Transparent CNT Heaters for Laminated Glass

Abstract

A CNT heater is described comprising, in sequence, a first substrate, an adhesive layer, a protective layer, a CNT network layer, and a second substrate. Resistive heating is generated when a current passes through the CNT layer. The system has been found to provide improved stability and stability.

Claims

1. A heater comprising, in sequence, a first substrate, an adhesive layer, a layer comprising a sulfonated tetrafluoroethylene based fluoropolymer-copolymer, a CNT network layer, and a cover layer comprising a protective coating and/or a second substrate; and further comprising a conductor electrically connected to the CNT network layer so that a current can be passed through the CNT layer during operation.

2. The heater of claim 1 wherein the cover layer comprises a second substrate.

3. The heater of any of claims 1-2 wherein the first substrate and the heater are transparent.

4. The heater of any of claims 1-3 wherein the adhesive layer comprises PVB.

5. The heater of any of claims 1-4 wherein the conductor comprises a copper braid.

6. The heater of any of claims 1-5 wherein the first and second substrates each comprise at least one layer comprising borosilicate glass.

7. The heater of any of claims 1-6 having stability such that, when exposed to ambient conditions for 14 days, there is a 10% or less increase in electrical resistance.

8. The heater of any of claims 1-7 wherein the CNT network layer comprises a length of at least 12 cm disposed between two electrical leads; wherein the heater has flexibility such that, having stability such that, if subjected to a strain of 6000 microstrain via a four-point bend test according to ASTM D6272A at a strain rate of 0.5 in/min, and released to no strain, there is less than a 5% change in electrical resistance.

9. The heater of claim 8 wherein, if subjected to ASTM D6272A, there is a 10% or less, preferably 5% or less, change in resistivity at a deflection of 0.75 inches.

10. The heater of any of claims 8-9 wherein the first and second substrates comprise fiberglass.

11. The heater of any of claims 1-10 wherein, if subjected to a strain of 6000 microstrain via a four-point bend test according to ASTM D6272A at a strain rate of 0.5 in/min, and released to no strain, there is less than a 10% change in average optical transmittance over the surface of the heater that has the underlying CNT layer.

12. A turbine blade or a vehicle window comprising the heater of any of the previous claims.

13. A defroster structure comprising, in sequence, a transparent substrate, an adhesive layer, a protective interlayer, a CNT network layer, a transparent cover layer, and further defined by a stability such that, when exposed to ambient conditions for 14 days, there is 10% or less increase in electrical resistance.

14. The defroster structure of claim 13 wherein the adhesive layer comprises PVB.

15. A method of making the heater or defroster of any of claims 1-14, comprising: applying a CNT network layer on a first substrate; applying a protective interlayer onto the CNT network layer; applying an adhesive layer and second substrate onto the CNT network layer to form a laminate such that the adhesive layer is disposed between the CNT layer and the second substrate; heating the laminate under vacuum that removes gas generated during the heating to form a heated laminate; and cooling the heated laminate.

16. The method of claim 15 wherein the heating comprises hot pressing the laminate under vacuum that removes gas generated during the hot pressing to form a pressed, heated laminate; and cooling the pressed laminate.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIG. 1 shows resistance as a function of time for laminates bonded in a convection oven and in an autoclave with reduced pressure. The autoclaved samples (the three data points showing the highest initial resistivity) showed the smallest initial increase in resistivity while the ordinary convection oven samples (the three data points showing the lowest initial resistivity) showed a larger initial increase in resistivity.

[0025] FIG. 2 is a schematic cross-sectional view of a transparent laminate before bonding.

[0026] FIG. 3 is a photograph of the transparent panel resulting from hot pressing the transparent laminate.

[0027] FIG. 4 is a plot of resistance versus time for panels prepared without a Nafion protective layer (the three samples showing a large increase in resistivity over time) and panels with a Nafion protective layer disposed between the CNT network layer and the PVB layer.

[0028] FIG. 5 shows the configuration used for flexural testing. Dimensions are shown in inches.

DETAILED DESCRIPTION OF THE INVENTION

[0029] CNT network layers of the present invention preferably contain at least 25 weight % CNT, in some embodiments at least 50 wt %, and in some embodiments 25 to 100 wt % CNT. The CNTs can be distinguished from other carbonaceous impurities using methods known to those skilled in the art, including NIR spectroscopy (“Purity Evaluation of As-Prepared Single-Walled Carbon Nanotube Soot by Use of Solution-Phase Near-IR Spectroscopy,” M. E. Itkis, D. E. Perea, S. Niyogi, S. M. Rickard, M. A. Hamon, H. Hu, B. Zhao, and R. C. Haddon, Nano Lett. 2003, 3(3), 309) Raman, thermogravimetric analysis, or electron microscopy (Measurement Issues in Single Wall Carbon Nanotubes. NIST Special Publication 960-19). The CNT network layer (again, prior to coating) preferably has little or no polymer (“polymer” does not include CNTs or carbonaceous materials that typically accompany CNTs—typical examples of polymers include polyurethane, polycarbonate, polyethylene, etc.); preferably the network layer comprises less than 5 wt % polymer, more preferably less than 1 wt %) The volume fraction in the network layer is preferably at least 2% CNTs, more preferably at least 5%, and in some embodiments 2 to about 90%. The remainder of the composite may comprise air (by volume) and/or other materials such as residual surfactant, carbonaceous materials, or dispersing agent (by weight and/or volume). “Substantially without polymer” means 5 weight % or less of polymer in the interior of a CNT film, preferably the film has 2 weight % or less of polymer, and still more preferably 1 weight % or less of polymer in the interior of the CNT film. This is quite different from composite materials in which CNTs are dispersed in a polymer matrix.

