Synthetic musks in the natural environment: Sources, occurrence, concentration, and fate-A review of recent developments (2010 – 2023)

on SM levels in different natural environments, a systematic review of their contemporary presence is lacking. This review aims to address this gap by summarising recent research developments on SMs in diverse natural environments, including river water, lake water, seawater, estuarine water, groundwater, snow, meltwater, sediments, aquatic suspended matter, soils, sands, outdoor air, and atmospheric particulate matter. Covering the period from 2010 to 2023, the review focuses on four SM categories: nitro, polycyclic, macrocyclic, and alicyclic. It systematically examines their sources, occurrences, concentrations, spatial and temporal variations, and fate. The literature reveals widespread detection of SMs in the natural environment (freshwater and sediments in particular), with

• Review of four types of synthetic musks (SMs) in the natural environment.• Polycyclic musks were the dominant group being studied, followed by nitro musks.• SMs displayed spatial variations and complex temporal changes.• SM degradation and transformation products may be more persistent and eco-toxic.• Modelling tools help to study SM fate when direct measurements are not practical.Synthetic musks (SMs) have served as cost-effective substitutes for natural musk compounds in personal care and daily chemical products for decades.Their widespread use has led to their detection in various environmental matrices, raising concerns about potential risks.Despite numerous studies on SM levels in different natural environments, a systematic review of their contemporary presence is lacking.This review aims to address this gap by summarising recent research developments on SMs in diverse natural environments, including river water, lake water, seawater, estuarine water, groundwater, snow, meltwater, sediments, aquatic suspended matter, soils, sands, outdoor air, and atmospheric particulate matter.Covering the period from 2010 to 2023, the review focuses on four SM categories: nitro, polycyclic, macrocyclic, and alicyclic.It systematically examines their sources, occurrences, concentrations, spatial and temporal variations, and fate.The literature reveals widespread detection of SMs in the natural environment (freshwater and sediments in particular), with Abbreviations: ADBI, celestolide; AHMI, phantolide; AHTN, tonalide; APM, atmospheric particulate matter; ASM, aquatic suspended matter; ATII, traseolide; ATTN, versalide; DNS, data not shown; DPMI, cashmeran; EB, ethylene brassylate; EXA, exaltolide/pentadecanolide/thibetolide; HHCB, galaxolide; HHCB-lactone, galaxolidone; LOD, limit of detection; LOQ, limit of quantification; MA, musk ambrette; MC, muscone; MK, musk ketone; MM, musk moskene; MT, musk tibetene; MX, musk xylene; OTNE, Iso E super; ROM, romandolide; SMs, synthetic musks; WWTPs, wastewater treatment plants.
The water solubility of SMs (e.g., 1.75 and 1.25 mg/L for galaxolide (HHCB) and tonalide (AHTN) at 25 • C, respectively) is lower than many other emerging micropollutants (e.g., antibiotics) (Klaschka et al., 2013;Tumová et al., 2019).However, the continuous input of SMs into the environment has made them a virtually ubiquitous group of contaminants.It was estimated that the daily input of SMs in Portugal was about 6.7 mg/person (Homem et al., 2015a).In 2007, the discharge of HHCB and AHTN from domestic wastewater into the aquatic environment was about 1.26 and 0.38 t in Shanghai in China, respectively (Zhang et al., 2008).In general, wastewater treatment plants (WWTPs) are considered important sources of SMs in the environment (Wong et al., 2019).Since SMs are generally lipophilic and resistant to biodegradation treatment, they tend to accumulate easily in WWTP sludges (Lee et al., 2010;Yang and Metcalfe, 2006).Subsequently, SMs may enter the environment through treated effluents and/or biosolids for soil/land application (Chen et al., 2014;Lee et al., 2016;Vallecillos et al., 2015b).
Generally, SMs are addressed together with other emerging micropollutants in published reviews (Arpin-Pont et al., 2016;Colas-Ruiz et al., 2023;Jiménez-Díaz et al., 2014).However, there are fewer comprehensive review studies specifically focused on SMs than other emerging micropollutants (e.g., antibiotics).Recent reviews have concentrated on detection methods for SMs (Katuri et al., 2021;Marchal and Beltran, 2016;Vallecillos et al., 2015a), their potential toxicity (e.g., on human health, biofilms, aquatic organisms) (Arruda et al., 2022;Kathryn et al., 2014;Luo et al., 2023a;Pinkas et al., 2017;Tumová et al., 2019), their behaviours in WWTPs (Homem et al., 2015b), and polycyclic musks (Liu et al., 2021).Rainieri et al. (2017) specifically discussed SMs in the marine environment, while Wang et al. (2023) and Diao et al. (2024) reviewed SMs in broad matrices (e.g., WWTPs, natural environment, biota).To the authors' best knowledge, there is no review specifically encompassing sources, occurrence, concentration, and fate of SMs in the natural environment.Therefore, this review aims to cover the research developments regarding the occurrence and concentration of SMs in various natural matrices (i.e., river water, lake water, seawater, estuarine water, groundwater, snow, meltwater, sediments, aquatic suspended matter (ASM), soils, sands, outdoor air, atmospheric particulate matter (APM)) by carefully examining the newest published studies on SMs from 2010 to 2023, utilizing scientific databases such as Web of Science, Elsevier®, ACS Publications®, Springer®, Scopus®, and Google Scholar.Sources and fate of SMs in the natural environment are discussed, and risk assessments of SMs are also briefly mentioned to provide a better understanding of the impact of SMs.Finally, prospects are outlined for further studies in this field.

Types of synthetic musks
From the perspective of chemical structure, SMs can be classified into four categories: nitro, polycyclic, macrocyclic, and alicyclic (Pinkas et al., 2017), as shown in Fig. 1.
The first SMs to be used as alternatives to natural musks were the nitro musks, dinitro-and trinitrobenzenes with additional alkyl, keto, or methoxy groups (Marchal and Beltran, 2016).Typical nitro musks include musk ketone (MK), musk xylene (MX), musk moskene (MM), musk tibetene (MT), and musk ambrette (MA).The production volume of nitro musks in the market was about 35 %, but their use decreased at a rate of about 5 % per year (Nakata et al., 2007;European Commission, 2006).Due to concerns about potential health risks associated with carcinogenicity, persistence, and bioaccumulation, MA and MT have been banned in the EU and China cosmetic industries.Additionally, MK and MX are restricted in personal care products in the EU, with maximum concentrations of 1.4 % and 1.0 % in fine fragrance, 0.56 % and 0.4 % in eau de toilette, and 0.042 % and 0.03 % in other products, for MK and MX, respectively (European Parliament, 2009).Although the application of nitro musks is being limited, they remain an important group of SMs and continue to be a hot topic because of their occurrence, concentration, toxicological risks (e.g., negative effects on fish embryos and development), and possible health issues (e.g., carcinogenicity of MX and its indirect photochemical transformation products, and disruption of MK and MX on the (supra-) hypothalamic-ovarian axis) (Gao et al., 2019;Wang et al., 2021;Eisenhardt et al., 2001).
Currently, polycyclic musks are the predominant group in use.Polycyclic musks were first developed in the 1960s and soon became substitutes of nitro musks.They comprise acetylated and highly methylated pyran, tetralin, and indane compounds (Marchal and Beltran, 2016).Commonly used polycyclic musks include HHCB, AHTN, celestolide (ADBI), traseolide (ATII), cashmeran (DPMI), phantolide (AHMI), and Iso E super (OTNE).Sometimes, HHCB-lactone (galaxolidone), a transformation product of HHCB, is detected along with HHCB.Polycyclic musks lead the production of SMs with a market share of about 61 % (Nakata et al., 2007).Recent data showed that HHCB and AHTN, which account for approximately 95 % of the polycyclic musk market in Europe, were produced or imported at rates ranging from 1000 to 10,000 t per year in the EU (Aminot et al., 2021;Clara et al., 2011).In China, the presence of HHCB and AHTH was about 73 % and 65 %, respectively, in personal care products (Luo et al., 2023a).Polycyclic musks have been widely detected in environmental matrices and most published research on SMs involves polycyclic musks (Federle et al., 2014;Liu et al., 2021;Reiner and Kannan, 2011).Typically, polycyclic musks are considered to be resistant to biodegradation and may pose potential risks to various organisms, e.g., oxidative stress and fish malformations (Liu et al., 2021;Wang et al., 2021).
Macrocyclic musks are cyclic 15-or 17-membered ring system compounds with alkyl and acetylated groups, such as muscone (MC), exaltolide (EXA) and ethylene brassylate (EB).Despite becoming more accessible in recent years, macrocyclic musks are not as prevalent as polycyclic musks due to the high cost of their synthesis (Colas-Ruiz et al., 2023;Vallecillos et al., 2015a).Since the uses of macrocyclic musks are almost exclusively limited to the perfume industry, they contribute only about 3-4 % of the market but are expected to prevail (Marchal and Beltran, 2016).Compared to nitro and polycyclic musks, research on the ecotoxicity and neurotoxicity of macrocyclic musks has just started, and has not been systematically concluded (Pinkas et al., 2017;Vallecillos et al., 2015a).
The newest SM group is the alicyclic musks.They are cycloalkyl esters or linear musks formed by modified cykloalkyl esters (Marchal and Beltran, 2016).Typical compounds include romandolide (ROM), helvetolide, and cyclomusk.Although alicyclic musks are currently rarely used in personal care products, they are expected to have a broader future application in industries due to their biodegradable properties and cost-effective manufacture (Marchal and Beltran, 2016;Vallecillos et al., 2015a).A few studies focusing on alicyclic musks in the environment have been reported in recent years (Arbulu et al., 2011;Llamas-Dios et al., 2021;Relić et al., 2017).

