditions; therefore relative losses throughout sample preparation will be equal.  In other words if 20ppm ISTD (final concentration in the extraction solution) is added in a sample, but when the sample extract is analyzed and the instrument signal corresponds to only 15ppm of ISTD, we known only 75% of actual ISTD concentration was detected by the instrument.  The use of an ISTD correction procedure will account for these losses in the analyte.  For the case just stated, if the instrument detects a signal corresponding to 12ppm analyte, the instrument will back calculate the actual concentration of analyte in the solution to be 16ppm (a correction of +25 percent). 

Experimental Procedures

The best way to ensure an equal concentration of internal standard across your samples is to deliver an equal amount of internal standard to all solutions prior to diluting.  Also, make sure that the final concentration of internal standard in your sample extract is the same concentration as internal standard used in making your calibration standards.  A given volume of internal standard solution is most accurately delivered by a Hamilton-type micro-syringe where the full volume (or near full volume) of the syringe is used. Choosing an internal standard for a GC-MS analysis of caffeine is problematic because most compounds that share structural similarities with caffeine thermally degrade before volatilizing.  4-Acetylpyriding emerges on the gas chromatogram as two separate peaks, 4-Acetylpyrdine and its hydrated derivative, making a standardized integration of the peak problematic.  Decyl-alcohol works as an internal standard, but is not ideal because it shares no structural similarity to caffeine.  Cyclizine would likely make an appropriate internal standard for a caffeine GC-MS analysis but is more costly.

Chemicals and Supplies:
- High-Resolution GC grade methanol
-Caffeine
-Ammonium chloride
-Ammonium hydroxide
-GC grade dichloromethane
-Sodium chloride
-High purity (99.999%) helium gas

Instrument Settings:
-Mode:  splitless
-Inlet temperature:  250 C
-Pressure:  10 psi
-Purge Flow:  50 mL/min
-Purge Time:  0.50 min.
-Total Flow:  54.0 mL/min.
- Injection Volume:  1.00 mL

Column Specifications:
-HP-5MS  5% Phenyl Methyl Siloxane
-Length:  30. m
-Diameter:  250. mm
-Film Thickness:  0.25 mm
-Flow Rate:  1.2 mL/min.
-Linear Velocity:  40 cm/sec.

Oven Settings:
-Initial Temp.:  50 C
-Initial Time:  2.00 min.
-Ramp:      15.0 degrees per minute to 260 C, hold for 2.00 min.
-Transfer Tube Temp.:  280 C

MS Parameters:
-Solvent Delay:  4.00 min.
-Ionization Source:  EI
-Temperatures:  MS Source 230 C; MS Quad 150 C

TABLE 7.1. Approximate Peak Retention Times

Compound

Retention Time
(min)

Caffeine

13.86

Decyl-alcohol

9.08

4-Acetylpyrdine

7.28

 

Breaking an Emulsion:

The liquid-liquid extraction used in this procedure has a tendency to form emulsions, mixtures of two immiscible liquids.  They generally appear as either a cloudy combination of liquids, or as bubbly pockets at the interface between the two liquids.  Emulsions interfere with the recovery of the extraction solvent and the analyte, and therefore are problematic in analytical analyses.  Emulsions can be resolved or “broken” several different ways.  A saturation of the aqueous phase with sodium chloride is a first defense against emulsion formation, but has not been shown to be sufficient with this procedure.  Sonicating solutions is another common solution, but did not prove completely successful here.  Centrifuging is another common solution.  A far less costly strategy for breaking emulsions is glass wool.  Glass wool can be used to break emulsions by packing the glass wool into the bottom 1cm-2cm of a transfer pipet, then filtering the emulsified layer through the packed pipet.  While glass wool does successfully break the emulsion, the likelihood of experimental loss makes it non-ideal.  If all else fails time will break the emulsion; allow solutions to sit over night.

