Microfabrication Of Bulk PZT Transducers By Dry Film .
IOP PUBLISHINGJOURNAL OF MICROMECHANICS AND . Micromech. Microeng. 22 (2012) 085017 (10pp)Microfabrication of bulk PZTtransducers by dry film photolithographyand micro powder blastingI Misri, P Hareesh, S Yang and D L DeVoe 1Department of Mechanical Engineering, University of Maryland, College Park, MD, USAE-mail: ddev@umd.eduReceived 28 March 2012, in final form 30 May 2012Published 5 July 2012Online at stacks.iop.org/JMM/22/085017AbstractA facile fabrication process for bulk PZT microsystems using dry film photoresist and micropowder blasting is presented. Bulk PZT and dry film photoresist etching characteristics areevaluated as a function of process parameters and mask dimensions using 127 μm thick PZTsubstrates. The resulting process simplifies microscale patterning of bulk PZT compared withexisting methods, with selection of suitable etching parameter providing excellent etch rate,selectivity and anisotropy. The technique is used to fabricate two different cantilevermicroactuator topologies based on piezoelectric d31 and d33 mode actuation, demonstrating thecapabilities of the patterning method for applications in bulk PZT microelectromechanicalsystems (MEMS).(Some figures may appear in colour only in the online journal)1. Introductionand MEMS applications [2], etching processes for thin filmPZT based on reactive ion etching (RIE) using high-densityplasma systems have been widely explored and optimized[3–8]. While RIE can provide reasonable etch anisotropy forthin film PZT, etch rates are generally below several hundrednanometers per minute [5, 7, 9, 10], far too slow for processingbulk PZT substrates. Additionally, the electrical propertiesof PZT can deteriorate during RIE due to a combination ofphysical and chemical damage [11, 12]. As an alternativeto RIE, wet chemical etching has been explored for PZTpatterning [12–15]. However, these processes suffer fromundesirable undercut of masked features [14, 16, 17] and therapid formation of insoluble etch products [12, 14] that renderthem impractical for bulk PZT patterning. While wet etchchemistries have been modified to optimize their performancewith thick film [18] and bulk [16, 19, 20] PZT, etch anisotropyremains low and multiple processing steps are required toremove etch residue [20].To overcome these limitations, various nonchemicalmethods for etching bulk PZT have been explored. Forexample, direct machining or milling using diamond toolshas been successfully employed for patterning bulk PZT.Mechanical machining using a wafer dicing saw has beendemonstrated for patterning 100 μm square pillars in bulk PZT[21], although this approach is limited in its ability to produceBulk processed piezoceramic lead zirconatetitanate (PZT;Pb(Zr1-xTix)O3) offers significant potential for the developmentof high-performance microscale piezoelectric transducers.A ferroelectricperovskite-type oxide, bulk PZT sheets aretypically manufactured from spray-dried layers of powderedpolycrystalline elemental oxides mixed with a binder andsintered in a mold to yield the desired shape. Unlikethin film PZT materials commonly used in piezoelectricmicroelectromechanical systems (MEMS), bulk processedmaterials offer exceptionally high and repeatable piezoelectriccoupling coefficients at low cost, with a variety ofcompositions available to meet specific requirements includinghysteresis behavior, dielectric constant, leakage current,ageing, density, frequency dispersion, and electromechanicalnonlinearities. More fundamentally, microsystems employingthick layers of bulk PZT can offer significantly higher arealenergy density than devices based on thin or thick film PZT.Microfabrication of bulk PZT demands etching processescapable of achieving high-resolution patterns in relativelythick substrate layers on the order of tens or hundreds ofmicrometers. Due to its use in both ferroelectric memory [1]1Author to whom any correspondence should be addressed.0960-1317/12/085017 10 33.001 2012 IOP Publishing LtdPrinted in the UK & the USA
