Accounting for intended use application in characterizing the contributions of cyclopentasiloxane (D5) to aquatic loadings following personal care product use: Antiperspirants, skin care products and hair care products

Chemosphere 93 (2013) 735–740

Contents lists available at SciVerse ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Accounting for intended use application in characterizing the contributions of cyclopentasiloxane (D5) to aquatic loadings following personal care product use: Antiperspirants, skin care products and hair care products Beta P. Montemayor a,⇑, Bradford B. Price b, Roger A. van Egmond c a

Canadian Cosmetic, Toiletry and Fragrance Association, 420 Britannia Road East, Suite 102 Mississauga, Ontario, Canada L4Z 3L5 Procter & Gamble, 11511 Reed Hartman Highway Cincinnati, OH 45241, USA c Unilever, Safety and Environmental Assurance Centre, Colworth Science Park, Sharnbrook, Bedford MK44 1LQ, United Kingdom b

h i g h l i g h t s " D5 in antiperspirants, skin and hair care products. " D5 volatilization, important considerations in assessing potential aquatic loadings. " Small fraction of D5 in leave-on personal care products destined to enter wastewater.

a r t i c l e

i n f o

Article history: Available online 24 November 2012 Keywords: Cyclopentasiloxane D5 Cyclic volatile siloxanes Aquatic contributions Personal care products

a b s t r a c t Decamethylcyclopentasiloxane, commonly known as D5 (cyclopentasiloxane) has a wide application of use across a multitude of personal care product categories. The relative volatility of D5 is one of the key properties attributed to this substance that provide for the derived performance benefits from the use of this raw material in personal care formulations. On this basis, rapid evaporative loss following use of many products comprising D5 is expected following typical use application and corresponding wear time. Studies were conducted on three key product categories containing D5 (antiperspirants, skin care products and hair care products) to characterize the amount of D5 that may be destined to ‘go down the drain’ following simulated typical personal care use scenarios. Marketed antiperspirants and skin care products were applied to human subjects and hair care products were applied to human hair tresses and subsequently rinsed off at designated time points representative of typical consumer cleansing and personal hygiene habits. Wash water was collected at 0, 8 and 24 h (antiperspirant and hair care analysis) and additionally at 4 h (skin care analysis) post product application and samples were analyzed by isotope dilution headspace gas chromatography/mass spectrometry (GC/MS) to quantify the concentration of D5 destined to be available to go down the drain in captured wash water. It is demonstrated that significant amounts of D5 in ‘leave-on’ application products evaporate during typical use and that the concentration of D5 available to go down the drain under such conditions of use is only a very small (negligible) fraction of that delivered immediately upon product application. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction Cyclic volatile methyl siloxanes (cVMSs), in particular, decamethyl cyclopentasiloxane (‘‘D5’’) are used across a wide spectrum

⇑ Corresponding author. Tel.: +1 905 890 5161x239; fax: +1 905 890 2607. E-mail addresses: [email protected] (B.P. Montemayor), [email protected] (B.B. Price), [email protected] (R.A. van Egmond). 0045-6535/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.chemosphere.2012.10.043

of personal care applications, including antiperspirants; skin, hair and nail care formulations; grooming products; sun protection products; personal lubricants; and fragrance and cosmetic preparations. cVMS such as D5 are highly volatile materials, with multimedia environmental fate models indicating that the dominant fraction of D5 released to the environment will partition to the atmosphere (Environment Canada, 2008; Brooke et al., 2011) where it is believed that atmospheric degradation dominates its fate (Atkinson, 1991; Navea et al., 2011).

