Encapsulants, which are chemically stable, compatible with cells, and exhibit high resistance against atmospheric gases, pollutants, radiation, mechanical stress, and so forth, are of great interest for photovoltaic (PV) module manufacturing, especially in terms of reliability [1–5]. Polymer film based on ethylene–vinyl acetate copolymer (EVA) has hitherto been the most widely used encapsulant in single-glass/backsheet c-Si modules . However, many results from research and applications indicate that EVA’s potential to release acetic acid under exposure to atmospheric water and/or ultraviolet radiation, along with related problems in reliability, restrict its application, especially for double-glass applications [7,8]. For building-integrated photovoltaic (BIPV) applications, the requirements for the mechanical stability and safety of PV modules are more essential. As a well-known thermoplastic interlayer for laminated safety glass, polyvinyl butyral (PVB), which has higher resistance against mechanical load, has been successfully implemented in BIPVs . The most critical disadvantage of PVB is its high sensitivity to hydrolysis, followed by a high water uptake, which results in a significant decrease of its adhesion to glass  and the formation of haze . In recent years, there has been increasing demand for an advanced encapsulation solution based on polyolefin elastomers (POE) arising from PV module manufactures —especially for double-glass c-Si modules and thin-film modules. Unfortunately, the water vapor transmission rate (WVTR) of POE, which is about a few grams per square meter per day according to technical data sheets, is not low enough. The edges of PV modules encapsulated with any of the abovementioned materials must therefore be additionally well sealed against moisture ingress [13,14]. Through our elaborate investigations and perennial application experiences, we report an ionomer-based self-breathing encapsulant film with no edge sealing for our Cu (In,Ga)Se2 (CIGS) thin-film PV modules.
《2. Material and methods》
2. Material and methods
The investigated encapsulation film, which has a multilayer structure consisting of poly(ethylene-co-acrylic acid) (EAA) and thermoplastic ionomer (metal-ion-neutralized EAA), is a commercial product provided by our partner company. It has a thickness of 0.7 mm, an EAA/ionomer volume ratio of 1:2, and a melting point of 98 °C.
In order to investigate their moisture absorption, the encapsulant samples and their corresponding raw material pellet samples were respectively placed in glass bottles and dried in a vacuum dryer (Goldbrunn 450, GOLDBRUNN, Germany) under about 2 kPa at 80 °C for one week. Although a certain amount of residual water should still exist, we define the samples after this treatment as dry samples in this paper. Subsequently, the samples were respectively stored in a climate chamber (VCL 4010, Vötsch Industrietechnik GmbH, Germany) for moisture absorption. Under every set of temperature and relative air humidity (details provided in Section 3.1), one encapsulant sample (~2 g), one EAA pellet sample (~10 g), and one ionomer pellet sample (~10 g) were kept in the chamber until their equilibrium moisture contents (EMCs) were achieved. In order to observe the moisture desorption, samples with saturated moisture contents were stored in a climate housing, where the temperature was kept at 22 °C and the relative air humidity was maintained between 8% and 10%, until a new EMC was reached.
Moisture absorption and desorption in the ionomer encapsulant samples were monitored by an analytical scale (ABT120-5DM, KERN GmbH, Germany). During moisture absorption under given conditions in the climate chamber, the samples were taken out of the chamber after a certain time interval and weighed using a scale as quickly as possible. After that, the samples were placed back into the climate chamber for further absorption. During moisture desorption, the encapsulant samples were weighed in situ in the climate housing. The moisture content M is therefore a function of the storage duration t and is proportional to the increase in sample weight :
where W(0) and W(t) represent the weight of the dry sample and its value at t, respectively.
《3. Results and discussion》
3. Results and discussion
《3.1. Moisture absorption》
3.1. Moisture absorption
Fig. 1 illustrates the time-dependent moisture absorptions of the ionomer-based encapsulant samples and their pellet raw materials under three climatic conditions: 30 °C and 70% relative humidity (RH; corresponding to an absolute humidity of 21.2 g·m-3 ); 50 C and 26% RH (21.5 g·m-3 ); and 50 °C and 70% RH (57.9 g·m-3 ). These can be well described using a simplified solution of the second Fick equation  with the following form:
where is the EMC and K is a constant that contains the diffusion coefficient and a geometric factor.
