For more information contact ABB Group

Thomas Buijs (ABB, Québec, Canada) & Simon Fischer (Krones AG, Neutraubling, Germany) – IFPAC 2016 Use of FT-NIR Spectroscopy for Fast Quality Analys...
Author: Carol Palmer
33 downloads 0 Views 1MB Size
Thomas Buijs (ABB, Québec, Canada) & Simon Fischer (Krones AG, Neutraubling, Germany) – IFPAC 2016

Use of FT-NIR Spectroscopy for Fast Quality Analysis of PET Packaging Material © ABB Group February 29, 2016 | Slide 1

www.abb.com/analytical For more information contact [email protected]

Manufacturing of PET packaging material 

Polyethylene terephtalate (PET) containers are commonly made by using injection molding to manufacture preforms, which in turn are processed in high-speed stretch-blow molding machines.



Between insertion and extraction the stretch-blow molding process is run in multiple cavities in parallel, yielding quasi continuous production of containers at rates currently ranging up to 81,000 per hour on a single stretch-blow molding machine.

© ABB Group February 29, 2016 | Slide 2

PET bottle forming sequence

Key attributes of PET packaging material 

Requirements for the container and PET resin derive from its contents and the method of filling:

Characteristic PET containers for carbonated (left), hot-filled (middle) and non-carbonated beverages (right) 

Carbonated beverages build up pressure in container  walls need to be stiff and creep-resistant.



Microbiologically sensitive beverages sometimes hot-filled to sterilize the container and closure  containers need to exhibit enhanced crystallization to yield the necessary thermal stability.



Containers for non-carbonated contents do not need to withstand high pressures and temperatures and can be very light weight. But limitations in injection molding can restrict the preform geometry which entails high stretching ratios and demands the use of a resin with adequate properties.

© ABB Group February 29, 2016 | Slide 3

Determination of PET packaging key characteristics 

The following properties are commonly measured in laboratory to assess PET containers characteristics: Property

Measure of

Information on

Laboratory method

Intrinsic viscosity

Chain length and molecular weight

- Mechanical properties (Stiffness and creep resistance) - Solution viscometry - Thermal properties - Melt flow index determination - Amount of energy required for processing - Final use of the resin

Moisture

Water content in resin

- Thermal properties (glass transition and cold crystallization temperatures) - Orientation and risk of stress cracking - Age and storage conditions

Crystallinity

Degree of structural order

- Polymer density - Density column - Mechanical properties (stiffness, density, permeation), - Electronic densimeter transparency and thermal stability - Differential scanning calorimetry - X-ray diffraction - Gel permeation chromatography

- Karl-Fischer titration - Thermogravimetry (Loss on drying) - Calcium hydride - Microwaves

Co-monomers content Additives and by-products concentration (DEG and IPA)

- Thermal properties and crystallinity - Barrier properties

- Gas chromatography after liquid extraction

Wall thickness

- Container stability - Performance and stability of the stretch-blow molding process

- Segment weight measurement - Hall effect probe - IR beam broad-band attenuation

© ABB Group February 29, 2016 | Slide 4

Thickness of the PET container

Issues associated with PET packaging testing 

The stretch-blow molding step is the last conversion step where the PET container takes shape.



Usually both process interfaces (preform and finished PET container) are covered by specifications which define the quality attributes and their respective test methods.  Regular checks are performed to ensure production quality.



Some tests require specialized equipment, time, repeated measurements, operator expertise and the availability as well as proper handling and storage of chemicals.  Depending on ease of implementation, some tests can only be carried out hourly, upon changing formats, once per shift or even less frequently.



On high-line output speeds the risk associated with detecting a defect too late is even higher as large production volume is handled between measurements.



An immediate insight into material properties and processing conditioning on both preform and finished PET containers can enable an improved understanding of the process, providing feedback on the selected process settings or indicating the need to make process adjustments.

