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

Survey of chemical substances in toys for animals

3 Analyses

• 3.1 Beilstein's test
  • 3.1.1 Method of analysis
• 3.2 FT-IR examinations
  • 3.2.1 Used equipment and techniques
  • 3.2.2 Results of Beilstein's test and IR analyses
• 3.3 Survey of further analyses for each type of products
  • 3.3.1 PVC products
  • 3.3.2 Rubber products
  • 3.3.3 Textiles
  • 3.3.4 Fur
  • 3.3.5 Other hard and half-stiff types of plastics
• 3.4 XRF screening
  • 3.4.1 Method of analysis
  • 3.4.2 Result of XRF screening
• 3.5 Migration of heavy metals
  • 3.5.1 Method of analysis
  • 3.5.2 Result of migration analysis
• 3.6 Cr(VI)
  • 3.6.1 Method of analysis
  • 3.6.2 Results of analyses
• 3.7 Phthalates in PVC
  • 3.7.1 Method of analysis
  • 3.7.2 Results of analyses
• 3.8 Screening for organic compounds
  • 3.8.1 Method of analysis
  • 3.8.2 Results of analyses
• 3.9 Quantitative organic analysis
  • 3.9.1 Method of analysis
  • 3.9.2 Results of analyses
• 3.10 Brominated flame retardants
  • 3.10.1 Method of analysis
  • 3.10.2 Results of analyses
• 3.11 Mercaptobenzothiazol (MBT)
  • 3.11.1 Method of analysis
  • 3.11.2 Results of analyses
• 3.12 Screening for azo dyes
  • 3.12.1 Method of analysis
  • 3.12.2 Results of analyses
• 3.13 Azo dyes
  • 3.13.1 Method of analysis
  • 3.13.2 Results of analyses
• 3.14 Formaldehyde analysis
  • 3.14.1 Method of analysis
  • 3.14.2 Results of analyses
• 3.15 Smell
  • 3.15.1 Method of analysis
  • 3.15.2 Results of analyses

As the information collected on materials used for manufacture of toys for animals was very limited, Beilstein's test and IR screening of the purchased products were performed in the first phase of the project.

3.1 Beilstein's test

Beilstein's test is a quick method for determination of halogens. The principle in the test is that volatile copper salts will colour a flame green due to the copper content. Copperhalides (F is excepted) are volatile, and only in very few other cases will the test give a positive reaction. If there are halogens in plastic, the plastic is probably a PVC plastic and could typically have been softened by a phthalate plasticizer.

3.1.1 Method of analysis

A microburner and a strong copper wire were used. The microburner must have full air intake (almost colourless flame). The copper wire is annealed, the hot wire is rubbed at the sample so that some of the sample melts on the wire. The wire is led into the outer zone of the flame. If the sample is lightened and burns, it should burn out outside the flame. The wire is again led into the flame and shortly before annealing, the green colour is obvious if the sample holds halogens.

3.2 FT-IR examinations

3.2.1 Used equipment and techniques

The FT IR analyses were made by a Nicolet Impact 400 FT IR spectrometer.

As many samples showed signs of silicone oil which has probably been used as release layer, these samples were washed in petrol ether to remove residues of the release layer etc. before the actual test preparation.

At first a screening analysis to estimate the type of material was made. Where a toy consisted of more than one type of material, the part of the toy that was considered largest/most considerable was analysed.

Different techniques were used according to the product. Flat, smooth materials were examined by use of ATR technique. Materials that were not smooth or flat were examined by rubbing a silicium carbide sandpaper against the sample and take the spectrum by DRIFT (diffuse reflectance) with the clean sandpaper as reference. Textiles were also examined by DRIFT technique with KBr powder as reference.

Both ATR and DRIFT are reflection techniques and the spectra become a little distorted compared to normal transmission spectra.

For identification of plastic types primarily electronic reference libraries were used (Hummel-Scholl or Sadtler Know-it-all) combined with FORCE Technology's general experience.

