| Front page | | Contents | | Previous | | Next |

Environmental and Health Assessment of Substances in Household Detergents and Cosmetic Detergent Products

4. Nonionic surfactants

Nonionic surfactants are surface-active compounds with hydrophobic and hydrophilic moieties. These surfactants do not ionize in aqueous solutions. Commercial nonionic surfactants are normally a mixture of homologuos structures composed of alkyl chains that differ in the number of carbons and with hydrophilic moieties that differ in the number of ethylene oxide (ethoxylate, EO), propylene oxide (propoxylate, PO) and butylene oxide (butoxylate, BO) units. Nonionic surfactants are widely used in consumer products like, e.g., laundry detergents, cleaning and dishwashing agents, and personal care products. Nonionic surfactants are also widely used in cleaning agents formulated for the industrial and institutional sector. By volume, the most important nonionic surfactants are included in the very versatile group of alcohol ethoxylates and alcohol alkoxylates.

4.1 Alcohol ethoxylates and alcohol alkoxylates

Alcohol ethoxylates (AE) are nonionic surfactants composed of a hydrophobic alkyl chain (fatty alcohol) which is combined with a number of ethoxylate, or ethylene oxide, units via an ether linkage. Alcohol alkoxylates (AA) normally contain both ethylene oxide (EO) and propylene oxide (PO) in their hydrophilic moiety, whereas butylene oxide (BO) is less frequently used. The abbreviation AA has been used to designate nonionic surfactants with a hydrophilic part containing PO (or BO), frequently in combination with EO. AE are used in many types of consumer and industrial products like, e.g., laundry detergents, all-purpose cleaning agents, dishwashing agents, emulsifiers, and wetting agents. AA are used as weakly foaming and foam-mitigating surfactants in household cleaning agents, dishwashing agents and cleaning agents designed for the food industry (Bertleff et al. 1997). Other applications of AA include textile lubricants, agricultural chemicals, and rinse aid formulations.

The nonionic surfactants described in this section include several chemical structures of which a few representative structures are given below.

Lineary primary AE, C₁₃ EO7:
CH₃-(CH₂)₁₂-O-(CH₂-CH₂-O-)₇H
Iso-C₁₃ branched primary AE, EO7:

Linear primary AE, C₁₃ EO10, end-capped with n-butylether
CH₃-(CH₂)₁₂-O-(CH₂-CH₂-O-)₁₀-CH₂-CH₂-CH₂-CH₂-OH

Linear primary AA, EO5, PO4:

The hydrophobic fatty alcohol usually contains 12-15 carbon atoms, but chain lengths of C_9-11 are also used.

4.1.1 Occurrence in the environment

During the 1980s non-quantitative methods for detection of AE were used together with analyses for nonionic surfactants (bismuth iodide active substances, BiAS) to determine the presence of AE in e.g. effluents from wastewater treatment plants. Today, the efforts are directed towards the development of new methods for specific determination of AE at low concentrations in environmental samples. Effluent concentrations of AE in wastewater treatment plants were in the order of 0.006-0.02 mg/l for AE concentrations of 0.19-0.91 mg/l in the influent. A higher AE concentration of 3.4 mg/l in the influent resulted in an effluent concentration of 0.04 mg/l (Holt et al. 1992). A recent environmental monitoring showed that the effluent concentrations of AE from municipal sewage treatment plants in the Netherlands varied between 0.0022 and 0.013 mg/l with an average value of 0.0062 mg/l (Matthijs et al. 1999).

The presence of AE in the aquatic environment has been reported for a Japanese river. The concentration of AE in the river water was below the detection limit of 0.005 mg/l, whereas the concentration in the sediment ranged from 0 to 1.0 mg/kg. A concentration of 0.004 mg/l for C_14-15 AE was observed in Ohio River, USA (Holt et al. 1992).

4.1.2 Environmental fate

Biodegradation pathways

Three different mechanisms have been proposed for the biological degradation of AE under aerobic conditions (Marcomini et al. 2000a, 2000b).

