Abstract
Disclosed is a ferroelectric material and methods for its use in capacitors that includes a polymer blend of at least two polymers, wherein the first polymer is a ferroelectric polymer and the second polymer has a low dielectric constant.
Claims
1. A ferroelectric material consisting of a polymer blend of at least two polymers, wherein the first polymer is a ferroelectric polymer and the second polymer has a low dielectric constant, wherein the second polymer is a polyphenylene ether having the following structure: ##STR00005## wherein the oxygen ether atom of one unit is connected to the benzene nucleus of the next adjoining unit, R.sub.1 and R.sub.2 are each individually a hydrogen, a halogen, a hydrocarbon radical, a halohydrocarbon radical having at least two carbon atoms between the halogen atoms and the phenyl nucleus, a hydrocarbonoxy radical, a halohydrocarbonoxy radical having at least two carbon atoms between the halogen atom and the phenyl nucleus, or a substituted or unsubstituted phenyl group, and wherein the polymer blend comprises a crystalline or semi-crystalline polymeric matrix of the ferroelectric polymer and a plurality of amorphous nanostructures comprising the second polymer, wherein the plurality of amorphous nanostructures are comprised within the polymeric matrix, and wherein the polymer blend comprises 1 to 8 wt. % of the second polymer.
2. The ferroelectric material of claim 1, wherein the polymer blend is a solution blend in which the at least two polymers have been dissolved in a common solvent.
3. The ferroelectric material of claim 2, wherein the common solvent is methyl-ethyl-ketone, cyclohexanone, hexanone, or a solvent that comprises a ratio of at least two solvents capable of dissolving both the first and second polymers.
4. The ferroelectric material of claim 1, wherein the polymer blend is a melt blend.
5. The ferroelectric material of claim 4, wherein the temperature used to obtain the melt blend is above the melting point and below the thermal degradation temperatures for each of the at least two polymers.
6. The ferroelectric material of claim 1, wherein the ferroelectric polymer is a co-polymer.
7. The ferroelectric material of claim 6, wherein the co-polymer is poly(vinylidene fluoride-trifluoroethylene) (P(VDF-TrFE)).
8. The ferroelectric material of claim 7, wherein the mole ratio of VDF to TrFE is about 80:20, 77:23, 75:25, 70:30, or 55:45.
9. The ferroelectric material of claim 1, wherein the dielectric constant of the second polymer is between about 1 to 3.
10. The ferroelectric material of claim 1, wherein the polyphenylene ether is poly(2,6-dimethyl-1,4-phenylene oxide).
11. The ferroelectric material of claim 1, wherein the ferroelectric polymer is polyvinylidene fluoride (PVDF), a poly(vinylidene fluoride-tetrafluoroethylene) co-polymer (P(VDF-TrFE)), or a poly(vinylidene-fluoride-co-hexafluoropropene) (P(VDF-HFP)).
12. The ferroelectric material of claim 1, wherein the material is a film.
13. The ferroelectric material of claim 12, wherein the thickness of the film is 20 nanometers to 10 microns.
14. The ferroelectric material of claim 12, wherein the film is a single monolayer film.
15. The ferroelectric material of claim 1, wherein the ferroelectric polymer is in crystalline or semi-crystalline form.
16. The ferroelectric material of claim 1, wherein the plurality of nanostructures are charge trap regions that are capable of storing charge.
17. The ferroelectric material of claim 1, wherein the size and number of the plurality of nanostructures are dependent on the amount by weight of the second polymer in the polymer blend.
18. The ferroelectric material of claim 1, wherein the plurality of amorphous nanostructures are nanospheres.
19. The ferroelectric material of claim 1, wherein the material is a liquid, a gel, or a melt.
20. The ferroelectric material of claim 1, wherein the material comprises from or 6 to 8% by weight of the second polymer.
21. A ferroelectric capacitor, wherein the ferroelectric capacitor comprises the ferroelectric material of claim 1, a first conductive material, and a second conductive material, wherein at least a portion of the ferroelectric material is disposed between at least a portion of the first conductive material and at least a portion of the second conductive material.
22. The ferroelectric material of claim 1 comprised in a printed circuit board.
23. The ferroelectric material of claim 1 comprised in an integrated circuit.
24. The ferroelectric material of claim 1 comprised in an electronic device.
25. The ferroelectric material of claim 1, wherein the a plurality of amorphous nanostructures has a size of less than 200 nm.
26. The ferroelectric material of claim 1, wherein the polymer blend comprises 1 to 4 wt. % of the second polymer.
27. A method of making the ferroelectric material of claim 1 comprising: (a) obtaining a solution comprising a solvent, the ferroelectric polymer, and the second polymer having a low dielectric constant, wherein the ferroelectric polymer and the second polymer having a low dielectric constant are dissolved in the solvent; (b) disposing the solution on a substrate; and (c) subjecting the solution to a heating or annealing step under conditions sufficient to obtain the ferroelectric material.
28. A method of making the ferroelectric material of claim 1 comprising: (a) obtaining the ferroelectric polymer and the second polymer having a low dielectric constant; (b) blending the ferroelectric polymer and the second polymer having a low dielectric constant in an extruder; and (c) melt extruding the ferroelectric polymer and the second polymer having a low dielectric constant under conditions sufficient to obtain the ferroelectric material.
29. A ferroelectric material consisting of a polymer blend of at least two polymers, wherein the first polymer is a ferroelectric polymer and the second polymer has a low dielectric constant, wherein the second polymer is a polyphenylene ether, and the polymer blend comprises 1 to 8 wt. % of the second polymer, and wherein the polymer blend comprises a crystalline or semi-crystalline polymeric matrix of the ferroelectric polymer and a plurality of amorphous nanostructures comprising the second polymer, wherein the plurality of amorphous nanostructures are comprised within the polymeric matrix.
30. The ferroelectric material of claim 29, wherein the ferroelectric polymer is a co-polymer.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
(1) FIGS. 1(a)-(b): Schematic 2-D (a) and 3-D (b) cross section of a ferroelectric capacitor incorporating the blend of a ferroelectric polymer and a second polymer having a low dielectric constant of the present invention.
(2) FIG. 2: Block diagram illustrating implementation of a circuit in a semiconductor wafer or an electronic device using ferroelectric capacitors of the present invention.
(3) FIG. 3: Block diagram showing an exemplary wireless communication system in which a ferroelectric capacitor of the present invention may be advantageously employed.
(4) FIG. 4: Chart showing hysteresis of a ferroelectric capacitor of the present invention (polarization over electric field).
(5) FIG. 5: Chart showing fatigue characteristics of a ferroelectric capacitor of the present invention (p* switching polarization over cumulative cycles).
(6) FIG. 6: Chart showing fatigue characteristics of a ferroelectric capacitor of the present invention (p^ non-switching polarization over cumulative cycles).
(7) FIG. 7: Chart showing fatigue characteristics of a ferroelectric capacitor of the present invention (dP over cumulative cycles).
(8) FIG. 8: Chart showing fatigue characteristics of a ferroelectric capacitor of the present invention (dP over cumulative cycles).
(9) FIG. 9: Chart showing breakdown characteristics of a ferroelectric capacitor of the present invention (breakdown field over resin loading).
(10) FIGS. 10(a)-(d): Charts showing thermal stability capabilities of a ferroelectric capacitor of the present invention
(11) FIGS. 11(a)-(c): (a) Schematic 3-D cross section of ferroelectric capacitors with phase separated blends of P(VDF-TrFE)-PPO sandwiched between Pt and Au electrodes. The morphology includes phase separated nanospheres of amorphous PPO, surrounded by P(VDF-TrFE) matrix. (b) Solutions with 0 wt %, 2 wt %, 4 wt %, 6 wt % and 8 wt % PPO (Left to Right) showing clear, homogenous and stable solutions. (c) Solution with 25 wt % PPO showing clear, homogenous and stable solution.