736

B.P. Montemayor et al. / Chemosphere 93 (2013) 735–740

Volatile silicones, including D5, have been measured in indoor dust (Lu et al., 2010), the atmosphere (McLachlan et al., 2010; Genualdi et al., 2011), water and wastewater (Sparham et al., 2008) and both river and estuarine sediments (Sparham et al., 2011). Recent studies have evaluated the occurrence and concentration of D5 in personal care products (Horii and Kannan, 2008; Wang et al., 2009; Lu et al., 2011). Wang et al. evaluated D5 concentrations in 252 consumer products in Canada and found detectable levels of D5 in 14% of the products tested with concentrations ranging from 0.02 mg g 1 to as high as 680 mg g 1 in an antiperspirant product. Other product forms with significant levels of D5 include skin creams and hair styling products. The most recent survey by Lu reported measureable concentrations of D5 in 64% of hair and body care products from Shanghai, China with concentrations as high as 1.1 mg g 1 in a hair care product. These results are consistent with the earlier report by Wang et al. since antiperspirant products were not reported in the survey by Lu. The occurrence of D5 in multiple environmental compartments and its presence as a significant ingredient in personal care products raises the question of the relative impact of personal care uses on the distribution of D5 in the environment (Brooke et al., 2011). Dermal absorption studies using stable isotopically labeled D5 (Plotzke et al., 2002; Reddy et al., 2007) or radiolabelled D5 (Jovanovic et al., 2008) demonstrate less than 0.2% of the applied D5 is absorbed through the skin. During these studies it was reported that the majority of the applied D5 had volatilized from the skin surface with >95% volatilization measured by carbon trapping and subsequent analysis. While these studies indicate dermal absorption is minimal, they may not accurately assess the fraction of D5 remaining on the skin surface over time across various product forms or when products are evaluated under more realistic consumer use conditions. Understanding the dominant pathway of D5 to the environment is important for estimating the relative contribution of personal care product uses of D5 to environmental compartment loadings. In a series of product application and wash off studies, the amount of D5 that would remain available to ‘go down the drain’ following typical use is measured. Since dermal penetration of D5 accounts for less than 0.2% of total mass balance, the fraction lost through evaporation can be calculated by difference. In this work, a variety of commercially available skin, hair and antiperspirant products were investigated in order to identify the types of personal care products that may contribute to overall aquatic D5 loadings.

America as well as to assess the impact of non-polar emollients on the volatilization of D5. Solid forms typically contain the highest levels of both emollients and D5 across this market category. Skin creams and lotions occur in a variety of formats including both hydroalcoholic and anhydrous solutions or various types of emulsions. To assess the broad spectrum of available skin care formats, three products were selected. First, an anhydrous solution (‘‘serum’’) comprising a very high percentage of D5 provided a format relatively low in other surfactants or emollients. In addition, an oil in water (‘‘water dominant’’) and a water in oil (‘‘oil dominant’’) emulsion format product were selected to assess if emulsion type or the relative magnitude of D5 content impacts the evaporative fate of D5 across this market category. For these products, the emulsion type was inferred based on declared ingredient order as per typical personal care product labeling convention (i.e. labeled in descending order of concentration). For hair products, four product formats were evaluated including three leave on styling (or dry conditioning products) plus a hair conditioner product intended for in shower application. Products comprising D5 within this market category are very diverse specifically in relative levels of D5 in product and overall composition, particularly with styling polymers that may impact on D5 volatilization. To span these considerations, a pump spray conditioning/ shine serum, pump spray styling form and gel-based styling product were selected. Furthermore, the in-shower, rinse-off hair conditioner was included to assess evaporative loss and direct discharge down the drain during in-shower use applications. Leave on serum styling or conditioning products typically contain the highest levels of D5 across this market category. 2.2. Analytical equipment and reagents Analytical measurements were performed using a ThermoFinnigan Trace-GC-Ultra/Polaris-Q MS system (West Palm Beach, FL USA) equipped with 4 mL gas sampling loop injector and a 30 m  0.25 mm  0.25 lm DB-FFAP column (Agilent, Wilmington, DE USA). Both D5 and [13C5]-decamethylcyclopentasiloxane (13C5-D5, 99.4% D5 purity by GC-FID) internal standard were provided by Dow Corning (Midland, MI, USA). 3. Methods 3.1. Product applications, wash off and sampling

2. Materials 2.1. Product selection A series of three evaporative fate studies were conducted (antiperspirants, skin care and hair care products). Commercially available personal care products containing D5 were used for all three studies. Products were selected to span the range of typical uses including both leave-on and in-shower applications. Products included underarm antiperspirants, several skin lotions, an inshower hair conditioner and post-shower hair styling products. These categories span the typical personal care product forms that may contain D5 and represent the major product categories where D5 was reported (Wang et al., 2009). Specific products were selected based on ingredient labels and a general understanding of product types. All products listed cyclopentasiloxane as an ingredient. Additional products used during the study included a sodium lauryl sulfate based liquid handsoap, cotton t-shirts and tap water. For underarm antiperspirant products, invisible solid and soft solid forms were selected to span the typical distribution of available product formats comprising this market category in North