Overview of published studies of synthetic musks in the natural environment from 2010 to 2023
Fig. 2 shows the number and distribution of studies reporting on SM research in the natural environment between 2010 and 2023 worldwide.The two main geographic areas in which research is conducted on SMs in the natural environment are China and Europe, followed by Singapore, the United States, and South Korea.In Europe, Spain has carried out more research than other European countries.In Africa, Tunisia (Necibi et al., 2016) is the only country researching on SMs in the natural environmental, but Yessoufou et al. (2016) have raised concerns about SM pollution in Benin.It is worth noting that approximately half of the selected published works reported SMs in natural water, followed by studies on SMs in sediments/ASM, outdoor air/APM, and soils/sands.

Sources of SMs in the natural environment
Generally, wastewater from municipal, agricultural, and industrial sources is regarded as the largest contributing source of SMs in the natural environment (Guo et al., 2013;Rainieri et al., 2017;Wang et al., 2018;Zeng et al., 2018c).Raw wastewater containing SMs enters WWTPs, but studies have shown that these compounds are not always effectively removed during the treatment process, leading to their presence in WWTP effluents and, subsequently, their release into the natural environment.In Germany, Klaschka et al. (2013)  sediments in aerated grit chamber tanks.
Since SMs have high octanol-water partition coefficients, they tend to absorb onto sludges or other biosolids during the treatment processes (Böhm and Düring, 2010;Wu et al., 2018).Hu et al. (2011b) measured various SMs in sewage sludges from Beijing WWTPs, finding concentrations of HHCB, AHTN, MK, ATII, and MX at 260-12590, 10-2560, 130-530, 15-300, and <3.3 ng/g dw, respectively.Ramos et al. (2019a) detected thirteen SMs (six polycyclic, two macrocyclic, and five nitro musks) in sludges in Portugal, observing that the highest concentrations of SMs were HHCB and AHTN at 71010 ± 4790 and 8863 ± 602 ng/g dw in summer.Therefore, effluents and sludges containing SMs may pose potential environmental risks when used for reclaimed water irrigation, sludge utilisation and landfills (Chase et al., 2012;Liu et al., 2021;Wang et al., 2013).There are more examples of SM levels in the wastewater and sludges in the summaries by Homem et al. (2015b) and Liu et al. (2021).
In addition to indirect sources like sludge discharges, it is also important to highlight the presence of direct input of SMs into the environment.This direct input occurs through human activities such as outdoor swimming, bathing, and other recreational activities, and can result in the introduction of SMs into water, sediment, soil, and sand environments (Arpin-Pont et al., 2016;Homem et al., 2017;Picot-Groz et al., 2018;Pintado-Herrera et al., 2017).Besides, emissions of SMs from products containing them (e.g., personal care products) and temperature-dependent volatilization from WWTPs during treatment of water can also contribute to SMs in the gaseous environment because SMs are semi-volatile organic compounds (Liu et al., 2021;McDonough et al., 2016;Wong et al., 2019).To date, SMs have been widely found in both outdoor environments (e.g., outdoor air, APM) and indoor environments (e.g., indoor air and dust) (Balci et al., 2020;Weinberg et al., 2011a;Xie et al., 2007).However, since the indoor environment (e.g., school, barbershop, hospital) is not the primary focus of this review study, the following discussions and comparisons do not include SMs in indoor air.

Synthetic musks in natural water
Table 1 provides an overview of the occurrences and concentrations of SMs in natural water from selected published studies, including river water, lake water, groundwater, seawater, estuarine water, snow, and meltwater.Notably, among the four categories of SMs, most investigations focused on polycyclic and nitro musks, while several works studied macrocyclic, and alicyclic musks, e.g., (Homem et al., 2016;Klaschka et al., 2013;Llamas-Dios et al., 2021).It is important to recognise that while some reported data come from studies primarily designed for the development of SM detection methods without conducting extensive discussions, e.g.(Cavalheiro et al., 2013;González-Hernández et al., 2021;Homem et al., 2016), the levels of SMs detected by applying these methods to real samples can still provide valuable insights.These data contribute to understanding the occurrences and concentrations of SMs in natural water environments, and help to assess potential risks and environmental impact.

River water
Most reported SM levels in the natural water environment are from river water.Some studies carried out simple concentration determination in their method validation.For example, up to 104, 2.8, 379, 61, and 78 ng/L of DPMI, ADBI, HHCB, AHTN, and MK were detected in northern Portuguese rivers, respectively (Celeiro et al., 2019).In terms of spatial analysis, published studies have emphasized the influences of WWTPs and high population density on the levels of SMs in rivers.Lee et al. (2010) studied four SMs in the rivers of Busan in South Korea.The overall concentrations of HHCB, AHTN, and MK were 0.10-13.92,0.03-2.80,and <LOQ (limits of quantification)-0.42 μg/L, respectively.
It was found that several sampling sites showed higher levels due to discharges from WWTPs and the presence of non-point sources.In the Molgora River in Italy, concentrations of HHCB, AHTN, and ADBI varied greatly, ranging from <0.05 to 1141 ng/L, <0.25 to 364.42 ng/L, and <0.25 to 23.4 ng/L, respectively.Samples collected downstream of WWTPs exhibited the highest SM concentrations (Villa et al., 2012).In the river samples from the North Canal River watershed in Beijing, China, HHCB and AHTN presented concentrations of 13.2-395 and 2.98-232 ng/L, respectively.High population density was considered an important influencing factor (Zhang et al., 2020).Additionally, Lange et al. (2015) found concentrations of 0.001-0.26μg/L of HHCB, 0.001-0.06μg/L of AHTN, and 0.003-1 μg/L of HHCB-lactone in the Ammer River in Germany.This is compared to concentrations below 0.005 μg/L of SMs in non-effluent-affected upstream water, with WWTP effluents identified as the main source of SMs.Similar patterns of higher SM concentrations in natural water near or downstream of WWTPs and densely populated areas have been observed in other rivers worldwide, for example, in the Han River and Geum River (South Korea) (Hong et al., 2021;Kim et al., 2022), the Ebro River (Spain) (García-Pimentel    (continued on next page) J. Li et al. et al., 2023), the Hudson River and Boulder/Fourmile Creek (USA) (Barber et al., 2013;Reiner and Kannan, 2011), the Iguassu River and Amazon River (Brazil) (Froehner et al., 2011;Rico et al., 2021), and rivers in Bavaria (Germany) (Klaschka et al., 2013), Jiaozhou Bay, Guangzhou, and Haikou (China) (Huang et al., 2020;Jiang et al., 2018;Su et al., 2023), Bangkok (Thailand) (Juksu et al., 2020), and Austria (Clara et al., 2011).
Comparing results from different studies conducted at the same sites can provide valuable insights into the variability of SM levels in the water environment.For instance, SM levels in the Ebro River in Spain were 0.49-1.72,1.40-26.2.0.34-0.37,n.d.(not detected)-0.55,and n. d.-0.80 ng/L for DPMI, HHCB, AHTN, MX, and MK, respectively (Ramírez et al., 2011).In a subsequent study conducted by the same authors, their concentrations lay within the ranges of n.d.-3.0, 3.1-16, 0.9-2.0,0.6-0.9, and 3.6-7.2ng/L, respectively (Ramírez et al., 2012), showing no significant difference except for MK.However, a more recent study by García-Pimentel et al. ( 2023) detected higher levels of HHCB (up to 143.5 ng/L) and AHTN (up to 52.8 ng/L) in Ebro River.Another example from Haihe River in China showed that, HHCB and AHTN were measured at 3.5-32.0and 2.3-26.7 ng/L by Hu et al. (2011a), considerably lower than that detected at 0.15-1.96and 0.06-3.39μg/L by Mu and Wen (2013), respectively.Such comparisons or differences can also be found in reported data from the Leça River (Portugal) (Homem et al., 2016;Ramos et al., 2019b) and the Nakdong River (Korea) (Lee et al., 2016(Lee et al., , 2014)).These comparisons illustrate that SM levels in natural river water might be influenced by the sampling time, as well as potential changes in pollution sources and the amounts of SM usage over time.Long-term monitoring of SMs in the river water environment may provide more comprehensive assessments, given their trace levels.

