A Survey of Ultrasonic Piezoelectrics for High Temperatures

46th Annual Review of Progress in Quantitative Nondestructive Evaluation QNDE2019 July 14-19, 2019, Portland, OR, USA

QNDE2019-1234

A SURVEY OF ULTRASONIC PIEZOELECTRICS FOR HIGH TEMPERATURES

B. Tittmann, C. Batista, and B. Reinhardt, Penn State University University Park, PA

ABSTRACT temperature but instead the melting temperature is given. Also Current transducers use primarily PZT5-H as piezoelectric listed is the conventional PZT-5H which is the commonly used element for ultrasound transmission and detection. This material piezoelectric in commercial applications. has a Curie-Weiss Temperature which limits its use to about 2100 C. Some industrial applications require much higher temperatures. Here we present a survey of piezoelectrics capable TABLE 1: SOME WELL-KNOWN PIEZOELECTRICS of operation at higher temperatures. Material Curie-Weiss Temperature (°C) Keywords: Piezoelectrics, high temperature, survey.

NOMENCLATURE PZT-5H 210 Piezoelectric Coefficient × 10 Keramos Lead Metaniobate 400 −12 𝑑𝑑33 � 𝐶𝐶� � 𝑁𝑁 Bismuth Titanate 685 1. INTRODUCTION urrently, ultrasonic non-destructive evaluation (NDE) is Aluminum Nitride 1600-1800 C employed periodically on passive high temperature (Melting Temperature) components, but continuous online monitoring has not been Lithium Niobate 1000 widely implemented. The need for continuous online monitoring is becoming more important with the need for high temperature infrastructure license extension. Additionally, ultrasound is a 2. SINGLE CRYSTAL PIEZOELECTRICS highly attractive NDE methodology given that it allows for In the category of the single crystals, both maximum inspection in optically opaque materials, such as liquid metal temperature and long term in-situ operation were investigated in coolants, steam generator piping and heat exchangers pipes. a comparison study. These are lithium niobate (LiNb03), Further applications may be found in materials research reactors aluminum nitride (AlN) and YCOB [YCa4O (BO3)3]. As shown where ultrasonic NDE can be used for in-situ analysis of in Table 2. all three materials exhibited stability in ultrasonic radiation effects on novel radiation hard materials currently performance through heat treatment of 9500 C for 24 hours and being developed. This paper presents a survey of piezoelectrics 10000 C for 48 hours. This “cook-and-look” testing has revealed for possible high temperature applications. The survey is significant changes in the dielectric properties and only small conveniently divided into several categories: single crystals, changes in the ultrasonic performance of lithium niobate. piezoelectric ceramics, composite ceramics and possible Dielectric changes of the observed magnitude would be radiation proof materials. The survey starts with several expected to have a noticeable effect on the ultrasonic relatively well-known high temperature piezoelectrics performance. However, the heat treatments were not equivalent summarized in Table 1 for comparison. Listed also are the during the dielectric and ultrasonic testing. It is quite likely that Curie-Weiss Temperatures, which are useful in that they limit the longer heat treatment caused a more pronounced change in the temperature to which a material can exhibit piezoelectricity. the dielectric properties the lithium niobite. Aluminum Nitride is not a ferroelectric so that there is no Curie

1 © 2019 by ASME Conventional piston-type transducers that send and receive ultrasonic waves typically use lead zirconate titanate for the active element and have backing and matching layers. In addition, they are usually coupled to the substrate through gel or adhesive. Harsh environments limit the types of couplants that can be used and curved surfaces present additional challenges. In contrast, spray-on transducers are bonded directly to the substrate, precluding the need for couplants. Spraying transducers onto curved surfaces is not substantially different from that on flat surfaces. No matching or backing layers are used in this work, but they could be used if deemed necessary. One advantage that spray-on transducers provide is the ability to design the transducer material for a specific operating temperature by mixing powders into a sol gel to create a composite (or alloy). Material selection is based primarily on combining Curie temperatures (Tc) and coupling coefficients (e.g. d33) of the constituents to achieve the desired overall piezoceramic properties. To maintain an in-field transducer at high signal-to-noise (SNR), the piezoelectric transducer material should have both a large coupling coefficient and a Curie temperature exceeding the transducer’s operating temperature. Micromechanical modeling enables prediction of overall properties based on the properties of the constituents.

