Environmental Science: Processes & Impacts blog

Bisphenol A (BPA) is one of the most used industrial chemicals worldwide. Since its introduction in the market in 1959, BPA production has increased steadily and it is forecasted to reach 7.3 million tons by the end of 2023.¹ BPA is used for a range of applications: from dental sealants to internal can coatings, electronic equipment and supermarket receipts.² “In the past years, concerns have been raised over the use of this compound due to its estrogenic effects that can be observed also at low BPA concentration, such as the ones found in the natural environment.^2,3 Thus, investigation of natural attenuation processes might help us developing strategies to reduce human exposure to this widespread chemical.

Several literature studies showed that manganese oxides (MnO_x)-mediated oxidation represents the main PBA degradation pathway in anoxic conditions.² This process produces a series of degradation products, including radicals that might couple to dissolved organic matter to form the so-called “bound residues”, unknown high molecular weight products whose long-term environmental risks are still debated.⁴ A detailed knowledge of the reaction mechanism will therefore allow to predict, and ideally prevent, the formation of degradation products that might be more hazardous than the parent compound.

In this context, Balgooyen et al. used stirred flow reactors to investigate the effect of influent concentrations on BPA degradation mechanism via birnessite (δ-MnO₂) oxidation. This research question was motivated by the hypothesis that higher influent concentrations might lead to a higher formation of bound residues. The results of this work are directly relevant for engineered water treatment systems that use MnO_x-coated sand,⁵ where contaminant inflow concentrations might change during time.

As a unique feature of this work, the authors used a combined approach based on the detection of both organic and inorganic reaction products. Specifically, they followed the formation of both hydroxycumil alcohol (HCA) and aqueous Mn(II). HCA is the main PBA oxidation product and is considered a proxy for bound residues formation, while Mn(II) is a reaction byproduct released in solution upon reduction of birnessite.

Unexpectedly, the two approaches gave opposite results: HCA yields were constant for the influent concentration range investigated, while Mn(II) yields decreased as the influent concentration increased. In order to explain their results, the authors hypothesized that Mn(II) was not an accurate proxy, as comproportionation and disproportionation reactions occurring at the mineral surface might alter aqueous Mn(II) concentrations. Using an elegant series of sorption and desorption experiments, Balgooyen et al. were able to confirm this hypothesis, leading to the conclusion that BPA oxidation mechanism in stirred-flow reactors is indeed independent from the influent concentration.

In addition to providing a valuable new piece of information for the complex puzzle of BPA cycling in anoxic conditions, the work of Balgooyen et al. teaches us something that has little to do with micropollutants or flow-through reactors: for a throughout study of a chemical mechanism, all reaction partners must be considered – no matter how many different analytical techniques you will have to use.

To download the full article for free*, click the link below:

Impact of bisphenol A influent concentration and reaction time on MnO2 transformation in a stirred flow reactor

Sarah Balgooyen, Gabrielle Campagnola, Christina K. Remucal and Matthew Ginder-Vogel

Environ. Sci.: Processes Impacts, 2019, 21, 19

DOI: 10.1039/c8em00451j

About the Webwriter:

Rachele Ossola is a PhD student in the Environmental Chemistry group at ETH Zurich. Her research focuses on photochemistry of dissolved organic matter in the natural environment.

Additional references

(1)        The Global Bisphenol A Market, https://www.researchandmarkets.com/reports/4665281/the-global-bisphenol-a-market (accessed May 26, 2019).

(2)        Im and Löffler, Environ. Sci. Technol. 2016, 50 (16), 8403–8416.

(3)        vom Saal and Hughes, Environ. Health Perspect. 2005, 113 (8), 926–933.

(4)        Barraclough et al. Environ. Pollut. 2005, 133 (1), 85–90.

(5)        Charbonnet et al., Environ. Sci. Technol. 2018, 52 (18), 10728–10736.