Lake water
For lake water, Guo et al. (2013) found that HHCB, AHTN, MK, and MX concentrations ranged from 3.1 to 212.04, 0.71 to 6.07, 0.13 to 0.80, and 0.12 to 0.58 ng/L in Taihu Lake in China, respectively.Although HHCB exhibited relatively high levels, the ecological risks posed by all the SMs investigated were low, even in the worst case.Spatial analysis further demonstrated that SMs may be transported from WWTPs and industrial wastewater from surrounding areas.In lake water of the nearshore and shoreline areas of lower Great Lakes, concentrations of up to 1.5, 3.9, 37, 1625, and 162 pg/L were measured for ADBI, AHMI, ATII, HHCB, and AHTN, respectively, while nitro musks were not detected (McDonough et al., 2016).Further analysis revealed significant correlations (p < 0.01) between population density within a 20 to 40 km radius of each site and SMs in the lake water.HHCB, AHTN, and OTNE were investigated in eight permanent mountain lakes of the French Pyrenees by Machate et al. (2022).Detection frequencies were 63 % to 81 %, with concentrations below 4.6 ng/L in all samples.In this research, the authors pointed out the potential impact of tourists and atmospheric deposition on other contaminants, but SMs were not discussed.Besides these studies, other research on lake water has also reported SM levels in the range of ng/L (Table 1).
Low occurrence of SMs have also been observed in estuarine water samples.In a study of nine SMs, only ATII was detected at 17 ± 1 ng/L in the estuarine water of the Gernika Estuary in Spain, and DMPI at 42 ± 11 ng/L in the Adour Estuary in France, respectively (Cavalheiro et al., 2013).Jo et al. (2021) found that, up to 6.6664, 0.2744, and 0.5193 μg/ L of HHCB, MK, and OTNE, respectively, were detected in estuarine water near the Han River of Korea, while the occurrences or concentrations of the other nine SMs were low.Relatively lower occurrences of SMs were also reported in UK estuarine water (Sumner et al., 2010) and in the estuarine water of Urdaibai, located in the Bay of Biscay, Spain.It is suggested that the low occurrences and concentrations of SMs in seawater and estuarine water might be due to dilution by tidal currents (Homem et al., 2016;Jo et al., 2021), and SMs may also be adsorbed by other marine substances, e.g., microplastics (Elseblani et al., 2023).a Data are displayed in their original units from the references.b n.d.: not detected.c <LOD: below limit of detection.d <LOQ: below limit of quantification.e DNS: data not shown.

Groundwater
Studies examining the presence of SMs in groundwater have been relatively limited, possibly due to the challenges associated with groundwater sampling.Ding et al. (2023) found HHCB at 111-144 ng/L in the groundwater near the Chaobai River in China.Chase et al. (2012) detected up to 56-72 ng/L of AHTN in the groundwater of Lubbock, Texas.Comparatively, concentrations of up to 687 ng/L of HHCB and 187 ng/L of AHTN were observed at a site in Central Canada (McGregor and Carey, 2019).Additionally, in Alava, Spain, HHCB and DPMI were detected in groundwater with concentrations ranging from n.d. to 338 ng/L and 354 to 573 ng/L, respectively (Arbulu et al., 2011).The same concentration range (308-543 ng/L for HHCB and n.d. to 365 ng/L for DPMI) was found in surface waters sampled at the same site, suggesting a potential relationship between SMs in groundwater and the natural recharge of the aquifer from the Alegría River (Arbulu et al., 2011).
Several researchers explored levels of other categories of SMs in groundwater.Llamas-Dios et al. ( 2021) investigated five nitro, five macrocyclic, and one alicyclic musks in the groundwater of the Guadalhorce River basin in Spain.They found up to 5 ng/L of MK and 7 ng/L of musk R1 (macrocyclic), while results for other SMs were not provided.Notably, high levels of HHCB (435-3130 ng/L) and AHTN (100-646 ng/L) were also detected in the groundwater.In a study conducted in Belgrade in Serbia, Relić et al. (2017) identified eleven alicyclic and two macrocyclic musks in groundwater.The concentration of total eleven alicyclic musks was 1.85 μg/L, while two macrocyclic musks (cyclopentadecanone and EB) were not detected.In comparison, the nearby Sava River had a total eleven alicyclic musks presenting at 3.158 μg/L.The authors suggested that the levels of these SMs depend on the water regime in the Sava River and recharges from it, as well as elimination mechanisms such as soil sorption.
The reviewed studies indicate that SMs can be present in groundwater, and the findings emphasise the importance of further research into the contamination of groundwater by SMs, as their presence in this vital water resource could have potential implications for water quality and human health.

Snow and meltwater
Several researchers have investigated SMs in environmental snow and meltwater, primarily in Italy and China.Hu et al. (2012) collected 42 snow samples from urban areas of Beijing and observed HHCB and AHTN in all samples with concentration ranges of 2.2-205.9ng/L and 6.5-754.1 ng/L, respectively.MX, MK, and ATII were only detected in some samples, ranging from 4.4 to 16.1 ng/L, 6.7 to 43.2 ng/L, and 8.4 ng/L, respectively.Additionally, AHTN concentrations were found to be positively correlated with dissolved organic compounds (p < 0.01).In a study by Villa et al. (2014), fresh snow and stream water were analysed for the presence of AHTN and HHCB in a glacial flow area of the Forni Glacier in Italy, with up to 5.07 ng/L and 9.83 ng/L of AHTN and HHCB observed in snow samples, respectively.These results support the authors' hypothesis of SM medium-distance atmospheric migration.The relatively low HHCB/AHTN ratio in snow samples was attributed to faster HHCB reaction kinetics with ⋅OH radicals during atmospheric transport.
However, a higher HHCB/AHTN ratio was found in glacier and nonglacier meltwater in the Adamello-Brenta Natural Park in Italy (Rizzi et al., 2022), with detected levels of 1.87-16.2ng/L and 0.48-2.60ng/L for HHCB and AHTN in glacier meltwater, respectively.The authors speculated that AHTN is photolyzed more rapidly than HHCB at this site, resulting in the progressive accumulation of HHCB on the glacial surface.Environmental meltwater was also sampled by Ferrario et al. (2017) to investigate HHCB (0.99-1.57ng/L) and AHTN (0.86-1.79 ng/ L) in meltwater of three Alpine glaciers.The study found that the two SMs are characterized by high and constant releases in the urbanized areas present in the Po River plain, with a potential for them to accumulate in meltwater in cold environments near densely anthropized areas.
From the published results shown above, it is evident that a variety of factors affect SMs in snow and meltwater, including atmospheric transport, photolysis, and urbanisation.Further investigations could be conducted to support these hypotheses.

