Anhydrous caffeine is found as a white crystalline powder or white crystals, in water it is sparingly soluble but becomes freely soluble when the water is boiling (Pharmacopoeia.com, 2019). Caffeine is an alkaloid xanthine derivative (Tarnopolsky, 2010) that acts as a psychostimulant and mild diuretic to humans and animals and is found within the beans, leaves and fruit of more than 60 plants, (Institute of Medicine (U. S.) et al., 2014), it is mostly consumed by humans as coffee or tea but can also be found in other foods and drinks that contain derived products of the cola nut or cocoa pods.
Phenanthrene is found as white crystals that are practically insoluble in water (Pharmacopoeia.com, 2019) it is a polycyclic aromatic hydrocarbon with three fused benzene rings, the name phenanthrene stems from the terms phenyl and anthracene (Emsbo-Mattingly and Litman, 2016), for industrial purposes phenanthrene is derived from coal tar, it has a variety of uses such as the production of dyes, drugs, pesticides and explosives.
The primary source of this PAH is combustion of fossil fuels and exhausts produced through industry and can be detected in cigarette smoke which then can lead to skin photosensitivity and can also be present in contaminated foods, water and soil (German Environment Agency, 2019).
HPLC stands for High Performance Liquid Chromatography and it is used for the identification and separation of two of more analytes. HPLC is a form of column chromatography where the sample (analyte) is dissolved in the mobile phase (a solvent) and is moved through the column which contains the stationary phase at high pressure.
Interactions between the sample, the mobile phase and the stationary phase occur due to their nature (normally this is polarity) and will determine the retention times of the analytes. As the sample runs, the analytes that have a greater affinity for the stationary phase will elute the column more slowly and therefore have the longest retention time. It then stands to reason that the analytes that have a greater affinity for the mobile phase elute more quickly and have a short retention time. When the mobile phase exits the column it passes through a detector such as a fluorimeter, a UV-absorbance detector or diode array detector; it is important that the most appropriate detector is selected and the correct wavelengths examined as this gives the optimal results from the HPLC, the detector sends the information it collects (quantity of analytes and the retention time) to a computer with the software that can then be used to analyse the data (Petrova and Sauer, 2017).
To use a HPLC to analyse standard solutions of caffeine and phenanthrene and produce calibration curves for each analyte and then use these calibration curves will then be used to calculate the concentrations of Drug A and Drug B in an unknown sample and identify them as either caffeine or phenanthrene.
The apparatus that was used was a HPLC with a vacuum degasser, autosampler and diode array detector coupled with a C18 Luna Column that was 250mm long by 4.6mm id with a particle size of 5?m by phenomenex. The analytical conditions that were adhered to was that the mobile phase was 10/90 water/methanol which was also the same as the sample diluent. The injection volume used for the HPLC was 10?l with a flow rate of 1ml/min and a run time of 5 minutes. The wavelength that detection occurred at was 254nm and the needle wash used to clean the HPLC needle was IPA (isopropyl alcohol).
Two stock solutions were prepared of the analytes caffeine and phenanthrene; the concentrations of these solutions were 0.1mg/ml and 0.05mg/ml respectively. A system suitability test was carried out to ensure the column was conditioned, the mobile phase (10/90 water/methanol) was run through the HPLC for 20 minutes at a flow rate of 1ml/min and the pressure was kept around <30psi. The plot for this run was observed on the computer and no injections were made until the baseline had settled. A blank dilutent sample was prepared and injected to make sure there were no interfering peaks at the expected retention times which were 3 minutes for caffeine and 7 minutes for phenanthrene.
A 1/10 dilution in a 50ml volumetric flask was then made of the stock solutions and once again a vial was filled and placed into the HPLC where 5 replicate injections of 10?l were performed; the peak areas were recorded and the %RSD for each analyte was calculated as well as the number of theoretical plates, the resolution and the tailing factor. The system suitability test showed that all the criteria was accepted, and it was confirmed that the HPLC was conditioned correctly.
From the stock solutions 5 serial dilutions were made up to 10ml in volumetric flask and injected to the HPLC and calibration curves were created for caffeine and phenanthrene; the lowest concentration is injected first to avoid any contamination from the more highly concentrated dilutions.
A table containing the volume of the analyte stock solution added before the dilutent was used to make each value up to 10ml.
Stock Concentration mg/ml Dilution 1
ml Dilution 2 ml Dilution 3 ml Dilution 4 ml Dilution 5 ml
Phenanthrene 0.05 5.0 1.0 0.5 0.1 0.02
A sample of an unknown concentration was then injected into the HPLC and the calibration curves were used to determine the concentration of the caffeine and phenanthrene within the unknown mixture from the peak areas (PHA302 Study Guide 2019-2020).
The system suitability test found that there were no interfering peaks of caffeine and phenanthrene and the following tables contain the results from the replicate injections that were used to carry out the calculations to determine the %RSD, the number of theoretical plates and tailing factors for each analyte as well as the resolution between the peaks.
