Chem 407
HPLC of Analgesics
(Aspirin, Caffeine, Acetaminophen)

A personal note. When you rupture a spinal disk as I appear to have done, you begin to appreciate the role of analgesics or pain medication. You also realize that the items on our list are not very strong and you are greatful for the expanded list of pain relievers. Unfortunately, we can't provide you with samples of the "really good stuff" for ordinary laboratory tests.

last edited: October 11, 2000
file called HPLC_remarks.htm

This instrument has not been giving us good results so it is time to do a little diagnostics.

Brief description of our system

• column: C18 or ODS (octadecylsilane) on 5 micron silica (Econosphere)
• column is 4.7 mm ID x 100 mm long.
• mobile phase: methanol/ water
  • you'll also find water/acetonitrile is often used with this type of column.
  • typically 50%,
  • flow rate 1 ml/minute
    • may be preferable to use methanol:1% acetic acid in water
  • detector: UV absorption (254 nm typically)
    • 254 nm is a strong Hg emission line
    • tunable detectors may perform better in the 260-275 nm region
  • Figure is from the Altec catalog.
    • it is for a C8 column which should behave similar to our C18.
    • actual retention times will probably range from 1-8 minutes.

  ---

  If you inject an aqueous sample (or just a sample of water) you see a characteristic signal. This is small, but larger than baseline noise. It looks like the derivative of a simple peak. That is, it first rises (about 0.02 Absorbance Unit) then drops below baseline by a comparable amount. The first rise occurs about 1 minute after sample injection and the full event lasts about 0.5 minutes.

  This is due to the difference in the refractive index of the water compared to that of the 50% alcohol. It makes it difficult to measure chromatographic peaks with a retention time below 1.5 minutes, especially if the peak amplitude is in the 0.05 AU neighborhood. The effect can be greatly reduced by preparing samples using a methanol/water mixture comparable to that used as the mobile phase.

  However, this solvent transient provides some useful information. It tells us that the dead time in the instrument is about one minute. That is, a sample remains in the column about 1 minute even if it is not retained at all by the stationary phase. It suggests that we will not be successful separating materials that have retention times of 1-2 minutes on this column. After all, the first minute represents no chromatography at all.

  This t₀ time can also be determined with a compound that has virtually no retention on the column. Uracil is a good candidate that also has a strong absorbance in the 250-275 nm region.

  ---

  Observations over the first two weeks:

  The aspirin samples I saw really had retention times in the neighborhood of 1.5 minutes. It was clear that the more dilute samples had absorption peaks that overlapped with the water transients. Varying the %methanol had relatively little effect on retention times. We would expect corrected retention times ( t_r- t₀) to vary by a factor of two for a 10-20% change in the methanol ratio.

  A more useful parameter is the capacity factor, k'

  k' = [( t_r)- (t₀)] / t₀.
  A value of zero means no chromatography is involved. Values of k'=0.25-5.0 are often convenient.

  In retrospect, it looks like there was little retention of sample by the column.

  What could be wrong?

  One possibility is that someone inadvertently changed the material in the fluid reservoirs on the instrument. We know that the methanol/water ratio should change the retention time (by perhaps a factor of five.) What would happen if both bottles were filled with methanol? This might shorten the retention time and attempts to change the polarity of the solvent would have no apparent effect.

  it's probably easier to replace the fluids than to test them.

  A more likely situation occurs when the column has seen a lot of use or abuse. Some materials of low polarity will be held by the C18 coating. Gradually, the column can become overloaded with such material and there's little available coating for normal chromatography. Usually one observes a slow deterioration of the column, with retention times slowly increasing. Most such components are slowly desorbed from the column by a prolonged cleaning cycle of pure methanol. However, some materials will remain permanently bound, ruining a column. (The symptom would be what we observe: that most materials show rapid peaks with nearly identical retention times.)

  A similar problem occurs if the column has suffered internal mechanical damage. If the column developed a channel, we'd find that the sample would largely flow through the column through that channel. Since the sample did not come into intimate contact with the stationary phase, there was little opportunity for normal chromatographic processes to occur. Again, we'd see relatively short retention times.