© ABB Group February 29, 2016 | Slide 5

FT-NIR analysis - Principles Measurement technique based on the absorption/reflection at different wavelengths of a near-infrared beam by a sample. Two measurement modes: 

Transmission (Liquids, clear solids)

Source radiation

Transmitted radiation

Sample



Reflection (Opaque solids)

Source radiation

Sample Reflected radiation

FT-NIR analysis - Principles -



PET bottles PET preforms PET pellets

Each peak in the FT-NIR spectrum has a well defined frequency (position), band shape and height. These three parameters are related to: 

Nature of the group of atoms that vibrate in a molecule (Example vibrations: C-H: 2,860 cm-1, N-H: 3,300 cm -1, O-H: 3,400 cm-1)



Atom group environment (linear/branched molecule)



Number of a particular group of atoms in a sample



FT-NIR provides information on chemical composition of samples. In addition the technique can also provide information on some physical attributes of the sample.



It therefore enables multi-property determination on a complex sample or mixture through a single measurement.

© ABB Group February 29, 2016 | Slide 7

FT-NIR analysis - PET characteristics determination 

FT-NIR is an ideal technique for fast, precise and non-destructive determination of several PET quality attributes such as intrinsic viscosity, moisture content, crystallinity, composition and wall thickness. Property

Traditional methods limitations

FT-NIR correlation

Intrinsic viscosity

- MFI not very useful for PET (susceptibility to hydrolytic degradation in the presence of moisture) - Solution viscometry time-consuming, requires hazardous solvents, results often lab- or operator-dependent

- FT-NIR spectrum contains information on sample molecular weight, chain length and conformation

Moisture

- Karl-Fischer time-consuming, requires reagents and operatordependent, variations in accuracy and precision. - Thermogravimetry sensitive to hydrolytic degradation at elevated temperatures - Calcium hydride method time-consuming

- Water has strong infrared signature (one of strongest absorbers)

Crystallinity

- Density-based methods are time-consuming, require operator training and care - DSC and X-ray require highly specialized equipment and an even higher level of operator training to interpret the results

- Crystallinity is largely determined by chain length, stereoregularity, presence of polar groups and degree of substitution on the chains. All those effects are present in the FTNIR spectrum

Co-monomers content

- Gas chromatography is time-consuming, requires column preparation and operator expertise

- FT-NIR spectra will carry qualitative and quantitative information on organic compounds, or compounds with C-H, O-H or N-H bonds, in concentration >0.5% w/w.

Wall thickness

- Segment weight measurement is fast and easy but somewhat - PET has direct NIR signature, therefore thickness variations will translate in absorption band intensities that can be directly inaccurate as resolution becomes that of the segments - Hall effect probe is very accurate if calibrated right but slow and correlated time consuming for many spots on a container - Broad-band IR beam attenuation is expensive to implement

© ABB Group February 29, 2016 | Slide 8

FT-NIR analysis - Accuracy 

For each property of interest, the FT-NIR analyzer is calibrated by establishing a numerical relationship between the NIR spectra of a set of representative samples and their corresponding assay values as measured by a reference method.



Once calibrated, the FT-NIR analyzer will typically exhibit a reproducibility similar to traditional laboratory methods, but with a superior precision (repeatability) and robustness.

© ABB Group February 29, 2016 | Slide 9

FT-NIR analysis - Accuracy 

Examples of validation results (comparison with traditional methods) for PET analysis with FT-NIR: Crystallinity (% - Bottles)

Intrinsic viscosity (dl/g - Preforms)

© ABB Group

Humidity (% - Preforms)

Wall thickness (mm - Preforms)

FT-NIR analysis - Precision 

Compared to traditional laboratory methods, an FT-NIR analyzer will exhibit similar accuracy (reproducibility) but superior precision (repeatability) due to the absence of sources of human errors and the auto-adjustment of analyser frequency against ultra-precise laser.