[0030] After the CNT network layer has been coated, it retains electrical conductivity provided by contacts between CNTs; it is preferably not a dispersion of CNTs in a polymer matrix. Typically, a cross-sectional view of the composite material will show a polymer layer that contains little or preferably no CNTs and a CNT network layer that comprises CNTs (and possibly other carbonaceous materials that commonly accompany CNTs, as well as surfactants) with little or no polymer. Preferably, a CNT network layer that has an overlying polymer coating comprises 50 mass % or less of the coating polymer within the CNT layer, more preferably 25 mass % or less, and still more preferably 10 mass % or less of the coating polymer within the layer. Preferably, a CNT layer comprises at least 25 mass % CNTs and carbonaceous materials, and preferably at least 50 mass % CNTs and in some embodiments 30 to 100 mass % CNTs. CNT networks and CNT fibers have very distinct rope-like morphology as observed by high resolution SEM or TEM. See for example Hu, L.; Hecht, D. S.; and Gruner, G. Nano Lett., 4 (12), 2513-2517 for CNT networks and U.S. Pat. No. 6,683,783 for images of CNT fibers. Because the CNT layers typically contain little or no polymer, they exhibit surface roughness, if characterized by AFM, associated with the CNT diameter and bundle size, in the range of 0.5 to 50 nm. Preferably, the coating composition contacts the surface of the CNT network layer but does not fill spaces within the network layer. Penetration of a coating into the CNT layer could also be determined by crosssection of the multi-layer sample and then analysis by methods such as SEM-EDS or XPS; the CNT layer is preferably substantially free from N-groups that are associated with the topcoat.

[0031] CNT layers have many contacts between CNTs and good conductivity that is, a resistivity less than 0.05 Ω.Math.cm, preferably less than 0.002 Ω.Math.cm. The sheet resistance of this layer should be less than 500 Ω/square, preferably less than 200 Ω/square, more preferably less than 50 Ω/square. The CNT layer is preferably substantially planar (similar to a sheet of paper or a nonwoven textile sheet, a few fibers may project from a planar layer). These are preferred characteristics of the CNT layer both before and after a glass sheet is applied over the CNT layer.

[0032] A CNT network in this invention can be prepared as a dispersion of CNTs applied directly to a substrate where the solvents used in the dispersion process are evaporated off leaving a layer of CNTs that coagulate together into a continuous network. The CNT network may be prepared from dispersions and applied by coating methods known in the art, such as, but not limited to, spraying (air assisted airless, airless or air), roll-coating, gravure printing, flexography, brush applied and spin-coating. The thickness of the CNT layer is preferably in the range from 0.005 μm to 300 μm, preferably in the range of 0.05 μm to 100 μm, more preferably in the range of 0.3 μm to 100 μm.

[0033] The thickness of a protective layer disposed between a CNT layer and an adhesive layer can be, for example, in the range from 0.05 μm to 1000 μm, preferably in the range of 0.5 μm to 500 μm, or in the range of 0.3 μm to 100 μm. The thickness of an adhesive layer in the laminate can be, for example, in the range from 0.1 μm to 1000 μm, preferably in the range of 0.5 μm to 600 μm, or in the range of 1 μm to 400 μm. The length of a CNT network layer between electrodes is preferably at least 5 cm, or at least 10 cm, or at least 20 cm, or at least 30 cm, or, in some embodiments, in the range of 5 to 100 cm. As is conventional, distances refer to average distances unless specified otherwise.

[0034] The CNT layer may include other optional additives such as p-dopants. P-dopants could include, but are not limited to, perfluorosulfonic acids, thionyl chloride, organic pi-acids, nitrobenzene, organometallic Lewis acids, organic Lewis acids, or Bronsted acids. Materials that function as both dispersing agents and dopants such as Nafion and hyaluronic acid may be present. These materials contain p-doping moieties, i.e., electron accepting groups, within their structure, often as pendant groups on a backbone. Generally, these additives will be present as less than 70% by weight of the CNT film, and in some embodiments as less than 50% by weight of the CNT film. Polymers and carbohydrates that function as both dispersing agents and dopants can be distinguished from other polymer materials, i.e. those functioning as only a dispersing agent or those functioning as a structural component. Because of the presence of electron accepting moieties, these materials can form a charge transfer complex with semiconducting CNTs, which p-dopes the semiconducting CNTs and raises the electrical conductivity. Thus, these dual dispersing agent/dopants can be tolerated at a higher mass percentage within the CNT layer than other types of polymer materials or surfactants. In some embodiments, the CNT layer contains no or essentially no dopants.

