Array of SunshineUHPLC can be used with photodiode array detection to measure UV-absorbing compounds in sunscreen. by Dave Thomas, Applications Manager, Thermo Fisher Scientific, San Jose, CA, USA
UHPLC measures UV-absorbing compounds in sunscreen. | Recent concerns about UVA and UVB radiation from the sun contributing to skin cancers have highlighted the importance of applying an effective sunscreen before exposure to these harmful rays. As sunscreens are categorized as drug products by the U.S. Food and Drug Administration (FDA), they must undergo rigorous testing before being released to market. Proposed changes to the FDA regulations governing the labeling of sunscreen products are also expected to come into force by the end of 2009, which will require manufacturers to supply information to regulators on the UVA screening provided by its products. As a result of regulatory concerns, an efficient means of analysis is needed that can provide fast and dependable measurements of UV-absorbing compounds.
There are many available techniques for analyzing these compounds, including thin layer chromatography (TLC), gas chromatography (GC) and high performance liquid chromatography (HPLC). However, the oily matrix of sunscreen lotions coupled with the high UV absorbance of the analytes makes ultra high performance liquid chromatography (UHPLC) the method of choice for analyzing sunscreens. By additionally using photodiode array detection, the complete absorption spectrum of each compound is obtained as it elutes. The sensitivity of HPLC/PDA is also sufficient to measure these compounds in environmental water samples, which facilitates research on exposure and environmental fate.
Sunscreens
.jpg)
Table 1 Performance of UHPLC method for sunscreens performed on Hypersil GOLD 1.9μm, 1 x 100mm column
a Capacity factor (k’) and Resolution (R) calculated according to Reference 4 b Detection limit (MDL) calculated by multiplying the standard deviation of the concentration determined for a low level standard by Students’ t for n = 7 replicates c for 1μL injection from 25μL sample loop, n = 30 replicates d for 5μL full loop injection, n = 30 replicates Click to enlarge
| Skin cancer accounts for almost half of all cancers in the U.S.—more than one million new cases were diagnosed in 2007 and more than 10,000 people died. A majority of skin cancers are associated with exposure to harmful radiation from the sun. The most damaging radiation occurs in two wavelength regions, designated UVB and UVA. Radiation from 290 to 320 nm (UVB, as in “burn”) is absorbed mostly near the skin surface, causing the immediate redness and pain of sunburn. Radiation from 320 to 400 nm (UVA) penetrates deeply into the skin and causes DNA damage to cells in the under layer of skin, increasing the risk of malignant melanoma. UVA radiation also causes much of the skin damage associated with aging.
Sunscreens are topical lotions that disperse UV-absorbing compounds in an oily base that may also contain emollients, skin moisturizers and light reflecting particles of zinc oxide. Common UV-absorbers added to sunscreens include oxybenzone (benzophenone), avobenzone, octinoxate (octyl methoxycinnamate), octisalate (octylsalicylate), homosalate and octocrylene.
These same additives that protect skin from sun damage may also have a dark side as endocrine disruptors, causing abnormal development of fish, frogs and other aquatic organisms. Research performed in Switzerland in 2006 demonstrated that the common sunscreen additives 4-methylbenzylidene camphor (4-MBC) and octocrylene were present in surface water and bioaccumulate in fish, and a study of six common sunscreen chemicals found that octyl-methoxycinnamate (OMC) and homosalate (HMS) exerted significant estrogenic activity, as measured by the increase in proliferation rates of human breast cancer cells (MCF-7 cells) grown in vitro.
Experimental
A Thermo Scientific Accela high-speed chromatography system was used for this analysis. A proportioned mobile phase was implemented. Reservoir bottle A of the Accela pump was filled with solvent and fresh HPLC-grade water. The solvent line was then purged with at least 30 mL of the water. A fresh bottle of HPLC-grade acetonitrile was then connected to reservoir B and purged the same way. A stock standard of 200 mg/L was used. A total of 10 mg of each neat compound was accurately weighed into a 50-mL volumetric flask and 25 mL of acetonitrile were added. The solution was vortexed, sonicated for 5 min, brought to volume with acetonitrile and mixed. A calibrated pipette was used to dilute the intermediate standard with mobile phase in volumetric glassware to 100, 50, 20, 5, 1, 0.2, 0.05 and 0.03 mg/L.