Fig. 1. Moisture absorption by EAA pellets, ionomer pellets, and ionomer-based encapsulant at (a) 30 °C and 70% RH (21.2 g·m-3 ); (b) 50 °C and 26% RH (21.5 g·m-3 ); and (c) 50 °C and 70% RH (57.9 g·m-3 ). The blue and red dashed lines are fitting curves according to Eq. (2), and the black dashed line is a curve-fit according to Eq. (3).
Since the scale of the encapsulant film thickness h is much smaller than its area, we may consider the film as an infinitely large plate; thus, the moisture only diffuses perpendicularly through the film surface into the film. The constant K is therefore a quotient of the diffusion coefficient D against 4h2 . Due to the multilayer structure, we further developed Eq. (2) for our case as follows:
where xEAA and xionomer = 1 – xEAA represent the content of the EAA and the ionomer, respectively. Using Eqs. (1) and (2), we graphically determined the key parameters to describe the moisture absorption process in both the raw materials and the encapsulant, as listed in Tables 1 and 2, respectively.
The moisture absorption in the investigated encapsulant is due to the fact that the COO–Me+ groups in the ionomer are highly polar, with a strong hydrophilic character. Kutsumizu et al.  concluded by analyzing sodium-neutralized EAA that each COO– Na+ ion pair is able to absorb three water molecules in its vicinity to build a tightly bonded primary hydration shell. As absorption proceeds, the excess water molecules locate around the primary hydration shell. Based on our results, shown in Table 1, we confirm that the polar ionomer is able to absorb more water molecules than the less polar EAA. For the same reason, the mobility of the water molecules trapped by the COO– Me+ ion pair in the ionomer is significantly suppressed, resulting in a much smaller diffusion coefficient (corresponding to a lower WVTR). In 2010, Kempe et al.  experimentally characterized the moisture permeation in laminated glasses with different encapsulants. Considering the moisture permeation as a one-dimensional diffusion-controlled process, they determined the diffusion coefficient of water in their investigated ionomer at 85 °C and 85% RH to be 1.25 × 10-4 mm2 ·s-1 (this result is calculated according to the data in Ref. ), which is more than one magnitude smaller than the values in EVA (4.01 × 10-3 mm2 ·s-1 ) and PVB (1.74 × 10-3 mm2 ·s-1 ), and is in reasonable agreement with our results in Table 2.
Table 1 EMC and constant K of EAA and ionomer pellets, and the adjusted R2 obtained by fitting the experimental results illustrated in Figs. 1(a)–(c).
Table 2 EMC of each of the components, , in the encapsulation film; diffusion coefficient of the water molecules DEAA and Dionomer ; and the adjusted R2 obtained by fitting the experimental results illustrated in Figs. 1(a)–(c).
At thermodynamic equilibrium, the partial Gibbs energy of a water molecule absorbed in the ionomer encapsulant is identical to that of a free water molecule in the surrounding air, namely ; where R and T represents gas constant and temperature, respectively. The thermodynamic activity of absorbed water , which is given by the ratio of the vapor pressure of the water in the encapsulant to the vapor pressure of pure water, is therefore equivalent to the relative air humidity, which corresponds to the ratio of the vapor pressure of the water in air pw to its saturation vapor pressure p0.w. An increase or decrease in air humidity will shift the thermodynamic equilibrium and result in further moisture absorption or desorption, respectively, until a new equilibrium is established.
The establishment of thermodynamic equilibrium is preceded by the homogenization of the absorbed water molecules in the material, which is a diffusion-controlled process. Under constant relative humidity, a higher temperature increases the diffusion coefficient of the water molecules, causing the absorption or desorption to reach its equilibrium faster (results shown in Figs. 1(a)–(c)).
《3.2. Moisture desorption》
3.2. Moisture desorption
As mentioned above, an encapsulant with saturated moisture content can release the absorbed water molecules if the air humidity of its surroundings decreases. This phenomenon is clearly indicated by the results shown in Fig. 2. As can be seen, although the EMC in the samples is initially unequal, the temporal developments of the moisture content are similar. This means that the kinetics of moisture desorption, and therefore the rate constant of desorption kdes, are independent of the amount of adsorbed water molecules. Hence, we describe this kind of water desorption by using the Kissinger analysis for the first-order homogeneous reaction  with the following form:
where represents a thermodynamic term and contains an activation energy for desorption Edes and Boltzmann constant k; and t' is the water desorption duration. Through mathematic fitting of the curves in Fig. 2, we determine that the water desorption rate constant of the investigated encapsulant samples is equal to 0.02304 h-1 .