[0035] The invention also includes methods for anti-icing, de-icing, and/or a heating system using the laminates described herein.

EXAMPLES

Example #1—Prepreg Glass

[0036] E glass prepreg 7781 material from Fibreglast was cut into 4″×4″ panels. One of the panels was placed onto non-stick aluminum foil. Two 6″ lengths of Hex Wik W55 copper braid was dipped into uncured Epoxy-Tek EJ2189 silver epoxy and placed on the prepreg glass panel 3″ apart to create two busbars on opposing sides of the heater area. The coated HexWik busbars were each covered with thin silicone films and allowed to cure at room temperature for 16 hours. A 3″×3″ area between the busbars was masked off and coated by spray application with SWCNT dispersion with 3-4 coats until a resistance of about 2 ohms per square was reached using a 4 point probe. The coated panel was allowed to air dry at room temperature for 16 hours. A second 4″×4″ prepreg 7781 glass panel from Fibreglast was placed over the coated panel. Eight total panels were prepared. Four panels were cured in an autoclave at 300° F. for 1 hour and the other four were cured in a convection oven at 300° F. for 1 hour. The translucent samples were tested for resistance at the busbars for 68 days. The data is shown in FIG. 1.

Example #2—Clear Borosilicate Glass

[0037] Borosilicate glass panels size 4″×4″× 1/16″ are cleaned with water and Dawn detergent, rinsed with water then rinsed with isopropyl alcohol then allowed to air dry. A 3″×3″ area is masked off with scotch tape on one of the panels. A #8 wire wound bar is used to drawn down a thin coating of SWCNT dispersion at 20.5 microns applied thickness. The CNT coating is allowed to air dry (about 30 minutes in ambient air) then the scotch tape mask is removed. A Nafion coating was brush applied manually over the CNT area leaving the edges exposed to allow the silver epoxy to be applied any still make contact with the CNT layer. The Nafion was allowed to air dry for a few minutes before the silver epoxy was applied.

[0038] Two busbars of silver filled epoxy resin (Epo-Tek® EJ2189-LV) are brush applied onto two opposing sides of the of the CNT square and extending to one edge of the glass panel. The silver epoxy will provide electrical contact to power the heater. While the silver epoxy is still liquid and before it cures, one film sheet (3.5″×3.5″) of the PVB film was applied over the heater area. The second glass panel was placed over the PVB film (Trosifol BGR15 0.38 mm thick) and the entire assembly is secured on the edges with binder clips. See FIG. 2.

[0039] The assembly was placed in a vacuum oven pre-heated to 170° C. Vacuum was pulled to 24 inch Hg for 60 minutes, then the vacuum was removed and the assembly remained in the oven for an additional 5 minutes. The glass laminate assembly was removed from the oven and allowed to cool to room temperature.

[0040] Applying the CNT as transparent conductor to the glass provided better visual appearance than applying to the polyvinyl butyral (PVB). The PVB films are textured and the CNT tends to pool in the divots creating a non-uniform heater both visually and electrically. Testing was performed with 2 different PVB interlayer films typically used for laminating automotive windshields:

Saflex RA41 (0.80 mm thick) from Eastman
Trosifol BG-R15 (0.38 mm thick, PVB) from Kuraray
Can apply CNT network layer to either films or glass. Resistance increased from 45-50Ω to about 150-220Ω after lamination. Processing parameters (temp, time and vacuum) were controlled to minimize air entrapment without disrupting CNT layer. An image of the film applied to the transparent substrate is shown in FIG. 3.

[0041] Good results for both prepreg and glass laminate substrate show reasonable thermal performance even after a cure cycle. No significant drift in resistance over time was observed with prepreg samples. As shown in FIG. 4, resistance increased significantly over time in glass laminate samples; however, application of a Nafion (a sulfonated tetrafluoroethylene based fluoropolymer-copolymer) protective barrier coating resulted in a highly stable laminate.

[0042] A thermal image of a glass sample at 167 Ω heating using 20 volts, 0.12 amps showed uniform heating over the panel with a temperature between 85 and 90 F.

[0043] Prepared glass laminate samples were heated in a vacuum oven. In some cases, the silver epoxy was observed to cure hard before full flow out of the PVB film preventing air escape. It is believed that glass treatment to improve flow or higher laminating temperature may reduce/eliminate this problem.

Example 3, Flexibility

[0044] A sample was prepared by first applying a primer (PPG primer 44GN072, low density epoxy primer) coating to a fiberglass (prepreg) substrate followed by applying a CNT network film via a dispersion, drying, applying a second primer coating (PPG primer 02Y04 chromium primer, low density epoxy primer), and sandwiching with a second fiberglass substrate. The primer layers were relatively thin compared to other layers. Copper braids provided electrical connections to the CNT layer. The laminate was bonded by hot pressing. The flexural stability of the laminate was tested in a four-point bend test using ASTM D6272 Procedure A at a strain rate of 0.5 inch/min flex to 6000 microstrain. The design of the test is shown in the FIG. 5. The samples had similar thermal uniformity and same resistivity (within 5%) before and after the 4 point bend test.