Plasticizers as phthalates are normally used in large quantities (30%) and will be visible in the spectra immediately. Often these substances will camouflage the spectra of basis polymer. Phthalates present in few percentages of another ester will usually not be visible. Other additives used in 0.1% to few percentages will not be detected during the screening analysis under normal conditions, unless they have absorptions in areas where basis polymer and possible plasticizer with certainty do not absorb.

Fillers with characteristic spectra, i.e. chalk, can be established at levels of 10-30%, while other fillers most often cannot be proven with certainty.

3.2.2 Results of Beilstein's test and IR analyses

The detailed results of the Beilstein's test and FTIR analyses are shown in Appendix B, Analysis of materials.

As it appears from Appendix B, Analysis of materials, the purchased toys for animals are made of a number of different types of material:

1. Soft and half-stiff plastic: PVC with phthalate as plasticizer
2. Rubber (natural rubber, latex, polyisoprene)
3. Textiles (polyesters, acrylonitrile, viscose/cotton)
4. Real fur
5. Hard plastic types (polypropylene, polystyrene, EPDM)

In the below table the products are divided according to type of material:

TABLE 3.1.

It is concluded that approx. 25% of the materials are made of PVC, approx. 35% is made of natural rubber/synthetic rubber, approx. 30% are textiles, excluding all the filling used for stuffing. Finally approx. 10% is a mixture of various hard types of plastic.

The preliminary examination did not check for (or observe) other harmful substances than phthalate, this could have been other plasticizers, heavy metals, dyes, amines, monomers, residues of solvents, etc. However, it was noted that some of the rubber products might hold a content of mercaptobenzothiazol (MBT).

3.3 Survey of further analyses for each type of products

According to prEN 71-9:2002, table 1, it is recommended that polymer materials are analysed for monomers, solvent migration and plasticizers, while leather products (including fur) should be analysed for dyes and primary aromatic amines. It is recommended that textile materials are analysed partly as leather, partly for their content of brominated flame retardants.

Based on the above FTIR screening and Beilstein's test as well as prEN 71-9:2002 it was decided, in agreement with the MST, to analyse each sample as stated in Appendix C, Survey of analyses.

3.3.1 PVC products

Heavy metals
All the PVC products are screened for their content of heavy metals by XRF, as it is not expected in advance that the other products will show a too large content of heavy metals.

Migration of heavy metals
Based on the heavy metals screening the products where the content of heavy metals exceeds the limits for migration as stated in DS/EN 71-3, table 1, were tested for migration of heavy metals in accordance with DS/EN 71-3.

Tin
The transparent PVC products are screened by XRF for their content of tin that may typically be added transparent PVC as stabiliser.

Phthalates
The FTIR analyses of the PVC samples all clearly demonstrated content of phthalates. The PVC samples are therefore also analysed for qualitative and quantitative content of phthalates by extraction and subsequent analysis by use of GC-MS.

3.3.2 Rubber products

Organic compounds
Products of natural rubber/synthetic rubber are all screened for volatile organic compounds by use of GC-MS and headspace technique.

Chromium and cadmium
Rubber products with transparent yellow, orange or red colours are screened by XRF for their content of chromium or cadmium. These two heavy metals have been used in exactly these colours.

Mercaptobenzothiazol (MBT)
This substance is typically used as accelerator in production of natural rubber.

A number of selected rubber products have been analysed by use of HPLC for release of MBT (migration to water according to prEN71-10).

Odours
One product had a characteristic odour, which was determined partly by GC-MS with headspace technique, partly by "trained" noses.

3.3.3 Textiles

Brominated flame retardants
All textiles are screened for their content of Br as indication of content of brominated flame retardants. Screening is made by XRF.

In samples with high Br content flame retardants, if any, are attempted identified and quantified by extraction and GC-MS analysis.

Formaldehyde
All textiles, including fill in teddy bears, are tested for their content of formaldehyde according to DS/EN ISO 14184-1.