Recent studies have elucidated the relations between the biodegradation mechanisms and the structure of the AE (Marcomini et al. 2000a, 2000b). The formation of PEG was observed only for a linear AE and an oxo-AE (composed of linear AE and monobranched AE with short 2-alkyl chains, i.e. 2-methyl-, 2-ethyl-, 2-propyl-, and 2-butyl-), whereas only carboxylated AE (with the carboxylic group at the polyethoxylic chain) were detected during biodegradation of a multibranched AE. The absence of carboxylated AE in the experiments with the linear and the monobranched (2-alkyl branched) oxo-AE indicates that the central scission (mechanism 1) was the primary mechanism for the biodegradation of linear and most monobranched AE in the examined commercial mixtures, whereas the multibranched AE was degraded via w -oxidation of the polyethoxylic chain (mechanism 3) (Marcomini et al. 2000a). Biodegradation of an oxo-2-butyl-substituted AE only resulted in carboxylated AE (mainly metabolites with the carboxylic group at the alkyl chain) suggesting that w -oxidation of the alkyl chain was the primary mechanism (mechanism 2). The results obtained with the 2-butyl-substituted AE show that a shift from the central scission to the w ,b -oxidation is introduced when the length of the 2-alkyl branch exceeds three carbon atoms (Marcomini et al. 2000b).

Far less is known about the biodegradation of AA and of end-capped AE. AA containing butoxylate (BO) or propoxylate (PO) groups in their hydrophilic moiety are degraded via cleavages of the hydrophilic chain, which may be either non-oxidative or oxidative like the degradation of PEG. A secondary carbon atom in the hydrophilic moiety, e.g. in PO groups, inhibits the oxidative route (Balson and Felix 1995). End-capped AE are degraded by a combination of w -oxidation of the hydrophilic chain and central hydrophobe-hydrophile scission. The w -oxidation is inhibited by the presence of PO in the hydrophilic chain, whereas the extent of central scission is determined by the degree of 2-alkyl branching (Balson and Felix 1995). The findings in the above-mentioned studies with 2-butyl-substituted AE (Marcomini et al. 2000b) further illustrate the effect of the length of the 2-alkyl substituent.

The anaerobic biodegradation of linear AE is apparently initiated by a stepwise release of C₂ units as acetaldehyde to form the corresponding shortened AE and, eventually, a fatty acid (Wagener and Schink 1988). This pathway was recently confirmed in anaerobic assays with a linear pure C₁₂ AE (with 8 EO) and a linear technical C₁₂ AE (with an average of 9 EO), as the first identifiable metabolites were shortened AE and subsequent metabolites included dodecanoic acid and acetic acid. No PEG was observed during the degradation of linear AE, which indicates that central scission of the AE molecule was not the biodegradation mechanism under the applied anaerobic conditions (Huber et al. 2000).

Effects of structure of AE on biodegradability

The biodegradability of the AE is relatively unaffected by the alkyl carbon chain length and the number of EO units but highly affected by the molecular structure of the hydrophobic chain (Dorn et al. 1993). The linear AE are normally easily degraded under aerobic conditions. Only small differences are seen in the time needed for ultimate degradation of linear AE with different alkyl chain lengths. AE with a typical alkyl chain (e.g., C₁₂ to C₁₅) will normally reach more than 60% degradation in standardized tests for ready biodegradability. The length of the EO chain determines the hydrophobicity of the AE and, hence, influences the biodegradability in terms of the bioavailability. Longer EO chains decrease the bioavailability of the AE due to increased hydrophilicity and molecular size, which limits the transport of the molecule through the cell wall (Balson and Felix 1995). For AE containing more than 20 EO units, a reduced rate of biodegradation has been observed (Scharer et al. 1979; Holt et al. 1992).

