Analyzing Artificial Sweeteners as Environmental Contaminants

The Column

Column, The Column-10-15-2018, Volume 14, Issue 10
Pages: 16–23

The Column spoke to Núria Fontanals, a senior researcher at the Department of Analytical and Organic Chemistry, Universitat Rovira i Virgili, Spain, about her work investigating the impact of artificial sweeteners as environmental pollutants using hydrophilic interaction chromatography (HILIC) and the broader role of HILIC in environmental analysis.

The Column spoke to Núria Fontanals, a senior researcher at the Department of Analytical and Organic Chemistry, Universitat Rovira i Virgili, Spain, about her work investigating the impact of artificial sweeteners as environmental pollutants using hydrophilic interaction chromatography (HILIC) and the broader role of HILIC in environmental analysis.

Q. When did the issue of artificial sweeteners as environmental contaminants first come to light?

A: To understand the presence of artificial sweeteners in the environment, we should first look at their uses and consumption historically.

Artificial sweeteners are high‑production-volume chemicals because they are used globally as tabletop sweeteners and food additives to sweeten diet beverages, pharmaceuticals, and some personal care products, including toothpastes. Artificial sweeteners approved by the European Union (EU) for these purposes include acesulfame, aspartame, cyclamate, saccharin, sucralose, and neohesperidin dyhydrochalcone (NHDC). Sweeteners with a natural origin, such
as stevioside and glycyrrhizic acid, have only been permitted in the EU since

The outlook for high-intensity sweetener consumption during 2016–21 varies by region and by sweetener. Considering this, the demand for sweeteners is flat or declining in North America and Western Europe although consumption continues to increase in Asia, the Middle East, and Africa. In contrast, the consumption of acesulfame, sucralose, and stevia extract is growing in all regions.

Although controversy surrounds the use of sweeteners, most of them have been approved in many countries. However, some of them are restricted or even banned in some countries, for example, saccharin, cyclamate, glycyrrhizic acid, and NHDC (1).

After years of using artificial sweeteners, recent studies have documented their widespread occurrence in the environment and it is for this reason that they have been included in the group of emerging organic contaminants (EOCs). The need for accurate and reliable analytical methods to determine sweeteners not only in food samples, but also in environmental ones has become necessary.

Q. How do wastewater treatment facilities affect the flow of artificial sweeteners into the environment?

A: As artificial sweeteners are rarely metabolized, they are mostly discharged through urine and faeces, passing unchanged through the human body and reaching the wastewater treatment plants (WWTPs) at relatively high concentrations. Thus, WWTPs are the main source of artificial sweeteners in receiving water bodies because most of them are not completely eliminated. Earlier studies (2,3) reported that artificial sweeteners are one of the most abundant detected EOCs in influents and effluents of WWTPs, with concentrations up to a thousand ng/L.

Among the different artificial sweeteners used as food additives, acesulfame, cyclamate, saccharin, and sucralose are recurrent in the different environmental compartments. Among them, a substantial variation in influent wastewater concentration (which is typically at µg/L levels) was noted for each treatment plant within the same region or among the different geographical regions. Other sweeteners, such as aspartame, NHDC, neotame, and alitame, are not frequently detected, and when they are, the concentration levels are much lower (from low μg/L to few ng/L depending on the WWTP and region).

As for the elimination in the treatment plants, on the one hand, cyclamate and saccharin are usually biodegraded more than 90% during biological wastewater treatment and they are, therefore, rarely detected, or only at low levels in effluents wastewaters. On the other hand, sucralose and acesulfame were consistently reported to be persistent during conventional WWTPs, and they pass through WWTPs mainly unchanged. In this sense, the level of sucralose reported in effluent of the different geographical regions is the same as those found in influent (2).

In fact, acesulfame and sucralose have been considered as ideal wastewater markers to trace contamination in aquatic environments because of their recalcitrance to transformation, persistence, high water solubility, low adsorption capacity on soils, high detection frequencies, and the high levels of concentration found in wastewaters (3).

As artificial sweeteners are not completely eliminated-in particular, acesulfame and sucralose-they are found in the receiving waters at concentration levels that agree with the removal efficiency in the treatment plant (4). The scenario with acesulfame is similar and it is worth noting that this sweetener was found at the highest concentration in the different environmental compartments analyzed. Nevertheless, several environmental studies on sweeteners are in agreement with respect to the distribution of these contaminants in the environment. This distribution exhibits different concentration values depending on the matrix, spatial locations, and the weather conditions (5).