(12) FIGS. 12(a)-(d): (a) AFM phase image of pure P(VDF-TrFE) film showing island like grains in the film. (b) AFM phase image of as spun blend films with 6 wt % PPO without annealing. (c) AFM phase image of blend films with 6 wt % PPO after annealing at 135 C., with increase in grain size of P(VDF-TrFE). (d) R.sub.rms (Left) and peak height (Right) of blend films as a function of PPO loading. The inset shows average size of the PPO nanospheres as calculated from the AFM phase images.
(13) FIGS. 13(a)-(e): AFM topography images of blend films with 0 wt % (a), 2% (b), 4% (c), 6% (d) and 8 wt % PPO (e).
(14) FIGS. 14(a)-(e): AFM 3-D topography images of blend films with 0 wt % (a), 2% (b), 4% (c), 6% (d) and 8 wt % PPO (e).
(15) FIGS. 15(a)-(e): AFM phase images of blend films with 0 wt % (a), 2% (b), 4% (c), 6% (d) and 8 wt % PPO (e).
(16) FIGS. 16(a)-(d): Cross section TEM images of pure P(VDF-TrFE) films (a) and blend films with 6 wt % PPO (b,c,d).
(17) FIGS. 17(a)-(d): Cross section Energy Filtered TEM (EFTEM) images of pure P(VDF-TrFE) films (a) and blend films with 6 wt % PPO (b,c,d).
(18) FIGS. 18(a)-(b): (a) Grazing incidence XRD spectra for pure ferroelectric P(VDF-TrFE) films and blend films with 8 wt % PPO. (b) FT-IR spectra of pure ferroelectric P(VDF-TrFE) thin film, pure PPO thin film and blend films with 4 wt % and 8 wt % PPO.
(19) FIGS. 19(a)-(d): (a) Polarization-Electric Field hysteresis loop measurements for blend films at 10 Hz as a function of amount of PPO. (b) Switching current response from blend films with 0 to 8 wt % PPO and with Platinum/Gold electrodes. The inset shows switching characteristics for blend film at 125 MV/m, with peak of dP/d(log(t)) vs. log(t) representing respective switching times. (c) Dielectric spectroscopy study with dielectric constant (left axis) and dielectric losses (right axis) for blend films with 0 to 8 wt % PPO and pure PPO films. (d) Temperature dependence of dielectric permittivity for devices with 0, 4 and 8 wt % PPO at 1 KHz.
(20) FIGS. 20(a)-(d): (a) Remnant Polarization and Coercive field as a function of temperature for pure P(VDF-TrFE) films and blend films with 8 wt % PPO. (b) Polarization-Electric Field hysteresis loop measurements for pure P(VDF-TrFE) films at 10 Hz as a function of temperature. (c) Polarization-Electric Field hysteresis loop measurements for blend films with 8 wt % PPO at 10 Hz as a function of temperature. (d) Current density-Electric Field measurements of blend films (0, 8 wt % PPO) at 0 C. and 60 C. A voltage of approximately 15 V(125 MV/m) was applied to pole the devices before measuring the leakage current in the devices. The inset shows the leakage current density of pure PPO device with PtAu electrodes.
(21) FIGS. 21(a)-(d): (a) Electrical fatigue properties showing relative polarization of blend films with 0 to 8 wt % PPO. The films were stressed at 100 MV/m and a frequency of 100 Hz and the PUND measurements were done at saturation fields of 125 MV/m and 100 Hz. (b) Polarization-Electric Field hysteresis loop measurements for pure P(VDF-TrFE) films and 6 wt % PPO films before (BF) and after fatigue (AF), characterized also at 100 Hz. (c) Current density-Electric Field measurements of pure P(VDF-TrFE) films and 6 wt % PPO films after fatigue up to 106 cycles at 100 Hz (d) Dielectric breakdown strength (Left) and Fatigue or polarization retention after 106 cycles (Right) as a function of amount of PPO in blend films.
(22) FIG. 22: Fatigue performance of blend films with 0 wt % to 8 wt % PPO content. The films were stressed at 100 MV/m and a frequency of 1 KHz and the PUND measurements were done at saturation fields of 125 MV/m and 1 KHz.
(23) FIG. 23: Thermal breakdown and delamination of top Au electrode during fatigue of pure P(VDF-TrFE) capacitors. The devices were fatigued at 100 Hz and 12 V (100 MV/m).
DETAILED DESCRIPTION OF THE INVENTION
(24) Historically fatigue has been a significant problem with the use of polymer ferroelectric memory. In particular, injection of charges from electrodes which are subsequently trapped at crystallite boundaries and defects, inhibit ferroelectric switching and lead to higher fatigue rates.
(25) However, an improved ferroelectric material for capacitors has been discovered that addresses the drawbacks of current polymer ferroelectric memory. In particular, fatigue characteristics in a ferroelectric capacitor may be improved by blending a polymer having a low dielectric constant to a ferroelectric polymer. The blend can be produced by having these polymers dissolved in a common solvent, by melt blend extrusion or by other methods known in the art that can create a homogenous polymer blend. As illustrated in non-limiting embodiments in the Examples, this blend can reduce charge trapping, thereby leading to improvements in fatigue and dielectric breakdown characteristics of capacitors using said material.
(26) These and other non-limiting aspects of the present invention are discussed in further detail in the following sections.
(27) A. Ferroelectric Capacitor
(28) FIG. 1(a) is a 2-D cross-sectional view of a ferroelectric capacitor (1) comprising a ferroelectric material (12) of the present invention. For the purposes of FIG. 1(a), the ferroelectric material (12) is in the form of a film or layer. The ferroelectric capacitor (1) can include a substrate (10), a lower electrode (11), a ferroelectric material (12), and an end electrode (13). The ferroelectric capacitor can be fabricated on substrates by sandwiching a ferroelectric material (12) between two conducting electrodes (11) and (13). Additional materials, layers, and coatings (not shown) known to those of ordinary skill in the art can be used with the ferroelectric capacitor (1), some of which are described below. FIG. 1(b) is a 3-D cross-sectional view of the ferroelectric capacitor and does not include a substrate (10).
(29) The ferroelectric capacitor in FIG. 1 is said to have memory because, at zero volts, it has two polarization states that do not decay back to zero. These polarization states can be used to represent a stored value, such as binary 0 or 1, and are read by applying an electric field. The amount of charge needed to flip the polarization state to the opposite state can be measured and the previous polarization state is revealed. This means that the read operation changes the polarization state, and can be followed by a corresponding write operation, in order to write back the stored value by again altering the polarization state.
(30) 1. Substrate (10)
(31) The substrate (10) is used as support. It is typically made from material that is not easily altered or degraded by heat or organic solvents. Non-limiting examples of such materials include inorganic materials such as silicon, plastic, paper, banknotes as well as SABIC substrates including Polyethylene terephthalate, polycarbonates, and polyetherimide substrates.
(32) 2. Lower Electrode and Upper Electrodes (11) and (13)
(33) The lower electrode (11) is made of a conductive material. Typically, the lower electrode (11) is obtained by forming a film using such a material (e.g., vacuum deposition, sputtering, ion-plating, plating, coating, etc.). Non-limiting examples of conductive material that can be used to form a film include gold, platinum, silver, aluminum and copper, iridium, iridium oxide, and the like. In addition, non-limiting examples of conductive polymer materials include conducting polymers (such as PEDOT: PSS, Polyaniline, graphene etc.), and polymers made conductive by inclusion of conductive micro- or nano-structures (such as silver nanowires). The thickness of the film for the lower electrode (11) is typically between 20 nm to 500 nm.