Figure 1: COMPARISON OF RESULTS OF HEAT 4. COMPOSITE CERAMICS TREATMENTS ON THREE SINGLE CRYSTAL PIEZOELECTRICS. The biggest difference between piezoelectric materials used in conventional transducers and spray-on piezoelectric The YCOB on the other hand exhibited a much less pronounced transducers is density/porosity. Pressure is an integral part of change in dielectric properties after heat treatment. It is expected forming fully dense piezoceramics, and it is not part of spray-on that YCOB is more stable at high temperatures than LiNbO3 processing. Thus, spray-on transducers have porosity that affects which is known to deplete its oxygen particularly at low oxygen their properties. On the positive side, it also provides strain partial pressure. tolerance to the piezoceramic, which is bonded to a metal substrate that is subject to temperature changes. The pioneers of 3. PIEZOELECTRIC CERAMICS spray-on piezoelectric transducer technology are Barrow and Kobayashi. Barrow’s group added powder to sol-gel to form In the category of thick ceramics sample preparation, poling, piezoelectric films thicker than 1-2 µm using a spin coating acoustic data, high temperature tests and the effect of protective methodology. Kobayashi’s group then adapted the powder/sol- aluminum oxide layer on both poling and temperature gel technique using a spray gun to deposit films on metal performance were studied. Bismuth titanate thick film substrates. Tittmann’s group has provided technological transducers performed well up to 600 0 C. Currently, tests are advancements on processing methods. ongoing with thick film transducers deposited on pipes and simulated casings for NDE with guided waves generated by These results showed that sol-gel fabricated both flat and curved arrays. Recently developed piezoelectrics PZT/Bi4Ti3O12 transducers can operate over a broad band of with high Curie temperatures are listed in Table II. frequencies. This specific transducer could effectively receive signals of frequencies in the low kilohertz range when operating TABLE 2: PIEZOELECTRIC in through-transmission mode. CERAMICS

Piezoelectric Materials Curie Temperature (⁰C) d33 (pC.N-1) Reference Prazeodymium Titanate >1550 0.5 1 Lanthanum Titanate 1461 2.6 2,5 Neodymium Titanate 1482 1.2 2,6,7 Strontium Niobate 1327 3 Calcium Niobate >1525 3

FIGURE 3: TEMPERATURE DEPENDENCE OF BI4TI3O12-LINBO3 Pulse-echo amplitude [1] FIGURE 4: PZT/ Bi4Ti3O12 AMPLITUDE VERSUS TEMPERATURE. [2] The transducer’s efficiency decreased when operating in pulse- echo mode, but a discernable signal was still observed as low as TABLE 3: CANDIDATE PIEZOELECTRIC MATERIALS 500 kHz. The thickness of this transducer was still relatively Material Transition Transition Structure thin, especially for low frequency operation. The broadband Temperature Type o nature of this transducer was very evident in its testing in that it C had a center frequency around 2.75 MHz but could still operate AlN 2826 melt Wurtzite effectively well below 1 MHz. The Bi4Ti3O12/Bi4Ti3O12 Bi3TiNbO9 909 Curie Perovskit transducer was also tested for low frequency operation, but it layered was considerably less efficient. The signal effectively LiNbO3 ~1200 Curie Perovskit disappeared at frequencies much below 1 MHz. This again Sr2Nb2O73 1342 Curie Perovskit shows the great advantage to the use of the PZT/Bi4Ti3O12 layered composite. The PZT/Bi4Ti3O12 has a much greater signal La2Ti2O7 1500 Curie Perovskit amplitude and is more broadband allowing it to operate at low layered frequencies and produce viable waveforms. Thicker GaPO4 970 α-β SiO2 PZT/Bi4Ti3O12 transducers may further enhance their Homeoty operation at low frequencies. Both the signal amplitude and RareEarthCa4(BO3)3 >1500 melt Oxyborat signal- to-noise ratio can be increased along with better Homeoty operation in the pulse-echo mode. ZnO 1975 melt Wurtzite

4. POSSIBLE PIEZOELECTRICS FOR RADIATION CONCLUSION ENVIRONMENT This survey is still incomplete but attempts to gather readily The most straightforward down selection parameter seems to available information in summarize it. The findings indicate be the transition temperature, which provides an upper limit on that PZT/BIT and BIT/LNO composite transducers functioned o the operating range of the piezoelectric material. In fact, a higher in pulse-echo mode until 675 and 1000 C, respectively. For the Curie temperature has been found to correlate with increased brush-on/paint-on LiNbO3/Ba TiO3 high temperature could be 0 fabrication and testing of high temperature radiation tolerance observed to about 700 C. and the primary effect of radiation damage in piezoelectric REFERENCES materials appears to be depolarization. With this in mind a table of candidate materials for longitudinal wave generation is [1] C. Searfass, “Fabrication and Characterization of Bismuth provided below in Table 3, however, this is only the first step. Titanate Thick Films Fabricated Using a Spray-On Technique The final column in Table 3 is of substantial importance as it has for High Temperature Ultrasonic Non-Destructive Evaluation.” been found that crystal structure plays a significant role in PhD thesis, The Pennsylvania State University, 2012. radiation tolerance of ceramics.[3] [2] C. E. Pheil, “Fabrication and testing of high temperature ultrasonic transducers.” B.S. Thesis, The Pennsylvania State University, 2012. {3] B. Reinhardt. “Nonlinear Ultrasonic Measurements in Nuclear Reactor Environments”. PhD thesis, The Pennsylvania State University, 2016.