Sample preparation
Samples of three commercially available sunscreens were purchased from local retail stores: Coppertone Water Babies SPF 50 Sunscreen Lotion, Coppertone Sunscreen Lotion SPF 45
Fig 1 Separation of UV absorbing sunscreens: A) on the Accela High Speed LC by reversed phase chromatography with photodiode array detection at 313 nm. Peak 1, oxybenzone; peak 2, avobenzone; peak 3, octinoxate; peak 4, octisalate; peak 5, homosalate; peak 6, octocrylene, 100 mg/L each. Conditions: Isocratic separation using 40:60 water: acetonitrile on Hypersil GOLD 1.9 μm, 2.1 x 100mm column at 45°C and 1000μL/min. B) Conventional separation used by a commercial laboratory, for comparison. Click to enlarge
| (Schering-Plough HealthCare Products) and Banana Boat Kids Sunblock Lotion SPF 30 (Sun Pharmaceuticals Corp.). Sample preparation for UHPLC analysis was initiated by transferring 0.2 g of a lotion to a 100-mL volumetric flask and adding about 5 mL of HPLC-grade water. The flask was then covered and vigorously shaken in order to mix contents. Approximately 50-mL acetonitrile and sonicate were added with occasional shaking for 10 min to dissolve or evenly disperse the lotion. The solution was brought to volume with acetonitrile; it was mixed and filtered through a 0.45-μm nylon syringe filter into a glass autosampler vial prior to analysis.
A sample of swimming pool water was collected from a local public pool and prepared using solid phase extraction (SPE). A HyperSep SPE cartridge (100-mg bed weight and 1-mL bed volume) was placed in a vacuum manifold and conditioned by washing with 2-mL methanol followed by 2-mL LC-MS-grade water. During conditioning, the vacuum was adjusted to produce a flow rate through the SPE cartridge of about 1 to 2 mL/min. At a rate of 1 to 2 mL/min, 20 mL of sample were passed through the filter. Analytes were eluted with 5 mL of acetonitrile into a clean 5-mL glass test tube. The extract was evaporated to dryness in a Caliper LifeSciences TurboVap LV concentration workstation at 35 C under a stream of nitrogen at 17 psi. The sample was then reconstituted in 200 μL of mobile phase, filtered into an autosampler vial and injected.
System preparation
Table 2 Concentration of UV-absorbing compounds determined in various matrices by High Speed Liquid Chromatography. Concentrations in sunscreens 1- 3 given in weight percent of original sample. “n.d.” indicates not detected. N = 3 replicates. Click to enlarge
| The UHPLC system was plumbed with precut and polished 0.005" ID high-pressure tubing and high-pressure fittings. For all tubing connections, it was ensured that the tubing end was square-cut and burr-free. While the high-pressure fitting was being tightened, the tubing was firmly pushed into the injection valve port in order to maximize peak efficiency. The pulse dampener was then primed, and the solvent lines were purged. It was important to verify that the pump was performing well by monitoring the pressure pulsation and by testing the pump proportioning accuracy.
Prior to analysis, the AS configuration entry of the instrument for “dead volume” was verified to be correct. This entry is the calibrated volume in μL written on the transfer tubing between the injection port and the injection valve. The entry for “loop size” was also verified. The operator had to fill the flush reservoir with 90:10 (v/v) acetonitrile:water and flush the syringe with solvent to purge any air bubbles from the syringe and tubing.
A Thermo Scientific HypersilGOLD, 1.9-μm 2.1- x 100-mm column was installed using a 10-cm-long precut and polished 0.005" ID high-pressure tubing, as well as a high-pressure fitting comprising a nut, back ferrule and front ferrule. The tubing was fully pushed into the column inlet when the high-pressure fitting was tightened.
For preparing the system’s detector, a 10-mm Thermo Scientific LightPipe flow cell was used. A short section of 0.005" PEEK backpressure tubing was added to the flow cell outlet to suppress bubble formation. Confirming the deuterium lamp had been formerly used for <2000 hrs was also necessary. The UHPLC system was equilibrated using direct control or a downloaded method. A method was created based on these operating conditions, and a sequence was generated to enable several injections of HPLC-grade water. The system was ready to run standards and samples when the peak-to-peak baseline oscillation was between 50 to 200 μAU/min (average of ten 1-min segments), and no significant peaks eluted in the retention time window of the analytes.
Results
Fig 2 Determination of UV absorbing compounds in a sunscreen lotion by reversed phase chromatography with photodiode array detection at 313 nm. Sample, Sunscreen 2 from Table 2. Peak 1, oxybenzone; peak 2, avobenzone; peak 3, octinoxate; peak 4, octisalate; peak 5, homosalate; peak 6, octocrylene, 100mg/L each. Conditions: Isocratic separation using 40:60 water: acetonitrile on Hypersil GOLD 1.9μm, 2.1 x 100 mm column at 45°C and 1000μL/min. Click to enlarge | The UHPLC separation of sunscreen agents provided better resolution, peak shape and runtime (Figure 1A) compared with the original method used by the commercial laboratory (Figure 1B). Particularly noteworthy was the six-fold improvement in runtime and sample throughput.
