X Link's benefits for mass production of PV modules

Hanwha Q CELLS GmbH (Germany) has installed X Link® at all of its production sites worldwide. Recently, the PV module manufacturer presented results of their X Link® evaluation for mass production at EUPVSEC 2015. The following article is a summary of Hanwha's technical paper [1].

The driving force was the elimination of the Soxhlet extraction – a common method in the solar industry for determination of the EVA gel content (GC) – and compare it with X Link®:

 X Link® Soxhlet extraction
Non-destructive+-
Test of standard laminate+-
Typical test duration60 s per measurement> 24 h incl. drying time 
Measurement accuracy

high systematic measurement accuracy
0.7% abs

low, especially at lower gel content values
2.5% abs

Accuracy independent of operator+-
Standardized test procedure+-
Without harmful chemicals+-


X Link®
 is based on a mechanical method (likewise micro-DMA): The tool is sensitive to the visco-elastic properties of EVA, which depend on the crosslinking degree. It calculates a value representing the ratio of elastic to viscous modulus called “LayTec cross-linking metric” (LXM) and thus a measure for the  cure state and mechanical properties. An empirical correlation between the LXM and GC [%] enables a conversion of the LXM-values into the common GC-values [%].

For accuracy evaluation, 48 samples out of one EVA piece were measured with Soxhlet extraction method by 4 operators. Another sample was measured 48 times with the X Link® at always the same point, also by 4 operators. The standard deviation of GC of the chemical method was determined as 2,5 %abs. and of X Link® as 0,7 %abs. So, X Link® is around three times more accurate than the Soxhlet method!

Furthermore, different X Link® devices, different material combinations as well as different wafer types and thicknesses were compared. The tests showed no significant differences.

X Link’s short test duration of less than 1 minute and the higher measurement accuracy enable a faster and more exact recipe evaluation, process- and quality control as well as laminator characterization. While every employee and every minute counts for the effectivity of a running mass production, a fast estimation of process quality is mandatory! With the Soxhlet it takes typically 24 hours, depending on the test setup, and the staff effort in this process takes 2-3 hours. In contrast to that, X Link® takes 10-20 min for GC-results of one module (10 measuring points per module) and 60 minutes for a complete map (Fig. 1).
Degree of cross-linking in LXM (LayTec cross-linking metric) and g.c. (gel content) mapped across a module with 60 points in less than 1 hour.
Fig. 1: Degree of cross-linking in LXM (LayTec cross-linking metric) and g.c. (gel content) mapped across a module with 60 points in less than 1 hour. Corresponding degree of 64% - 85% measured with Soxhlet extraction method.

Fig. 2 shows the generated GC values during the process control of a production line within 6 months. Illustrated are only values of one laminator with one lamination recipe. After 2.5 months the Soxhlet system was replaced by X Link®. After this, a smaller distribution of measured values can be seen, which indicates a higher measurement accuracy of X Link®.

Gel content results of Soxhlet and X Link during the process control of a production line within 6 months
Fig. 2: Gel content results of Soxhlet and X Link® during the process control of a production line within 6 months

In addition, an exact gel content area distribution of single modules or the complete laminator heating plate is possible. Fig. 3 shows the LXM area distribution of a complete laminator heating plate. Four modules with five measuring points per module were tested for this overview. As expected, the highest LXM-values are measured in the middle of the heating plate. The reason is the laminator set-up that generates the highest lamination temperatures in this area. Additionally, high LXM-values can be also seen in the middle of each laminate because of the warpage of glass during the first minutes of lamination. In this time, the bottom of the glass is hotter than the top. As a result, the glass edges bend up and lose contact to the heating plate.

X Link allows a precise characterization of the heating plate and, thereby, an early detection of possible failures
Fig. 3: LXM area distribution of laminator heating plate: 2 x 2 laminate-matrix with five measuring points per laminate

Unusual GC distributions can indicate defects like inhomogeneity of the heating plate or other problems with the laminator. Another possibility would be an inhomogeneous distribution of additives in the EVA-foil. X Link® allows a precise characterization of the heating plate and, thereby, an early detection of possible failures.

For further details please ask us for the poster presentation and technical paper via mail@laytec.de

[1] S. Kunath, M. Vogel, M. Mette, E. Malguth: Evaluation of gel content measurement method for mass production of glass-foil based crystalline silicon PV modules, to be published in EUPVSEC 2015 Proceedings