J. Micromech. Microeng. 22 (2012) 085017I Misri et alcomplex patterns. Single-point milling of PZT has also beenevaluated, but rapid wear of the diamond cutting tools anddamage to the PZT surface significantly limits this approach[22]. Ultrasonic micromachining, in which an abrasive slurryis used to remove material through microchipping induced byhigh-frequency acoustic energy injected through a mechanicalprobe [23], may be used for micropatterning ceramic materials.However, this is a serial process with limited throughput,and with constraints on the range of shapes and patternresolution that may be achieved based on the dimensionsof the ultrasonic probe [24]. While PZT etching has beensuccessfully demonstrated using a wafer-scale ultrasonicprocess in which a micromachined metal template is usedfor parallel patterning of multiple features [10], templatefabrication is complex and time consuming, and uniformpattern transfer between the template and PZT surface ischallenging. In addition, template wear during ultrasonicmachining can be significant [24], further complicating theprocess. Another bulk PZT etching technique that has beenexplored is excimer laser ablation, which has been shownto yield moderate etch rates with excellent etch anisotropy[25, 26]. However, it remains a slow serial process that is notsuitable for batch-scale processing. Laser ablation also suffersfrom redeposition of etch debris surrounding the etching zone[26], and thermal shock that can produce significant changesin the morphology and electromechanical properties of thesubstrate material [17, 27].Micro powder blasting is a mechanical etching processwhereby a focused jet of micron-scale abrasive particles isused to ablate material from a substrate, often using a metalor polymeric mask to selectively pattern regions of the surface[28]. While micro powder blasting is a serial process, etchingoccurs over a relatively large region of the surface typicallymeasured in tens of mm2 due to lateral spreading of the abrasiveparticles from the nozzle, and etch rates are sufficientlyhigh to make the process suitable for batch manufacturingby rastering the nozzle across the surface for large-areapatterning. Micro powder blasting has been widely adapted formicromachining, and offers much potential as a technique forrapid microstructuring of complex patterns in materials whichare not amenable to traditional etching methods. While micropowder blasting has been widely studied as a method to patternglass [29–32], its use toward bulk PZT etching has not beenextensively explored. Previous work on micro powder blastingbulk PZT has employed masks fabricated from metal [33, 34]or elastomers [34, 35], with these masking techniques capableof realizing features with moderate resolutions on the order of75–100 μm [33, 36]. Because the metal or elastomer layers arefabricated separately from the PZT substrate, mask alignmentcan be challenging, particularly in the case of metal maskswhich lack the optical transparency of silicone elastomers.Furthermore, unwanted gaps between the mask and PZTsubstrate can allow abrasive particles to rapidly undercut themask, leading to degradation of patterning resolution. Whileelastomer masks based on polydimethylsiloxane (PDMS)elastomer have shown excellent etch selectivity during PZTmicro powder blasting [35], issues of mask alignment andresolution limits remain, and this approach requires significanteffort for mask fabrication. Finally, the use of a separatelyaligned mask layer prohibits the formation of PZT geometriesrequiring discrete piezoelectric elements fully separated fromthe neighboring substrate, since this would require that themask itself be discontinuous.The motivation for the present work is threefold. First,we seek to elucidate the relationships between micro powderblasting process parameters and PZT etching characteristics,which has not previously been explored in detail. Unlikeamorphous glass, which has been extensively evaluatedfor micro powder blasting etch performance, PZT is apolycrystalline ceramic with relatively large grains and adensity over twice that of silica, and thus may be expected toexhibit unique etching characteristics. Key process parametersexplored in this work include particle size, nozzle pressureand nozzle-to-substrate distance, with etch rate and etchanisotropy evaluated as a function of these parameters andspace resolution. Our second motivation is to demonstratephotolithographic masking of bulk PZT using dry filmphotoresist, yielding a facile method for achieving precise andhigh-resolution features in PZT using a masking material thatis compatible with micro powder blasting, while avoiding thedisadvantages associated with physical masks fabricated frommetal or molded elastomer. Dry film photoresist designed formicro powder blasting is readily available, inexpensive andrequires minimal equipment for preparation beyond a standardUV flood exposure system or mask aligner, enabling a simpleone step process for direct photolithographic patterning ofPZT sheets prior to etching. Here we demonstrate the useof dry film photoresist as an effective method for greatlysimplifying