Table 1 summarizes the results obtained with regards to the performance of the method, including peak resolution, linear calibration range, dynamic range, limits of detection, and precision of retention time and peak area.
For most of the analytes, the peak area precision for partial loop injections was better than 1% RSD and for full loop injections was better than 0.15% RSD. For the best precision and accuracy, a full loop injection from a calibrated sample loop should be used. Minimum detection limits obtained with a 1-μL partial loop injection were on the order of 0.1 to 0.8 mg/L, which are more than sufficient for analyzing commercial sunscreen products. For analyzing environmental water samples, the SPE procedure provided a 100-fold concentration of the sample. If necessary, a larger injection volume could also be used to further lower the detection limits.
Figure 2 shows a chromatogram of Sunscreen 2, and Table 2 summarizes the concentration of UV absorbers determined in each sample. The results from the commercial products agreed well with the nominal values on the product label with none deviating by >10%. For the pool water sample, the detection limits after SPE were 50x lower than for direct injection. In the sample of pool water, only oxybenzone was detected at a concentration of 10 μg/L.
The UV absorption spectra collected by the PDA detector provided three important benefits. First, the spectrum of each compound could be stored in a library and later matched with unknown samples to identify and confirm analytes present in a sample. Second, the peak purity index automatically detected poorly resolved peaks, such as degradation products. Third, because sunscreens are formulated to protect against specific regions of the solar spectrum, the spectrum of each agent clearly identified it as a UVB, UVA or broad-spectrum absorber.
Fig 3 Spectral data of UV absorbing compounds in Sunscreen 2. Obtained on the Accela High Speed LC by reversed phase chromatography with photodiode array detection. Peak 1, oxybenzone; peak 2, avobenzone; peak 3, octinoxate; peak 4, octisalate; peak 5, homosalate; peak 6, octocrylene. Conditions: Isocratic separation using 40:60 water: acetonitrile on Hypersil GOLD 1.9μm, 2.1 x 100mm column at 45 °C and 1000μL/min.Click to enlarge | Figure 3 shows the spectral data displayed by the Thermo Scientific ChromQuest for Sunscreen 2. It was simple to determine from either the isoabsorbance plot or the 3-D plot which compounds protect against UVA (absorbing at <320 nm), which protect against UVB (absorbing at >320 nm), and which protect against a broad spectrum of wavelengths.
UHPLC coupled with PDA is a powerful method for resolving six UV-absorbing components of sunscreens in only 7 min with retention time and peak area precision better than 1% RSD. The UV absorbance spectrum of each compound is acquired as it elutes and can be matched against a stored spectral library to aid in compound identification. The resolution of adjacent peaks is improved even as method runtime is reduced from 45 to 7 min, while minimum detection limits obtained with a 1-μL partial loop injection were on the order of 0.1 to 0.8 mg/L.
Conclusion
Measuring UV-absorbing compounds in sunscreens is an important process, as these products are linked with the disruption of the endocrine system in both humans and aquatic organisms. The use of UHPLC surpasses the analytical capabilities of traditional analytical techniques such as TLC, GC and HPLC for difficult matrices such as sunscreens.
Coupled with PDA, UHPLC is capable of analyzing the UV-absorbing compounds found in sunscreens within minutes. The high retention time, increased peak area precision and improved resolution are easily achieved, while the sensitivity of UHPLC/PDA is also sufficient to measure UV-absorbing components of sunscreens in environmental water samples.
For more information about the Accela high-speed chromatography system, please call 866-463-6522 or visit www.thermo.com/accela.
References
1 http://www.fda.gov/Cosmetics/ProductandIngredientSafety/SelectedCosmeticIngredients/ucm127406.htm
2 http://www.webmd.com/skin-problems-and-treatments/news/20090521/fda-wrapping-up-sunscreen-label-changes
3 Armstrong, B.K., and A. Kricker. How much melanoma is caused by sun exposure? Melanoma Research, 1993: 3:395-401.
4 Hans-Rudolf Buser, Marianne E. Balmer, Peter Schmid, and Martin Kohler. Environ. Sci. Technol., 2006, 40, (5), pp 1427–1431.
5 Schlumpf M, Cotton B, Conscience M, Haller V, Steinmann B, Lichtensteiger, W (2001). In vitro and in vivo estrogenicity of UV screens. Environmental Health Perspectives 109(3): 239-244.
|