*Article free to access until the 30th June 2019

Polychlorinated biphenyls (PBCs) are a wide class of compounds with numerous everyday applications such as electrical insulators, cooling fluids, plasticizers and flame retardants – just to name a few. However,  back in the 1970s evidence started accumulating on their environmental persistency and on their toxicity as human carcinogens. In 1978, PBCs production was terminated and an international ban followed their inclusion in the Stockholm convention of Persistent Organic Pollutants. Interestingly though, PBCs are still present in our houses and schools today. A recent study conducted in rural and urban schools in the US measured indoor PBCs concentrations one to two orders of magnitude higher than outdoor values.¹ Another study measured PBCs in residential homes and found kitchen cabinets to act as an indoor source of these semivolatile compounds.² Considering their adverse health effects and their widespread occurrence, non-invasive, easy-to-use and cheap detectors are needed to monitor indoor PBCs levels.

In this respect, passive samplers represent a valid alternative to conventional sampling techniques. They consist of a disc of polymeric material placed into a protective shell. After the sampler is deployed in the environment, semivolatile compounds diffuse into the chamber and get absorbed onto the polymer. After a certain exposure time, the passive sampler is withdrawn from the field, and the absorbed compounds are extracted and quantified. The “on-the-sampler” concentration (C_sampler) is then used to obtain environmental exposure values.

However, using passive samplers in an accurate and reliable manner is challenging. One of the most critical but elusive parameters is the sampling rate (R_s), which represents the volume of air sampled per unit of time and is required to correctly convert C_sampler into exposure data. In outdoor applications the sampling rate is commonly measured using a “depuration compound”, an isotopically-labelled version of the species of interest that is adsorbed onto the polymeric disc before deploying the sampler in the field. The sampling rate is simply estimated from the loss of the depuration compound. This technique is effective, but the toxicity of these reference molecules makes it unsuitable for indoor applications. Another possibility involves the calibration of the passive sampler before its use, but this approach is time consuming and requires the use of an additional independent sampling method (for instance, an active air sampler).

Alternatively, R_s can be estimated with mathematical models. These models have already been developed for outdoor applications and allow the estimation of R_s from the wind speed data. Starting from this point, Herkert and Hornbuckle in their most recent publication hypothesized that these same models, if adjusted appropriately, can provide R_s from the indoor airflow data. To test their ideas, they set out a two-phase study with the final aim of providing practical recommendations for an accurate use of passive samplers in indoor environments.

In the first phase, they measured the sampling rate of thirty-eight PBCs congeners in a school room using a combination of passive and active samplers, and compared the results with the modelled values. The predicted R_s values were obtained from the room-averaged wind speed, a parameter that can be easily measured with an anemometer. Their results showed that the difference between the empirical and the simulated values was on overall less than 25%, demonstrating that mathematical models represent a reasonably good method to access sampling rates.

In a second phase, they investigated how the position of the passive sampler within the room influenced the value of the sampling rate. They observed that location did matter, as different zones of the room experienced different air flow. Specifically, fluid dynamics simulation of a typical room showed that samples placed close to the walls (< 30 cm), the ceiling (< 30 cm), the air diffuser (< 50 cm) or placed on surfaces experience unrepresentative wind speeds, while open or closed doors seem to have a minimal effect. They thus concluded that R_s can be modelled accurately if the passive samplers are placed appropriately, opening up this technology for use in indoor settings.

To download the full article for free*, click the link below:

Effects of room airflow on accurate determination of PUF-PAS sampling rates in the indoor environment

Nicholas J. Herkert and Keri C. Hornbuckle

Environ. Sci.: Processes Impacts, 2018, 20, 757

DOI: 10.1039/c8em00082d

About the Webwriter:

Rachele Ossola is a PhD student in the Environmental Chemistry group at ETH Zurich. Her research focuses on photochemistry of dissolved organic matter in the natural environment.

References in article:

(1)        Marek et al., Environ. Sci. Technol. 2017, 51 (14), 7853–7860.

(2)        Herkert et al., Environ. Sci. Technol. 2018, 52 (9), 5154–5160.

*Article free to access until the 1st of January 2019