Azo dyes
According to prEN71-10, annex A, strongly dyed textiles have, at first, been screened for their coming off on different textile materials. The colours where the coming off was observed, are further analysed according to EN ISO 14362 for content of azo dye and/or aromatic amines.

3.3.4 Fur

Chromium
Three of the samples are made of real fur. These three samples are analysed for their content of heavy metals by XRF and migration of heavy metals according to DS/EN 71-3. Finally the three furs have been analysed quantitatively for content of Cr (VI) by a spectro photometric method used for examination of leather products for personal safety devices (DIN 53314).

3.3.5 Other hard and half-stiff types of plastics

Further to identification of types of material only a few analyses have been made within this group.

3.4 XRF screening

3.4.1 Method of analysis

X-LAB 2000 instrument (Spectro) was used for the X-ray analyses (XRF). For quantification of the content, the programme TURBO-QUANT was used. By this technique all elements larger or equal to no. 11, Sodium, can be analysed. The minimum quantity to be determined depends on matrix and element, but is for certain elements <10 ppm.

No actual sample preparation was made. The sample was either placed directly in the instrument, or a piece has been cut/carved out of approximately 5 cm x 5cm. These test samples have been analysed directly in the instrument.

The analysis is a surface analysis which means that the analysis goes to a depth of approximately 100μm, depending on the material.

3.4.2 Result of XRF screening

The results of the XRF screening appear from Appendix D, XRF results. Thirty-five products have been screened. In one sample two XRF analyses were made. The requirements for migration of heavy metals according to DS/EN 71-3 can be seen in Appendix D, XRF results, and all samples in which the XRF result for a heavy metal exceeds the requirement for migration are stressed.

Heavy metals
The below table 3.2 shows the total of seven samples (out of thirty-five) where the result shows a content of heavy metal exceeding the requirements for migration. These seven samples have therefore been further analysed according to DS/EN 71-3, see section 3.5, Migration of heavy metals.

In the selection of products the guidelines in DS/EN 71-3 for adjustment of the analytical result have been taken into consideration. "The adjusted analytical result" is obtained by deducting a correction from the analytical result.

E.g. is the analytical correction for Ba 30%. The found value for Ba of 1011 ppm for Toy 34 will thus corrected be considerably below the requirement of 1000 ppm. For this reason no migration analysis was made on Toy 34.

TABLE 3.2 – Content of "EN 71-3 heavy metals", ppm, by XRF analysis

*Maximum migration from materials from toys in mg/kg

Tin
Organic tin compounds are often used as stabilisers in particularly transparent PVC.

Out of the two samples of transparent PVC analysed for content of tin (Sn), one showed a large content; see table 3.3.

TABLE 3.3 – Content of Sn, ppm

No attempt has been made to identify the tin component in detail.

Chromium/Cadmium
Chromium (Cr) and cadmium (Cd) have previously entered into the composition of yellow/red/orange colours.

Nine samples, all yellow, red or orange, have been screened for their content of Cr and/or Cd, see table 3.4.

The found contents of Cr and Cd are so low that they do not indicate the use of Cr and Cd-containing pigments in the analysed samples.

TABLE 3.4 - Content of Cr and Cd, in ppm

Nickel
DS/EN 71-3 comprehends no requirements for the content of Ni.

However, the nickel statutory states requirements for both maximum content and release of nickel under certain conditions. A comparison between the values found in the samples analysed, see table 3.5, and the requirements in the nickel statutory (which is being revised at present), where the requirement for content of nickel in e.g. ear stickers is maximum 500 ppm, it can be concluded that the content found is considerably below the allowed value.

TABLE 3.5 – Content of Ni, ppm

Flame retardants
As it appears from Appendix D, XRF results, a content of bromine (Br) was also found in the three products, see table 3.6. A content of Br may indicate a content of brominated flame retardant. Antimon (Sb) is also used as flame retardant, and is used together with brominated flame retardants as well. The content of Sb is therefore also stated in table 3.6. Yet nothing indicates the use of Sb in the analysed samples. Therefore the mentioned products were analysed further to identify the content and type of flame retardants, see section 3.10 for analysis for brominated flame retardants.