Q. Artificial sweeteners do not bioaccumulate in humans. Does this mean these substances do not pose a risk to human populations?

A: As a result of their use as food additives, there is an ongoing discussion about their potential adverse health effects on humans. Based on the current literature, risk of artificial sweeteners to induce cancer is considered to be negligible (6). However, to prevent potential danger to humans, EU regulations set an upper limit on the concentration of artificial sweeteners in foods and beverages (7). Furthermore, it has established an acceptable daily intake between 5 mg/kg and 50 mg/kg of body weight per day. These levels (5–50 mg/kg) are far from the measured concentrations (ranging up to µg/L levels) of some artificial sweeteners in surface water, groundwater, and drinking water. Therefore, with the levels found in the environment, the potential adverse health effects in humans can be considered negligible.

Q. What impact do artificial sweeteners have on the ecosystem?

A: So far, there is little information reported on the toxicity of artificial sweeteners to aquatic organisms and human health at environmentally relevant concentrations (8). Most of these studies have focused on sucralose and they conclude that this sweetener does not alter the survival, growth, or reproduction of aquatic organisms at levels above those measured is surface waters. They also highlight that this compound may not cause toxicity to aquatic organisms at concentrations lower than 1000 mg/L (8), with this value being higher than the concentrations reported in the aquatic environment. Toxicity studies of sucralose on Lemna gibba, Daphnia magna, Pseudokirchneriella subcapitata, and Danio renio revealed no toxic effects. However, some other studies found that sucralose alters the behavioural response of Daphnia magna and Gammarus spp., which needs to be taken into account because they are modifications of the normal behaviour (9).

Consequently, continuous discharge and simultaneously chronic exposure to artificial sweeteners and their transformation products and other EOCs may increasingly pose a risk to aquatic ecosystems (2). Furthermore, it should be considered the risk associated to the contribution in the toxicity of each of the different contaminants in the same environmental compartment; but the information in this regard is very limited, which is a recurrent pending issue in risk assessment.

Q. The nature of environmental analysis means that target analytes are often contained within complex matrices. What methods did you use to clean up your samples? What advice could you offer to other researchers who are likely to encounter the same issues?

A: The clean-up strategy is linked to the extraction techniques used. One of the extraction techniques more used for liquid samples is solid-phase extraction (SPE), whereas for solid samples it is pressurized liquid extraction (PLE).

In detail, for SPE, one way to remove organic matter and other interferences includes a washing step using an organic solvent. However, in nonselective sorbents (either silica- or polymeric-based) using the organic solvent wash, a significant decrease in the recovery values of the analytes is observed as a result of the disruption of the nonspecific interactions between the target analytes and the sorbent. In contrast, an aqueous cleanup is just able to remove salts and other ions (which may greatly interfere during hydrophilic interaction chromatography [HILIC] separation) in the matrix, but not the high organic matter content usually found in complex samples. A washing step consisting of 10 mL of water was applied in the SPE when artificial sweeteners were determined by HILIC–high-resolution mass spectrometry (HRMS) (10). A compromise is to add small percentages, that is, up to 10% of organic modifier to the aqueous solution, however, this strategy is detrimental to the partial elution (losses) of the polar compounds and limited effectiveness of the cleanup. In fact, a 5 mL washing solution based on 10% methanol in water provided the best compromise between washing of interferences and elution of the analytes when the artificial sweeteners were determined by SPE/reversed-phase liquid chromatography (LC)–MS/MS (5).

When the target analytes are ionizable, using mixed-mode ion-exchange sorbents of the opposite charges of the target compounds is a suitable option because during the cleanup based on organic solvent the target analytes still remain ionically retained (11).

In the case of solid samples extracted using PLE, one of the most straightforward strategies includes in-cell and on-cell PLE cleanup. In the in-cell cleanup, a sorbent (typically, alumina, silica, and florisil, among others) is placed in the extraction cell with the sample instead of inert material to retain interfering substances. In the case of on-cell cleanup, a solvent with complementary properties to the one used in the extraction is passed through the sample before the extraction itself.

Another strategy is to load the extract from PLE to SPE using similar sorbents to the ones described to clean up liquid samples. This usually involves the complete or partial evaporation of the PLE extract, which is much more tedious and time‑consuming. In spite of this, this is currently the most commonly used clean‑up strategy. Furthermore, as a result of the complexity of the solid environmental samples, more than one clean-up step is usually required and different clean-up strategies are combined (12).

