Experimental investigation of methods to reduce outgassing from core materials in the fabrication process of transparent vacuum insulation panels

. The notion that modern buildings should strive to be net-zero energy buildings (NZEBs) is widely accepted. In order to improve window’s insulation in existing buildings, structured-core transparent vacuum insulation panels (TVIPs) are proposed. TVIPs mainly consist of the structured core material, the low-emissivity film, and the transparent gas barrier envelope. TVIPs have high insulation performance and are inexpensive to manufacture and can be easily installed. However, it is necessary to overcome the issue of preventing the pressure rise inside TVIP after vacuum sealing. The authors constructed an experimental setup by applying the pressure-rate-of-rise-method to establish the fabrication process for TVIPs that prevents the pressure rise. In this experiment, a gas barrier film with a straw was used as the vacuum chamber. This could reproduce the pressure increase in the TVIP after sealing and the gas flow rate in the TVIP is evaluated. The experimental results showed that the internal pressure of TVIP could be reduced to about 1 Pa after vacuuming while heating for 8 hours, coating the core material, and sealing the getter material.


Introduction
High insulation is necessary to convert existing buildings to ZEH and ZEB, and in particular, insulation of window surfaces and other openings is indispensable. The author's laboratory has been studying Transparent Vacuum Insulation Panels (TVIP), which are made by vacuum-sealing a frame-type core material in a transparent gas barrier bag to provide a vacuum layer with heat-insulating performance. Conventional vacuum insulation materials are opaque and cannot be applied to windows. In addition, vacuum glass windows are expensive. TVIP is inexpensive to manufacture and can be easily installed, for example, by attaching it directly to the window glass of an existing building, and it has the potential to improve the thermal conductivity of single-pane glass to less than 1.5 W/(m 2 -K). In addition, the manufacturing cost is expected to be less than one-third compared to the conventional vacuum glazed window. On the other hand, the realization of TVIP involves the issues of preventing pressure increase inside the TVIP after vacuum sealing and long-term durability. In this study, we conducted an experiment to quantify gas emission by the Pressure-Rate-of-Rise Method.

Outline of experiment
In order to quantify the change in outgassing rate after vacuum sealing, the Pressure-Rate-of-Rise Method is experimentally applied [1]. In the vacuum apparatus shown in Figure 1. The valve between the TVIP and the vacuum pump is closed when the TVIP has been evacuated, and the vacuum pump is turned off. The total gas flow rate Q[Pa-m 3 /s] may be calculated using the following equation when the rise in pressure is measured.
Here, V [m 3 ] is the internal volume of the TVIP and the pipe between the TVIP and valve, and dp/dt [Pa/s] is the rate of pressure rise. In this study, the apparatus is configured as shown in Figs. 2 and 3, and the TVIP with a gas barrier envelope with a straw was used as the vacuum chamber for experiments on the Pressure-rateof-rise method. This made it possible to measure the pressure inside a full-scale TVIP, which was not possible in previous paper [2] using the Throughput method. It also enabled quantification of the reduction in gas emissions due to heating during the evacuation at E3S Web of Conferences 396, 04008 (2023) https://doi.org/10.1051/e3sconf/202339604008 IAQVEC2023 high vacuum pressure, the use of coated core materials, and the attachment of adsorbents (Getters). The gas barrier envelope with straw is the same as the one used in previous papers except for the straw part, therefore, the pressure changes after closing the valve and stopping the vacuum pump can almost reproduce the pressure change in the TVIP after vacuum sealing. The vacuum gauge is a Pirani vacuum gauge that was connected to a data logger and has a measurable pressure range from 5.0×10 -2 to 1.0×10 5 Pa with a measurement accuracy of ±20 %. A rotary pump and a turbomolecular pump were used simultaneously. The pumping speed is 50 L/m for the rotary pump and 50 L/s for N2 and 40 L/s for H2 for the turbomolecular pump. The standard configuration of the TVIPs used in the experiments are shown in Figure  4.  When a frame-type core material is used, the core material is covered with a low-emissivity film and sandwiched between two sheets of glass. In the case of peak type core material, the core material is covered with a low-emissivity film and sandwiched between glass and acrylic plate. The peak-type core material was newly produced by using a 3D printer. Although this peak core material is not transparent, it is possible to produce a transparent peak-type core material by using the transparent material.