ABB FT-NIR LAB

Example of repeatability tests for a laboratory FT-NIR instrument calibrated against an un-biased laboratory method

Example of repeatability tests for a laboratory FT-NIR instrument calibrated against a laboratory method with systematic bias

X

ABB FT-NIR LAB

X

FT-NIR analysis - Ease of implementation 

PET sample analysis with FT-NIR is performed within a few seconds, as an alternative to slower laboratory methods.



No specific sample or instrument preparation is required, measurements are totally nondestructive. The operator just places the intact bottle or preform in the analyzer sampling compartment or on an alignment jig and follows on-screen instructions.



Analysis with FT-NIR does not require any chemical reagent, waste disposal or consumable.



The analyzer can be used by operators without analytical background.



Multiple PET sample properties can be determined simultaneously via a single measurement.

© ABB Group February 29, 2016 | Slide 12

FT-NIR analysis - Flexibility 

The FT-NIR analyzer is a flexible tool that can be used for measurements on intact PET samples (preform or container) but also powders, films, pellets or resins.



Measurements on various sample types are performed with a set of swappable sampling accessories that can be placed in the analyzer’s universal sampling compartment.

© ABB Group February 29, 2016 | Slide 13

FT-NIR analysis - Reliability 

The FT-NIR analyzer features a complete set of metrology self-validation diagnostics that are permanently run in the background, so that the customer can be fully confident that all validated measurements are performed with a fully operational instrument.



For each FT-NIR assay, a statistical indicator is automatically calculated and allows to assess the quality of the sample recognition by the system. Automatic warnings are displayed for sample formulations that are unique or significantly different from the calibration database. This serves as an indication for the user that the FT-NIR calibration must be adjusted to include some samples of this formulation and further enhance the robustness of the model.

© ABB Group February 29, 2016 | Slide 14

FT-NIR analysis - Robustness 

The pre-aligned source module with electronic stabilization is designed to operate for 10 years without replacement, and the solid state laser-based metrology module has a 20-year lifespan. The FT-NIR analyzer is therefore virtually maintenance-free and allows end-user to significantly reduce maintenance and consumable costs compared to traditional laboratory methods.



Overall FT-NIR analysis enables to significantly reduce laboratory analyses costs while improving product consistency and laboratory throughput.

© ABB Group February 29, 2016 | Slide 15

PET analysis with FT-NIR – Example 1 Summary 

Fast determination of crystallinity of 88 polyethylene intact bottles with a custom calibration.



The instrument was also used for quantification of water content and bottle wall thickness, as alternative to Karl Fisher coulometer titration that takes 1 hour per sample. Method  Instrument: ABB MB3600-CH80  Sampling technique: intact bottles placed on sampling port.  Analysis performed at room temperature  Resolution: 8 cm-1  Analysis time: 12 seconds / sample  Chemometrics model: Partial Least Squares  Accuracy: 0.7%

© ABB Group February 29, 2016 | Slide 16

Calibration Performance

Property

Range (%)

R2

Crystallinity

23.8 – 32.1

0.91

SECV (%)

0.7

PET analysis with FT-NIR – Example 2 Use of FT-NIR pre-loaded global calibration 

Use of ABB MB3600-CH80 FT-NIR analyzer with global calibration model for %crystallinity on customer samples that are not included in the global calibration.



The FT-NIR crystallinity predictions on 67 samples are tested against density column values.

© ABB Group February 29, 2016 | Slide 17

PET Crystallinity Determination – Example 3 Analysis of crystallinity variations at different positions on PET bottles 

The global crystallinity calibration of the ABB MB3600-CH80 was used to measure crystallinity in different points (neck, shoulder, center) of several PET bottles



The repeatability (precision) of FT-NIR determinations is excellent



The FT-NIR results also show consistent and reproducible crystallinity pattern accross all bottles



In contrast the density column results do not show reproducible pattern for all bottles

© ABB Group February 29, 2016 | Slide 18

© ABB Group February 29, 2016 | Slide 19

Suggest Documents