These studies were designed to reflect as realistically as possible, typical conditions of use. Consequently, the approach selected throughout these studies included product application in accordance with typical consumer behavior and ‘wash off’ scenarios, designed to collect rinse water (including residual D5) at various collection times post-application. Washing times and methods were selected to mimic typical washing/bathing procedures that would ultimately lead to D5 contributions to wastewater. Treatments and treatment areas were randomly designated, with wash off and collection only occurring once at a designated single time point. All product application and subsequent water collection was performed by Hilltop Research at their Canada Research Center (Winnipeg, MB, Canada) in the early autumn. The Research Center managed all subject recruiting and informed consent policies, subject pre-screening, product application, washing and large volume sample collection. Between application and sampling, subjects were permitted to leave the facility and perform typical daily tasks with the exception of washing the product application area or removing provided clothing (t-shirt) that might come into direct contact with the application area. Wash off rinse water sub-sampling and analysis was performed by P-O Laboratories (Dorval,

B.P. Montemayor et al. / Chemosphere 93 (2013) 735–740

QC, Canada) who provided on-site staffing at the Research Center to subsample and manage the analytical samples. A series of quality control procedures were used across all studies to minimize laboratory background levels of D5. All subjects, clinical and analytical staff were prohibited from using products containing D5 for at least 24 h before the studies and throughout the study. Additional controls included the physical separation of product application and washing rooms with separate, dedicated staffing to minimize any cross-contamination. At each time point, study design included random panelists with no product application (placebo control) and laboratory basin/bottle blank controls to assess laboratory background levels. Sample collection bottles (Freund Container, Lisle, IL, USA) were pre-cleaned with detergent and equipped with foil-lined closures to minimize potential contamination or loss of D5 during sampling, transport and analysis. 3.2. Antiperspirant product testing Antiperspirant testing (application) was conducted in accordance with standard clinical investigation techniques for underarm antiperspirant products as defined by US FDA (2006). Commercially available test product (0.4 g) was applied to each axilla of the human subject consistent with standard clinical application techniques. Each subject was provided with a clean, long-sleeved, cotton t-shirt to wear until the specified wash off time. To mimic typical shower/ bathing habits, time points of 8 (±0.5) and 24 (±1) h were designated for sampling and rinse water collection. At the designated wash time, subjects returned, the t-shirt was removed from subject and the application area washed with two sequential wash–rinse steps each using 1 g of liquid handsoap with a total of 900 mL of water. The combined wash and rinse water (1800 mL total) was collected in a dedicated 2 L glass bottle that was immediately capped. A slight headspace was required in the collection bottle to accommodate the foaming of the handsoap:water mixture. Each product and sampling time was initially designed to include 20 subjects and three blank controls. Subjects who did not arrive within the scheduled sampling time window were removed from the study. To assess initial D5 application levels, 0.4 g of product was applied to a forearm area and washed off within 10 min of application. Physical transfer of D5 to clothing (provided t-shirt) of five subjects was assessed by cutting out (using a nominal 400 cm2 area template) from the t-shirt region in contact with the underarm and using a concentrated surfactant wash solution (liquid handsoap) to assess residual D5 that might have been transferred to the clothing material (and therefore available for contribution to waste water following laundering of clothing). Cut out shirt material was placed in a pre-cleaned 150 mL glass jar with 1 g of liquid handsoap. The liquid handsoap solution was preferred over commercial laundry detergent since prior analytical assessments confirmed the solubility and stability of D5 in the system and found the matrix free of background interferences. The jar was filled with water, capped and agitated by hand shaking for approximately 5 min to solubilize any product residue and then allowed to sit at room temperature for approximately 4 h to permit further solubilization and equilibration. Subsequently, a sub-sample of the tshirt wash water was collected for analysis. 3.3. Skin care product testing Clinical application techniques for skin care products have been defined for sunscreen products (US FDA, 1999). Consistent with these approaches, a 50 cm2 area was marked on a subject’s forearm and 100 mg of the test product was uniformly applied over the area by trained clinical staff. Similar to the antiperspirant testing protocol, each subject was provided with a clean, long-sleeved cotton tshirt to wear until the specified wash off time. Since hand and face

