Refine
Document Type
- Journal article (6)
Language
- English (6)
Is part of the Bibliography
- yes (6)
Institute
- Life Sciences (6)
To correctly assess the cleanliness of technical surfaces in a production process, corresponding online monitoring systems must provide sufficient data. A promising method for fast, large-area, and non-contact monitoring is hyperspectral imaging (HSI), which was used in this paper for the detection and quantification of organic surface contaminations. Depending on the cleaning parameter constellation, different levels of organic residues remained on the surface. Afterwards, the cleanliness was determined by the carbon content in the atom percent on the sample surfaces, characterized by XPS and AES. The HSI data and the XPS measurements were correlated, using machine learning methods, to generate a predictive model for the carbon content of the surface. The regression algorithms elastic net, random forest regression, and support vector machine regression were used. Overall, the developed method was able to quantify organic contaminations on technical surfaces. The best regression model found was a random forest model, which achieved an R2 of 0.7 and an RMSE of 7.65 At.-% C. Due to the easy-to-use measurement and the fast evaluation by machine learning, the method seems suitable for an online monitoring system. However, the results also show that further experiments are necessary to improve the quality of the prediction models.
Hyperspectral imaging and reflectance spectroscopy in the range from 200–380 nm were used to rapidly detect and characterize copper oxidation states and their layer thicknesses on direct bonded copper in a non-destructive way. Single-point UV reflectance spectroscopy, as a well-established method, was utilized to compare the quality of the hyperspectral imaging results. For the laterally resolved measurements of the copper surfaces an UV hyperspectral imaging setup based on a pushbroom imager was used. Six different types of direct bonded copper were studied. Each type had a different oxide layer thickness and was analyzed by depth profiling using X-ray photoelectron spectroscopy. In total, 28 samples were measured to develop multivariate models to characterize and predict the oxide layer thicknesses. The principal component analysis models (PCA) enabled a general differentiation between the sample types on the first two PCs with 100.0% and 96% explained variance for UV spectroscopy and hyperspectral imaging, respectively. Partial least squares regression (PLS-R) models showed reliable performance with R2c = 0.94 and 0.94 and RMSEC = 1.64 nm and 1.76 nm, respectively. The developed in-line prototype system combined with multivariate data modeling shows high potential for further development of this technique towards real large-scale processes.
The in-line control of curing during the molding process significantly improves product quality and ensures the reliability of packaging materials with the required thermo-mechanical and adhesion properties. The choice of the morphological and thermo-mechanical properties of the molded material, and the accuracy of their determination through carefully selected thermo-analytical methods, play a crucial role in the qualitative prediction of trends in packaging product properties as process parameters are varied. This work aimed to verify the quality of the models and their validation using a highly filled molding resin with an identical chemical composition but 10 wt% difference in silica particles (SPs). Morphological and mechanical material properties were determined by dielectric analysis (DEA), differential scanning calorimetry (DSC), warpage analysis and dynamic mechanical analysis (DMA). The effects of temperature and injection speed on the morphological properties were analyzed through the design of experiments (DoE) and illustrated by response surface plots. A comprehensive approach to monitor the evolution of ionic viscosity (IV), residual enthalpy (dHrest), glass transition temperature (Tg), and storage modulus (E) as a function of the transfer-mold process parameters and post-mold-cure (PMC) conditions of the material was established. The reliability of Tg estimation was tested using two methods: warpage analysis and DMA. The noticeable deterioration in the quality of the analytical signal for highly filled materials at high cure rates is discussed. Controlling the temperature by increasing the injection speed leads to the formation of a polymer network with a lower Tg and an increased storage modulus, indicating a lower density and a more heterogeneous structure due to the high heating rate and shear heating effect.
Employing diffuse reflection ultraviolet visible (UV–Vis) spectroscopy we developed an approach that is capable to quantitatively determine flux residues on a technical copper surface. The technical copper surface was soldered with a no-clean flux system of organic acids. By a post-solder cleaning step with different cleaning parameters, various levels of residues were produced. The surface was quantitatively and qualitatively characterized using X-ray photoelectron spectroscopy (XPS), Auger electron spectroscopy (AES), Fourier transform infrared spectroscopy (FTIR) and diffuse reflection UV–Vis spectroscopy. With the use of a multivariate analysis (MVA) we examined the UV–Vis data to create a correlation to the carbon content on the surface. The UV–Vis data could be discriminated for all groups by their level of organic residues. Combined with XPS the data were evaluated by a partial least squares (PLS) regression to establish a model. Based on this predictive model, the carbon content was calculated with an absolute error of 2.7 at.%. Due to the high correlation of predictive model, the easy-to-use measurement and the evaluation by multivariate analysis the developed method seems suitable for an online monitoring system. With this system, flux residues can be detected in a manufacturing cleaning process of technical surfaces after soldering.
An epoxy compound’s polymer structure can be characterized by the glass transition temperature (Tg) which is often seen as the primary morphological characteristic. Determining the Tg after manufacturing thermoset-molded parts is an important objective in material characterization. To characterize quantitatively the dependence of Tg on the degree of cure, the DiBenedetto equation is usually used. Monitoring polymer network formation during molding processes is therefore one of the most challenging tasks in polymer processing and can be achieved using dielectric analysis (DEA). In this study, the morphological properties of an epoxy resin-based molding compounds (EMC) were optimized for the molding process using response surface analysis. Processing parameters such as curing temperature, curing time, and injection rate were investigated according to a DoE strategy and analyzed as the main factors affecting Tg as well as the degree of cure. A new method to measure the Tg at a certain degree of cure was developed based on warpage analysis. The degree of cure was determined inline via dielectric analysis (DEA) and offline using differential scanning calorimetry (DSC). The results were used as the response in the DoE models. The use of the DiBenedetto equation to refine the response characteristics for a wide range of process parameters has significantly improved the quality of response surface models based on the DoE approach.
Monitoring of molding processes is one of the most challenging future tasks in polymer processing. In this work, the in situ monitoring of the curing behavior of highly filled EMCs (silica filler content ranging from 73 to 83 wt%) and the effect of filler load on curing kinetics are investigated. Kinetic modelling using the Friedman approach was applied using real-time process data obtained from in situ DEA measurements, and these online kinetic models were compared with curing analysis data obtained from offline DSC measurements. For an autocatalytic fast-reacting material to be processed above the glass transition temperature Tg and for an autocatalytic slow-reacting material to be processed below Tg, time–temperature–transformation (TTT) diagrams were generated to investigate the reaction behavior regarding Tg progression. Incorporating a material containing a lower silica filler content of 10 wt% enabled analysis of the effects of filler content on sensor sensitivity and curing kinetics. Lower silica particle content (and a larger fraction of organic resin, respectively) favored reaction kinetics, resulting in a faster reaction towards Tg1. Kinetic analysis using DEA and DSC facilitated the development of highly accurate prediction models using the Friedman model-free approach. Lower silica particle content resulted in enhanced sensitivity of the analytical method, leading, in turn, to more precise prediction models for the degree of cure.