The biodegradation of branched AE tends to be slower than biodegradation of linear AE. One important observation that may explain this fact is that the b -oxidation is hindered by the branching of the alkyl chain (Holt et al. 1992; Balson and Felix 1995). Furthermore, branching at the C-atom forming the internal ether linkage may hinder the hydrophobe-hydrophile scission (Balson and Felix 1995). The biodegradability of AE depends on degree and structure of the branching. The general trend is that the biodegradation decreases considerably with an increasing branching of the carbon chain (Kaluza and Taeger 1996). A highly branched C₁₃ AE, prepared from 2-propyl-C₁₀ and 2-pentyl-C₈ with 46% branching, was not readily biodegradable in the DOC die-away screening test as only 50% of the initial DOC was removed during 28 days (Kaluza and Taeger 1996). The structure of the backbone in the carbon chain also affects the biodegradability. Swisher (1987) found that one single internal methyl group had no effect on the biodegradation compared to the entirely linear AE, whereas two methyl groups decreased the degradation rate markedly, especially if the methyl groups were located at the same carbon resulting in a quaternary structure. The rate of biodegradation of monobranched AE is strongly influenced by the length of the side chain. Although the 60% pass level was fulfilled in the CO₂ evolution test (but not the 10-day window), the degradation of an oxo-2-butyl-substituted AE occurred more slowly than the degradation of an oxo-AE blend containing 2-methyl-, 2-ethyl-, 2-propyl-, and 2-butyl side chains. The degradation of the 2-butyl-substituted AE showed a time profile similar to that of a multibranched AE (Marcomini et al. 2000b). Kaluza and Taeger (1996) compared the biodegradability of branched AE based on different carbon chains (all with 7-8 EO units). They found that an iso-C₁₃ AE based on propylene tetramer (four internal methyl groups) did not pass a test for ready biodegradability, whereas an iso-C₁₃ based on butylene trimer (three internal methyl groups) did. The ultimate degradation of iso-C₁₀ based on propylene trimer (three internal methyl groups) also complied with the criteria for ready biodegradability. Kravetz et al. (1991) studied the degradation of a C_11-15 AE based on propylene and containing four internal methyl groups as well as a C_10-14 AE containing three internal groups (both with 7 EO units). The structures of the two substances were complex as they both contained a quaternary carbon. None of the branched AE passed the criteria for ready biodegradability and no difference in the degradation rates for the two substances was observed.

Aerobic biodegradability

Linear C_12-18 AE, containing 5-14 EO units, are ultimately degraded under aerobic conditions. The degradation rate of AE containing more than 20 EO units is slower, although an extensive primary degradation may take place for AE containing up to 50 EO units (Birch 1984). Only a few studies report the fate of AE in wastewater treatment plants. Average concentrations of 0.33 mg/l (0.19-0.47 mg/l) in the influent and 0.009 mg/l (0.006-0.012 mg/l) in the effluent of C_14-15 EO7 indicate a removal of 97-98% of the AE during wastewater treatment (Holt et al. 1992). Data on the ultimate aerobic biodegradability of linear AE are shown in Table 4.1. As described previously, the aerobic biodegradation of branched AE depends on the structure of the hydrophobic carbon chain. In general the biodegradability decreases with increasing branching of the alkyl chain, but also the number of internal methyl groups and the presence of quaternary carbon atoms affect the biodegradability of AE. Normally, AE containing a quaternary carbon atom are not readily biodegradable (Table 4.2).

Table 4.1
Ultimate aerobic biodegradability of linear AE.

Table 4.2
Ultimate aerobic biodegradability of branched AE.

The biodegradability of AA generally decreases with an increasing number of PO units in the hydrophilic part. The results summarized in Table 4.3 confirm this trend as, e.g., the C_12-18 AA containing 6 PO units did not pass the level required for ready biodegradability whereas the same alcohol containing 2 PO units attained 83% ThOD in the closed bottle test (Schöberl et al. 1988). The general trend for linear AA is that, apparently, there is a limit of 6-7 PO units in order to qualify for primary biodegradation (Balson and Felix 1995). However, increased branching of the carbon chain determines the construction of the hydrophilic part of the surfactant, as fewer PO units can be tolerated in branched AA in order to comply with the requirement for primary biodegradability (Naylor et al. 1988). Naylor et al. (1988) showed that a primary biodegradation > 80% was achieved for the most linear AA (20% branching) containing up to 3.5 PO units, an AA with 1 internal methyl group containing up to 2.0 PO units, and an AA with 2 internal methyl groups containing up to 0.4 PO units.