Temporal variations of SM levels in natural water environment
The temporal variations of SM levels in the natural water environment differ between published results.Peng et al. (2017) investigated temporal changes of four SMs in river water in the Guangzhou area in China.In the dry season (winter), concentrations of up to 753, 126, 43.8 and 98.3 ng/L were detected for HHCB, AHTN, MX, and MK, respectively, while concentrations decreased to 685, 74.0, 41.6, and 18.4 ng/L in the wet season (summer).Although higher SMs levels were statistically observed, likely due to slower river flow in the dry season, there was no significant difference in total or individual concentrations between the two seasons (p > 0.05).Lu et al. (2015) sampled six SMs in different months in the Songhua River in northeastern China.Temporal profiles were found to be seasonal, with higher SM levels in the cold season (specific data not shown).Lower concentrations of SMs in the summer were also observed by Pintado-Herrera et al. ( 2020) investigating six SMs in the surface water of Cadiz Bay, Spain.
However, contrasting findings were reported by Yao et al. (2018), who monitored AHTN and HHCB temporal changes in the Pearl River and Yangtze River in China.In the Pearl River, AHTN presented 6.01-17.2ng/L in the wet season (summer) but 0-10.6 ng/L in the dry season (winter).Similar patterns appeared in the Yangtze River, with HHCB levels of <0.94-2.64ng/L in the dry season and higher levels of 7.86-10.3ng/L in the wet season (7.86-10.3ng/L) (Yao et al., 2018).In a comprehensive study by Homem et al. (2022), almost all the DPMI, ADBI, AHMI, HHCB, AHTN, and MK detected showed higher concentrations in the summer than in the winter in the Ave, Leça, Antuã, and Cértima rivers in Spain (Table 1).Besides, the overall concentrations of AHTN in seawater in the Yellow Sea and the East China Sea were also found to be higher in summer (23.7-38.2ng/L) than in winter (19.0-24.8ng/L) (Hua et al., 2022).Additionally, HHCB and AHTN exhibited relatively higher levels in the summer river water of the Adige River catchment in Italy, although in this study the authors hypothesised that more winter tourists might lead to stronger pollution (Villa et al., 2020).Research on temporal changes of SMs can also been found in other studies, e.g., (Lange et al., 2015;Villa et al., 2012;Vimalkumar et al., 2021;Wang and Kelly, 2017a).
Generally, in the cold (dry) season, smaller runoff decreases the natural dilution effects, and various degradation rates of SMs could be slower than those in the hot (wet) season because of the lower temperature.Such effects lead to higher SM levels in the water.However, it should be noted that the effects of the direct input of SMs from human activities (e.g., swimming) and emissions of SMs from WWTPs into receiving waters at different times cannot be omitted in the hot (wet) season and may significantly influence the results (Lu et al., 2015).These factors can make it difficult to predict the exact temporal patterns of SMs in natural water environments.Thus, their temporal variations can be complex.In addition, it should be mentioned that most studies investigating temporal changes in SMs focus on surface water.Further exploration is needed to investigate other types of natural water environments.

Summary
From Table 1, it is evident that research on surface water, especially river water, comprises a significant portion of published studies.Factors such as WWTPs and nearby densely populated areas, which are considered important pollution sources, influence the spatial distribution of SMs.Variations in SM levels within the same sampling zone suggest possible changes in pollution status and usage patterns over time.Although SMs exhibit relatively low occurrence and concentration, they have been detected widely in seawater, estuarine water, and groundwater.This underscores the complexity of SM pollution in natural water environments worldwide, as these waters are typically perceived to be less contaminated than river and lake water.Furthermore, SMs have been found in snow and the meltwater of glaciers, indicating the atmospheric transport of SMs to high-altitude areas.Additional investigations could be carried out in other regions.Temporal patterns of SMs in the natural water environment are more complex than spatial changes.They are influenced by various factors such as hydro-dilution, temperature, and changes in SM inputs during different seasons.
Comparing SM levels in natural water environments worldwide is challenging due to variations in methodologies across studies.To address this, we have compiled and presented the highest reported concentrations of SMs in Fig. 3.
Of the nitro musks, MK and MX displayed the highest occurrence, with mean highest detected concentrations of 63 ng/L and 84 ng/L, respectively, while other nitro musks were infrequently detected.Furthermore, although OTNE and EXA (macrocyclic musk) were only sporadically studied, their concentrations could exceed 1 μg/L (Arbulu et al., 2011;Llamas-Dios et al., 2021;Pintado-Herrera et al., 2020).This suggests that other types of SMs may warrant investigation in the future.

Synthetic musks in sediments and aquatic suspended matter
Table 2 summarises the SM types and concentrations in sediments and ASM environments.Polycyclic and nitro musks have been the primary compounds of interest in these studies.Similar to studies in the water environment, some research that focused on developing analytical methods used sediments and ASM for method application, e.g., (Chiaia-Hernandez et al., 2013;Necibi et al., 2016;Zeng et al., 2018a;Zhang et al., 2015aZhang et al., , 2015b)).Such works focused on factors like method accuracy, precision, and achieving better LOD and LOQ, but in-depth discussions on SMs were usually not emphasized.

Sediments
More than twenty studies were conducted in China, covering locations such as the Yellow Sea and the East China Sea (Hua et al., 2022), Chaohu Lake (Tian et al., 2021), Suzhou Creek (Song et al., 2015;Wang et al., 2010), Taihu Lake (Zhang et al., 2013), Haihe River (Mu and Wen, 2013), Jiaozhou Bay (Jiang et al., 2018), Songhua River (Lu et al., 2015), Liangtian River (Sang et al., 2012), and Huangpu River (Wang et al., 2018) Generally, concentrations of SMs were found at levels in the range of ng/g dw.Wang et al. (2018) investigated HHCB, AHTN, MX, and MK concentrations and their spatial distributions in the Huangpu River and its tributaries in Shanghai.The mean concentrations of the four SMs were 4.32, 0.339, 7.48, and 1.79 ng/g dw, respectively.Higher SM concentrations were observed in samples near chemical industrial parks and densely populated urban areas, suggesting domestic wastewater and municipal sewage as the main sources of SMs.
In the study by Lyu et al. (2020) on Chaohu Lake and six nearby rivers, lake sediments showed concentrations ranging from <0.94 to 9.97 ng/g dw for DPMI, AHMI, ADBI, HHCB, MK, and AHTN, while river sediments had concentrations ranging from <0.94 to 65.8 ng/g dw.Besides industrial and domestic sources, swimming and other recreational activities were found to contribute to the input of SMs into rivers, but a low ecological risk was observed for aquatic organisms.
Interestingly, Tian et al. (2021) also investigated SM levels in sediments of Chaohu Lake and nearby rivers, revealing much higher levels of     2018c).Since the sampling time and sites of Pearl River sediments in these studies were not identical, the discrepancies shown indicate that SMs levels might be greatly influenced by the sampling strategy and potential changes in pollution sources.Therefore, long-term monitoring and deeper investigations are necessary to draw comprehensive conclusions.
In the USA, Sapozhnikova et al. (2010) investigated sediment samples collected from three tidal tributaries of Chesapeake Bay.The highest concentrations of HHCB and AHTN were 1.6-9.2 and 0.14-8.0ng/g dw in the three rivers, respectively, with the most urbanized tributary, Magothy River, presenting the highest SM levels.In another study conducted by Subedi et al. (2014), mean concentrations of HHCB and AHTN in New Bedford Harbour sediments were 12 and 6.3 ng/g dw, respectively.Concentrations of HHCB and AHTN in the upper harbour were 21 and 31 times higher than the concentrations found in the lower harbour, with WWTP discharge identified as the main source.In comparison, in the upper Hudson River, higher HHCB and AHTN concentrations were detected in sediments (72.8-388 and 113-544 ng/g dw), as well as in water (3.95-25.8 and 5.09-22.8ng/L), fishes (<1-125 and <1-32.8ng/g, lipid weight), and zebra mussels (10.3-19.3 and 42.2-65.9ng/g, lipid weight) (Reiner and Kannan, 2011).
In Europe, concentrations of 7.0-12.9and <4 ng/g dw were observed for AHTN and HHCB in coastal lagoon sediments of Sacca di Goro, Italy (Casatta et al., 2015).Combi et al. (2016) also found a higher level of AHTN (up to 24.3 ng/g dw) than HHCB (up to 16.0 ng/g dw) in the sediments of the Adriatic Sea in Italy.The estimated burdens in the whole Adriatic basin were 424 kg for AHTN and 275 kg for HHCB, with similar total annual accumulation (~140-210 kg/year).In Austria, Clara et al. (2011) studied the concentrations of six SMs in river sediments downstream of WWTPs.Up to 20, 120, and <10 ng/g dw were measured for AHTN, HHCB, and ATII, respectively, whereas neither DPMI, ADBI, nor AHMI was detected.The authors also found that the highest concentrations were observed in the samples with the highest proportion of treated wastewater.In France, nine SMs were studied in various sediments near the North Atlantic Ocean, with total concentrations ranging from 0.1 to 14.3 ng/g dw in Capbreton Canyon and 0.1-3.1 ng/g dw in continental shelf surface sediments, and several canyon locations were examined as specific accumulative areas (Azaroff et al., 2020).
In comparison, SM levels in sediments were much higher in Southeast Asia (Table 2).Concentrations of up to 32,600-43,800 and 1600-3500 ng/g dw, respectively, were detected for HHCB and AHTN in the central area of Jakarta, Indonesia.These elevated levels were attributed to significant SM input from the city centre and their subsequent accumulation in the sediments.The authors also expressed concerns about the potential risks of SMs to benthic invertebrates (Dsikowitzky et al., 2020).Similarly, high concentrations of SMs were found in riverine and estuarine sediments in Bangkok, Thailand (Juksu et al., 2020).AHTN, HHCB, and MX exhibited the highest concentrations at around 10 3 ng/g dw and HHCB-lactone exceeded 10 4 ng/g dw in riverine sediments.Additionally, concentrations of around 10 2 and 10 3 ng/g dw were found for AHTN and MX in estuarine sediments, respectively.Consequently, medium-high and low-high ecological risks were calculated in riverine and estuarine environments, respectively.In addition to Thailand and Indonesia, surveys were conducted in Singapore, where SMs were found at ng/L levels (Wang andKelly, 2018, 2017b;Zhang et al., 2015b;Zhang and Kelly, 2018).The relatively higher concentrations of SMs in Southeast Asia might be attributed to increased SM loads (e.g., perfume) in the warm climate, the dense population at sampling sites, and the use of SMs in religious activities (e. g., incense burning).However, more data are needed to enable a a Data are displayed in their original units from the references.
b n.d.: not detected.c <LOD: below limit of quantification.d <LOQ: below limit of detection.e DNS: data not shown.
comprehensive comparison between Southeast Asia and other regions worldwide.Furthermore, SM levels were sporadically reported in sediments of other areas, such as Bizerte Lagoon in Tunisia (Necibi et al., 2016), Todos os Santos Bay and the north Salvador coast in Brazil (Sotão Neto et al., 2020), and the Geum River in South Korea (Kim et al., 2022).Detected concentrations were generally at the ng/g level (Table 2).