the overall PZT patterning process while providingexceptionally high line and space resolution compared topreviously explored methods for patterning bulk PZT by micropowder blasting. Finally, to demonstrate the combination ofdry film photoresist patterning and micro powder blastingfor bulk PZT microfabrication, two different cantilevermicroactuator topologies are demonstrated using this process,namely a longitudinally actuated d33 mode cantilever unimorphcomprising a single layer of patterned PZT with alignedinterdigitated electrodes fabricated on one side of the bulkPZT substrate, and a transverse-actuated d31 mode cantilevermultimorph consisting of a PZT/glass composite structurewith continuous electrode layers on each side of the PZT.These particular actuator designs were selected to demonstratethe utility of the presented fabrication approach toward devicetopologies useful for a wide range of bulk piezoelectricmicrosystems.2. Experimental procedures2.1. PZT masking and etchingAll experiments were performed using 7.2 cm square PZTsheets (PSI-5A4E, type 5A, Piezo Systems Inc.) suppliedwith thin film nickel electrodes deposited on both sides.Micro powder blasting was performed using a commercialabrasive blaster (AccuFlo AF-10, Comco Inc.) configuredwith a 1.17 mm diameter nozzle. Aluminum oxide (Al2O3)2
J. Micromech. Microeng. 22 (2012) 085017I Misri et alparticles (Comco) with reported mean sizes of 10 μm and25 μm were used as the abrasive powder. Measurements ofthe supplied particles revealed actual size distributions of11.6 3.7 μm and 25.5 4.7 μm, respectively. To improvethe mechanical robustness of the brittle PZT during processing,each piezoelectric sheet was temporarily bonded to a glassplate using a desktop laminator (PL-1200hp, ProfessionalLaminating Systems) at 110 C with a 14 μm thick layer ofdry film photoresist (DF-2014, Engineered Materials Systems)used as a sacrificial adhesion layer. Before bonding the PZTsheet to the dry film photoresist, the glass/resist substratewas baked at 100 C on a hotplate. This method of bondingthe substrates was found to be very robust, yielding highlycoplanar surfaces with no trapped bubbles between the layers.A second dry film photoresist specifically designed for usein micro powder blasting (RapidMask High Tack, 100 μmthick, Ikonics) was then applied to the exposed PZT surfaceto serve as an etch mask during micro powder blasting, withlamination performed at room temperature. The photoresistwas patterned with arrays of trenches of varying line and spacewidths using a contact mask aligner (EVG 620, EV Group)with a total UV dose of 20.7 mW cm 2. The RapidMaskphotoresist is designed to become embrittled upon exposureto UV radiation, allowing the exposed resist to be rapidlyremoved during the initial stage of micro powder blasting,eliminating the need to employ a chemical developer toremove the exposed photoresist from the substrate prior toPZT etching. For micro powder blasting of the masked PZT,the blasting nozzle was fixed to an adjustable stand within asealed chamber, allowing the nozzle-to-substrate distance tobe controllably adjusted. Upon etching, the exposed brittleareas of the dry film photoresist were rapidly ablated, whilethe unexposed regions served to mask the underlying PZTfilm from the abrasive particles. Following a timed etch, boththe masking and adhesion photoresist layers were removed byan overnight soak in acetone. The released PZT chips werefinally rinsed with DI water and dried in vacuum. For timedetch tests, the resulting etch depths were measured by opticalprofilometry (Eclipse LV-100, Nikon). Etch anisotropy wasdetermined from the ratio of maximum etch depth to half ofthe difference between the initial mask opening dimension andthe final etch pit width measured at the top surface of the PZTsubstrate.using parameters selected based on the results of the etchcharacterization study, with a nozzle pressure of 415 kPa andblasting distance of 8 cm using 25 μm particles. After etchingthrough the PZT/glass composite, the remaining dry filmphotoresist mask was mechanically removed with tweezersin a completely dry process. This process was found to beeasy to implement, allowing for a complete set of devices tobe fabricated at the wafer scale within 1 h.While conceptually simpler than the d31 modemultimorphs, the d33 mode unimorph cantilever actuatorsemploy interdigitated electrodes on the upper surface of thePZT and thus required an additional masking step for electrodepatterning. Both top and bottom nickel electrodes were firstremoved from an as-purchased PZT sheet by wet etching