737

washing is typically performed at a higher frequency than shower and bathing, wash off time points included 4 (±0.5), 8 (±0.5) and 24 (±1) h. At the designated wash time, subjects returned, the t-shirt was removed from subject and the application area was washed with two sequential wash–rinse steps each using 1 g of liquid handsoap with a total of 900 mL of water. The total, combined wash and rinse water solution (1800 mL total) was collected in a dedicated 2 L collection bottle that was immediately capped. For each product and sampling time, a total of at least 20 subjects and three blank controls were collected. To assess initial D5 application levels, four panelists per product were randomly selected to have the application area washed off within 10 min of application. As with the antiperspirant protocol, physical transfer of D5 to clothing (provided t-shirt) was assessed by collecting t-shirt lower sleeves (nominal 500 cm2 area template) from the t-shirt region in contact with the forearm with sample preparation similar to that described under Section 3.2. 3.4. Hair care product testing Given the wide variety of hair types, lengths and personal hair care habits, testing hair products on human subjects introduces many uncontrollable variables. To minimize variability, in vitro human hair tress testing methods have been developed to assess product performance. These methods provide a more controlled environment to assess D5 evaporation from products applied to hair. Virgin brown European human hair tresses in a 2 or 4 g, flat configuration were obtained from De Meo Brothers Inc. (Passaic, NJ, USA) and pre-washed with a sodium lauryl sulfate solution and dried prior to use. For the two pump spray product forms, 4 g tresses were treated with two pumps applied to the middle of the tress. For the gel styling form, 0.4 g product was hand applied to 2 g tresses. To mimic shower/bathing habits, wash off time points of 8 (±0.5) and 24 (±1) h were selected. For each product/ sampling time point 15 separate tresses were prepared. At the designated wash time, hair tresses were washed manually while wearing nitrile gloves with two sequential wash–rinse steps each using 1 g of liquid handsoap with a total of 900 mL of water. The combined wash and rinse water (1800 mL total) was collected in a dedicated 2 L glass bottle that was immediately capped. Initial samples were prepared by washing the treated tress within 10 min of product application. The rinse-off conditioner product was applied to pre-wet, 2 g tresses at a rate of 1 g/tress and gently spread throughout the hair. Since this product is typically applied and rinsed off in a short period of time, product was applied and then washed off with the aid of water and liquid handsoap within 1–3 min of application using two sequential wash–rinse steps. Each wash–rinse step involved manually massaging 1 g of liquid handsoap into the tress and rinsing with a total of 900 mL of water, collecting the entire wash and rinse into a dedicated bottle. Both wash–rinse steps for a tress where combined immediately capped. The 1–3 min time window between application and rinse-off is consistent with typical use instructions for in shower hair conditioning products. Although detergent or shampoo is not typically used following a conditioner, the liquid handsoap was used to assure complete collection and stabilization of D5 in the collected solution. A dosing control for the rinse-off conditioner was prepared by directly adding 2 g of product into a 2 L bottle and adding 2 g of liquid handsoap and 1800 mL water. 3.5. Analytical sampling and measurement of D5 in wash off rinse water The concentration of D5 in all samples was determined using isotope dilution headspace GC/MS as described by Sparham et al.

738

B.P. Montemayor et al. / Chemosphere 93 (2013) 735–740

(2008). Briefly, 1800 mL samples of washwater containing 1 g L 1 of liquid handsoap used to wash the treated area and solubilize the D5 and other residual product were collected in pre-cleaned 2 L narrow-mouth glass bottles equipped with foil-lined closures (as described under Sections 3.2–3.4). Bottles were tightly capped, mixed and sub-sampled within 5 min of collection. Analytical samples (5.0 g) were rapidly weighed into a 20 mL headspace vial, followed by 20 lL of internal standard solution (0.3 lg 13C5-D5 in isopropanol) and immediately crimp-capped with foil-lined butyl rubber septa. Vials were frozen and transported on dry ice to the laboratory for headspace analysis. To avoid breakage due to ice expansion, vials were placed on a slight incline during freezing. The use of foil-lined closures on all containers reduced potential contamination or loss of D5 to polymeric surfaces. The same analytical approach was used for fabric wash samples, however sample volumes were only 150 mL. Prior to analysis, samples were removed from the freezer and thawed at room temperature for approximately 15 min. Samples were then placed on the autosampler where they were held at 25 °C for 60 min, then heated to 60 °C for 20 min and sampled by direct headspace sampling. Calibration standards were prepared in the same water:handsoap matrix with the same internal standard dosing. D5 was monitored at m/z 355 and 13C5-D5 internal standard at m/z 360, providing a limit of detection of 0.07 lg D5/ L and a limit of quantitation of 0.23 lg D5/L. Method precision was evaluated across multiple days showing a 1.8% pooled RSD at a nominal 10 lg L 1 concentration. The analytical method was optimized for low level detection and samples in the range from 0.25 to 1000 lg L 1 were quantified using a eight point quadratic calibration curve with 1/X weighting. Some high level samples show response above the preferred calibration region. To quantify these higher level samples, up to seven additional standards were added and a point-to-point calibration was constructed across the higher concentration range. 4. Results Regardless of product tested or sample collection time, all subject wash samples were prepared for analysis in a standard format of 1800 mL of tap water with 2 g of liquid handsoap. Since the volume is constant, the measured concentration of D5 in the water samples directly correlates with the residual amount of D5. Therefore, amounts of residual D5 in samples are reported using the directly measured wash water concentrations. 4.1. Antiperspirant product testing