Aquatic suspended matter
Studies focusing on SM levels in ASM have been relatively limited and are often included as part of broader environmental assessments.Most of these studies reported SM concentrations in units of ng/g dw, except Zhang et al. (2015a) who used absolute concentration in the aqueous phase as pg/L to calculate eight SMs in the coastal seawater ASM around Singapore Island.The highest mean concentrations of ADBI, HHCB, AHTN, and MK were 59.6, 536, 101, and 35.4 pg/L, respectively, while AHMI, ATII, MA, and MX were not detected.In comparison, the highest mean concentrations for the same eight SMs in the seawater was 31.6-21,800pg/L.Similarly, seven SMs were measured in the Tamar Estuary in UK, but only HHCB and AHTN were detected, at low concentrations of 11-29 and 1-11 ng/g dw for ASM and 11-17 and 2-10 ng/g dw for sediments, respectively, showing no significant difference and reflecting the SMs input from nearby WWTP effluents (Sumner et al., 2010).However, much higher concentrations were found in the Molgora River in Italy.Concentrations of HHCB, AHTN, and ADBI reached up to 17,993, 4321 and 23.4 ng/g dw in suspended sediments, while the highest concentrations were 1141, 364.42, and 23.4 ng/L in the water phase, respectively (Villa et al., 2012).This may be ascribed to the fact that the Molgora River is located in a densely populated area, and intense human activity contributed to the high concentrations.The authors also noted that reduced river flow may contribute to higher contamination.Apart from these regions, SMs in ASM were also investigated in Singapore and China, with varying concentrations reported (Table 2) (Mu and Wen, 2013;Wang and Kelly, 2017a;Zhang and Kelly, 2018).It is worth noting that, although SMs in the aqueous suspended phase are not typically the primary focus of research, their high concentrations indicate a potential migration of SMs with ASM.

Summary
On a global scale, SMs in sediments and ASM were surveyed more in China and Europe from 2010 to 2023, followed by North America and Southeast Asia.Typically, detected SMs showed ng/g dw levels, while they exhibited relatively high concentrations in the sediments of several Southeast Asian areas, likely due to various reasons such as increased SM usage in warm climates and thriving religious activities.Additionally, similar to the natural water environment, some researchers also noted the impact of WWTPs and dense populations on the levels of SMs.However, there is still an inadequacy in ASM investigation as they may be a reservoir involved in SM migration in the environment.

Soils
SMs in natural soils generally exhibit low concentrations (Table S1).Chen et al. (2011) tested natural (uncultivated and unpolluted) field soils collected from the surface layer (0-20 cm in depth) of an arboretum area, and found that the concentrations of HHCB and AHTN were both <0.001 mg/kg dw (baseline).A comprehensive study of SM levels in various types of soils (e.g., agricultural, garden, industrial, and school yard) found low occurrences of twelve target SMs with the highest concentrations below 10 ng/g dw (Ramos et al., 2021a).Similar results were also reported by Domínguez-Morueco et al. (2018) monitoring nitro and polycyclic musks in petrochemical, chemical, urban soils, and soils away from pollution sources.No significant differences between sampling areas were statistically observed.Zheng et al. (2019) found that the total concentrations of seven SMs in farmland soils from three Chinese northeast provinces (Jilin, Liaoning, Heilongjiang) ranged from 2.40 to 12.2 ng/g dw.HHCB and AHTN were the dominant pollutants detected in all samples, constituting 99.35 % of the total SMs.The authors noted no serious environmental impact from SMs.Other studies showing low levels of SMs in natural soils include those by Chane et al. (2023), Fernandes et al. (2022), Vecchiato et al. (2021) and Wong et al. (2019).
Several studies have investigated SM levels in soils with external inputs.In one systematic research by Chen et al. (2014), HHCB and AHTN levels in various soil types were examined after applying biosolids containing HHCB and AHTN (2950 and 1400 μg/kg, respectively).The HHCB and AHTN mean concentrations were n.d.-29.0 and 0.4-67.5 μg/ kg dw for various soil types (paddy soil/silt loam, red soil/loam, fluvoaquic soil/clay loam) after different treatments (none, one, and yearly biosolid application).Two SMs showed low to medium ecological risks.In another study investigating SM residue levels after bioremediation by white rot fungi combined with phytoremediation (Zea mays) in biosolidamended soils, HHCB and AHTN were found to be 39.8-66.5 and 15.3-22.3μg/kg dw, respectively, in different treated soils, after application containing 4161 and 1053 μg/kg dw biosolids of them, but no environmental risk was found (Chane et al., 2023).
Regarding SMs in the soils after irrigation, by using reclaimed water (HHCB and AHTN at 200-251 and 65-130 ng/L, respectively), HHCB and AHTN in two public parks in Beijing reached concentrations of 1.33-2.55and 1.91-3.92ng/g, respectively, compared to 0.233-0.435and 0.330-0.659ng/g after using public water (SM levels below LOD) for irrigation (Wang et al., 2013).This study suggested that it would take 243 and 666 years respectively for HHCB and AHTN to accumulate in soils to cause potential ecosystem harm.In a one-year study, Biel-Maeso et al. ( 2019) used WWTP effluents (770-5700, 271-3990, 1640-6980 and 1310-6000 ng/L for HHCB, AHTN, ATII, and OTNE, respectively) for soil irrigation.In summer, soil levels were 3.7-221.8,<LOQ-8.5, 0.8-5.2,n.d., and 6.1-66.0ng/g dw.Higher concentrations were observed in winter, ranging from 338.6 to 651.4,9.8 to 51.2, 18.1 to 50.4,0.5 to 8.6, and 89.4 to 589.4 ng/g dw for HHCB, AHTN, ATII and OTNE, respectively.This study noted that these variations might be due to weaker biodegradation under lower temperatures.In comparison, in a one-year study by Chase et al. (2012), no difference was observed in nitro and polycyclic musks between soils inside and outside a pivot irrigation area irrigated with WWTP effluents (up to about 4000 ng/L for single SM) for over 70 years (Table S1).The authors assumed that this may be because of the runoff and microbial degradation.