Experimental conditions
Experimental conditions are shown in Table 1. Table. 1 Implemented experimental boundary conditions H, C, and G in the table stand for Heating, Coating, and Getter agent. The empty samples in A, H, and J are composed of only the gas barrier envelope without the core and other materials from the standard TVIP configuration in order to quantify outgassing from the envelope and gas penetration through gaps. In case B, a frame type core material was used, and the configuration was standard. The evacuation time after reaching 2 Pa was for 30 minutes. In case C, the double-stretched PET plates were used instead of glasses. In case D, heating was started at 50 °C on the top surface and 70 °C on the bottom surface simultaneously with the start of evacuation, and stopped when the valve was closed and the vacuum pump was stopped. In case E, organic polysilazane was used as coating agent and the core material were dipped in the organic polysilazane, and dried before being placed in the drying chamber. Then, the effect of organic polysilazane coating was verified. In a previous study [2], it was confirmed that the gas release rate was reduced by applying an organic polysilazane coating to small specimen pieces. In case F, the getter agent, which consists of alloy and CaO (Calcium Carbonate), was enclosed. In case G, I, K, and N, all of the simultaneous heating, the coating core material, and the enclosing getter agent were carried out. The vacuuming time after reaching 2 Pa was 30 minutes for G, 4 hours for I, and 8 hours for K, respectively. G, I, and K were compared to verify the effect of changing the vacuuming and heating time on the outgassing. Case L is K minus the simultaneous heating and M is K minus the getter agent. The effects of the simultaneous heating and the enclosing getter agent on the reduction of gas emission were verified by comparing K to L and M, respectively. In case N, the core material was changed from the frame type to the peak type and the other conditions were the same as case K. For K to N, the pressure was recorded not only for 30 minutes after the valve was closed, but also for several days after the valve was left open and the pressure was recorded on the day it was checked.
In the experiment applying the Pressure-rate-of-rise method, the total gas flow rate was determined by using the obtained internal pressure data per second, Qtotal is the total gas flow rate [Pa-m 3 /s], V is the volume [m 3 ], Δt is the time [s], and (p1 -p2) is the pressure change between Δt [Pa]. The volume V differs between empty and otherwise. In the empty case (A, H, and J), is equal to the volume V1 of the cross-shaped vacuum pipe to which the vacuum gauge and valve are connected. When a TVIP with a core material is connected, the TVIP's volume V2 was determined from the thickness of 3 mm and side lengths of 151 mm 151 mm for the frame type and from the thickness of 2.5 mm and side lengths of 150 mm 150 mm for the peak type. Then the volume V was the sum of the volume of the cross-shaped vacuum pipe V1 and TVIP V2 (V = V1 + V2). t is set to 1 second, and a moving average of 60 seconds is obtained. The gas flow rate inside the TVIP QTVIP was calculated by subtracting Qtotal for empty case (A or H or J) from the total gas flow rate Qtotal obtained in each experiment. However, if the calculated QTVIP is lower than 0 (QTVIP < 0), QTVIP is regarded as 0 (QTVIP = 0). suggests that the outgassing increases in the case of double-stretched PET plates. From cases D, E, and F, it was confirmed that the simultaneous heating, the coating core material, and the enclosing getter agent can reduce the pressure increase and QTVIP. The reduction by the simultaneous heating or the enclosing getter agent were more effective than the coating core material in the case of a 30-minute vacuuming. From G, the combination of the simultaneous heating, the coating core material, and the enclosing getter agent reduced the pressure increase and QTVIP the most. The gas flow rate QTVIP dropped and reached almost zero by 500 s for case D and case G, where the simultaneous heating was conducted, indicating that more gas has been desorbed from the core material due to the simultaneous heating. Next, the effects of changing the evacuation and heating time regimes on the outgassing were examined. Figure 6 shows (a) pressure change in case A, G, H, I, J, and K and (b) Qtotal (case A, H, and J) or QTVIP (case G, I, and K). The gas flow rate Qtotal (case A, H, and J) varies with the vacuuming time even in the empty case and is almost constant during the 30-minute measurement. A comparison of G, I, and K confirms that the pressure increase is suppressed by extending the vacuuming and heating time. Especially in K, the internal pressure was suppressed to approximately 1 Pa. This pressure met the performance requirements for TVIPs, and the end result was significantly closer to TVIP implementation. In cases, G, I, and K, the initial QTVIP after closing the valve became smaller as the vacuuming and heating time was increased.

