Femtosecond Laser Additive Manufacturing of Multi-Material Layered Structures

applied sciences

Article Femtosecond Laser Additive Manufacturing of Multi-Material Layered Structures

Shuang Bai and Jian Liu * PolarOnyx, Inc., 2526 Qume Drive, Suites 17 and 18, San Jose, CA 95131, USA; [email protected] * Correspondence: [email protected]

Received: 15 December 2019; Accepted: 27 January 2020; Published: 3 February 2020

Abstract: Laser additive manufacturing (LAM) of a multi-material multi-layer structure was investigated using femtosecond ﬁber lasers. A thin layer of yttria-stabilized zirconia (YSZ) and a Ni–YSZ layer were additively manufactured to form the electrolyte and anode support of a solid oxide fuel cell (SOFC). A lanthanum strontium manganite (LSM) layer was then added to form a basic three layer cell. This single step process eliminates the need for binders and post treatment. Parameters including laser power, scan speed, scan pattern, and hatching space were systematically evaluated to obtain optimal density and porosity. This is the ﬁrst report to build a complete and functional fuel cell by using the LAM approach.

Keywords: laser; additive manufacturing; femtosecond laser; ﬁber laser; YSZ; fuel cell; SOFC

1. Introduction Laser additive manufacturing (LAM) [1–6] uses powder bed fusion (PBF), feedstock, or powder injection systems to build three dimensional (3D) parts. To date, the PBF AM system is the best approach compared with other AM approaches because it can achieve excellent density (>99%) and mechanical strength. Solid oxide fuel cells (SOFC) [7–10] provide high efficiency of energy conversion and little pollution to the environment. Although conventional methods can make good quality SOFCs, they are limited in part by geometry (i.e., planar or tubular) and flexibility. Moreover, post heat treatment is usually required for binder-based AM to achieve reasonable density and mechanical strength. This adds significant cost, time, and process complexity, especially for making large sized parts. It is even more challenging to make multi-material multi-layer parts, in the case of SOFC, which has at least three basic layers: the anode, electrolyte, and cathode. Therefore, LAM could be a promising approach to make SOFCs of complex structures and shapes without the need for binders. In our previous publication [11], we developed an AM process to make a solid YSZ layer with greater than 99% density. Micron sized powders were used with a femtosecond (fs) laser PBF to directly manufacture YSZ samples without binders and post heat treatment. Process control parameters were systematically studied including laser power, scanning speed, and hatching space. Microstructure and mechanical properties of the fabricated parts were investigated and demonstrated to have a comparable performance with conventional methods. In this paper, we report on our research to directly make multi-material and multi-layer structures without binders and post-process. This is the essential step to build a SOFC. To our knowledge, this is the first publication demonstrating a working basic three-layered SOFC cell using LAM technology.

2. Materials and Methods A high power mode-locked Yb-doped ﬁber laser (Laser-Femto, Inc., San Jose, CA, USA) was used in the experiment. It delivers up to 250 W of average power at a pulse repetition rate (PRR)

Appl. Sci. 2020, 10, 979; doi:10.3390/app10030979 www.mdpi.com/journal/applsci Appl. Sci. 2020, 9, x FOR PEER REVIEW 2 of 8