Sands
Homem et al. ( 2017) assessed the seasonal variation of ten SMs in beach sands from the Oporto coastal region in Portugal.In summer, concentrations of HHCB, AHTN, EXA, and MK reached up to 25, 3.1, 3.8, and n.d.ng/g dw, respectively, while in winter, they were 27, 2.0, 2.5, and 0.70 ng/g dw.The authors suggested that more intense solar radiation in summer may promote SMs photodegradation.Additionally, three target SMs were considered not to pose risks to the environment.Since many outdoor recreational activities are associated with beach environments, more studies on SMs in sands should be carried out in the future.

Summary
Based on the discussed research, it appears that SMs in soils and sands are generally found at low levels.Preliminary risk assessments conducted by researchers suggested low ecological risks associated with them (Zheng et al., 2019;Wang et al., 2013;Homem et al., 2017).However, it is important to note that the number of studies conducted on SMs in soils and sands is relatively limited.More systematic studies and assessments will still need to be done to further understand the environmental impact of SMs.Additionally, results from Chase et al. ( 2012) demonstrate a possible SM accumulative pattern in soils after years of irrigation that differs from other researchers' observations.This also highlights the importance of long-term monitoring of the impacts of irrigation on soil.

Synthetic musks in outdoor air and atmospheric particulate matter
Researchers have shown interests in gaseous SMs, and SMs have been detected in outdoor air, even in remote areas, indicating potential atmospheric transport and air-water exchange of SMs, for example, median concentrations of gaseous HHCB and AHTN were 4 and 18 pg/m 3 in the Arctic, 28 and 18 pg/m 3 in the North Sea area, respectively (Xie et al., 2007).Table S2 summarises levels of SMs in outdoor air and APM from selected published works.

Outdoor air
Most published studies have investigated polycyclic musks.Wong et al. (2019) detected twenty-one SMs in different areas of Canada but, unlike Xie et al. (2007), found no SMs in the Arctic sampling.After analysing six polycyclic musks (DPMI, ADBI, AHMI, ATII, HHCB, AHTN), it was revealed that their total levels in the air of WWTPs and indoor areas were significantly higher than in urban and rural air.The median level of six SMs was about three times higher at the three most urban sites than at the three most rural sites and their concentrations along the urban-rural transect were positively correlated with population density.Temporal analysis showed that SM concentrations in the air were generally higher during the warm season than the cold season, due to greater SM volatilization and release during warm weather.It is also worth noting that several macrocyclic musks (e.g., Musk MC-4, EB) were also detected in the air of WWTP zones (Table S2).
In another study investigating the air and water of the lower Great Lakes, average total concentrations of five gaseous polycyclic musks were <LODs to 3.2 ng/m 3 on the western shoreline of Lake Erie in Toledo (McDonough et al., 2016).Significant correlations between population density and recorded SMs in the air were observed.Air-water fluxes of HHCB and AHTN were 11-341 and (− 3)-28 ng/m 2 /day, respectively.It was also estimated that volatilization may cause the loss of 64-213 kg/year of dissolved SMs from the lakes.Melymuk et al. (2014) also calculated the loadings of HHCB and AHTN from Toronto to Lake Ontario.With total levels of two SMs ranging from 0.002 to 3.5 pg/ m 3 , air-water transfer accounted for 10 % or 71 ± 39 kg/year of SMs total loading to the lake.
In terms of nitro musks, Ramírez et al. (2010) reported levels of 1.6-4.0 and 0.8-1.5 ng/m 3 for MX and MK, respectively, in the urban and suburban outdoor air in Spain.

Atmospheric particulate matter
SMs have also been found in APM.For instance, J. Sánchez-Piñero and coworkers conducted a series of measurements of the occurrences and levels of polycyclic and nitro musks in the urban atmospheric PM 2.5 in Vigo in Spain from 2021 to 2023.These investigations showed low occurrences and concentrations of SMs (Table S2).In general, important pollution point sources, such as WWTPs and landfill sites, contribute to SM accumulation in both outdoor air and APM.Weinberg et al. (2011a) observed that up to 1362 and 211 pg/m 3 of HHCB and AHTN, respectively, were present in the APM at WWTP aeration tank areas, compared to 11 and 16 pg/m 3 in the APM away from the WWTPs.In the outdoor air, their concentrations were 5157-407,194 and 304-65,063 pg/m 3 in the WWTP aeration tank areas, and 63-831 and 10-131 pg/m 3 in the reference sites.However, in another work by the same authors, HHCB and AHTN levels were lower in the APM at landfill sites than in the reference sites (specific discussion not provided), while higher concentrations were still found in the outdoor air at landfill sites (Weinberg et al., 2011b).Given this, the authors suggested WWTPs as sources of target SMs in the gaseous phase.
Regarding the seasonal pattern, HHCB and AHTN adsorbed on APM were quantified in Hanoi in Vietnam (Duong et al., 2019).Up to 24.54 and 10.76 ng/m 3 for HHCB and AHTN were detected in the dry season, which is higher than the 11.82 and 3.608 ng/m 3 in the rainy season.However, this work did not provide a deep discussion on the reason.Here, it is speculated that the dry climate may lessen the hydro-dilution, subsequently causing stronger volatilization of SMs to the gaseous phase.Additionally, the authors considered Hanoi to be a hotspot for such compounds because the city is highly populated.Other published works on SMs in APM can be found in Table S2.

Summary
The existence of SMs in outdoor air and APM indicates the phase exchange of SMs in the natural environment, which is in accordance with the observations of SMs in snow and meltwater coming from atmospheric transport (Section 5.5).Source analysis revealed the potential links between SMs and WWTPs, landfill sites, and densely populated areas.Several investigations also showed temporal changes in SMs in the gaseous phase.Polycyclic musks were the dominant SMs found, while other SMs were typically detected in low occurrences or concentrations.Nevertheless, the presence of macrocyclic musks in WWTP air (Table S2) might suggest their possible atmospheric migration.
In recent years, bioaccumulation of SMs has also been widely found in aquatic and soil biota.For instance, high biota-sediment accumulation factors of 2.5 and 4.0 for common carp, 1.5 and 1.9 for silver carp, and 1.6-2.4 and 1.6-2.6 for crucian carp, respectively, for HHCB and AHTN in the Haihe River (Hu et al., 2011a(Hu et al., , 2011b)).In Chaohu Lake, DPMI and HHCB showed a trend toward trophic magnification in the freshwater food web in a study examining six SMs in fifteen aquatic species (fish, shrimp) (Lyu et al., 2021).Moreover, HHCB and AHTN not only posed potential risks to soils amended with a commercial sewage sludge-based organic fertilizer but were also found in Micro-Tom tomatoes grown with bioconcentration factors ranging from 0.2 (AHTN) to 23 (HHCB) (Ramos et al., 2021b).Bioaccumulation of SMs in plants and vegetables was also reported in other studies, e.g., MacHerius et al. (2012); Ramos et al. (2020); Wang et al. (2013).Consequently, SMs in the environment may finally enter human bodies through the food chain, posing potential health risks (MacHerius et al., 2012;Rainieri et al., 2017;Zhang et al., 2013).
Although SMs are only partially biodegradable, SM biodegradation can occur in the natural environment (Chase et al., 2012;Ramírez et al., 2012).The biodegradability coefficients of HHCB and AHTN are 0.71 and 0.023 (Liu et al., 2021), and macrocyclic and alicyclic musks are considered more biodegradable (Sanganyado et al., 2020;Wang et al., 2023).In a study by Wong et al. (2019) in the Toronto urban-rural area, no detectable SMs were found with the reason being ascribed to the relatively high volatility and fast biodegradation rate of the tested SMs.In another study by Wang et al. (2013) on soils irrigated with reclaimed municipal wastewater, biodegradation of HHCB and AHTN was assumed to take place only in the top 20 cm of the surface layer, with both calculated first-order biodegradation rate constants at 0.0039 d − 1 .However, different findings have been reported.Klaschka et al. (2013) evaluated a much faster OTNE primary 50 % degradation time of one day in a biodegradation simulation test according to OECD TG 314, even though OTNE is registered as not readily biodegradable according to the European Chemicals Agency.
It should be noted that HHCB-lactone, a transformation product of HHCB, was found in the natural environment, sometimes in high concentrations (Tables 1 and 2).It is a polar and recalcitrant compound showing greater polarity and higher affinity for water (Tasselli et al., 2021).The biotransformation metabolites of HHCB and AHTN were investigated with Myrioconium sp.strain UHH 1-13-18-4 and Clavariopsis aquatica, two mitosporic fungi isolated from freshwater environments (Martin et al., 2007).The fungi converted HHCB and AHTN into various products via initial hydroxylation at different carbon positions.Further metabolism caused the subsequent formation of diketone, peroxide, and O-methylated derivatives.The isolated laccases were able to oxidize HHCB and AHTN and catalysed the HHCB to HHCB-lactone.The biodegradation of HHCB-lactone was also reported by Vallecillos et al. (2017) using the enzyme laccase from Trametes versicolor and the redox mediator 2,2′-Azino-bis(3-ethylbenzothiazoline-6-sulfonic acid).More than 90 % of HHCB-lactone was removed after 144 h, while other SMs ranged from 42 % to >70 %.However, compared to the SM biodegradation research using the WWTP techniques (Homem et al., 2015b), there is still a need for SM biodegradation investigations in natural matrices, especially for SMs other than polycyclic musks.