The results for the 5 replicate injections were as follows.
0.01 136945 0.005 1005857
0.01 136955 0.005 1005855
0.01 136940 0.005 1005901
0.01 136900 0.005 1005866
0.01 136950 0.005 1005850
There are also several other parameters that are required for the system suitability calculations and they are expressed in the table below.
Width at half peak height (W0.5) 2.5 3
Width at 0.05 peak height (W0.05) 0.25 0.3
Distance from centre line of peak to front slope (f) 2.25 3.25
Average retention time in minutes (tR) 3.1375
%RSD stands for relative standard deviation and it is a statistical measurement that shows the spread of data taking into account the mean and expressing it as a percentage, it is normally used to inspect the variation within the data in this case it was peak area (Helpcomputerguides.com, n.d.). The equation used to calculate this is %RSD = (Standard Deviation/Mean) X 100. These %RSD values fall within the acceptance criteria which states that %RSD ? 2.0% (PHA302 Study Guide 2019-2020).
%RSD Caffeine = (21.96588264/136938) X 100 = 0.01604%
%RSD Phenanthrene = (20.51097267/1005865.8) X 100 = 0.00204%
A theoretical plate is a hypothetical zone/style where two phases are in equilibrium; this may also be known as an equilibrium stage, ideal stage or theoretical tray. It can be calculated using N = 5.45(tR/W0.5)2 where tR is mean retention time in seconds and W0.5 is the width of the peak at half the peak height. These N values fall within the acceptance criteria which states that N > 1600 (PHA302 Study Guide 2019-2020).
N Caffeine = 5.45(188.25/2.5)2 = 30901.9905
N Phenanthrene = 5.45(469.938/3)2 = 133731.9328
The resolution is a quantitative measure of how well two peaks are differentiated in the chromatogram, it is the difference in retention time in seconds of the peaks divided by the total width of both peaks, the equation is RS = 2?tR/WA+WB. This RS value falls within the acceptance criteria which states that RS ? 1.5 (PHA302 Study Guide 2019-2020).
RS = 2(469.938 188.25)/(5.5+6.5) = 46.948
The tailing factor is a measure of peak tailing and is defined as the distance from the front slope of the peak to the back slope of the peak divided by the distance from the centre line of the peak to the front slope multiplied by two. All these measurements are made at 0.05 of the peaks maximum height and is calculated with the equation T = W0.05/2f where f is the distance from the centre line of peak to its front slope. These T values fall within the acceptance criteria which states that T ? 2.5 (PHA302 Study Guide 2019-2020).
T Phenanthrene = 0.3/(2 X 3.25) = 0.4615
All of the system suitability checks determined that the HPLC was conditioned and therefore was suitable to be used for the practical. The calibration curves of concentration against peak area were then plotted using the following results.
A table showing the concentration, peak area and retention time for the serial dilutions of caffeine.
Concentration, mg/ml Peak area RT, min
0.001 13778 3.12
0.005 68893 3.117
0.01 136943 3.132
0.025 337547 3.141
0.05 705835 3.148
0.1 1385014 3.167
A graph showing the relationship between the concentration of caffeine and peak area.
A table showing the concentration, peak area and retention time for the serial dilutions of phenanthrene.
Concentration, mg/ml Peak area RT, min
0.0001 49616 7.832
0.0005 285515 7.837
0.0025 664656 7.832
0.005 1005855 7.831
0.025 4515300 7.826
0.05 9021277 7.836
A graph showing the relationship between the concentration of phenanthrene and peak area.
The peak areas of the drugs within the unknown mixture are shown in this table.
Injection number Drug A Drug B RT, min Peak area RT, min Peak area
1 3.15 138647 7.833 2796116
2 3.158 140977 7.825 2805687
3 3.15 141451 7.829 2799183
4 3.15 139303 7.836 2806052
5 3.145 143806 7.815 2813547
This table shows the mean peak area for each drug, the standard deviation and the %RSD for each unknown analyte.
Standard deviation 2022.4251 6773.0049
%RSD 1.4360062 0.2415379
For Drug A the mean peak area is similar to what we have typically seen for caffeine, therefore we can use the equation from the calibration curve for caffeine to determine the concentration of caffeine within the unknown sample. This can be solved as follows;
140836.8 = 1389595.7x 1103.6283
1389595.7x = 140836.8+1103.6283
For Drug B the mean peak area is similar to what we have typically seen for phenanthrene therefore we can use the equation from the calibration curve for phenanthrene to determine the concentration of caffeine within the unknown sample. This can be solved as follows;
2804117 = 177268490x + 135201.253
177268490x = 2804117 135201.253
The LOD (Limit of Detection) is the lowest concentration of analyte that will be distinguishable from the limit of blank at which it will actually be possible for detection to occur (Armbruster and Pry, 2008), the LOD is usually accepted as 3 times the noise of the signal as this is the lowest measurable response (PHA302 Study Guide 2019-2020). The LOQ (Limit of Quantification) is the lowest concentration that an analyte can be quantitatively detected with stated accuracy and precision however it depends on predetermined acceptance criteria and performance requirements (Vashist and Luong, 2018); this is usually accepted as 10 times the noise (PHA302 Study Guide 2019-2020).