Table 2 Skin products – measured concentration of D5 in wash water from treated and control subjects (lg L 1).

a

Time

Treated subject mean (±95% CI)

N

Control subject mean

Oil dominant lotion

Initial 4h 8h 24 h

6650 (±2860) 1.7 (±0.2) 0.84 (±0.1) 1.1 (±1.7)

3 24 22 21

3.1 0.85 0.75 0.78

0.03 0.01 0.02

Water dominant lotion

Initial 4h 8h 24 h

2480 (±2960) 4.2 (±0.9) 2.4 (±0.45) 2.3 (±0.39)

4 23 24 23

2.8 0.95 0.72 1.1

0.21 0.12 0.11

Anhydrous serum

Initial 4h 8h 24 h

24 800 (±19 500) 1.65 (±0.2) 0.92 (±0.38) 1.49 (±0.77)

4 24 24 24

4.1 0.85 0.60 1.0

0.01 0.01 0.01

Based on initial mean and upper 95% CI at time – see Section 5.

ual mean D5 concentrations from untreated control subjects ranged from 22 lg L 1 at initial sampling to 0.7 lg L 1 at the 24 h sampling time. These values correspond to an overall study background concentration. Higher levels during initial sampling are consistent with cross-contamination from treated subjects being washed concurrently in the facility. T-shirt wash samples concentrations ranged from 2 to 36 lg L 1, however the t-shirt water volume was only approximately 10% of the subject wash water volume, so transfer to fabric was determined to be negligible. 4.2. Skin care product testing Residual D5 concentrations measured in wash water associated with the skin product testing are summarized in Table 2. Across the study, residual mean D5 concentrations from untreated control subjects ranged from a maximum of 6.2 lg L 1 to as low as 0.30 lg L 1. Similar to the antiperspirant study, t-shirt wash samples also showed low residual levels of D5. T-shirt wash water concentrations ranged from a maximum of 37.3 lg L 1 for a single subject at the 4 h sampling to a typical value of 0.75 lg L 1 at 24 h. Since the water volume is approximately 10% of the wash volume, it was determined that the amount of D5 residuals transferred to t-shirt material is a small fraction of initial values. 4.3. Hair care product testing Residual D5 concentrations measured in wash water from the hair tress testing are summarized in Table 3. The residual

Residual D5 concentrations measured in wash water for underarm antiperspirant product testing are summarized in Table 1. Since typical bathing will use larger volumes of water, the concentration values in Table 1 will not directly measure wastewater concentrations, but provide a constant, relative comparison of D5 as a function of the time the product was on the subject. During this study, resid-

Table 3 Hair products – measured concentration of D5 in wash water from treated and control hair tresses (lg L 1).

Table 1 Antiperspirant products – measured concentration of D5 in wash water from treated and control subjects (lg L 1).

a

Residuala (%)

Product

Time

Spray serum

Initial 22 600 (±16 200) 8h 2.7 (±0.41) 24 h 0.79 (±0.13)

4 1.4 14 1.3 15 0.64

0.01 0.004

Initial 90 200 (±55 000) 8h 2.0 (±0.68) 24 h 0.68 (±0.08)

4 1.5 15 0.93 15 0.94

0.001 0.001

Initial 135 000 (±57 300) 8h 1.6 (±0.37) 24 h 0.60 (±0.08)

4 1.8 15 0.82 15 0.48

0.001 0.001

Dose 18 400 (±33 700) 3 min 5730 (±1530)

4 0.6 15 2.1

40

Product

Time

Treated subject mean (±95% CI)

N

Control subject mean

Residuala (%)

Spray styling

Invisible solid AP

Initial 8h 24 h

38 020 (±11 750) 153 (±167) 12 (±15)

22 21 21

22.1 8.7 0.7

0.8 0.07

Gel styling

Soft solid AP

Initial 8h 24 h

468 500 (±261 750) 137 (±110) 26 (±1)

22 22 19

11.6 20.0 0.75

0.05 0.01

Rinse-off conditioner

Based on initial mean and upper 95% CI at time – see Section 5.

a

Treated tress mean (± 95% CI)

N

Control tress mean

Residuala (%)

Product

Based on initial mean and upper 95% CI at time – see Section 5.