(34) The upper electrode (13) can be disposed on the ferroelectric material (12) by thermally evaporating through a shadow mask. The material used for the upper electrode (13) can be conductive. Non-limiting examples of such materials include metals, metal oxides, and conductive polymers (e.g., polyaniline, polythiophene, etc.) and polymers made conductive by inclusion of conductive micro- or nano-structures such as those discussed above in the context of the lower electrode (11). The upper electrode (13) can be a single layer or laminated layers formed of materials each having a different work function. Further, it may be an alloy of one or more of the materials having a low work function and at least one selected from the group consisting of gold, silver, platinum, copper, manganese, titanium, cobalt, nickel, tungsten, and tin. Examples of the alloy include a lithium-aluminum alloy, a lithium-magnesium alloy, a lithium-indium alloy, a magnesium-silver alloy, a magnesium-indium alloy, a magnesium-aluminum alloy, an indium-silver alloy, and a calcium-aluminum alloy. The film thickness of the upper electrode (13) is typically between 20 nm to 500 nm.
(35) 3. Ferroelectric Material (12)
(36) The ferroelectric material (12) can be interposed between the lower electrode (10) and the upper electrode (13). In one instance, the material (12) can be obtained from a blend of a ferroelectric polymer and a polymer having a low dielectric constant, wherein the polymers have been solubilized in the same solvent or solvent system. In the examples, the ferroelectric polymer that was used was P(VDF-TrFE), and the polymer with a low dielectric constant that was used was a polyphenylene ether sold under the trade name Noryl SA90 Resin by SABIC Innovative Plastics Holding BV (SABIC) (Pittsfield, Mass., USA). Noryl SA90 Resin has the following structure.
(37) ##STR00003##
The Noryl SA90 Resin is in powedered form. It has an intrinsic viscosity (IV) of about 9 ml/g, a molecular weight (MW) of about 1700 g/mol, a glass transistion temperature (Tg) of 135 C., and a specific gravity of 1.02 g/cm.sup.3. However, and as noted above, other types of ferroelectric polymers and low dielectric constant polymers are also contemplated as being useful. By way of example only, all of the various SABIC PPO* Resins and Noryl Resins (e.g., polyphenylene ether and high impact polystyrene, polyphenylene ether and polystyrene, etc.) are also contemplated as being useful in the context of the present invention. For instance, SABIC's Noryl SA9000 Resin can be used, which has the following structure:
(38) ##STR00004##
The Noryl SA9000 Resin is in powedered form. It has an intrinsic viscosity (IV) of about 9 ml/g, a molecular weight (MW) of about 1700 g/mol, a glass transistion temperature (Tg) of 154 C., and a specific gravity of 1.02 g/cm.sup.3. The amount of the polymer with a low dielectric constant that can be added to the ferroelectric material (12) can vary so as to achieve a desired result or fatigue characteristic of the resulting capacitor (1). For instance, the amount of the polymer with a low dielectric constant can range from 1 wt % to 50 wt % based on the total weight of the ferroelectric material (12). This allows for the creation of capacitors to have pre-defined fatigue characteristics in a relatively easy and scalable manner. Without wishing to be bound by theory, it is believed that by blending a polymer having a low dielectric constant with a ferroelectric polymer results in a blend that includes good charge trap regions and does not allow charge carriers to move freely through the film. This directly affects thermal stability of the blend, thus leading to reliably switching the polarization even at elevated temperatures close to the Curie temperature. These characteristics can be modified by increasing or decreasing the amount of the low dielectric constant polymer present in the ferroelectric material (12).
(39) The blend film layer can be deposited by obtaining a solution that includes a solvent and the polymers of the present invention solubilized therein. The blend solutions is prepared in a common solvent which dissolves both ferroelectric polymer and polymer having the low dielectric constant. Non-limiting examples of such solvents include Methyl Ethyl Ketone, Di-methylformamide, Acetone, Di-mehtyl sulfoxide, Cyclohexanone, Tetrahydrofuran (THF). The solution can be deposited by doctor blade coating, spin coating, meniscus coating, transfer printing, ink jet printing, offset printing, screen printing process, dip coating, casting, bar coating, roll coating, wire bar coating, spraying, screen printing, gravure printing, flexo printing, offset printing, gravure offset printing, dispenser coating, nozzle coating, capillary coating, etc. Alternatively, and as explained above, other processes such as melt blend extrusion can also be used to create the polymer blend
(40) As illustrated in FIG. 1(b), the ferroelectric material (12) can be such that the ferroelectric polymer forms a crystalline or semi-crystalline polymer matrix (12a). The polymer having a low dielectric constant can phase separate from the matrix (12a) such that a plurality of separated regions (12b) are formed within the matrix that include the polymer having the lower dielectric constant. The polymer within these regions (12b) can be in amorphous form. The regions (12b) can have a variety of sizes (e.g., nano-sized, micro-sized, etc.) and shapes (e.g., spherical, substantially spherical, etc.) and can be distributed throughout the polymeric matrix (12a). The plurality of regions (12b) are capable of acting as charge-trap regions to store charge (e.g., positive charge, negative charge, etc.). The plurality of regions (12a) can also reduce or prevent charge carriers to move freely throughout the polymeric matrix (12a). As explained above and in the examples, these features can improve the thermal stability of the ferroelectric material (12) and lead to reliable switching of the polarization even at elevated temperatures close to the Curie temperature. Such features can be modified or tuned as desired by increasing or decreasing the amount of the low dielectric constant polymer that is included the ferroelectric material (12).
(41) B. Embodiment of Process for Producing Ferroelectric Capacitor
(42) For example, with reference to FIG. 1, a ferroelectric capacitor (1) can be fabricated on a silicon substrate (10) by disposing a ferroelectric material (12) between two conducting electrodes (11) and (13). A Pt-coated silicon substrate can be used and cleaned with acetone, IPA and DI water prior to device fabrication. A 2 wt % pure solution of P(VDF-TrFE) can be prepared by dissolving 20 mg/mL of the P(VDF-TrFE) copolymer in methyl-ethyl-ketone (MEK) solvent. Blend solutions of P(VDF-TrFE) and the Noryl SA90 Resin can be made by adding different amounts of Resin to the P(VDF-TrFE) solution. The amount may vary as desired to alter the concentration of the Noryl SA90 Resin from 1 wt % to 50 wt %. Various substitutions, modifications, additions, and/or rearrangements within the spirit and/or scope of the underlying inventive concept will become apparent to those skilled in the art from this disclosure. For example, the concentration of P(VDF-TrFE) can also be varied from 0.1 wt % to 50 wt % depending on the desired thickness of the film.
(43) The ferroelectric material (12) can be spun from the solution by spin coating at 4000 rpm for 60 seconds. The spin coating can be varied from 500 rpm to 8000 rpm and 10 seconds to 100 seconds. The film can then be annealed on a hotplate for 30 minutes at 80 C. prior to annealing in a vacuum furnace at 135 C. for 4 hours to improve the crystallinity. Annealing time can be varied from 30 minutes to 8 hours. Finally, a top gold electrode can be thermally evaporated through a shadow mask.
(44) The process of the present invention can efficiently produce high-performance ferroelectric capacitors in large-scale quantities.
(45) C. Applications for Ferroelectric Capacitor
(46) Any one of the ferroelectric capacitors of the present invention can be used in a wide array of technologies and devices including but not limited to: smartcards, RFID cards/tags, memory devices, non-volatile memory, standalone memory, firmware, microcontrollers, gyroscopes, acoustics sensors, actuators, microgenerators, power supply circuits, circuit coupling and decoupling, RF filtering, delay circuits, and RF tuners. If implemented in memory, including firmware, functions may be stored in the ferroelectric capacitors as one or more instructions or code on a computer-readable medium. Examples include computer-readable media encoded with a data structure and computer-readable media encoded with a computer program. Computer-readable media includes physical computer storage media. Combinations of the above should also be included within the scope of computer-readable media.
(47) In many of these applications thin films of ferroelectric materials are typically used, as this allows the field required to switch the polarization to be achieved with a moderate voltage. Although some specific circuitry has been set forth, it will be appreciated by those skilled in the art that not all of the disclosed circuitry is required to practice the disclosure. Moreover, certain well known circuits have not been described, to maintain focus on the disclosure.
