Identifying DEHP using heat and spectroscopy

Materials World magazine
2 Mar 2016

Dr G Kaiser and Dr M Schöneich discuss the determination of high boiling plasticisers in PVC by direct TG-FT-IR coupling.

Most additives used in polymers to achieve the desired material properties are plasticisers, which affect the brittleness, hardness or the application temperature range of a given plastic compound. For a long time, DEHP (diethylhexyl phthalate or bis(2-ethylhexyl) phthalate), also called DOP (dioctyl phthalate), was one of the most frequently used plasticisers in PVC, employed in consumer products such as floor coverings, medical devices, seals, cables or packaging. 

For many years, DEHP was classified as a category 1B reproductive toxicant, in accordance with the REACH process but, in June 2015, the European Commission added it to the list of restricted substances under its Restriction of Hazardous Substances (RoHS). Manufacturers, importers and distributors have until 2019 to manage compliance issues with the new restrictions. It is, therefore, beneficial to have an analytical tool to discover and clearly identify DEHP. 

Fourier Transform Infrared Spectroscopy (FT-IR) can be found in many laboratories and the coupling between an FT-IR spectrometer and a thermobalance is a widely used combination for compositional analysis. DEHP, however, is a substance with a high boiling point (385°C) and therefore can cause problems within the classical transfer line approach. The evolved gases can get stuck with negative side effects, such as retardation, which is why it is advantageous to have a direct coupling as realised in the Netzsch PERSEUS system.

Thermal behaviour of plasticised PVC

Plasticised polyvinylchloride (PVC-P) generally contains quite a high content of plasticisers – proportions of up to about 35% are common. There is no chemical interaction between the phthalate and the polymer, it is just a physical mixture. As a consequence, migration of the plasticiser, especially in contact with lipophilic compounds such as food or blood, can take place. One factor that additionally affects the migration tendency is the molecular weight of the plasticiser used – low molecular weight (LMW) substances have high migration tendencies and DEHP is an LMW plasticiser.

When heated, polyvinyl chloride (PVC) typically decomposes in two steps. The main decomposition products under inert conditions, such as in a nitrogen atmosphere, are hydrogen chloride (HCl) and chlorinated hydrocarbons – produced because of cracking of the polymer back bone. Depending on the molecular weight of the added plasticiser, its release can either start prior to the pyrolysis of the polymer or simultaneously. 

Extraction of a 2D spectrum at 319°C from Figure 2 at the point of the maximum FT-IR intensity (related to the red line), and comparison with library spectra reveals HCl as well as DEHP as major gaseous constituents (Figure 3).

The evaporation of substances during heating in an open crucible starts much sooner compared with the boiling point specified in literature – therefore a temperature range for the DEHP release of around 320°C sounds realistic. 

Integration of an FT-IR band that is exclusively related to DEHP (here the range between 920 and 990 cm-1) finally leads to the results in the DEHP trace (shown in Figure 1, red dot-and-dashed curve) which represents the course of the absorption intensity of this specific band as a function of temperature.

Under the selected conditions, the evolvement of the plasticiser takes place in the interval from 220°C to 400°C – the plasticiser release completely superimposes the polymer decomposition within the first mass-loss step. The peak temperature of the DEHP trace profile is just a few degrees higher than the peak temperature of the Gram-Schmidt plot and, again, a distinct indication for the short time lag between evaporation and analysis of the high-boiling chemical compound.  

A short transfer line is of substantial advantage when high-boiling, meaning condensable evolved gases should be investigated using thermogravimetry coupled to FT-IR (TG-FT-IR). Such effective hyphenation system opens the door to track DEHP in PVC plastic, allowing companies to fulfill regulatory stipulations. 

1. TG-FT-IR experiment on PVC-P (see above)

Sample mass – 10.65mg, heating rate – 10K/min, atmosphere – N2 (up to 850°C), crucible – Al2O3, open.

The black line represents TG, the green dashed line represents the first derivation of the TG curve versus time (DTG), the blue line represents the Gram-Schmidt plot, which corresponds to the sum of all registered FT-IR intensities and is automatically inserted in the measurement  graphics while the red dot-and-dashed line represents the DEHP trace.

2. Detection and identification of DEHP:

3D presentation of all registered FT-IR spectra related to the TG-FT-IR experiment on PVC-P. The complete measurement, including the combustion of the carbon black under oxidising conditions above 850°C.

3. TG-FT-IR experiment on PVC-P: 

Comparison between the 2D spectrum extracted at about 319°C (black, correlates to the red line in Figure 2 at the maximum of the absorbance units) and library spectra of HCl (blue) as well as of
DEHP (red).

In the case shown here, the used PVC-P is stable up to around 220°C (derived from the TGA curve in black), followed by two TGA steps at 312°C and 467°C (DTG minimum each), split into a mass loss of 74% and a second one of nearly 16%. This decomposition behaviour is basically mirrored by the Gram-Schmidt plot (blue, from the FT-IR), but it seems that the Gram-Schmidt profile is even more structured and indicates two further small peaks prior and after the principal one. 

The temperature of the main peak of the Gram-Schmidt curve is only around 7K higher than the minimum temperature of the corresponding DTG peak. Having a heating rate of 10K/min in mind, this difference correlates to a delay time of 44 seconds between the generation of the gas in the thermobalance and the registration in the FT-IR. 

An overview over the individual FT-IR spectra is depicted in the 3D diagram in Figure 2. Because of the data exchange between the NETZSCH and the Bruker software packages as part of the PERSEUS combination, the graphic automatically contains the temperature information (z-axis) and the mass-loss curve, coming from the thermobalance (in the z-y-area, red curve). 

Dr Gabriele Kaiser studied Chemistry in Erlangen, Germany and started working at NETZSCH-Gerätebau after receiving her PhD in Physical Chemistry in September 1991. Currently, she holds the position as Head of Scientific and Technical Communication Department.

Dr Michael Schöneich studied Chemistry and received his PhD in Dresden, Germany. Three years ago, he took up his position as Application Scientist in the Applications Laboratory at the headquarters of NETZSCH-Gerätebau in Selb, Germany.