Experimental results and discussion
The effects of the simultaneous heating and the enclosing getter agent on the gas flow rate are discussed under the longer vacuuming and heating conditions. The result of changing the type of core material was discussed. Figure 7 shows (a) pressure change in case K, L, M, and N and (b) QTVIP.  Compared to case K, where all of the simultaneous heating, the coating core material, and the enclosing getter agent were carried out, case L, where the simultaneous heating was omitted, and case M, where the enclosing getter agent was omitted, showed the larger increase in pressure. The results show that simultaneous heating can significantly reduce the outgassing from core material and that the enclosing getter agent reduces the gas flow rate over a long period of time due to the adsorption. Furthermore, the combination of the simultaneous heating and the enclosing getter agent has a synergy effect. Eventually, it was shown that both of the simultaneous heating and the enclosing getter agent were necessary to stabilize the pressure at a low value. For case N, in which the core material was changed, the pressure was a little higher than in case K. However, it was confirmed that the pressure stabilized after 600 seconds, as in case K. Figure 8 shows a comparison of the integrated gas emissions for each condition when condition B is set to 1. In the 30-minute evacuation times from A to G, except for C, the integrated gas emission can be reduced by heating, getter agent, and coating. In the case of G, which is a combination of the three, the accumulated gas release was reduced to 1/19 of that in B. In the case of I, which is a 4-hour vacuum draw, the accumulated gas release was reduced to 1/149 of that in B, and in the case of K, which is an 8-hour draw, the accumulated gas release was reduced to 1/300 of that in B. Comparing K with L (without heating), M (without getter agent), and N (peak type), L is 26 times, M is 23 times, and N is 2 times higher integrated gas release than K. The pressure was measured in cases K to N not only 30 minutes after the valve was closed, but also the next day after several days had passed. Table 2 shows the measured pressure, according to the number of elapsed days after closing the valve and stopping the vacuum pump, from case K to case N. It was confirmed that case K and case N, with all of the simultaneous heating, the coating core material, and the enclosing getter agent were able to maintain a vacuum pressure of less than 4 Pa even after about 5 days. On the other hand, the pressure increased in case L and case M. These results indicated that both the simultaneous heating and the enclosing getter agent were necessary to stabilize the pressure at a low value as explained in the previous section. The effective thermal conductivity of the vacuum layer in case K was measured by applying the heat flux meter method, and 8.9 mW/(m-K) was obtained. Table 2 Measured pressure according to the number of elapsed days after closing the valve and stopping the vacuum pump from case K to case N

Conclusion
The following is a summary of this report.
1) The effects of vacuuming time, heating during vacuuming, use of coated core material, and addition of getter agent in a full-scale TVIP were verified by the accumulation method. 8 hours of heating, getter, and coating reduced the pressure increase to about 1 Pa, and TVIPs that could maintain the required pressure were fabricated.
2) Double-type vacuum insulation material was prepared and its performance was evaluated. It was confirmed that the insulation performance remained the same even when double-layered. In the future, we will start a test to measure the performance (long-term durability test performance) of the double-layer vacuum insulation material using this double-layer vacuum insulation material.