TABLE 3.6 - Content of Br and Sb, in ppm

3.5 Migration of heavy metals

3.5.1 Method of analysis

The applied method is DS/EN 71-3. This method simulates contact with gastric juice when a material has been swallowed. Soluble substances are extracted from the sample under simulated conditions that the sample remains in contact with the gastric juice for a period after it has been swallowed.

Sample preparation for most of the samples followed DS/EN 71-3, 8.2. The sample preparations for the three fur mice followed DS/EN 71-3, 8.6.

All samples were analysed according to DS/EN 71-3, 8.1.2, where the analytical result is stated to be "The adjusted analytic result" obtained by deducting a correction from the analytical result. Table 3.7 shows the results above the limit in DS/EN71-3, before and after correction.

3.5.2 Result of migration analysis

Table 3.7 shows requirements as well as results.

TABLE 3.7 - Migration, ppm

*Maximum migration from toy material in mg/kg.

Results for the three fur mice (Toy 28, Toy 39 and Toy 43) are also shown as "adjusted analytical result", that is corrected with the stated analytical correction. Only in these three samples migration exceeds the allowed limit for toys for children.

3.6 Cr(VI)

3.6.1 Method of analysis

The applied method is DIN 53314, which in a modified version is used for determination of chromate in leather used for personal safety devices.

The principle of the method is that the sample is extracted in a buffer with neutral pH and the content of Cr(VI) is then determined colorimetric by a spectrophotometer.

3.6.2 Results of analyses

The three samples of real fur, Toy 28, Toy 39 and Toy 43, have been analysed.

None of the samples demonstrated a content of Cr(VI).

The detection limit for the analysis is 3 mg Cr(VI)/kg.

Cr(III) is used for leather tanning, and if the tanning process is controlled correctly, no Cr(VI) is formed. (MST, survey no. 3, 2002 "Examination of the content of Cr(VI) and Cr(III) in leather articles in the Danish market").

3.7 Phthalates in PVC

3.7.1 Method of analysis

50 mg of the sample are weighed in tiny pieces in a 20 ml screw-capped glass. The samples are extracted with CH₂Cl₂ at indoor temperature overnight. Dissolved PVC, if any, is precipitated by addition of methanol.

The sample is centrifuged and the extract is analysed by gaschromatography with mass spectrometric detector (GC-MS). Butyl-hydroxy-toluen (BHT) is used as internal standard.

For the GC-MS analyses is used Varian Saturn 2000 iontrap system.

The detection limit is considerably below the found levels. The uncertainty of the quantification is approximately 10% relative.

3.7.2 Results of analyses

All thirteen PVC-containing products have been analysed. The content of phthalates varies between 11 and 54 weight %.

Table 3.8 demonstrates the found content of phthalates. Results have been sorted according to their content of diisononyl phthalate (DINP) respectively diethylhexyl phthalate (DEHP). Note that DINP is a mixture of isomers with many components.

TABLE 3.8 – Phthalate content in weight %

Further to the main component (DEHP or DINP) a small content (< 1%) of compounds as methyl-(2-ethylhexyl)-phthalate was also found in many of the samples. No attempt has been made to quantify this content.

Very small quantities of nonylphenol were found in several of the samples. The content has not been quantified.

Nonylphenole is used for e.g. production of antioxidants for rubber and plastics, surface active substances and in plasticizers for PVC.

3.8 Screening for organic compounds

3.8.1 Method of analysis

Approximately 0.5 g of the sample is cut into small pieces. The samples are placed in a closed test tube.

The samples are analysed by GS-MS using the headspace technique.

The samples are heated at 50°C for ten minutes and shaken intermittently.

Hereafter 1000 μl of the air above the sample (headspace) is injected in the GC.

The Varian Saturn 2000 ion-trap GC-MS system is used for the GC-MS analyses. By the headspace technique only substances with a certain vapour pressure are observed.