The removal of alcohol propoxylates in German wastewater treatment plants has been reported to be in the range of 73-81% (Holt et al. 1992). If the AA is terminated with PO units the degradation is highly influenced by the branching because the w -hydrophile oxidation is inhibited by the presence of PO. In this case, the level of branching determines the biodegradation which proceeds by the hydrophobe-hydrophile scission (Balson and Felix 1995). This was confirmed by a study showing that the primary biodegradability of an AA containing 6 EO units and 6.5 PO units was 97% for 20% 2-alkyl branching and 10% for 100% 2-alkyl branching (Balson and Felix 1995). Balson and Felix (1995) showed a primary degradation of 83-97% for a C_9-11-AE capped with an alkyl group and 80-99% primary degradation of a C_9-11-AE capped with an aryl group. Data describing the ultimate biodegradability of end-capped AE are sparse. A C_12-14 EO9 and a C_12-18 EO10, both end-capped with n-butylether, were confirmed to be readily biodegradable (Table 4.3).

Table 4.3
Ultimate aerobic biodegradability of end-capped AE and AA.

Anaerobic biodegradability

Most of the relatively few studies of the anaerobic biodegradability of AE have been performed with linear AE. Anaerobic biodegradation tests have been performed with various inocula like, e.g., anaerobically digested sludge (Steber and Wierich 1987; Salanitro and Diaz 1995; Madsen et al. 1995; 1996a) and anoxic sediments (Wagener and Schink 1987; Madsen et al. 1995, 1996a; Federle and Schwab 1992). Anaerobic biodegradability tests with diluted digested sludge have either been performed by use of screening methods (e.g., ECETOC 1988; ISO 1995) or by use of ¹⁴C-labelled model compounds (e.g., Steber and Wierich 1987). Since the concentration of surfactant in the screening test may inhibit its degradation by anaerobic bacteria, the results from studies using ¹⁴C-labelled compounds are generally considered to be of higher value. The results indicate that linear AE are normally mineralized in anaerobically digested sludge. The mineralization observed in experiments with ¹⁴C-labelled surfactants suggests that almost complete degradation of linear AE may be expected in anaerobic digesters and that the lower mineralization observed in the screening test was caused by inhibition (Table 4.4). AE end-capped with butylether were either partially mineralized or not degraded in the ISO 11734 screening test (Table 4.4; Appendix). Linear AE were also degraded in anoxic sediments, where a lower mineralization was observed at 22° C compared to the mineralization at higher temperatures (Table 4.5).

Table 4.4
Ultimate anaerobic biodegradability of AE in digested sludge.

Table 4.5
Ultimate anaerobic biodegradability of AE in sediments.

Bioaccumulation

Bioaccumulation of AE in aquatic organisms has been determined only for fish. The majority of the very few data is based on studies with ¹⁴C-labelled compounds that do not allow the distinction between the parent compound and metabolites. Because AE are metabolized in aquatic organisms, the bioconcentration factor for the parent compound may well be overestimated in experiments in which ¹⁴C-labelled model surfactants are used. By use of ¹⁴C-labelled surfactants, whole body concentration ratios have been estimated for four different AE in fish (Table 4.6).

Table 4.6
Whole body BCF values of AE in fish.

* BCF values based on radioactivity measurements.

Tolls (1998) combined ¹⁴C-techniques and chemical analysis and showed that the parent AE (C₁₃ EO8) was rapidly eliminated by transformation into metabolites, which were eliminated at a slower rate. The bioconcentration factors for C₁₂ EO8 and C₁₃ EO8 were 12.7 and 29.5-55.0, respectively, when the AE were monitored by chemical analysis. The BCF values for C₁₃ EO4 and C₁₄ EO4 were 232.5 and 237.0, respectively. The influence of the hydrophobe chain length was illustrated by BCF values of 56.7 to 135.2 for C₁₄ EO8 and 387.5 for C₁₆ EO8. AE with a relatively high number of EO units, i.e. C₁₄ EO11 and C₁₄ EO14 did not bioaccumulate in fish as indicated by the BCF values of < 5 and 15.8 (Tolls 1998; Table 4.6). The data in Table 4.6 indicate that the more hydrophobic AE (e.g. C₁₃ EO4, C₁₄ EO4, and C₁₆ EO8) have a moderate bioaccumulation potential.