B.P. Montemayor et al. / Chemosphere 93 (2013) 735–740

concentrations of D5 in blank tress controls ranged from 0.38 to 3.3 lg L 1 across the entire study. 5. Discussion Based on previous mass balance studies (Plotzke et al., 2002; Reddy et al., 2007), dermal absorption is expected to account for less than 0.2% of the total applied D5. To assess the dominant pathways to the environment, we assume this fraction is a relatively constant, negligible portion of the total mass of D5. For each product form, the residual percentage of D5 on the skin can be estimated by taking the mean D5 concentration in wash water at time t and dividing it by the mean concentration of the initial loading or dose control samples. To add an additional degree of conservatism, the upper 95% confidence level estimate of the mean at time t can be used to calculate the residual percent D5 available to be washed down the drain. In all three studies the initial samples show a high level of D5 indicative of minimal post-application loss. It is also noted that the untreated control concentrations trend higher at the initial sampling and trend lower over time. This trend is consistent with the presence of high amounts of D5 in the application and washing laboratories during the initial sampling. Since control and treated subjects were randomly interspersed throughout the product application and washing, all subjects were in close proximity with high D5 products. 5.1. Antiperspirant product testing The underarm product testing provides a scenario where products contain high concentrations of D5, the application area is typically covered by clothing throughout the day, resulting in minimal air circulation around the site of product application. This scenario would likely lead to the slowest, most conservative estimates of evaporation rates and potentially the highest amount of residual D5 remaining on skin. In the market survey (Wang et al., 2009), D5 concentrations in antiperspirant products were typically in the range of 20–45% D5. The initial sampling values in Table 1 confirm that high levels of D5 occur in these products. Considering the 1.8 L sample volume and nominal mean values of 38 mg L 1 for the invisible solid product, the total D5 present in the initial forearm wash is 68 mg or approximately 17% of the applied product mass. Considering the soft solid form, a similar calculation would indicate >100% of the applied product mass. For antiperspirant products, typical consumer habits are to apply product either post-bathing or at a time not immediately before bathing. To assess the realistic fraction of D5 available to go down the drain, the personal care products exposure assessment data of Loretz et al. (2006) indicates that the US consumer use frequency of shampoo, body wash and antiperspirant products averages between 1.1 and 1.4 uses per day with a mean antiperspirant use of 0.59–1.0 g/use day. Given this context, the product application rate of 0.4 g/underarm represents typical use and the 8 h time point between product application and wash-off is considered conservative. With typical 1–1.5 bathing events per day, the 24 h time point for underarm product wash off is likely most relevant. Considering the data in Table 1, the residual percent D5 at the 8 h time point relative to the initial sampling is 0.8% for the invisible solid form and 0.05% for the soft solid. Subsequent sampling at 24 h shows an additional order of magnitude decrease in residual D5. Using the same calculation approach, residual D5 levels are only 0.07% for the invisible solid and 0.01% for the soft solid. At 24 h, residual D5 levels are comparable to the laboratory background controls. While the product forms appear to contain differ-