Photodegradation
Photodegradation is another process affecting SMs in the environment.Direct photodegradation of SMs has been reported in various studies (Gao et al., 2017;Liu et al., 2021;Vallecillos et al., 2015a).More intense solar radiation generally promotes direct photodegradation (Homem et al., 2017).Butte et al. (1999) studied the photochemical degradation of MT, MK, MX, MA, and MM and found that they were all photochemically degradable.MT could also be converted under outdoor natural sunlight.Canterino et al. (2008) found that, under solar light, MT was slowly degraded when suspended in water, and three photoproducts (3,3,5,6,3,3,5,6,and 3,3,5,6,-nitro-3Hindolinone) were identified.The half-life of direct photodegradation products in the summer and winter was predicted to be 1-1.5 and 6-10 h, respectively, and their environmental persistence increased with increasing latitudes and during the cold seasons.Recently, Luo et al. (2023b) identified a new bi-radical species formed during the photochemical degradation of AHTN in water, generated from intramolecular H-abstraction.The photolysis mechanism of AHTN was photoenolization followed by cyclization, and the bioconcentration factor of photoenol was 13-fold higher than that of AHTN.In field research, Homem et al. (2017) attributed higher levels of SMs in beach sands in winter to more intense solar radiation in summer, and Chase et al. (2012) observed a decrease in SM levels from WWTP effluent to reservoir because of possible photolysis.It should be noted that, in addition to studies on direct photodegradation of SMs conducted in the lab or under simulated conditions, more investigations in the real natural environment are needed.Buerge et al. (2003) considered indirect photolysis of HHCB and AHTN by reactive oxygen species to be less important because they were mainly photolyzed by direct photolysis.However, the ⋅OH-initiated indirect photochemical transformation mechanism of AHTN in the aquatic environment was theoretically analysed in a modelling study by Gao et al. (2016).Indirect transformation was found to be more important than direct transformation.• OH-addition pathways were dominant at low temperatures (<~287 K), whereas H-abstraction was the dominant pathway at high temperatures.The bioconcentration factors and aquatic toxicities to fish of all transformation products from ⋅OH-addition pathways were up to eight times higher than AHTN.In addition, the indirect photodegradation of MX during ⋅OH-initiated transformation process was systematically studied using quantum chemistry (Gao et al., 2019).Compared to single-electron transfer and • OH-addition pathways, MX can be exclusively transformed via H-abstraction pathways from its methyl group.The dehydrogenation intermediates could transform into cyclic, aldehyde, and demethylation products, and all the transformation products exhibited decreased aquatic toxicity.However, these products were still classified as toxic or very toxic and exhibited carcinogenic activity during • OH-initiated transformation.Both studies suggested paying further attention to the indirect SM photochemical products.Using density functional theory, the multichannel mechanism of • OH radical-initiated atmospheric degradation reactions of HHCB has also been investigated (Y.Li et al., 2018).It is shown that ⋅OH-addition and H-abstraction reactions are competitive pathways for HHCB.Epoxide, dialdehyde, alcohol ketone, cyclolactone compounds, and HO 2 radicals are the dominant products in the presence of O 2 /NO.At 298 K, the calculated total rate constant of OH-initiated degradation is 2.71 × 10 − 11 cm 3 molecule − 1 s − 1 , and the atmospheric lifetime of HHCB is 10.09 h.The proposed medium-range transport for HHCB in the atmosphere is in accordance with the findings of Villa et al. (2014).

Other pathways
Partitioning between solid and aqueous phases leads to SM deposition in sediments, then SMs may potentially come into contact with groundwater.Additionally, through irrigation and long-term infiltration, SMs can enter soils from water sources, thereby contaminating groundwater as well (Ding et al., 2023;Ramos et al., 2021a).The calculated logK DOC of ADBI, AHMI, ATII, HHCB, and AHTN onto humic acid were between 3.32 and 3.67 (Böhm and Düring, 2010).These physicochemical lipophilic properties of SMs make adsorption easy.The study by Wang et al. (2010) on the distributions of SMs in Suzhou Creek in China shows that the ratio of HHCB/AHTN concentration was much higher in surface water than sediments, indicating stronger adsorption of AHTN in sediments than HHCB.Besides, soil colloids have huge specific surface areas, producing strong adsorption abilities that firmly bind SMs (Chen et al., 2011;Liu et al., 2021).In addition, through adsorption, SMs in the gaseous phase can also be adsorbed on APM (X.Li et al., 2018).
Because SMs are semi-volatile compounds, phase exchanges including air-water and air-soil exchanges occur through atmospheric deposition and volatilization.The median net air-sea volatilization fluxes for HHCB and AHTN were 27 and 14 ng/m 2 /d in the North Sea, respectively (Xie et al., 2007).In comparison, their calculated air-water fluxes were 11-341 and (− 3)-28 ng/m 2 /d in the lower Great Lakes (McDonough et al., 2016), showing AHTN to be more variable.Through simulation, the calculated input amounts of HHCB and AHTN in soils after irrigation with volatilization process would be 5-6 times lower than those without volatilization, and volatilization of AHTN was less significant than HHCB (Wang et al., 2013).Besides, the detection of SMs on high-altitude glaciers also revealed the air-water exchange process, as well as atmospheric transport (Villa et al., 2014).Exchanges between phases of SMs in the natural environment were also mentioned and briefly discussed in other studies, e.g., Lange et al. (2015) and Wong et al. (2019).
Generally, hydro-dilution decreases SM levels in the natural environment.Spatial analysis in water reveals that concentrations of SMs decline with increasing distance from sources (e.g., WWTPs) (Peng et al., 2017).Some temporal research shows that SM levels were lower in wet/ rainy seasons, probably due to more intense rainfall (Yao et al., 2018;Duong et al., 2019).However, rainfall may sometimes increase SM levels in surface water and sediments.In a study by Wang and Kelly (2017a) on the influence of rainfall on SMs levels in surface water and sediments in Singapore, the concentrations of SMs were lower during dry periods than rainy periods.The authors speculated that rainfall may directly influence SM atmospheric deposition and/or land-surface runoff, consequently increasing SM levels.The discrepancies between findings may suggest a deeper investigation.

Ratios of synthetic musks
Ratios of SMs can provide insights into their fate and behaviours in the natural environment, which are influenced by area usage, the effectiveness of WWTP treatment, and environmental processes (Wang et al., 2023).Generally, the ratio of HHCB/AHTN is higher in natural water than in sediments due to a higher photolysis rate in water and stronger adsorption in sediments for AHTN (Buerge et al., 2003;Wang et al., 2010), while reported gaseous HHCB/AHTN ratios varied in studies (McDonough et al., 2016;Weinberg et al., 2011b;Wong et al., 2019).Lee et al. (2010) found that measured AHTN levels fit very well with the estimated values in the surface waters predicted from the HHCB concentration, and fitted MK levels were slightly lower, revealing a similar fate for HHCB and AHTN but a relatively different fate for MK in the water environment.Significant positive correlations between HHCB and AHTN concentrations in water and sediment samples (R 2 = 0.9039, 0.9584, p < 0.05) were calculated in the Songhua River Basin (Lu et al., 2015).Results also exhibited that HHCB and AHTN had similar tendencies to partition between aqueous and solid phases due to similar soil organic carbon sorption coefficients.
HHCB-lactone/HHCB ratios were also used to analyse the HHCB degradation potential in various matrices.The calculated ratios were 0.6-2.5 and 0.8-2.0 in the Chaohu Lake river and lake sediments, which were influenced by sources and environmental physicochemical and hydrological conditions (Tian et al., 2021).The ratios of 0.15-0.64 in the Guangzhou waterways were calculated by Su et al. (2023).Lange et al. (2015) found significant seasonal variations of HHCB-lactone/ HHCB ratio in a German river and attributed the reason to the increasing microbiological or chemical degradation of HHCB at higher temperatures.Besides exploring the similarity of SM fate, ratios of SMs can also be employed as a source discriminator and tracer of distance from the emission source, as well as assessing the degree of transformation and degradation during transportation (Zeng et al., 2018c).