In the study by Tolls (1998) the BCF values ranged from < 5 to 387.5, whereas the uptake rates (k1) varied from 330 to 1660 (l x kg x d^-1) and the elimination rates (k2) varied from 3.3 to 59 (d^-1). According to the guideline on bioaccumulation studies in fish (OECD 305) the time to 95% steady state conditions can be estimated by the equation t₉₅ = 3.0/k2. Using this equation, the t₉₅ for the AE investigated by Tolls (1998) range from 1.2 to 22 hours. The results obtained by Tolls (1998) indicate that the time to steady state and the BCF for AE increase with decreasing length of the ethoxylate chain (e.g., t₉₅ for C₁₃ EO8 = 2.4 h and BCF = 30-55, and t₉₅ for C₁₃ EO4 = 17.1 h and BCF = 233).

The achievement of steady state conditions for AE (C_9-11 EO6, C_12-13 EO6.5, and C_14-15 EO7) after a relatively short exposure period has also been illustrated by Lizotte et al. (1999) who observed that ‘steady state’ mortality occurred within 240 hours of exposure in the higher exposure concentrations. At the lower exposure concentrations with C_9-11 EO6 and C_14-15 EO7, the mortality continued, however, throughout the treatment period. For an illustration of the time needed for achievement of maximum toxicity a comparison of toxicity data for AE obtained in short-term and long-term studies is presented in Table 4.7.

Table 4.7
Effects of different exposure periods on the toxicity of AE to fish.

* Parentheses indicate 95% confidence limits.

The data in Table 4.7 indicate that an increase of the exposure period did not lead to lower effect concentrations (LC50) and that maximum toxicity of the AE was achieved after a relatively short exposure period. However, the AE examined by Lizotte et al. (1999) did not include relatively hydrophobic types like, e.g., C₁₃ EO4, C₁₄ EO4, and C₁₆ EO8, for which BCF values above 100 have been determined (Table 4.6).

4.1.3 Effects on the aquatic environment

Many studies have been performed to determine the toxic effects of AE towards aquatic organisms. Extrapolation from laboratory toxicity tests to the environment is obviously not easy for readily biodegradable surfactants, because biodegradation of the compounds in the sewers and in wastewater treatment plants is expected to alter the composition of isomers and homologues. The toxicity of a linear C_12-15 EO9 and a branched C_11-15 EO7 was investigated after treatment in a continuously activated sludge reactor (Kravetz et al. 1991). Both AE were degraded to products that were not acutely toxic. A higher chronic toxicity was observed for the effluent from the branched AE than from the linear AE. The degradation products were not identified but it was believed that the EO-chain was shortened and, hence, more toxic AE metabolites would have been produced. Garcia et al. (1996) investigated whether the toxicity of AE (C₁₂) was affected by a broad-range or a narrow-range EO distribution. The AE with the narrow-range distribution were less toxic than were the AE with the broad-range EO distribution when the surfactants contained more than 8-10 EO, whereas no differences were observed for a lower degree of ethoxylation. The AE with narrow-range and broad-range EO distribution differed by the presence of a lower amount of free fatty alcohols in the AE with the narrow-range EO distribution. The following paragraphs describe the toxicity of AE and AA towards algae, invertebrates, and fish.

Algae

Algae constitute the group of aquatic organisms which appears to be the most sensitive to AE. The acute toxicity of linear and branched AE to algae is in the same range with EC50 values from 0.05 to 50 mg/l. Besides the differences in chemical structure, the reason for the variation may be due to different test conditions and different test species. For the linear types, the toxicity increases with increasing hydrophobe chain length (comparison of C₁₃ EO7-8 and C₁₅ EO7-8, Table 4.8) and decreasing EO chain length (comparison of C_12-14 with 4-13 EO, Table 4.8). The toxicity of AE to algae tends to decrease with increasing degree of branching (Table 4.9). Based on the low EC50 values (£ 1 mg/l), the linear AE of C_12-15 EO6-8 are considered as very toxic to algae. When the degree of branching is low (£ 25%), the branched types are also considered very toxic to algae. A C_12-14 EO9 end-capped with an n-butyl-group was very toxic to a non-specified alga as the EC50 was 0.3 mg/l (Schöberl et al. 1988).