Normal-phase chromatography is the original version of HPLC but is not often used now unless the reverse-phase results are inaccurate. With normal-phase chromatography the stationary phase is more polar than the mobile phase and the interactions between the analytes and column have mostly polar characters in the form of hydrogen boning or dipole-dipole interactions to name a few (NAGY and V?KEY, 2008). In reverse-phase then the opposite is true, where the stationary phase is less polar than the mobile phase and the interactions between the analytes and the column have mostly non-polar characters.
The column that was used in this experiment had a particle size of 5?m; and over time these silica particles have been improved to give the best size, shape and purity. The smaller a particle is more efficient it makes the HPLC column in regard to speed, sensitivity and resolution of the peaks however, this comes at the expense of the high pressures that the column can withstand (Henry, 2014).
The term gradient elution within HPLC means that the composition of the polar to non-polar compounds of the mobile phase are changed during the time that the chromatogram is running, gradient elution is faster, but it is more complex (Robards, Haddad and Jackson, 2004). Isocratic elution then refers to when the concentration of the mobile phase remains constant or nearly constant, but the composition will not change (Garcia-Lavandeira, Martinez-Pontevedra and Cela, 2011).
In reverse phase HPLC the mobile phase is normally a water/aqueous solution with an organic modifier; within reversed phase the weakest solvent is water as it is the most polar and repels any hydrophobic analytes into the stationary phase the most compared to other solvents and therefore increases retention times. When an organic modifier is added it is less polar and the hydrophobic analyte that was previously repelled strongly by the mobile phase will start to move into the mobile phase and will elute earlier than before as the repulsion of the analyte by the mobile phase has been reduced; as the solvent becomes more organic that retention time of the analytes will continually be reduced (Reversed Phase Chromatography – CHROMacademy, n.d.).
Armbruster, D. and Pry, T. (2008). Limit of Blank, Limit of Detection and Limit of Quantitation. The Clinical Biochemist Reviews, 29(Suppl 1):, pp.S49S52.
Emsbo-Mattingly, S. and Litman, E. (2016). Polycyclic aromatic hydrocarbon homolog and isomer fingerprinting. Standard Handbook Oil Spill Environmental Forensics, pp.255-312.
Institute of Medicine (U. S.), Pray, L., Yaktine, A. and Pankevich, D. (2014). Caffeine in Food and Dietary Supplements: Examining Safety: Workshop Summary. National Academies Press.
Garcia-Lavandeira, J., Martinez-Pontevedra, J. and Cela, R. (2011). A Binary-Like Approach for the Computer Assisted Method Development of Isocratic and Programmed Ternary Solvent Elutions in Reversed-Phase Liquid Chromatography. Journal of Chromatographic Science, 50(1), pp.33-42.
German Environment Agency, G. (2019). Phenanthrene German Environmental Specimen Bank. [online] Umweltprobenbank.de. Available at: [Accessed 5 Nov. 2019].
Helpcomputerguides.com. (n.d.). Excel Percent Relative Standard Deviation %RSD. [online] Available at: [Accessed 7 Nov. 2019].
Henry, R. (2014). Impact of Particle Size Distribution on HPLC Column Performance. Special Issues Volume 32, [online] (Issue 4), pp.12-19. Available at: [Accessed 5 Nov. 2019].
NAGY, K. and V?KEY, K. (2008). Separation methods. Medical Applications of Mass Spectrometry, pp.61-92.
Pharmacopoeia.com. (2019). Login – British Pharmacopoeia. [online] Available at: [Accessed 29 Oct. 2019].
Pharmacopoeia.com. (2019). Login – British Pharmacopoeia. [online] Available at: [Accessed 29 Oct. 2019].
Petrova, O. and Sauer, K. (2017). High-Performance Liquid Chromatography (HPLC)-Based Detection and Quantitation of Cellular c-di-GMP. c-di-GMP Signaling, pp.33-43.
Reversed Phase Chromatography – CHROMacademy. (n.d.). [ebook] p.10. Available at: [Accessed 6 Nov. 2019].
Robards, K., Haddad, P. and Jackson, P. (2004). High-performance Liquid ChromatographyInstrumentation and Techniques. Principles and Practice of Modern Chromatographic Methods, pp.227-303.
Tarnopolsky, M. (2010). Caffeine and Creatine Use in Sport. Annals of Nutrition and Metabolism, 57(s2), pp.1-8.
Ulster University Study Guide, Module PHA302, 2019-2020.
Vashist, S. and Luong, J. (2018). Bioanalytical Requirements and Regulatory Guidelines for Immunoassays. Handbook of Immunoassay Technologies, pp.81-95.