(48) Improvements in the dielectric breakdown of the ferroelectric capacitor of the present invention also allow for improved energy storage characteristics which can be used in many applications. As the polymer with a low dielectric constant (e.g., Noryl SA90 Resin, Noryl SA9000 Resin, Noryl PPO* Resins, etc.) wt % is increased for the ferroelectric material of the present invention, the electric breakdown field also increases. As a result, pursuant to the energy density equation below, the maximum volumetric energy density stored in the ferroelectric capacitor also increases:
U.sub.max=E.sub.BD.sup.2.sub.0.sub.r E.sub.BD=Electric breakdown field strength (V m.sup.1) .sub.0=permittivity of free space=8.85410.sup.12 F m.sup.1 .sub.r=Dielectric permittivity.
(49) These energy storage improvements can be incorporated in numerous applications including energy/charge storing devices for electric flashes, defibrillator, electromagnetic forming, Marx generators, pulsed lasers (including TEA lasers), pulse forming networks, radar, as well as elements in electric circuits including frequency tuners and filters in power supplies. In addition, these ferroelectric capacitors may be used in electronic devices containing volatile memory which require a power supply to prevent the loss of information if a power source is changed or removed.
(50) FIG. 2 is block diagram illustrating implementation of an integrated circuit in a semiconductor wafer or an electronic device according to one embodiment. In one case, a ferroelectric capacitor (1) with improved fatigue characteristics as discussed above may be found in a wafer (20). The wafer (20) may be singulated into one or more dies that may contain the ferroelectric capacitor (1). Additionally, the wafer (20) may experience further semiconductor manufacturing before singulation. For example, the wafer (20) may be bonded to a carrier wafer, a packaging bulk region, a second wafer, or transferred to another fabrication facility. Alternatively, an electronic device (26) such as, for example, a personal computer, may include a memory device (24) that includes the ferroelectric capacitor (1). Additionally, other parts of the electronic device (26) may include the ferroelectric capacitor (1) such as a central processing unit (CPU), a digital-to-analog converter (DAC), an analog-to-digital converter (ADC), a graphics processing unit (GPU), a microcontroller, or a communications controller.
(51) FIG. 3 is a block diagram showing an exemplary wireless communication system (40) in which an embodiment of the disclosure may be advantageously employed. For purposes of illustration, FIG. 3 shows three remote units (42), (43), and (45) and two base stations (44). It will be recognized that wireless communication systems may have many more remote units and base stations. Remote units (42), (43), and (45) include circuit devices (42A), (42C) and (42B), which may comprise integrated circuits or printable circuit boards, that include the disclosed ferroelectric capacitor. It will be recognized that any device containing an integrated circuit or printable circuit board may also include the ferroelectric capacitor disclosed here, including the base stations, switching devices, and network equipment. FIG. 3 shows forward link signals (48) from the base station (44) to the remote units (42), (43), and (45) and reverse link signals (49) from the remote units (42), (43), and (45) to base stations (44).
(52) In FIG. 3, remote unit (42) is shown as a mobile telephone, remote unit (43) is shown as a portable computer, and remote unit (45) is shown as a fixed location remote unit in a wireless local loop system. For example, the remote units may be mobile phones, hand-held personal communication systems (PCS) units, portable data units such as personal data assistants, GPS enabled devices, navigation devices, set upper boxes, music players, video players, entertainment units, fixed location data units such as meter reading equipment, or any other device that stores or retrieves data or computer instructions, or any combination thereof. Although FIG. 3 illustrates remote units according to the teachings of the disclosure, the disclosure is not limited to these exemplary illustrated units. Embodiments of the disclosure may be suitably employed in any device which includes the ferroelectric capacitor (1).
(53) Although the present disclosure and its advantages have been described in detail, it should be understood that various changes, substitutions and alterations can be made herein without departing from the technology of the disclosure as defined by the appended claims. Moreover, the scope of the present application is not intended to be limited to the particular embodiments of the process, machine, manufacture, composition of matter, means, methods and steps described in the specification. As one of ordinary skill in the art will readily appreciate from the disclosure, processes, machines, manufacture, compositions of matter, means, methods, or steps, presently existing or later to be developed that perform substantially the same function or achieve substantially the same result as the corresponding embodiments described herein may be utilized according to the present disclosure. Accordingly, the appended claims are intended to include within their scope such processes, machines, manufacture, compositions of matter, means, methods, or steps.
EXAMPLES
(54) The present invention will be described in greater detail by way of specific examples. The following examples are offered for illustrative purposes only, and are not intended to limit the invention in any manner. Those of skill in the art will readily recognize a variety of noncritical parameters which can be changed or modified to yield essentially the same results.
(55) For each example, a ferroelectric capacitor was fabricated on silicon substrates by sandwiching a blend film of a ferroelectric polymer [P(VDF-TrFE)] obtained from Piezotech S.A., France) and a polyphenylene ether polymer (Noryl SA90 Resin from SABIC Innovative Plastics Holding BV (Pittsfield, Mass., USA, was used in the Examples) between a first conducting electrode coated with Pt and a second conducting electrode made of Au. A 2 wt % pure solution of P(VDF-TrFE) was prepared by dissolving 20 mg/mL of the P(VDF-TrFE) copolymer in methyl-ethyl-ketone (MEK) solvent. Blend solutions of P(VDF-TrFE) and the polyphenylene ether were made by adding different amounts of the polyphenylene ether to the P(VDF-TrFE) solution. The polyphenylene ether was solubilized in each prepared solution. The amount varied from 4 mg to 17.39 mg to vary the concentration of the polyphenylene ether from 2 wt % to 8 wt %.
Example 1
Hysteresis (Polarization Versus Electric Field) of Non-Volatile Ferroelectric Memory Capacitors Using Blend Thin Films of Ferroelectric Polymer P(VDF-TrFE) and Polyphenylene Ether
(56) The concentration of the polyphenylene ether for the ferroelectric capacitors was varied from 0 to 8 wt %. With increasing wt % of the polyphenylene ether the polarization steadily decreases and the coercive field increases but still shows a typical hysteresis loop (see FIG. 4). Even with concentrations up to 8 wt % polyphenylene ether, there are two distinguishable polarization states (+Pr and Pr) needed for memory applications. The thin films were spun from the solution by spin coating at 4000 rpm for 60 seconds. The films were then annealed on a hotplate for 30 minutes at 80 C. prior to annealing in a vacuum furnace at 135 C. for 4 hours to improve the crystallinity. Finally, the upper gold electrode was thermally evaporated through a shadow mask.
Example 2
Fatigue Characteristics of Blend Films with 0 to 6 Wt % Polyphenylene Ether Content
(57) The fatigue test was carried out at 100 Hz with an applied field of 100 MV/m. The plot shows the total switching polarization (P*) versus cumulative cycles for the ferroelectric capacitor. With increasing polyphenylene ether content the fatigue improves. Pure P(VDF-TrFE) ferroelectric capacitors show 52% retention of polarization after 10.sup.6 cycles but it improves up to 90% upon using 6% polyphenylene ether (see FIG. 5).
Example 3
Fatigue Characteristics of Blend Films with 0 to 6 Wt % Polyphenylene Ether Content
(58) The fatigue test was carried out at 100 Hz with an applied field of 100 MV/m. The plot shows the total non-switching polarization (PA) versus cumulative cycles (see FIG. 6). The increase in non-switching polarization with number of cycles is due to injection of charges from electrodes which are subsequently trapped at crystallite boundaries and defects, inhibiting ferroelectric switching. Upon adding polyphenylene ether, the devices show lower rise in this non-switching polarization with number of cycles. Thus this leads to the better fatigue performance of our blend films.
Example 4
Fatigue Characteristics of Blend Films with Polyphenylene Ether
(59) A fatigue test was carried out at 100 Hz with an applied field of 100 MV/m for blend films of P(VDF-TrFE) and polyphenylene ether, in which the amounts of polyphenylene ether varied from 0% to 6% (see FIG. 7). FIG. 7 shows the relative remnant polarization (dP=P*P^) versus cumulative cycles. With increasing polyphenylene ether content the fatigue improves considerably. Pure P(VDF-TrFE) ferroelectric capacitors show only 20% retention of polarization after 10.sup.6 cycles but it improves up to 58% upon using 6% polyphenylene ether.