To summarize, the strategy to clean up the matrix should be considered in terms of cleaning (reducing the matrix effect), but also in maintaining the recovery of the target compounds.


Q. Do you have any advice to chromatographers starting out using HILIC in environmental analysis? Do you think there are any reservations or misconceptions about switching to methods using HILIC stationary phases?

A: Before starting to work with HILIC I had heard that the rule of thumb of HILIC was that the elution order for the same group of analytes is opposite to the one achieved by reversed-phase LC. However, this is not true. It is true that the elution strength is opposite to that of reversed‑phase LC because normally, the separation started with a high organic content (up to 98%) and this percentage is decreasing as the chromatographic run progresses. However, it should be noted that during HILIC separation there are several interaction mechanisms involved because it usually displays models of multimodal interactions. Thus, before starting with the chromatographic separation, one should carefully devise the structure of the compounds to be separated as well as the feasible interaction points between the stationary phase and the compound.

The characteristics of stationary phase, including the types such as bare silica, amino, diol, or zwitterionic, and the parameters of the mobile phase, such as the proportion of organic–aqueous solvent, pH, and the type and concentration of additive salts highly affect the retention and selectivity of the HILIC separation. These parameters should be carefully considered during the optimization of the HILIC performance. Moreover, their influence must always be assessed experimentally for any particular separation using either univariate or multivariate analysis.

Another issue that one should consider is when the HILIC column is connected to the mass spectrometer detector because of the variances during the ionization of the compounds. Therefore, I advise the HILIC user to re-optimize the analytical conditions if they have to analyze complex samples.

One limitation found in HILIC is the long times required for the equilibration of the stationary phase to obtain reproducible retention times. Thus, this should be kept in mind when optimizing any HILIC‑based method. Strategies to improve this limitation or the use of stationary phases addressing this issue should be explored in the future (13).

Q. In 2015 your research group published two papers on the determination of artificial sweeteners in aquatic environments. One used HILIC–HRMS while the other used reversed-phase LC–MS/MS. What are the advantages and limitations of each method?

A: The general advantages and limitations as commented above are all mainlly translated to this particular comparison of the reversed-phase LC- (5) and HILIC- (10) based methods to determine a group of sweeteners from complex environmental samples.

The chromatographic separation using HILIC was achieved within a short elution time (about 8 min) (10) compared with the elution time (11 min) in reversed-phase LC (5). Nevertheless, one should also account for the equilibration time needed in HILIC to obtain reproducible separations. Thus, in the HILIC method the column was equilibrated for 8 min with a total analysis time of 25 min, whereas the total analysis time for the reversed-phase LC method was 15 min with only a couple of minutes of equilibration time.

Apart from this, because HILIC delivers multimodal interactions, the optimization of the separation for a certain group of compounds might take longer than the optimization of the same group of compounds in reversed-phase LC. Moreover, if the optimization is not performed carefully it may end in failure. For example, in the study where the reversed-phase LC method was selected to separate the group of artificial sweeteners, the HILIC approach was preliminarily evaluated, which in this case was not able to successfully separate them (5).

One of the greatest advantages in the HILIC method developed for the artificial sweeteners over the reversed-phase LC method was the possibility of directly injecting the organic extract obtained from SPE into the HILIC–HRMS system (10). In both methods, the SPE ended with 5 mL of methanolic extract. This extract was directly injected in the LC instrument with the HILIC column because the separation started with 98% of acetonitrile. In contrast, the extract was evaporated to dryness and reconstituted with 2 mL (for surface waters) and 5 mL (for sewage waters) of aqueous-based solution in the reversed-phase LC method because the separation started with 5% of acetonitrile (5). The elimination of the evaporation step highly simplified the sample treatment procedure.

Regarding the matrix effect, the values in the form of ion suppression in all instances obtained in the reversed-phase LC–MS/MS method ranged from -6 to -72 (5), whereas the values in the HILIC–HRMS method were lower (ranging from -9 to -42) (10). Nevertheless, these matrix effect values cannot be strictly compared because the source of the samples was different and the instruments, as well as its ionization source configuration, were not the same.

As for the limits of detection achieved in both methods, they are of low ng/L. Again, this comparison is not pertinent because of the different detectors used in the two methods.