inferric chloride. The sheet was then temporarily bonded to asilicon handle wafer using a layer of RapidMask photoresist.Electrodes were patterned by depositing a lift-off photoresist(LOR-7A, MicroChem) followed by deposition of a positivephotoresist (1813, Shipley). After photolithography of bothresists, a 500 nm aluminum film was deposited by e-beamevaporation. Stripping of the remaining photoresist resulted inremoval of aluminum from unexposed regions of the wafer.RapidMask photoresist was laminated to the PZT surface,followed by UV exposure of the resist through a mask alignedto the metal patterns. Etching was performed using the samepowder blasting parameters as the d31 mode actuators. Beforetesting, individual chips were diced, and released from thesilicon handle wafer by an overnight soak in acetone. The fullfabrication process flow for the unimorph actuators is shownin figure 1(b).3. Results and discussion3.1. Micro powder blasting etch characterizationAn ideal etching process would provide perfect anisotropywith a high etch rate and infinite etch selectivity relativeto the masking layer. In RIE processes routinely used forMEMS fabrication, etching characteristics may be tuned bya combination of chemical and physical material removalmechanisms to achieve the desired process results. In contrast,micro powder blasting is a purely physical etching process,with momentum transfer between the abrasive particles andsubstrate material serving as the sole removal mechanism.Thus, beyond the mechanical properties of the substratematerial, the key parameters that affect etching rate are particlesize, nozzle pressure and nozzle-to-substrate distance. Toevaluate the performance of micro powder blasting for bulkPZT patterning, a sequence of experiments were conducted todetermine the impact of these parameters on PZT etch rate andetch anisotropy as a function of mask dimensions, as well theirimpact on the relative etch rate of the dry film photoresist maskused for photolithographic patterning of the PZT substrate.Example results of a short PZT etch performed witha dry film photoresist mask are shown in figure 2. Thescanning electron microscope (SEM) image in figure 2(a)reveals the photoresist surface roughened by a 1 s etchat a blasting distance of 7cm and pressure of 550 kPa2.2. Cantilever actuator fabricationTo fabricate d31 mode multimorph cantilever actuators, a1 mm wide hole was first powder blasted through an 80 μmthick glass cover slip to serve as an access port for makingelectrical contact with the anchored bottom electrode of thePZT cantilever. A layer of DF-2014 dry film photoresist waslaminated to the glass, followed by a second pass throughthe laminator with the PZT sheet placed on top of thephotoresist to permanently bond the multilayer structure.RapidMask photoresist was laminated at room temperatureon top of the PZT sheet as an etch mask. Patterning ofthe photoresist was performed in a contact mask aligner asdescribed previously. Micro powder blasting was performed3
J. Micromech. Microeng. 22 (2012) 085017I Misri et al(a)(a)(b)(b)(c )(c )(d )(d )(e)(e)(f )(f )(a)(b)Figure 1. Overview of the bulk PZT cantilever microfabrication processes for (a) d31 mode PZT/glass multimorph and (b) d33 mode PZTunimorph actuators.(a)(b)(c )Figure 2. (a) Electron micrographs of 150 μm wide etch pits with patterned photoresist after a 1 s etch at 550 kPa nozzle pressure and 7 cmnozzle distance using 25 μm alumina particles, and (b) the partially etched PZT surface following photoresist removal. (c) A close-up SEMimage of an etched sidewall with a measured surface roughness of Ra 1 μm, well below the roughness of the native PZT at Ra 2–3 μm.4
J. Micromech. Microeng. 22 (2012) 085017I Misri et al(b)(a)Figure 3. PZT etch rate as a function of nozzle-to-substrate distance for different nozzle pressures using (a) 10 μm and (b) 25 μm particles.The dashed lines indicate that etch rates above 60 μm s 1 were observed, preventing accurate measurements at the given nozzle distances.