(60) Another fatigue test was carried out at 1 kHz with an applied field of 100 MV/m for blend films of P(VDF-TrFE) and polyphenylene ether, in which the amounts of polyphenylene ether varied from 0% to 8% (see FIG. 8). PUND measurements were performed at saturation fields of 125 MV/m and 100 Hz. FIG. 8 shows the relative remnant polarization (dP=P*P^) versus cumulative cycles. With increasing polyphenylene ether content the fatigue improves considerably. Pure P(VDF-TrFE) ferroelectric capacitors show only 54% retention of polarization after 10.sup.6 cycles but it improves up to 80% upon using 8% polyphenylene ether.
Example 5
Dielectric Breakdown Field with Polyphenylene Ether Loading Wt %
(61) Capacitors under different polyphenylene ether loadings were all of equivalent thicknesses. An initial DC pulse (20 V) was applied to pole all devices before breakdown measurement. DC bias was applied at a ramp rate of 3 V/s until device failure. Plotted data points show average breakdown field for 10 devices/sample including error bars. As the plotted data points show, as polyphenylene ether loading wt % increased, the electric breakdown field also increases (see FIG. 9).
Example 6
Thermal Stability of P(VDF-TrFE)-Polyphenylene Ether Blends
(62) A study was conducted to determine the thermal stability of pure P(VDF-TrFE) (70/30 mol %) and blends of P(VDF-TrFE) (70/30 mol %) and 8 wt % polyphenylene ether content in a capacitor. The devices were studied for the ability to switch polarization at an applied field of 125 MV/m and frequency of 10 Hz. FIG. 10(a) shows the remnant polarization and coercive field as a function of temperature. In general it shows a slight increase in polarization and decrease in coercive field with increasing temperature. The pure P(VDF-TrFE) capacitors showed very poor thermal stability when compared with the blend capacitors. FIG. 10(b) shows that pure P(VDF-TrFE) at temperatures of 60 C. or above, the hysteresis loops displayed a resistive leaky behavior making it impossible to accurately determine polarization in these films. At higher temperatures leaky curves were observed especially in the negative bias regime, indicative of surface breakdown at one of the electrode/ferroelectric interfaces.
(63) By comparison, the blend of P(VDF-TrFE) (70/30 mol %) and 8 wt % polyphenylene ether showed much better thermal stability compared to pure P(VDF-TrFE) capacitors (FIG. 10(a)). The polarization increased as a function of temperature while the coercive field decreases upto 80 C. (FIG. 10(c)). At higher temperatures above 80 C. which are very closely to Curie temperature of the copolymer the polarization drops sharply to 0, well known for thin film ferroelectric capacitors. To further understand this, a study was performed on the leakage current of pure P(VDF-TrFE) films and blend films as a function of temperature. Leakage of ferroelectric capacitors based on the copolymer has been well studied and shows a relatively high leakage for thin films around 100 nm. The introduction of TrFE is very effective in obtaining the ferroelectric phase in the copolymer but also leads to larger leakage current. If an electric field to switch polarization is applied, current leakage occurs easily at the TrFE monomer because two fluorine atoms opposite from the carbon atoms of the TrFE monomer cause a current leak path. This is evident for thin films P(VDF-TrFE) capacitors with high TrFE content and several studies have reported large leakage issues. FIG. 10(d) shows that at saturation fields of 125 MV/m pure P(VDF-TrFE) capacitors show leakage current density of in excess of 10.sup.6 A/cm.sup.2 at room temperature. On the other hand pure polyphenylene ether films display low leakage current of 10.sup.8 A/cm.sup.2 even at high fields 300 MV/m further highlighting the excellent insulating properties of PPO. The blend devices with 8 wt % polyphenylene ether show similar leakage currents with slightly lower currents on the negative bias. Without wishing to be bound by theory, it is believed that this is due to the current conduction through the more leaky majority ferroelectric phase. Notably, a drastic improvement in leakage current of the blend films compared to pristine P(VDF-TrFE) films at higher temperatures of 60 C. was observed. This leakage current density of the blend films does not change much with temperature and is an order of magnitude lower than P(VDF-TrFE) films at 60 C. This is likely due to the highly insulating amorphous nanospheres of polyphenylene ether in the blend films acts as good charge trap regions and do not allow charge carriers to move freely through the film. This directly affects thermal stability of these blend films, thus leading to reliably switching the polarization even at elevated temperatures close to the Curie temperature.
Example 7
Additional Data of Prepared Films and Devices
(64) High performance polymer memory films were fabricated using blends of ferroelectric poly(vinylidene-fluoride-trifluoroethylene) (P(VDF-TrFE)) and highly insulating Poly(p-phenylene Oxide) (PPO). The blend films spontaneously phase separate into amorphous PPO nanospheres embedded in a semicrystalline P(VDF-TrFE) matrix. Using low molecular weight PPO with high miscibility in a common solvent (i.e., Methyl Ethyl Ketone), spin cast blended films with low roughness (R.sub.rms4.92 nm) and nano-scale phase separation (PPO domain size <200 nm) were produced. These blend devices display highly improved ferroelectric and dielectric performance with low dielectric losses (<0.2 up to 1 MHz), enhanced thermal stability (upto 353 K), excellent fatigue endurance (80% retention after 10.sup.6 cycles at 1 KHz) and high dielectric breakdown fields (360 MV/m).
(65) A. Materials and Methods
(66) Sample Preparation:
(67) The polymer blend thin films were fabricated on Platinum coated silicon substrates. Prior to device fabrication, the substrates were cleaned by ultra-sonication in Acetone, Isopropanol and DI water respectively. P(VDF-TrFE) (70/30 mol. %) obtained from Piezotech S.A, France was dissolved in anhydrous Methyl Ethyl Ketone (MEK) at a concentration of 20 mg/mL to make a 2 wt % solution. High purity low molecular weight Polyphenylene Oxide (Noryl SA90 PPO) (M.sub.n1800) obtained from Saudi Basic Industries Corporation (SABIC) was dissolved in 10 mL P(VDF-TrFE) solutions by varying the amounts (4.08 mg, 8.22 mg, 12.76 mg, 17.39 mg, 27.27 mg) to make 2 wt % to 8 wt % P(VDF-TrFE)-PPO blend solutions. All the different concentrations of PPO formed clear homogenous solutions stable even after a few weeks. The filtered polymer blend films were spun in a nitrogen filled glove box, at 4000 rpm for 60 seconds followed by a soft bake for 20 min at 70 C. The films were then annealed in vacuum at 135 C. for 4 hours to improve the crystallinity of the P(VDF-TrFE) phase. The thickness of the blend films was 12010 nm as measured by a Dektak profilometer, and did not change much with increasing PPO concentrations. To complete the device, 80 nm Gold (Au) was thermally evaporated through a shadow mask to define the top electrodes.
(68) Characterization of Films:
(69) All current-voltage measurements were carried out in air ambient using Keithley 4200 semiconductor characterization system, while Polarization-Voltage and fatigue tests were done using the Premier Precision II ferroelectric tester (Radiant Technologies Inc.). Surface morphology and roughness for the blend films was studied using Atomic Force Microscopy (Agilent 5400). Cross section morphology of the devices was studied using Transmission Electron Microscopy (Titan ST) and operated at an accelerating voltage of 300 kV. Energy Filtered TEM analysis was done to elementally map carbon in the polymer blend films. The crystallinity and inter-planar spacing of polymer chains was evaluated using Grazing Incidence X-ray Diffraction (Bruker D8 Discover) while the bonding and dipole orientation was analyzed using Fourier-transform infrared spectroscopy (FT-IR, ThermoScientific Nicolet iS10).