2. Materials and Methods A high power mode-locked Yb-doped fiber laser (Laser-Femto, Inc., San Jose, CA, USA) was Appl.used Sci. in 2020the ,experiment.10, 979 It delivers up to 250 W of average power at a pulse repetition rate (PRR)2 of of 8 80 MHz with a central wavelength at 1030 nm. The output pulses were compressed to have a full- ofwidth-half-maximum 80 MHz with a central (FWHM) wavelength pulse duration at 1030 nm.of 800 The fs. output The laser pulses beam were was compressed guided through to have a agalvanic-mirror full-width-half-maximum enabled scanner (FWHM) and pulseraster-scanned duration ofon 800 the fs. powder The laser surface. beam wasThe guidedlaser beam through was afocused galvanic-mirror by a lens with enabled a 100 scanner mm long and focal raster-scanned length and had on thea FWHM powder beam surface. diameter The laserof 25 beamμm on was the focusedtarget. Considering by a lens with the aloss 100 through mm long all focal of the length components, and had athe FWHM maximum beam power diameter on ofthe 25 powderµm on thesurface target. was Considering about 170 W. the An loss acoustic–optical through all of themodulator components, (AOM, the Gooch maximum & Housego, power Melbourne, on the powder FL, surfaceUSA) was was used about in 170the W.laser An beam acoustic–optical path to modula modulatorte the laser (AOM, power Gooch and & pulse Housego, format. Melbourne, The Galvano FL, USA)laser scanner was used was in themounted laser beam on a pathmotorized to modulate stage to the adjust laser powerthe focal and spot pulse position format. relative The Galvano to the laserpowder scanner surface. was It mounted was synchronized on a motorized with stage the AO to adjustM to scan the focal the laser spot positionbeam following relative tothe the designed powder surface.scan patterns, It was synchronizedcommanded by with a computer the AOM to[11]. scan Th thee AM laser process beam following needs a good the designed balance scanbetween patterns, the commandedpeak power byand a average computer power. [11]. The Proper AM processpeak power needs can a good give balancea good betweeninitiation the for peak melting, power while and average power.power helps Proper maintain peak power the melting can give process. a good initiationThe 80 MHz for melting, is the maximum while average we can power obtain helps for maintainhigh PRR the matching melting process.with 200 The W 80average MHz ispower. the maximum Too high we a canpeak obtain power for highwill generate PRR matching too much with 200ablation. W average power. Too high a peak power will generate too much ablation. Planar SOFCs are operated at high temperatures (800–1000 °C) using a thin ceramic yttria- Planar SOFCs are operated at high temperatures (800–1000 ◦C) using a thin ceramic yttria-stabilized stabilized zirconia (YSZ) electrolyte and supporting Ni–YSZ anode layers [7–10]. YSZ 6has a 9.2 × zirconia (YSZ) electrolyte and supporting Ni–YSZ anode layers [7–10]. YSZ has a 9.2 10− /K thermal −6 × expansion10 /K thermal coeffi expansioncient, 1.8 W coefficient,/m k thermal 1.8 conductivity, W/m.k thermal andmelting conductivity, temperature and melting greater thantemperature 2780 C. · ◦ Lanthanumgreater than strontium2780 °C. La manganitenthanum strontium (LSM) or mangan lanthanumite (LSM) strontium or lanthanum cobalt oxide strontium (LSC) cobalt is used oxide as a(LSC) cathode is used layer. asStainless a cathode steel layer. or ceramicStainless interconnects steel or ceramic is used interconnects for electric currentis used routing.for electric In acurrent SOFC, arouting. combination In a SOFC, of a gaseous a combination fuel (e.g., of hydrogen a gaseous or hydrocarbonfuel (e.g., hydrogen fuels) and or anhydrocarbon oxidant gas fuels) (e.g., oxygenand an inoxidant the air) gas goes (e.g., through oxygen electrodesin the air) goes and thethrough reaction electrodes occurs and on thethe anodereaction and occurs cathode, on the respectively. anode and Thecathode, actively respectively. charged carrier The actively species charged go through carrier an ion species conducting go through electrolyte an ion to conducting generate an electrolyte electronic current.to generate A high an electronic density of current. YSZ is required A high indensity order toof physicallyYSZ is required separate in theorder gaseous to physically fuel from separate oxygen tothe e ffigaseousciently fuel produce from electricity.oxygen to efficiently produce electricity. Figure1 1 gives gives a a schematic schematic drawing drawing for for a a typical typical SOFC. SOFC. Table Table1 shows 1 shows the the design design specifications specifications of theof the SOFC SOFC for for each each layer. layer. Once Once a cell a iscell made is made and tested,and tested, the process the process is repeated is repeated to make to themake next the cell next to stackcell to on stack the previouson the previous cell. The cell. cathode The cathode (LSM), electrolyte(LSM), electrolyte (YSZ), or(YSZ), anode or (Ni–YSZ) anode (Ni–YSZ) can be used can be as aused supporting as a supporting substrate insubstrate a three layerin a cellthree unit. layer In ourcell case, unit. we In will our focus case, on we an will anode focus supported on an anode SOFC. supported SOFC.