(b)(a)Figure 4. Etch selectivity for PZT over dry film photoresist for (a) 10 μm and (b) 25 μm particles.using 25 μm alumina particles, resulting in a 33 μm deepetch pit within the 150 μm wide regions of the PZT surfaceexposed to abrasive impact, as shown in figure 2(b) followingphotoresist removal. A higher magnification image of theetched PZT sidewall shown in figure 3(c) reveals a relativelysmooth surface with an average roughness (Ra) below 1 μm asmeasured by AFM. This roughness compares favorably withthe roughness of the as-purchased PZT sheets, which typicallyvaries between 2 and 3 μm.As shown in figure 3, the PZT etch rate was found tobe highly dependent on each of the process parameters, withetch rate following an inverse power relationship with nozzleto-substrate distance and approximately linear scaling withnozzle pressure, with the larger 25 μm Al2O3 particles yieldingetch rates twice those of the 10 μm particles for the same etchconditions. Note that due to high etch rates at small nozzledistances and high nozzle pressures, accurate measurementsof etch rate could not be performed under certain processconditions as indicated by dashed lines in figure 3. Whilesimilar overall behavior was observed for etching of thephotoresist mask, the photoresist etch rate was significantlyless sensitive to nozzle distance. As a result, the etch selectivity,defined as the ratio of PZT etch rate (ERPZT) to photoresistetch rate (ERPR), was greatly improved as the nozzle distancewas reduced (figure 4). Surprisingly, for moderate and largenozzle distances, the photoresist etch rate was maximum atthe intermediate value of nozzle pressure. As a result, theminimum etch selectivity occurred at 415 kPa, a trend thatwas observed for both the 10 μm and 25 μm particles. Overall,etch selectivities ranging from 4 to 10 were achieved. Whilethis selectivity is low compared to masks based on elastomericsilicone [35], it is sufficient for patterning a wide range ofPZT microsystems with thicknesses below 1 mm when using100 μm thick dry film photore
Microfabrication of bulk PZT demands etching processes capable of achieving high-resolution patterns in relatively thick substrate layers on the order of tens or hundreds of micrometers. Due to its use in both ferroelectric memory [1] 1 Author to whom any correspondence should be addressed. and MEMS applications [2], etching processes for thin .
100107 mobil almo 525,bulk 100110 mobil dte 24,bulk 100112 mobil almo 527,bulk 100121 mobil vacuoline 128,bulk 100123 mobil dte light,bulk 100126 mobil vactra oil no 1,bulk 100158 mobil almo 529,bulk 100159 mobil pegasus 701,bulk 100166 mobil pegasus 701 sae 30,bulk 100167 mobil almo 532,bulk 100386 mobil
Pressure Systems, Inc. KPSI Level and Pressure Transducers User’s Manual i www.PressureSystems.com Our Company Pressure Systems is an ISO9001:2000 certified U.S. manufacturer of submersible and non-submersible (above ground) pressure transducers for environmental, industrial and municipal applications.File Size: 1MBPage Count: 34Explore further1% BFSL Submersible Level Transducer KPSI 700 Series .www.te.comKPSI Transducers, Pressure Systems, Inc. WQPwww.wqpmag.comKPSI Submersible Pressure Transducers Available Online .www.te.comRecommended to you b
TCEQ, Emission Inventory of Bulk Gasoline Terminals and Bulk Gasoline Plants Final, August 15, 2013 4 consumer's motor vehicle gasoline tank delivered at retail gasoline stations. For purposes of this inventory, only those evaporative losses that occur within the bulk terminal or bulk plant fence
Pangaea Logistics Solutions Ltd ('Pangaea' or 'the Company') is a US publicly listed (NASDAQ: PANL) dry bulk shipping company with expertise in the operation of bulk carriers and bulk logistics services. The group includes Americas Bulk Transport, Nordic Bulk Carriers, Phoenix Bulk Carriers, and Seamar Management.
Use the HV/100 “MONITOR” output to monitor the high voltage on the oscilloscope. Sweeping this voltage sweeps the grating position using a small piezoelectric actuator (PZT). A good starting value of the knob of the coarse DC offset of the PZT controller can be about 4.5. While
consists of a 3"x 1.75"x.008" Morgan Matrox PZT-5A13 piezoceramic encapsulated using a polymer film. The PZT-SA piezoceramic mechanical and electro-mechanical properties are listed in Table 1. MSC/NASTRAN is used to model the structure; Figure 2 shows a finite element representation of the cantilevered beam indicating the location of
The electromechanical impedance model of a Lead Zirconate Titanate (PZT)-driven structural system. Equation (1) is the mathematical expression used to illustrate the relationship between the electric admittance (the reciprocal of impedance) and the frequency based on the piezoelectric wave equation [83]: Y iwa #T 33 (1 id)
Organizations have to face many challenges in modern era. The same is the position in schools and collages as they are also organizations. To meet the challenges like competition, efficient and economical uses of sources and maximum output, knowledge of management and theories of management is basic requirement. Among Management Theories, Classical Management Theories are very important as .