(70) B. Results
(71) 1. Morphology
(72) A schematic of the nanoscale phase-separated polymer blend devices is shown in FIG. 11(a). The active single-layer is a blend film spin-cast from a solution of ferroelectric P(VDF-TrFE) and insulating PPO from a common solventMethyl Ethyl Ketone. The morphology of the blend films includes a phase separated nanospheres of amorphous PPO, embedded in a semicrystalline, ferroelectric P(VDF-TrFE) matrix. The P(VDF-TrFE)-PPO blends are highly miscible and stable in a large range of compositions from 0 to 25 wt % of PPO content. FIG. 11(b) shows clear homogenous solutions from 2 wt % to 8 wt %. Solutions with up to 25 wt % PPO were made and remained stable up to one month (FIG. 11(c)). FIG. 11(a) also shows the chemical structure of P(VDF-TrFE) and PPO. PPO i.e. Poly(p-phenylene oxide) is an aromatic polyether with oxygen connected to aromatic aryl groups. Ethers are slightly polar in nature as the COC bond angle in the functional group is about 110 degrees, and the CO dipole does not cancel out. The presence of two lone pairs of electrons on the oxygen atoms makes hydrogen bonding with water and other polar molecules possible. It is believed that hydrogen bonding between the electronegative oxygen in PPO and electropositive hydrogen in P(VDF-TrFE) and the electropositive hydrogen in the methyl group of PPO and electronegative fluorine in P(VDF-TrFE) leads to the miscibility of these blends.
(73) The surface morphology and phase separation of 120 nm thick polymer blend films spun on Pt/Si substrates was characterized using atomic force microscopy (AFM as shown in FIG. 12. FIG. 12(a) shows the typical morphology of P(VDF-TrFE) thin films with crystalline grains about 80-100 nm in size. FIG. 12(b) shows spun cast blend films with 6 wt % PPO. These films were not subjected to any annealing process. The blends phase separate into amorphous PPO nanospheres (140 nm in size) randomly distributed throughout the films surrounded by the semi crystalline P(VDF-TrFE) matrix. The AFM study confirmed that phase separation for these blends was spontaneous and not thermally stimulated. FIG. 12(c) shows the phase image of the same blend film after annealing at 135 C. for 4 hours. After annealing, an increase in grain size of the semi crystalline P(VDF-TrFE) was observed, which is indicative of higher crystallinity. However, there was no significant change in the microstructure of PPO nanospheres or the roughness of the blend films. AFM measurements were performed as a function of blending ratio from 0 to 8 wt % PPO content. By increasing the amount of PPO the average lateral size of the PPO nanospheres increased in a linear fashion from 80 nm in a 2 wt % films, 105 nm (4 wt %), 140 nm (6 wt %) and 165 nm for 8 wt % film as depicted in the inset of FIG. 12(d). This was calculated from the AFM phase images of blend films with different PPO loadings. Furthermore the number of PPO nanospheres decreased with increasing PPO content (see FIGS. 12-15), i.e. the phase separation coarsened with increasing PPO content. These observations appear to rule out that the solidification process is due to nucleation and growth. It is believed that the phase separation is due to spinodal decomposition in which the separation occurs uniformly throughout the film and not at distinct nucleation sites. The surface roughness of ferroelectric thin films can be a key parameter when fabricating ferroelectric memory. High surface roughness leads to non-uniform electrical field across the active layer and possibly poor yield and reproducibility for ferroelectric capacitors and low mobility and low ON/OFF ratios in ferroelectric transistors (see Naber et al., Adv. Mater. 2010. 22, 933; Asadi et al. Adv. Funct. Mater. 2011, 21, 1887; Khan et al. Adv. Funct. Mater. 2012, 23, 2145). The measured surface roughness from the topography images of these blend films show relatively smooth films with an increase in roughness from 2 nm for pure P(VDF-TrFE) films to 5 nm for a blend film with 8 wt % PPO. As seen from FIG. 12(d), with increasing amounts of PPO the peak height of the amorphous PPO nanospheres increased hence leading to increased roughness. This can be further optimized and improved using techniques like temperature assisted wire-bar coating which has recently been used to fabricate smooth, polymer blend thin films. (see Li et al. Macromolecules. 2012, 45, 7477; Li et al. Adv. Funct. Mater. 2012, 22, 2750). FIG. 13, FIG. 14, and FIG. 15 provide additional AFM topography images, AFM 3-D topography images, and AFM phase images, respectively, of blend films with 0, 2, 4, 6, and 8 wt % PPO.
(74) The cross section morphology and phase separation was characterized through transmission electron microscopy (TEM) as seen in FIG. 16. FIG. 16(a) shows the cross section TEM image of a 120 nm thick pure P(VDF-TrFE) thin film. FIGS. 16(b)-(d) show the cross sections of 6 wt % PPO blend film sandwiched between Pt and Au electrodes at different locations. The TEM images confirm the morphology of these blend films consists of phase separated nanospheres of amorphous PPO, surrounded by a ferroelectric P(VDF-TrFE) matrix as seen by AFM. The observed phase separation was seen in multiple locations throughout the film, as shown in FIGS. 16(c) and (d). FIGS. 17(a)-(d) provide energy filtered cross section TEM images of the films with 0 and 6 wt % PPO.
(75) 2. Crystal Structure and Orientation
(76) FIG. 18(a) shows the Grazing Incidence X-ray diffraction (GIXRD) spectra which was used to study the crystal structure of pure ferroelectric P(VDF-TrFE) and the polymer blends with PPO. Pristine P(VDF-TrFE) films spun on Pt electrodes and annealed at 135 C. exhibit a peak centered at 219.76, characteristic of the ferroelectric phase and reflection from the (110) and (200) planes (see Park et al., IEEE. T. DIELECT. EL. IN. 2010, 17, 1135; Khan et al., Org. Electron. 2011, 12, 2225; Tajitsu et al., Jpn. J. Appl. Phys. 1987, 26, 554). The inter-planar distance was calculated to be approximately 4.14 and is consistent with earlier reports (see Khan et al., Org. Electron. 2011, 12, 2225; Bhansali et al., Org. Electron. 2012, 13, 1541). The broad peak full-width-half-maxima)(FWHM1.59 is typical of a semicrystalline polymer like P(VDF-TrFE) comprising of crystalline lamella and amorphous regions. Highly crystalline ferroelectric thin films are advantageous, as only the -crystalline regions in the films give rise to ferroelectricity because the dipole moments in the amorphous regions will be random and cancel out each other (see Naber et al. Adv. Mater. 2010, 22, 933). X-ray diffraction is the primary technique to determine crystallinity of semicrystalline polymers and has been previously used for P(VDF-TrFE) thin films (see Tajitsu et al., Jpn. J. Appl. Phys. 1987, 26, 554). The determination of the degree of crystallinity implies use of a two-phase model, i.e. the sample is composed of crystals and amorphous regions and no regions of semi-crystalline organization. The diffraction peak observed could be well resolved into two peaks, C (Crystalline) and NC (Non crystalline). A gaussian function was used to obtain the best fitting. The degree of crystallinity can be calculated from the ratio of area under C to total area under C+N. The calculated degree of crystallinity for pure P(VDF-TrFE) was 74%, typical of very thin (100-200 nm) P(VDF-TrFE) films (see Zeng et al., Chinese. J. Polym. Sci. 2009, 27, 479-485).
(77) FIG. 18(b) shows the XRD peak for blend films with 8% PPO content. The blend films with PPO exhibit a peak slightly shifted to the right at 220 for 8 wt % PPO films, indicating a smaller polymer chain inter-planar distance of 4.08 . The polymer blend films phase separate and thus it is possible that with increasing PPO content there is more stress on the P(VDF-TrFE) phase leading to closer chain packing or smaller interplanar distance. Furthermore, the XRD peaks indicate that the polymer blend films have smaller crystallite size compared to the pristine P(VDF-TrFE) films with a larger FWHM2. Lower crystallinity for blend films with increasing PPO content was observed, and was approx 62% for 8 wt % PPO films.