Evaporation/Concentration of the Extraction Solvent
         One of the advantages of using organic solvents is that the final extraction volume can be concentrated.  This results in the concentration of the analytes and improves the detection limit.  The final step in the procedure below calls for concentration of the extraction solvent.  Such a procedure is briefly given here.

         Procedures:

Creating a calibration curve:

  1. Make up solutions containing approximately 15ppm internal standard and a caffeine concentration of approximately 0.5, 1, 3, 5, 10, 15, 20, 25, and 50ppm.
  2.  Create a new GC method using the GC and MS conditions given above.
  3. Inject samples from low to high concentration and create calibration curve using mass spectra software.

Make up an Ammonium Buffer Solution:

      1. Add a few scoops of ammonium chloride to 15-20mL of High Resolution-GC grade methanol in a 25mL collection vial.
      2. Place a calibrated pH electrode in the collection vial.
      3. Add 28% ammonium hydroxide to the collection vial in single drop increments, until the solution reaches a pH of 9.5.  If the pH exceeds 9.5, add more ammonium chloride to lower the pH.

Extraction of Caffeine from Urine:
The next step involves a liquid-liquid extraction of caffeine from a bio-aqueous layer (urine) into a methanol and dichloromethane solvent mixture.  A high purity helium gas stream is used to evaporate the dichloromethane/methanol solvent, and then the remaining material is dissolved into HR-GC grade methanol.

  1. Collect a sample of urine 0.5 to 1 hour after the consumption of a drink containing caffeine.
  2. Quantitatively transfer 2mL of urine to a 10mL collection vial.
  3. Add ~100mg sodium chloride, 200µL ammonium buffer solution
  4. Add a volume of your chosen internal standard to make the concentration equal to that used in the calibration curve, keeping in mind that the entire solution will eventually be dissolved in 200 µL.
  5. Add 5mL of dichloromethane and methanol in a 9:1 volume ratio to the collection vial.
  6. Mix vials vigorously for 2 minutes.
  7. If necessary, use previously discussed emulsion techniques to break the emulsion.
  8. Place the recovered organic layer in a KD vial and the vial in a warm water bath. (Warm water is sufficient, no hot plate is needed)
  9. Use a high purity helium gas stream to evaporate the solvent.  Once half the solvent has evaporated, use a transfer pipet to wash the side of the KD vial with the remaining solvent.  Repeat this process again once half the remaining solvent has evaporated.  Continue to evaporate to dryness.
  10. Dissolve the remaining residue in 200µL High Resolution-GC grade methanol.
  11. Using 100µL vial inserts in a standard automatic sampler vial, inject methanol in the GC-MS.
  12. Analyze the results using MS computer software for identification and concentration using appropriate dilution factors.

    When blowing off solvent with a gas stream, there is a delicate balance between needle height and gas pressure.  The gas should maintain a small indentation on the surface of the solvent.  Submersing the needle will contaminate both the needle and your sample and too high a pressure will blow the solvent out of the KD vial.  It is best to set a needle height, then start the gas pressure at 0 and increase the pressure slowly.

Results:

         Caffeine concentrations vary depending on caffeine intake, but usually are in the parts per million range.  The procedure can be used in a variety of class experiments.  The simplest is to extract each students’ urine for a range of levels.  A more interesting experiment is to have one or more students drink a cup of coffee or other relatively high dose of caffeine and follow the clearance of caffeine from their body with time. 

Note that many other peaks are present in mass spectra of urine extracts.

References:

Mulé, S.J., and G.A. Casella. "Confirmation and Quantitation of Cocaine, Bezoylecgonine, Ecgonine Methyl Ester in Human Urine by GC/MS." Journal of Analytical Technology Vol.12 (1988): 153-55.

Thuyne, W.Van, and F.T. Delbeke. "Distribution of Caffeine Levels in Urine in Different Sports in Relation to Doping Control Before and After the Removal of Caffeine from the WADA Doping List." Int J Sports Med 2006 27: 745-50.


Frank's Homepage

©Dunnivant & Ginsbach, 2008