The detection limit is approximately 10 μg/g for BHT, but depends on the substance.

3.8.2 Results of analyses

The twenty identified elastomer products have been analysed. See the results in Appendix E, Screening of organic compounds.

The only organic compounds found in large quantities in this analysis is BHT (butylated hydroxytoluene) and alifatic hydrocarbon with tetradecan in the largest quantity.

Further to this, N-butyl benzensulfonamid and cyclohexanon have been detected in several samples and one single not-identified amine compound in one sample.

3.9 Quantitative organic analysis

3.9.1 Method of analysis

As the only measurable quantities found in the samples were BHT (butyl hydroxytoluen) and tetradecan, it was chosen to quantify by use of the headspace technique.

3.9.2 Results of analyses

The content of BHT in selected samples, in accordance with 3.8.2, determined by the headspace technique.

Uncertainly of this method is 50% relative.

The detection limit is 0.005 μg/g.

TABLE 3.9

The content of tetradecane in selected samples determined by the headspace technique.

TABLE 3.10

3.10 Brominated flame retardants

3.10.1 Method of analysis

GC-MS
0.5 g sample is extracted in 9 ml dichlormethan and is placed on ultrasound for 30  minutes.

2,4,6-tribrom-anisol (TBA), 100 g/sample is used as internal standard.

The extracts have been analysed by the Varian Saturn 2000 GC-MS system.

HPLC, decabromdiphenylether
Waters HPLC system with UV detector at 220 og 240 nm.

RP-C18 column with 100% methanol eluent 1 ml/minute.

0.5 – 0.6 g sample is extracted with 9 ml dichlormethane in ultrasound for 30 minutes. The extract is filtered and injected without further treatment.

3.10.2 Results of analyses

GC-MS
As mentioned in the XRF-analyses, a content of Br was found in three samples, Toy 3, Toy 31.2, and Toy 32. Therefore these samples have been analysed further by the above GC-MS method.

This technique has not made it possible to identify or quantify flame retardants, if any, in the three samples.

A frequently used flame retardant, decabromdiphenylether, could not be detected by this system and has therefore been analysed by the HPLC method.

HPLC, decabromdiphenylether
The detection limit by the applied method (without concentrating) is 0.25 g/ml, corresponding to 4 g/g textile or 4% of the expected content in sample 31. Sample 31 had the smallest content of bromine by the XRF examination.

Decabromdiphenylether was not detected by this method in any of the three samples.

3.11 Mercaptobenzothiazol (MBT)

3.11.1 Method of analysis

According to the FTIR spectra, nine of the rubber products indicated a possible content of mercaptobenzothiazol (MBT). It was chosen to select the migration procedure in EN 71-10, §6 followed by an HPLC analysis of the extract.

Pieces of the products with a surface of approximately 10 cm² were shaken with 25 ml distilled water for at least 1 hour. The filtered migration liquid was then analysed by HPLC on an RP18 column with UV detection at 320 nm ("Microbial and photolytic degradation of benzothiazoles in water and wastewater ", Dissertation von M.Sc. Hafida Kirouani-Harani, Fakultät III der Technischen Universität Berlin, 2003).

An external standard was used.

3.11.2 Results of analyses

The results of the nine tested rubber products are stated in the below table 3.11.

There is a considerable uncertainty about the actual surface area due to the shape of the samples. An attempt was made to have 10 cm², but some were approximately 25 cm² (the areas have been taken into consideration in the calculations).

The uncertainty of the calculation is approximately 6 μg/sample.

TABLE 3.11 – MBT release

n.d. = not detected

As it appears, five samples release significant quantities of MBT.

3.12 Screening for azo dyes

3.12.1 Method of analysis

Stain effect according to prEN 71-10:2002, Annex A (slightly modified EN ISO 105 E04).

The sample material in pieces of approximately 5 cm x 15 cm were analysed for their staining by applying acid and caustic artificial sweat. The samples were tested for staining at 37C on several different textiles by contact while they were being dipped in the two kinds of "sweat".