(78) The presence of PPO in thin films of the polymer blend was verified using transmission mode Fourier Transform Infra-Red (FTIR) spectroscopy. FIG. 18(b) shows the absorbance bands at 1288 cm.sup.1 and 850 cm.sup.1 associated with CF.sub.2 symmetric stretching vibration and are characteristic bands of the trans-zigzag formation ( phase) (see Khan et al., Org. Electron. 2011, 12, 2225; Prabu et al., Vib. Spectrosc. 2009, 49, 101; Reynolds et al., Macromolecules. 1989, 22 1092). Other major peaks identified are the 1400 cm.sup.1 band characteristic of the CH.sub.2 wagging vibrations, 1186 cm.sup.1 band characteristic of asymmetric stretching of CF.sub.2 and the 880 cm.sup.1 band related to the rocking CH.sub.2 vibration (see Khan et al. 1989). All these peaks were common in both pristine P(VDF-TrFE) and P(VDF-TrFE)-PPO blended films. A few additional peaks were identified in the blend films at 1605 cm.sup.1 characteristic of CCC symmetric stretching in the benzene ring, 1473 cm.sup.1 from CCC asymmetric stretching and 1020 cm.sup.1 from CO stretching confirming the preseence of ether group in PPO (S. B. Gajbhiye, IJMER. 2012, 2, 941; Kim et al., Nanotechnology. 2005, 16, S382; Shimizu et al., Biosci. Biotechnol. Biochem. 2001, 65, 990). Notably, the methyl functional groups were not deted in the FTIR spectra which might be due to some overlapping with other peaks or due to the poor resolution of the FTIR equipment. FTIR analysis confirms the presence of PPO in these polymer blend films but does not suggest any interaction or bonding between the PPO and P(VDF-TrFE) chains.
(79) 3. Ferroelectric and Dielectric Performance of Blend Films
(80) FIG. 19(a) shows the polarization-electric field hysteresis loops for P(VDF-TrFE)-PPO blend devices. The devices measured at 10 Hz exhibit well-saturated hysteresis curves and pure P(VDF-TrFE) capacitors show a remnant polarization (P.sub.r) of 7.3 C/cm.sup.2 and coercive field of 625 MV/m. With increasing PPO content, a monotonic decrease in remnant polarization and increase in coercive fields is observed. Blend films with 8 wt % PPO exhibit a remnant polarization (P.sub.r) of 4.93 C/cm.sup.2 and coercive field of 675 MV/m. This effect can be attributed to the decrease in crystallinity of the films upon adding PPO, as seen from the x-ray diffraction peaks in FIG. 18(a).
(81) In ferroelectric capacitors, the difference between switching and non-switching current should be maximized to be able to distinguish the 0 from 1 memory state. FIG. 19(b) shows that with increasing PPO content the switching current gradually decreases. But even with high amounts of PPO content up to 8 wt %, the blend capacitors display good switching current density 15 A/cm.sup.2; comparable to reports of pure P(VDF-TrFE) based ferroelectric capacitors (see Lee et al., Appl. Phys. Lett. 2009, 94, 093304; Noh et al., Appl. Phys. Lett. 2007, 90, 253504). At the same time the switching time characteristics of the P(VDF-TrFE)-PPO blend films were measured, which can be obtained by a time domain measurement of the charge density or polarization (P) response. Switching times (.sub.s) are estimated from the time of the maximum of dP/d(log t) vs. log (t) plot and are plotted in the inset of FIG. 15(b) (see Furukawa et al., IEEE. T. DIELECT. EL. IN. 2006, 13, 1120). At applied fields of 125 MV/m, pure P(VDF-TrFE) capacitors devices exhibit switching times of 0.19 ms while devices with 8 wt % PPO have similar switching times of 0.21 ms. Thus switching times do not vary significantly with increase in PPO content, another important aspect for ferroelectric memories.
(82) An advantage of using PVDF-based ferroelectric polymers is their high capacitance or high permittivity coming from their ability to polarize under an applied electric field. This makes it possible to fabricate devices with low operating voltages using them as a dielectric layer. A gate dielectric with high permittivity reduces the operating voltage of OTFTs effectively without the need for thickness reduction (see Jung et al., J. Appl. Phys. 2010, 108, 102810). Thus, it is important to characterize the effect of PPO on dielectric dispersion of these blend films. FIG. 19(c) shows the dielectric dispersion and the loss factor (tan ) of pure P(VDF-TrFE), pure PPO and blend ferroelectric capacitors. The P(VDF-TrFE) copolymer films exhibit a dielectric constant of 11 at 100 Hz, comparable to other reports in literature (see Khan et al., Appl. Phys. Lett. 2012, 101, 143303). A gradual decay of the dielectric permittivity (E.sub.r) is observed for pure P(VDF-TrFE) capacitors, consistent with the dielectric response of a polar polymer dielectric where the dipoles cannot respond to the applied field at high frequencies. On the other hand PPO exhibits a dielectric constant of 3 and a dielectric response which is independent of frequency, typical of low dielectric constant polymer dielectrics. In such materials the electronic polarization is the majority contributor to the overall permittivity and its response to the frequency of the applied field is almost instantaneous. FIG. 19(c) also shows the dielectric dispersion of the blend capacitors. With increasing amounts of PPO, the permittivity gradually drops but is relatively high for low voltage electronic applications; an 8 wt % blend film has a .sub.r9.3. FIG. 19(c) also shows the dielectric losses (tan ) calculated from the ratio of the imaginary and real part of the dielectric constant indicating power dissipation from the dielectric layer. An ideal dielectric would be one with high permittivity and low losses for electronic applications. The blend films with 8 wt % PPO show lower dielectric losses (0.17 at 1 MHz) compared with the baseline pure P(VDF-TrFE) films (0.21 at 1 MHz) resulting from the good insulating and low power dissipation properties of the PPO phase. Thus with small amounts of PPO, it is possible to maintain relatively high permittivity in the blend films and at the same time lower the dielectric losses.
(83) FIG. 19(d) shows the temperature dependence of the dielectric permittivity of the blend films. The dielectric permittivity of the ferroelectric capacitors increases with temperature, reaches a maximum, then decreases. This behavior is typical of ferroelectric materials which when subjected to heating-cooling cycle undergo a ferroelectric-to-paraelectric phase transition at the Curie temperature (T.sub.c). The curie temperature for our pure P(VDF-TrFE) films is approx. 115 C., consistent with reports in literature (Bhansali et al., Org. Electron. 2012, 13, 1541; Ducharme et al., Nature. 1998, 391, 874). Notably, blend films with increasing amount of PPO did not show any change in curie temperature.
(84) 4. Thermal Stability of Blend Films
(85) Large scale integration of ferroelectric memory based on the copolymer P(VDF-TrFE) remains a challenge due to its poor thermal stability (see Li et al., Nat. Mater. 2013, 1). The thermal stability of pure P(VDF-TrFE) and blend capacitors with 8 wt % PPO content were studied. The devices were evaluated based on their ability to switch polarization at an applied field of 125 MV/m and a frequency of 10 Hz. FIG. 20(a) shows the measured remnant polarization and coercive field vs. temperature. In general, a slight increase in polarization and decrease in coercive field is observed with increasing temperature, since the elevated temperatures supply some of the required energy to switch the dipoles (see Zhang et al., J. Phys. D. Appl. Phys. 2011, 44, 155501). Thermal stability of PVDF and PVDF based ferroelectric polymers is directly related to the Curie temperature of these polymers as these polymer undergo a ferroelectric-paraelectric transition at the Curie temperature (see Bhansali et al., Org. Electron. 2012, 13, 1541). Notably, a rapid deterioration in stability of the pure P(VDF-TrFE) thin film capacitors at only 50 C. was observed. This is consistent with other reports for thin film P(VDF-TrFE) capacitors that the polarization decreases notably at 50 C. and rapidly deteriorates at even higher temperatures (see Li et al., Nat. Mater. 2013, 1). This was surprising as it is still way below the curie temperature of 110-120 C. for a 70/30 molar ratio copolymer (see Ducharme et al., Nature. 1998, 391, 874; Zhang et al., J. Phys. D. Appl. Phys. 2011, 44, 155501). By comparison, blend films with 8 wt % PPO, showed much better thermal stability compared to pure P(VDF-TrFE) capacitors. The devices perform well upto 80 C. which are closer to curie temperature of the copolymer. The improvement in thermal stability likely did not come from a change or increase in Curie temperature of the blend films as seen in FIG. 19(d). FIG. 20(b) show the hysteresis loops for pure P(VDF-TrFE) capacitors at different temperatures. It was observed that at temperatures of 60 C. or above, the hysteresis loops displayed a resistive leaky behavior making it impossible to accurately determine polarization in these films. At higher temperatures it was noticed very leaky curves especially in the negative bias regime, indicative of surface breakdown at one of the electrode/ferroelectric interfaces. By comparison, the blend films with PPO show better saturated curves at high temperatures as can be seen in FIG. 20(c).