739

ent levels of D5, the absolute mass of residual D5 available to be washed down the drain at 8 and 24 h appears to be independent of form. These results indicate an order of magnitude lower residual percentage of D5 on the skin after 24 h than previous reported from the in vitro evaluation of an antiperspirant formulation (Jovanovic et al., 2008). 5.2. Skin care product testing Skin products were applied to levels consistent with use volumes designated for sunscreen product testing. This level is consistent with the habits and practices data of Loretz et al. (2005,2006) for face cream or liquid foundation makeup where average product use ranged from 0.54 to 1.2 g per application over a typical face and neck area. When D5 was found in body lotions (Wang et al., 2009), concentrations were below 4%. Since one of the products tested in this study is an anhydrous D5 serum product, it is likely that survey did not include some specialty facial moisturizers or cosmetics that typically contain higher concentrations of D5. Initial D5 wash water concentrations in Table 2 correlate with the expected concentrations of D5 in the product showing anhydrous > oil dominant > water dominant products. The low number of replicates used for the initial sampling leads to wide confidence intervals on the mean; however the mean is clearly higher immediately post-application than at the 4, 8 or 24 h time points. Table 2 summarizes the level of D5 for all three products at the various samplings. Calculating the residual percent of D5, we see that after 4 h wear time, at most 0.21% of the initially measured concentration of D5 is available to be washed down the drain. This fraction appears highest for the aqueous dominant emulsion product (0.21%) and an order of magnitude lower for both the oil dominant emulsion (0.03%) and anhydrous (0.01%) products. Wash water concentrations at the 4 h time point are similar across all three product forms and the higher fractional residue is primarily attributed to the lower dosing concentration. Levels measured after 8 or 24 h on the forearm trend slightly lower then at 4 h and are comparable to the background levels observed with the untreated control subjects. 5.3. Hair care product testing Hair products applied to tresses included both post-shower applied treatment/styling forms and an in-shower conditioner with D5 as a listed ingredient. Based on the product survey (Wang et al., 2009), the post-shower styling and conditioning products are expected to be the dominant Hair products containing D5. While not an exhaustive market survey, no shampoo or in-shower conditioning products containing D5 were reported by Wang et al. and only one in-shower conditioner product was identified based on ingredient labels for use in this study. The concentration of D5 found in styling products were typically in the range of 0.02– 1.7%. Based on ingredient labels, it appears some serum products may contain much higher concentrations since D5 is the dominant ingredient. The selection of only the 8 and 24 h washing times was based on habits and practices data that found average shampoo use of 1.1 applications/day and average hairspray use to be 1.5 uses/ day (Loretz et al., 2006). Within 8 h the residual percent D5 associated with all three post-shower products decreased to levels comparable to background controls. Considering only the styling and serum products, the residual D5 levels at 8 h is less than 0.02% of the initial applied amount. This indicates that evaporation of D5 from the hair tresses is rapid relative to the typical washing frequency. For the in-shower conditioner product, direct discharge to wastewater is expected to be significant. This is confirmed when comparing the tresses washed within 1–3 min of product application with

740

B.P. Montemayor et al. / Chemosphere 93 (2013) 735–740

the dosing control. Using the 3 min sampling relative to the dose control, the residual percent D5 is 40% of the initially applied dose. This indicates that for in-shower use some evaporation occurs; however a significant fraction of D5 in product will still be available for input into waste water. Given the low replication in the loading control and the corresponding high relative variance, this 40% residual percentage is only an approximation. Based on these results, it is reasonable to conclude that, for product used during bathing, a significant fraction of the D5 present in the product will be discharged to the drain with a relatively small fraction evaporating during product use.

6. Conclusions Evaporation of D5 during typical consumer use of personal care products is an important consideration in determining environmental exposures. Considering typical washing habits, only a very small fraction (significantly less than 1%) of the D5 used in products applied outside of a shower or bath is available to wash down the drain. For these leave-on post-shower applications, direct release to the atmosphere is the dominant pathway. For products used while showering or bathing, approximately 40% or more of the D5 is expected to be washed down the drain. While evaporative loss is observed in this use scenario, the relative distribution between evaporation and drain discharge are comparable.

Acknowledgements These studies were commissioned by the Canadian Cosmetic, Toiletry and Fragrance Association (CCTFA) on behalf of the global personal care products industry with business interests in cVMSs through an independent industry-sponsored research grant. A special thanks to the generous contributions of collective members of the CCTFA and their corresponding trade association partners in Europe and the United States, the Silicone Environmental, Health and Safety Council of North America (SEHSC), representing the producers of these raw materials, and the Allied Beauty Association (ABA), representing professional brands with interests in these ingredients. Also, thank you to the CCTFA Siloxanes Technical Working Group (TWG) for their input into the design of these studies; and to The Procter and Gamble Company, Unilever and Dow Corning Ltd., for their technical contributions in support of these studies. Lastly, thank you to our laboratory partners, Hill-Top Research Inc., and PO Laboratories Inc., for their professional support and contributions which made these studies possible.