The degree of staining is estimated on a scale from 1 – 5, where 5 means "not staining".

3.12.2 Results of analyses

See Appendix F, Screening for azo dyes.

Seven samples/partial samples with transparent colours were selected for analysis for azo dyes. Out of the seven analysed samples, three samples showed class 4 under some conditions, while the rest of the samples showed class 4-5 or 5.

The analysis does not reveal if it is due to azo dyes, but exclusively if the colours stain.

For further analysis for content of azo dyes the three samples were picked that got the classification 4 one or more times: Toy 16 yellow, Toy 16 green and Toy 31 green.

3.13 Azo dyes

3.13.1 Method of analysis

The analyses have been made according to SS-EN 14362-1: 2003 (without extraction of fibres) or SS-EN 14362-2:2003 (with extraction of fibres).
Toy 31 green is tested according to SS-EN 14362-1:2003, while Toy 16 yellow and Toy 16 green is tested according to SS-EN 14362-2:2003.

The principle of the method is that azo dyes, if any, are reduced to aromatic amines by a citrate-buffer, pH 6.00 and sodium dithionite.

Detection of aromatic amines is made by GC-MS, and quantification is made by use of internal standard and calibration solvents of amines.

Sample quantity: 1.0 g per sample

Detection limit per amine: 20 mg/kg

Toy 31 green held two colours, a green and a beige. The two colours were analysed together as one sample.

3.13.2 Results of analyses

No traces of aromatic amines could be detected in any of the analysed dye samples. Thereby it can be concluded that the three samples held no azo dyes.

3.14 Formaldehyde analysis

3.14.1 Method of analysis

Formaldehyde determination in textiles are made according to EN/ISO 14184-1:1998.

The principle of the analysis is that the content of formaldehyde, if any, is extracted from the textile sample with water at 40°C. The quantity of formaldehyde is determined spectrophotometrically after reaction with acetylacetone for formation of a yellow colour. Calibration takes place in a series of calibrations of reference solutions of formaldehyde in water.

The detection limit is 5 mg/kg. Uncertainty at this level is also 5 mg/kg.

3.14.2 Results of analyses

From the results in table 3.12 it can be concluded that only samples 3, 6 and 8 contain traces of formaldehyde.

TABLE 3.12

3.15 Smell

3.15.1 Method of analysis

Approximately 0.5 g of the sample is cut into small pieces. The samples are put into a closed sample bottle.

The samples are analysed by GS-MS using the headspace technique.

The samples were heated at 50°C for ten minutes and shaken at regular intervals.

Then 1000 μl of the air above the sample is injected (headspace) by the GC.

The Varian Saturn 2000 ion-trap GC/MS system is used for the GC-MS analyses. By the headspace technique only substances with a certain vapour pressure can be observed.

To supplement the above method, also "trained noses" have been used for determination/characterisation of the smell.

3.15.2 Results of analyses

One sample, Toy 34, had a very penetrating and characteristic smell.

GC-MS
Trace amounts of cyclohexanone, a C-14 hydrocarbon and a not identified amine were found.

"Nose" method

The smell was characterised as a mixture of liquorice allsorts, caramel, and particularly vanilla.

A rubber manufacturer informed that vanillin (4-hydroxy-3-methoxybenzaldehyd) is used as smelling substance in rubber.

No traces of vanillin could be found by the GC-MS analysis. However, this does not exclude that the smell in the toys derives from added vanillin.

By the internet and in literature, different values for the smell limit for vanillin have been found. The smell limit for vanillin in air is stated to be 1.1x10-8 ppb (http://cira.ornl.gov/documents/vanillin.pdf) or 2•10-7 mg/m3 (Am.Ind.Hyg.Assoc., March 1986).

Table 3.13 shows a short summary of the gathered Results of analyses.

TABEL 3.13 – Summary of results

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

Version 1.0 June 2005, © Danish Environmental Protection Agency