  Ideally, we would keep with the instrument a small folder of typical results of a test standard on a new column. We could then quickly run one of those test samples to see that the instrument and column were performing normally. We had such a folder for the HPLC but it is lost. It's time to prepare some new information to serve as such an instrument record.

  A test sample used by Altec has five components. It is run at 65% acetonitrile / 35% phosphate buffer (at pH 3.2) This pH will protonate the carboxylic acid and ensure consistent retention times for the amines. Approximate capacity factors are given.

  • uracil (t ₀)
  • pyridine (k'=1.25)
  • phenol (k'=0.5)
  • N,N diethyl-m-toluamine (k'=2.3)
  • 4-butylbenzoic acid (k'=1.8)
  • toluene (k'=2.2)
  The figure (from Altech catalog) is similar, but uses uracil, phenol, benzaldehyde, diethyl toluamide, toluene and ethyl acetate.

  ---

  About two weeks ago I ordered a replacement column for the HPLC. Unfortunately, I was unable to get it installed when it arrived. Dr. Den Besten should be installing it by 10/12/200 so we should be able to get some decent results this week.

  ---

  Actually there are several ways to protect HPLC Columns and this is a good point to discuss them.

  A Guard Column is a very short column that is introduced between the sample chamber and the analytical column.

  • The Guard Column contains the same stationary phase as the main column, but generally is a coarser particle (say 10-25 microns compared to a 5 micron column.)
  • The guard column is also only about 1-2 cm in length. It has virtually no effect on the chromatogram (except for a small increase in peak width.)
  • However, anything that would bind irreversibly to the analytical column will bind instead to the cheaper guard column.
  • (The Guard column can be replaced for $10-40, the analytical column for $250-300.)
  Another way is to use SPE (Solid Phase Extraction) to process any questionable samples.
  • The SPE is a short open column, usually containing the same stationary phases as the analytical column.
  • Many SPE's are basically 1 cm. of column packing on the bottom of a plastic syringe barrel.
  • The SPE has relatively large particle size so liquid flows freely or with a slight vacuum assist.
  • Typically an aqueous sample might be drawn through the column.
  • The organic components are bound to the solid phase.
  • The component of interest is then desorbed with 2-3 ml of methanol and this becomes the sample.
  • Any component that would bind irreversibly to the stationary phase remains on the disposable SPE cartridge.
  • Each SPE unit is treated as disposable and costs approximately $1-2.

  ---

  Revised Objectives, HPLC

  First, let us see that the column and instrument are working as expected. It might be useful if several earlier groups contribute to this final week's work, so you have some useful data and you see the proper bevavior of the instrument. I'd rather continue to have the class work another day on this project if it means we get some repectable results.

  • Prepare some samples of aspirin, acetaminophen and caffeine in the neighborhood of 250 ppm.
    • 250 mg/liter, use water as the solvent for the stock solution.
  • If available, prepare a 250 ppm solution of uracil.
  • If others are participating, let's add a sample of salicylic acid for comparison.
  • Prepare one working dilution of each (about 75 ppm)
  • Prepare one mixture that contains 50-100 ppm of each component.
    • These working solutions should be made in 50% methanol; keep these volumes small, preferably no more than 100 ml.
  • Run a chromatogram on each working standard at 50% methanol, 1 ml/min.
    • If any of the retention time are greater than 10 minutes, change the %methanol to reduce the retention time.
  • Instead of running each pure compound several times, chromatograph the mixture several times, varying
    • the % Methanol
    • the flow rate at one methanol/water ratio.
    • calculate k' for each of the runs
    • calculate N = 16 (peak width)²/ [(t_r-t₀)]²
  • I'd also like to see if we could try one run with methanol and 1% acetic acid in water as the two components. (Does pH change the chromatography?)
  • (I'd rather have a good report characterizing the instrument even if we don't get around to analyzing pills.)
  • If there is time, I'd still like to examine the question of a calibration curve and a detection limit for one or more of these species. (Plot of peak height vs concentration)