(86) To further understand this the leakage current of pure P(VDF-TrFE) films and blend films as a function of temperature was studied. Leakage of ferroelectric capacitors based on the copolymer has been well studied and shows a relatively high leakage for thin films around 100 nm (see Khan et al., Org. Electron. 2011, 12, 2225; Fujisaki et al., Appl. Phys. Lett. 2007, 90, 162902). The introduction of TrFE is effective in obtaining the ferroelectric phase in the copolymer but also leads to larger leakage current. If an electric field was applied to switch polarization, current leakage occurs easily at the TrFE monomer because two fluorine atoms opposite from the carbon atoms of the TrFE monomer cause a current leak path (see Fujisaki et al., Appl. Phys. Lett. 2007, 90, 162902). This is evident for thin films P(VDF-TrFE) capacitors with high TrFE content and several studies have reported large leakage issues. FIG. 20(d) shows that at saturation fields of 125 MV/m pure P(VDF-TrFE) capacitors exhibit leakage current density in excess of 10.sup.6 A/cm.sup.2 at room temperature. On the other hand pure PPO films display low leakage current of 10.sup.8 A/cm.sup.2 even at high fields 300 MV/m further highlighting the insulating properties of PPO. The blend devices with 8 wt % PPO show similar leakage currents with slightly lower currents on the negative bias. It is believed this is due to current conduction through the more leaky majority ferroelectric phase. Notably, a drastic improvement in leakage current of the blend films compared to pristine P(VDF-TrFE) films at higher temperatures of 60 C. was observed. This leakage current density of the blend films does not change much with temperature and is an order of magnitude lower than P(VDF-TrFE) films at 60 C. It is believed that the highly insulating amorphous nanospheres of PPO in the blend films act as good charge trap regions and do not allow charge carriers to move freely through the film. This directly affects thermal stability of these blend films leading us to reliably switch the polarization even at elevated temperatures.
(87) 5. Fatigue Endurance and Breakdown Strength of Blend Films
(88) Polarization fatigue is generally described as the reduction of amount of switchable polarization with repeated switching cycles. The fatigue performance of the blend films with common Pt and Au electrodes under relevant conditions (see Zhu et al., IEEE. T. DIELECT. EL. IN. 2010, 17, 1172; Zhu et al., J. Appl. Phys. 2011, 110, 024109) was determined. FIG. 21(a) shows the fatigue performance of the P(VDF-TrFE)-PPO blend film capacitors up to a million (10.sup.6) cycles. A bipolar triangular waveform with an electric field of 100 MV/m and 10 ms pulse width (100 Hz) was applied to fatigue the devices. The polarization was characterized periodically with a Positive-Up-Negative-Down (PUND) measurement at 125 MV/m at the same frequency. A gradual improvement in the fatigue performance of the blend devices with increasing PPO content was observed. With approx. 8 wt % PPO, the devices retain 60% of the polarization after 10.sup.6 cycles, which is a significant and surprising improvement from pure P(VDF-TrFE) capacitors which only retain 20% of the polarization. FIG. 21(b) shows the hysteresis curves at 100 Hz measured before and after fatigue cycles. Polarization decreases sharply from 7 C/cm.sup.2 to only 1 C/cm.sup.2 for pure P(VDF-TrFE), while for 6 wt % PPO films the polarizations drops marginally from 5.6 C/cm.sup.2 to 3.6 C/cm.sup.2. The devices were also fatigued at a higher frequency of 1 kHz which is close to the maximum frequency at which a switch in the polarization of copolymer films can occur. Even at higher frequencies the films with 8 wt % PPO show excellent polarization retention of 80% after 10.sup.6 cycles compared to 54% for pure P(VDF-TrFE) films (FIG. 22).
(89) To further understand the fatigue mechanism the current-voltage (leakage) characteristics of the pure and blend film devices after fatigue were compared. FIG. 17(c) shows high leakage current through fatigued P(VDF-TrFE) film while the films with 6 wt % PPO content show much lower leakage after fatigue (FIG. 16(d)). This suggests that the high number of trapped charges in pure P(VDF-TrFE) films causes poor fatigue performance. The leakage current of the P(VDF-TrFE) thin films also shows an S shaped behavior at high fields, exhibiting small negative differential resistance. This indicates current instability in the film; a situation in which a homogeneous current distribution becomes unstable and decays into filaments (see Zeller, IEEE. T. ELECTR. INSUL. 1987, 22, 115). The local charge and current densities are larger; leading to electrical thinning of the film. This is the reason for the lower coercive fields seen for the blend films after fatigue (FIG. 17(b)). This can also lead to a vastly increased thermal stress leading to electrode delamination also reported in literature especially with the use of unreactive metals such as Au (Zhang et al., Phys. D. Appl. Phys. 2011, 44, 155501). This was also observed during the fatigue testing of the pure P(VDF-TrFE) capacitors, where in some devices the top Au electrode delaminates due to the high thermal stress (FIG. 23). Thus continuous fatigue of thin film P(VDF-TrFE) ferroelectric capacitors leads to dielectric aging and a film close to breakdown. By comparison, blend films with PPO show only a slight increase in leakage current after fatigue, due to the good insulating and charge trapping properties of the PPO nanospheres which results in better fatigue endurance. These highly insulating nanospheres in the blend films act as good charge trap regions and do not allow charges to get trapped in the ferroelectric film, thereby improving fatigue performance. In a follow up study, the dielectric breakdown strength of the blend films was measured by using short time tests where the sweeping DC voltage was applied at a ramp rate of 3 V/s to reach device failure between 10-20 seconds was performed. FIG. 21(d) shows that with increasing PPO content, the breakdown strength of these films improves from 225 MV/m to 360 MV/m for 0% PPO to 8 wt % PPO content, respectively. The PPO in the blend films with its good insulating properties as well as its inherently high dielectric breakdown strength helps improve the breakdown strength of these P(VDF-TrFE)-PPO blend ferroelectric memory devices.
(90) 6. Conclusion
(91) Ferroelectric memory from polymer blends of phase-separated ferroelectric P(VDF-TrFE) and highly insulating amorphous Polyphenylene oxide (PPO) were fabricated. The morphology of these blend films includes phase separated nanospheres of amorphous PPO, surrounded by a crystalline ferroelectric P(VDF-TrFE) matrix. The highly insulating amorphous nanospheres of PPO in the blend films acts as good charge trap regions and do not allow charge carriers to move freely through the film. This directly affects the ferroelectric and dielectric performance of the devices. The blend devices display highly improved ferroelectric and dielectric performance with low dielectric losses (<0.2 up to 1 MHz), enhanced thermal stability (up to 353 K), excellent fatigue endurance (80% retention after 10.sup.6 cycles at 1 KHz) and high dielectric breakdown fields (360 MV/m). The blend devices provide a solution to some of the important limitations of ferroelectric memory based on the copolymer, making ferroelectric memory devices based on these blends more suitable for flexible and transparent electronic applications.