Q. Are there any other areas of environmental analysis where you think HILIC could be useful?

A: Last year we published a review (13) on the determination of EOCs in environmental samples using HILIC with a mass spectrometry detector. In this review, there is a summary of the most relevant and sorted applications in this field. In summary, as most of the EOCs have polar properties, HILIC has proven to be an excellent alternative to determine several EOCs whose polar properties caused problems in retention, separation, and detection when using conventional reversed-phase LC methods. Furthermore, the advances in HILIC might anticipate its continuing and further application in future studies in the environmental field.

As one of the advantages of HILIC is the possibility to directly inject organic extracts, another possible field is to analyze those polar compounds that are lost during evaporation, and that the injection of them in pure organic solvent in reversed-phase LC appears in the form of distorted peaks.

Q. What is your group focusing on at the moment?

A: The research group has a transversal research line based on the development of analytical methods to determine trace levels compounds (mostly contaminants) in different environmental samples to figure out the contamination. For solid samples, PLE and QuEChERS are typically used, whereas for liquid samples SPE is mainly used. In addition, novel materials with high capacity or selective properties are developed, evaluated, and exploited in different sorptive extraction techniques. The sample treatment step is followed by a separation technique (gas chromatography [GC], reversed-phase LC, HILIC, or capillary electrophoresis [CE]) connected to mass spectrometer detectors (QqQ and orbitrap as analyzers) to enhance the selectivity and the sensitivity of the method.

These methods allow the determination of EOCs, including drugs, sweeteners, high-production-volume chemicals, musks, and plastic additives, among others, at trace levels from samples such as surface and sewage water, sludge, biota, biological fluids, or air. Additionally, from the occurrence of these EOCs, the risk assessment in humans is being evaluated.


Project: CTQ2017-88548-P (MCIU/FEDER).


  1. K.-L. Ring and M. deGuzman, High-intensity sweeteners (2017) Available from:
  2. T. Ngoc Han, M. Reinhard, and K.Y.-H. Gin, Water Res.133, 182–207 (2018).
  3. F.Y. Lim, S.L. Ong, and J. Hu, Water9, 143–169 (2017).
  4. R. Loos, B.M. Gawlik, K. Boettcher, G. Locoro, S. Contini, and G. Bidoglio, J. Chromatogr. A1216, 1126–1131 (2009).
  5. P. Arbeláez, F. Borrull, E. Pocurull, and R.M. Marcé, J. Chromatogr. A1393, 106–114 (2015).
  6. M. Mukhopadhyay and A. Mukherjee, Food Chem. Toxicol.38, 75–77 (2000).
  7. European Parliament and Council Directive 2003/115/EC of 22 December 2003 Amending Directive 94/35/EC on Sweeteners for Use in Foodstuffs (2003).
  8. K.E. Tollefsen, L. Nizzetto, and D.B. Huggett, Sci. Total Environ.438, 510–516 (2012).
  9. M. Núñez, F. Borrull, E. Pocurull, and N. Fontanals, J. Chromatogr. A1479, 32–39 (2017).
  10. D. Salas, F. Borrull, N. Fontanals, and R. Marcé, Anal. Bioanal. Chem.407, 4277–4285 (2015).
  11. N. Fontanals, F. Borrull, and R.M. Marcé, in Advanced Separations by Specialized Sorbents, E. Dragana, Ed. (CRC Press, New York, USA, 2015), pp. 245–283.
  12. M. Núñez, F. Borrull, N. Fontanals, and E. Pocurull, Trends Anal. Chem. 97, 136–145 (2017).
  13. D. Salas, F. Borrull, N. Fontanals, and R. Marcé, Trends Anal. Chem.94, 141–149 (2017).

Núria Fontanals received her Ph.D. at Universitat Rovira i Virgili (URV) in 2005 in the chromatography and environmental applications research group under the supervision of Francesc Borrull and Rosa M. Marcé, following two postdoctoral periods. The first one in the Department of Analytical Chemistry in the University of Lund (Sweden) with J.A. Jonsson; and, then in the polymer research group in the University of Strathclyde (Glasgow, UK) with David C. Sherrington and Peter A.G. Cormack. In 2007, she reincorporated in the same research group at URV, where in collaboration with the Strathclyde research group, she leads a research line devoted to the development of novel materials to be applied in different analytical methods in order to enhance the sensitivity and selectivity in the determination of challenging compounds in complex samples. Moreover, she has extensive experience in the development of sample treatment procedures followed by liquid chromatography with tandem mass spectrometry and high resolution mass spectrometry methods for environmental and food analysis. From this research career has emerged one patent, six book chapters, 70 peer-reviewed papers, and more than 60 communications in conferences.