References Atkinson, R., 1991. Kinetics of the gas-phase reactions of a series of organosilicone compounds with OH and NO3 radicals and O3 at 297 K. Environ. Sci. Technol. 25 (5), 863–866. Brooke, D.N., Crookes, M.J., Gray, D., Robertson, S., 2011. Environmental Risk Assessment Report: Decamethylcyclopentasiloxane. Environment Agency of England and Wales, Bristol. Environment Canada, Health Canada, 2008. Screening assessment for the, challenge decamethylcyclopentasiloxane (D5). 2011. Genualdi, S., Harner, T., Cheng, Y., MacLeod, M., Hansen, K.M., van Egmond, R., Shoeib, M., Lee, S.C., 2011. Global distribution of linear and cyclic volatile methyl siloxanes in air. Environ. Sci. Technol. 45 (8), 3349–3354. Horii, Y., Kannan, K., 2008. Survey of organosilicone compounds, including cyclic and linear siloxanes, in personal-care and household products. Arch. Environ. Contam. Toxicol. 55 (4), 701–710. Jovanovic, M.L., McMahon, J.M., McNett, D.A., Tobin, J.M., Plotzke, K.P., 2008. In vitro and in vivo percutaneous absorption of C-14 octamethylcyclotetrasiloxane and C-14-decamethylecyclopentasiloxane (C-14-D-5). Regul. Toxicol. Pharmacol. 50, 239–248. Loretz, L.J., Api, A.M., Barraj, L.M., Burdick, J., Dressler, W.E., Gettings, S.D., Han Hsu, H., Pan, Y.H.L., Re, T.A., Renskers, K.J., Rothenstein, A., Scrafford, C.G., Sewall, C., 2005. Exposure data for cosmetic products: lipstick, body lotion, and face cream. Food Chem. Toxicol. 43 (2), 279–291. Loretz, L.J., Api, A.M., Barraj, L., Burdick, J., Davis, D.A., Dressler, W., Gilberti, E., Jarrett, G., Mann, S., Laurie Pan, Y.H., Re, T., Renskers, K., Scrafford, C., Vater, S., 2006. Exposure data for personal care products: hairspray, spray perfume, liquid foundation, shampoo, body wash, and solid antiperspirant. Food Chem. Toxicol. 44 (12), 2008–2018. Lu, Y., Yuan, T., Yun, S.H., Wang, W., Wu, Q., Kannan, K., 2010. Occurrence of cyclic and linear siloxanes in indoor dust from China, and implications for human exposures. Environ. Sci. Technol. 44 (16), 6081–6087. Lu, Y., Yuan, T., Wang, W., Kannan, K., 2011. Concentrations and assessment of exposure to siloxanes and synthetic musks in personal care products from China. Environ. Pollut. 159 (12), 3522–3528. McLachlan, M.S., Kierkegaard, A., Hansen, K.M., van Egmond, R., Christensen, J.H., SkjÃth, C.A., 2010. Concentrations and fate of decamethylcyclopentasiloxane (D5) in the atmosphere. Environ. Sci. Technol. 44 (14), 5365–5370. Navea, J.G., Young, M.A., Xu, S., Grassian, V.H., Stanier, C.O., 2011. The atmospheric lifetimes and concentrations of cyclic methylsiloxanes octamethylcyclotetrasiloxane (D4) and decamethylcyclopentasiloxane (D5) and the influence of heterogeneous uptake. Atmos. Environ. 45 (18), 3181– 3191. Plotzke, K.P., Utell, M.J., Looney, J.R., 2002. Absorption, distribution and elimination of 13-C-D5 in humans after dermal administration. US EPA document. 84030000008. Reddy, M.B., Looney, R.J., Utell, M.J., Plotzke, K.P., Andersen, M.E., 2007. Modelling of human dermal absorption of octamethyltetracyclosiloxane (D-4) and decamethylcyclopentasiloxane (D-5). Toxicol. Sci. 99 (2), 422–431. Sparham, C., van Egmond, R., O’Connor, S., Hastie, C., Whelan, M., Kanda, R., Franklin, O., 2008. Determination of decamethylcyclopentasiloxane in river water and final effluent by headspace gas chromatography/mass spectrometry. J. Chromatogr. A 1212 (1-2), 124–129. Sparham, C., van Egmond, R., Hastie, C., O’Connor, S., Gore, D., Chowdhury, N., 2011. Determination of decamethylcyclopentasiloxane in river and estuarine sediments in the UK. J. Chromatogr. A 1218 (6), 817–823. US FDA, 1999. Sunscreen Drug Products for Over-the-Counter Human Use-General Testing Procedures. US Food and Drug Administration. May 21, 1999, 21CFR352.72(e). US FDA, 2006. Guidelines for Effectiveness Testing of Antiperspirant Drug Products. US Food and Drug Administration. April 1, 2006, 21CFR350.60. Wang, R., Moody, R.P., Koniecki, D., Zhu, J., 2009. Low molecular weight cyclic volatile methylsiloxanes in cosmetic products sold in Canada: implication for dermal exposure. Environ. Int. 35 (6), 900–904.