The effect of the carbon chain length and structure on the toxicity to algae was examined for two AA containing 6 EO and 3 PO-groups (Bertleff et al. 1997). It was observed that the toxicity increased with an increasing carbon chain length and that branching of the carbon chain reduced the toxicity (Table 4.10).

Table 4.8
Effects of linear AE to algae.

* Parenthesis indicate 95% confidence interval.

Table 4.9
Effects of branched AE to algae.

Table 4.10
Effects of AA and end-capped AE to algae.

Invertebrates

The acute toxicity of AE to invertebrates varies with EC50 values from 0.1 mg/l to more than 100 mg/l for the linear types and from 0.5 mg/l to 50 mg/l for the branched types. The toxicity is species specific and may vary between 0.29 mg/l to 270 mg/l for the same linear AE (Lewis and Suprenant 1983). The most commonly used invertebrates for testing are Daphnia magna and Daphnia pulex, and they are also among the most sensitive invertebrates to AE. Apparently, the toxicity of AE to invertebrates was not related to hydrophobicity as it is the case for algae. Some AE are very toxic to invertebrates, i.e., linear AE of C_12-15 EO1-8 and branched AE with a low degree of branching, i.e. < 10-25%. Branching of the alkyl chain reduces the toxicity of AE to invertebrates as also observed for algae. This effect of branching is evident by comparison of the toxicity of the linear C₁₃ AE and C₁₃ AE containing more or less branched alkyl chains (Tables 4.11-4.12). The toxicity of commercial AE was recently determined by using a sperm cell toxicity test with the sea urchin Paracentrotus lividus. The EC50 obtained in this test have proven to be closer to chronic data for all the tested AE. Whereas a fully linear C₁₂ AE exhibited an EC50 of 0.96 mg/l in the sperm cell toxicity test, a fully monobranched C₁₂ AE exhibited an EC50 of 4.0 mg/l. In this case, the alkyl side chain reduced the toxicity of the C₁₂ AE by approximately a factor of 4 (Marcomini et al. 2000c). A C_12-14 EO9 end-capped with an n-butyl-group was toxic to daphnids as the acute and chronic EC50 values were 1-2 mg/l and 0.3 mg/l, respectively (Schöberl et al. 1988). Schöberl et al. (1988) report that an AA with 2-5 EO and 4 PO was toxic to daphnids as the EC50 ranged between 2.4 and 6.0 mg/l.

Table 4.11
Effects of linear AE to invertebrates.

* Parentheses indicate 95% confidence intervals.

Table 4.12
Effects of branched AE to Daphnia magna.

* Parentheses indicate 95% confidence intervals.

Fish

The acute toxicity of AE to fish varies with LC50 values from 0.4 mg/l to more than 100 mg/l for the linear types and from 0.25 mg/l to 40 mg/l for the branched AE (Tables 4.13-4.14). For linear AE the toxicity increases with decreasing EO units (comparison within C_12-15 EO7-9 and within C_14-15 EO 7-11). C_12-15 AE containing 7-11 EO groups are considered to be very toxic to fish (EC/LC50 £ 1 mg/l). There are only few data on the toxicity of branched AE to fish and only oxo-C_9-15 EO2-10 is considered very toxic. A C_12-14 EO9 end-capped with an n-butyl-group was toxic to fish (species not specified) as the EC50 was 0.5-4.6 mg/l (Schöberl et al. 1988). Schöberl et al. (1988) report that an AA with 2-5 EO and 4 PO was toxic to fish as the LC50 ranged between 0.7 and 5.7 mg/l.

Table 4.13
Effects of linear AE to fish.