Fig. 2. Moisture desorption of the ionomer-based encapsulant, whose moisture content was saturated at room temperature under two different air humidities. Moisture desorption was realized by storing these samples at room temperature with significantly lower air humidity. The normalized moisture contents, according to the saturated value, are plotted against the storage duration and fitted with a black dashed curve using Eq. (4).
《3.3. Absorption isotherm》
3.3. Absorption isotherm
Fig. 3 schematically describes the influence of RH on the moisture content at equilibrium in the ionomer pellet samples according to our current experimental results. The red dashed line is a guideline for the eye, based on the Hailwood–Horrobin equation :
A, B, and C represent the temperature-related constants, respectively.
Although a more accurate fitting is lacking, the red dashed line in Fig. 3, with A = 0.5278, B = 2.6516, and C = 4.0345, indicates a significant increase of EMC in the ionomer if the relative air humidity is greater than 60%. The line further serves as a guideline for the manufacture of both the encapsulant and the module, in order to control air humidity during production.
Fig. 3. Sorption isotherm of water molecules for ionomer pellets at 50 °C. The results at 26% RH and 70% RH are graphically determined and listed in Table 1. The result at 7.5% RH was calculated through our previous work (not published). Results are fitted with a red dashed curve using Eq. (5).
《3.4. Self-breathing encapsulant》
3.4. Self-breathing encapsulant
In regard to the spontaneous moisture absorption and desorption, as discussed above, it is reasonable to consider the investigated ionomer-based encapsulant as a self-breathing material that depends on air humidity: During the daytime, when there is comparative higher relative air humidity, the encapsulant absorbs moisture. Thanks to the low diffusion coefficient, as listed in Table 2, most of the absorbed water molecules are concentrated on the very outer edge of the PV module. During the night, the lower air humidity reverses the direction of the thermodynamic process; thus, the encapsulant releases moisture. Therefore, as long as the edge area is wide enough (which is also essential for insulation distance), moisture can scarcely reach the cells under conventional application conditions, and no edge sealing is required.
Compared with other conventional moisture-absorbing encapsulants such as EVA and PVB, we emphasize that the selfbreathing function of the ionomer-based encapsulant exhibits two practical and essential advantages: ① No hydrolysis reaction exists, and moisture absorption and desorption in the ionomerbased encapsutant are completely reversible; and ② the diffusion coefficient of water in the ionomer is significantly lower than that in EVA and PVB, as discussed in Section 3.1. Although water diffusion in EAA is comparatively faster, the EMC of EAA, which is about 0.02%, is negligible. Hence, as the presented encapsulant is a combination of these two components, the water penetration depth in it will be much smaller than that in EVA and PVB under the same conditions. According to the results in Ref. , 10 h is roughly sufficient for water molecules to penetrate a distance of 10 mm in EVA and PVB at 85 °C and 85% RH. As certificated by TÜV Rheinland, we report that, after a standard 1000 h of damp heat treatment, the power degradation of our CIGS modules, which have an edge width of 8.4 mm and use the ionomer-based encapsulant, is less than 5%, which fulfills the regulation of the international electrotechnical commission standards (IEC 61730).
In the present work, we have introduced a promising ionomerbased encapsulant, which is able to spontaneously absorb and/or desorb moisture. This process is thermodynamically driven by the RH of the environment. Furthermore, since the temperaturedepended diffusion coefficient of water molecules is relatively low due to the polar nature of the COO– Me+ groups in the ionomer, the encapsulant is able to "inhale” moisture when the air humidity is high, and "exhale” it when the air humidity decreases, without letting the moisture ingress into the cells.
The support from our partner company in providing the ionomer-based encapsulant samples and the corresponding raw materials as well as the permission for material analysis is fruitfully acknowledged. Special thanks are due to Dr. Gernot Oreski from Polymer Competence Center Leoben GmbH (Austra) for the helpful discussions.
《Compliance with ethics guidelines》
Compliance with ethics guidelines
Miao Yang, Raymund Schäffler, Tobias Repmann, and Kay Orgassa declare that they have no conflict of interest or financial conflicts to disclose.