Modelling approaches in exploring the fate of synthetic musks in natural environment
In recent years, modelling tools have been employed to study and predict the fate of SMs when direct tests are difficult to conduct.As mentioned above, researchers have computationally explored the indirect photochemical transformation of SMs in water and the atmosphere (Gao et al., 2019(Gao et al., , 2016;;Y. Li et al., 2018).Apart from them, Wang et al. (2013) used a water and solute fate and transport software package, HYDRUS-1D, to simulate the long-term accumulation of HHCB and AHTN in soils at a public park in Beijing.The transport simulation indicates that the inputs through reclaimed wastewater irrigation generally caused the accumulation of HHCB and AHTN in the surface soil layer only, while soil downward transport, biodegradation, and plant uptake were relatively less important.If the volatilization process was taken into account, concentrations of HHCB and AHTN in the 0 to 10 cm soil layer were simulated at 0.776 and 0.343 ng/g, respectively.Bu et al. (2022) explored the degradation rates of HHCB and AHTN in the North Canal River watershed in Beijing, using a level III fugacity model combined with a least-squares method.The degradation rates were 4.16 × 10 − 3 h − 1 (t 1/2 = 167 h) and 1.68 × 10 − 2 h − 1 (t 1/2 = 41.3 h) for HHCB and AHTN, respectively.A coupled model of HYDRUS-GMS was proposed by Ding et al. (2023) to predict the fate of HHCB in long-term infiltration from river receiving reclaimed water to groundwater.In the vadose zone, HHCB increased at an accumulative rate of 6.1 ng/ g year − 1 with infiltration time extension, posing ecological risk after 53 years of infiltration.The potential risk will extend to groundwater and diffuse along the groundwater flow direction with a migration rate along the horizontal direction at 0.03396 m/d.It was also calculated that the complete biochemical decomposition of HHCB will take approximately 0.38 years through metabolites within a 182.65-meter distance, and the major biochemical metabolisms were enzymatic hydrolysis, ring opening, and decarboxylation.Regarding gaseous SMs, Villa et al. (2014) utilized the OECD POV and LRTP screening tool model for the analysis of potential atmospheric transport of HHCB and AHTN.The modelling results displayed a possible medium-range atmospheric transport, and the theoretical findings were supported by the experimental results.
To sum up, SMs undergo various pathways in the natural environment.Biodegradation and photodegradation lower SM levels but may generate more toxic products, deserving deeper investigations.Adsorption, phase exchange, and hydro-dilution also play roles in affecting SM fate.Some differing findings indicate the complexity of fate and the potential influence of specific research systems during these processes.However, there is still a lack of investigations under real environmental conditions, especially for macrocyclic and alicyclic musks.Perhaps using SM ratios may help to predict more SM compounds.Additionally, since more studies employ modelling approaches, understanding the fate of SMs in the natural environment is likely to improve.

Future prospects
Polycyclic musks have been extensively studied.Although nitro musks have been partially restricted, their high occurrences and concentrations suggest their continued prevalence.In comparison, macrocyclic and alicyclic musks are the least investigated but several studies have revealed their existence in the environment, such as EXA and ROM.As market trends and regulatory measures evolve, it is essential to monitor and better understand the occurrence and behaviours of these types of SM in natural matrices.Besides, in recent years, awareness of the risks of non-musk fragrances (e.g., salicylates) has also been raised (Patel et al., 2020;Vecchiato et al., 2017Vecchiato et al., , 2018)).Comparisons between these and SMs in the natural environment could also be conducted.
From 2010 to 2023, most SM research has been concentrated in the Northern Hemisphere (Fig. 2).Expanding studies to other regions, including South America, Oceania, the Middle East, Africa, Antarctica, and different parts of Asia, Europe, and North America, as well as understanding SM distribution and relationships with other factors (e.g., population), would provide a more comprehensive understanding of the global contamination situation.More investigation on SM contamination in diverse natural matrices across various regions can also help to reveal geographical variations and sources of contamination.
The behaviours of SMs have been explored mainly for polycyclic musk, and there is less focus on other SM types.Insights into the mechanisms of biodegradation and photodegradation under real natural conditions are still inadequate, but this may be improved by employing modelling tools.Some differing findings between studies (e.g., temporal and distribution patterns, impacts of photodegradation and hydrodilution) further indicate the need for long-term investigation and comparisons between different situations and influences.Therefore, deeper and more long-term investigations on SM fate should be undertaken under various natural conditions for a broader range of SM compounds.
The degradation and transformation products of SMs should receive more attention.HHCB-lactone has been detected in surface water and sediments, sometimes in high concentrations (Tables 1 and 2).Some photoproducts were also considered to show higher persistence and ecotoxicity than the parental SM compounds, but they are currently not widely monitored or surveyed.As more SM pathways are studied and discovered, the risks of degradation and transformation products could be further studied, understood, and elucidated.
This review shows that SMs have been widely detected in natural matrices near WWTPs, implying the inadequacy of current SM treatment technologies.Given that WWTP discharges are significant sources of SMs in natural environments, proper WWTP effluent and sludge treatment may considerably lessen SM pollution.Therefore, there is a need for more effective treatment technologies for SM removal in WWTPs.

Conclusions
This review has provided insights into the sources, occurrence, concentration, and fate of SMs in the natural environment from studies between 2010 and 2023.SMs enter the natural environment from direct (e.g., wastewater discharges) and indirect (e.g., human recreational activities) sources, and SMs have been extensively detected in various natural matrices, including river water, lake water, seawater, estuarine water, groundwater, snow, meltwater, sediments, ASM, soils, sands, outdoor air, and APM, indicating their widespread presence in the natural environment.While polycyclic musks, particularly HHCB and AHTN, are commercially dominant and frequently detected, other SM categories have also been found, underscoring their prevalence.The calculated mean highest detected concentrations of HHCB and AHTN were found to be 1177 and 242 ng/L in natural water, and 1448 and 237 ng/g dw in sediments and ASM, respectively.Spatial and temporal changes in SM levels were also explored in some studies but the change patterns are typically complex, so long-term investigations are required.SMs undergo various processes in the natural environment, such as bioaccumulation, biodegradation, photodegradation, adsorption, phase exchange, and hydro-dilution.SM levels decrease during biodegradation and photodegradation, but degradation and transformation products may be more persistent and eco-toxic.Besides, fate complexity and possible influence of specific research systems may cause differing results between studies.The exploration of SM fate (e.g., indirect photodegradation, long-distance atmospheric transport) in the natural environment could be assisted with the help of modelling tools.Finally, future prospects, such as conducting broader surveys of different SM types, gaining a deeper understanding of fate related to SMs under real natural conditions, and monitoring the degradation and transformation products of SMs, are suggested.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 2 .
Fig. 2. Number and distribution of studies of SMs in the natural environment from 2010 to 2023.

Fig. 3 .
Fig. 3. Highest concentrations of SMs in natural water reported in the literature from 2010 to 2023.(Data only from studies showing the highest concentration of specific SM compounds are included.Data from the literature showing minimum only, not detected, below limit of detection (LOD) and/or LOQ, mean, median or no specific data are excluded.Star: mean value.Outliers of each box are values outside the range of ±1.5*IQR, where IQR is the interquartile range defined as the upper quartile minus the lower quartile).

Fig. 4 .
Fig. 4. Highest concentrations of SMs in sediments and ASM reported in the literatures from 2010 to 2023.(Data only from studies showing the highest concentration of specific SM compounds are included.Data from the literature showing minimum only, not detected, below LOD and/or LOQ, mean, median or no specific data are excluded.Star: mean value.Outliers of each box are values outside the range of ±1.5*IQR, where IQR is the interquartile range defined as the upper quartile minus the lower quartile).

Fig. 5 .
Fig. 5. Fate of SMs in the natural environment.

Table 1
Concentrations of SMs in the natural water from selected published studies.

Table 2
Concentrations of SMs in the sediments and ASM from selected published studies.

Table 2
(continued ) (continued on next page) J.Li et al.