Figure 1.1. Solid oxide fuel cell (SOFC)(SOFC) design architecture. A single cell unit include cathode (LSM), electrolyte (YSZ), andand anodeanode (Ni–YSZ).(Ni–YSZ). InIn thisthis paper,paper, thethe anodeanode layerlayer isis thethe supportingsupporting substrate.substrate. Table 1. Solid oxide fuel cell (SOFC) structure specifications. Table 1. Solid oxide fuel cell (SOFC) structure specifications. Component Material Thickness Porosity Component Material Thickness Porosity Anode Ni–YSZ 0.3–0.6 mm ~40% ElectrolyteAnode Ni–YSZ YSZ 5–100.3–0.6µm mm <~40%5% Electrolyte ConductingYSZ 5–10 μm <5% Cathode 10–50 µm ~30% Cathode Conductingceramic ceramic 10–50 μm ~30% Appl. Sci. 2020, 9, x FOR PEER REVIEW 3 of 8 Appl. Sci. 2020, 10, 979 3 of 8 Appl. Sci. 2020, 9, x FOR PEER REVIEW 3 of 8 3. Results and Discussion 3.3. Results and Discussion 3.1. Fabrication of Multi-Layer and Multi-Material Structures 3.1. Fabrication of Multi-Layer and Multi-Material Structures 3.1. FabricationThe Ni–YSZ of Multi-Layersupporting layer and Multi-Material was first made Structures with the LAM. Similar to YSZ LAM [11, the process windowTheThe Ni–YSZNi–YSZis small supporting supportingand a perfect layerlayer match waswas firstfirst between mademade with withthe thelasertheLAM. LAM. power Similar Similar and to toscan YSZ YSZ speed LAM LAM [ 11[11,is ],critical. the process The windowfeasibilitywindow isis small ofsmall melting and and a perfectaNi–YSZ perfect match powdersmatch between between was the laser successfullythe powerlaser power and demonstrated scan and speed scan is critical.speedwith 80is The critical.MHz feasibility PRRThe femtosecondoffeasibility melting Ni–YSZof laser melting powdersAM andNi–YSZ waswithout successfullypowders mixing wasany demonstrated binsuccessfullyders with with thedemonstrated 80 Ni–YSZ MHz PRR powders. femtosecondwith 80Figure MHz laser 2 (left)PRR AM showsandfemtosecond without the experimental mixing laser AM any bindersandresult without of with porosity the mixing Ni–YSZ (or anyrelative powders. binders density) Figurewith as the2 a (left) functionNi–YSZ shows ofpowders. the laser experimental power Figure at a2 result fixed(left) scanofshows porosity speed the experimental (orof 300 relative mm/s. density) result A high of as densityporosity a function of (or Ni–YSZ of relative laser powercan density) be atachieved a as fixed a function scan at 130 speed ofW laserlaser of 300 power power mm/ s.at and Aa fixed high 300 mm/sdensityscan speed scan of Ni–YSZ speed.of 300 Lowermm/s. can be densityA achieved high density(high at 130 porosity, of W Ni–YSZ laser specif power canications be and achieved 300 of the mm atNi–YSZ/s 130 scan W speed.anode laser power Lowerare in Tableand density 300 1) (highcanmm/s be porosity, scanachieved speed. specifications by Lower either increasingdensity of the (high Ni–YSZ the porosity,scan anode speed specif are to in 500ications Table mm/s1) canof or the reducing be Ni–YSZ achieved the anode bylaser either arepower. increasingin Table Figure 1) 2thecan (right) scanbe achieved speedshows to bythe 500 either microscopic mm increasing/s or reducing photos the scan of the the speed laser melted power.to 500 Ni–YSZ mm/s Figure or layers2 reducing (right) under shows the various laser the power. microscopic scan Figurespeed photosconditions2 (right) of shows the at a melted fixed the powermicroscopic Ni–YSZ of layers130 W.photos under The resultsof various the showmelted scan that speedNi–YSZ high conditions density layers Ni–YSZ under at a fixed varioushad power a better scan of surface 130speed W. Thesmoothnessconditions results showat (witha fixed that much power high density clearerof 130 W. Ni–YSZimage) The results hadthan a better showthe surfacethatlow highdensity smoothness density Ni–YSZ. Ni–YSZ (with Moreover, muchhad a clearerbetter we surface image) also demonstratedthansmoothness the low density(with that it much Ni–YSZ.was possible clearer Moreover, to image) control we also thethan demonstratedlaser the power low (Figuredensity that it 2) wasNi–YSZ. to possibleobtain Moreover, a tosupport control layerwe the laserandalso powerfunctionaldemonstrated (Figure layer 2that) in to one it obtain was process. possible a support Figure to layercontrol 3 (left) and theshows functional laser an power example layer (Figure in with one 2)the process. to combined obtain Figure a support functional3 (left) layer shows anode and layeranfunctional example and support layer with in the anode one combined process. layer. functional ItFigure has 35 3 to(left) anode 40 showsvol% layer of an andNi example from support Energy with anode the Dispersed layer.combined It hasX-ray functional 35 spectroscopy to 40 vol% anode of (EDX)Nilayer from and measurement. Energy support Dispersed anode Figure layer. X-ray 3 (right) It spectroscopyhas 35are to examples 40 vol% (EDX) of measurement.Nianode from samples Energy Figure withDispersed 3excellent (right) X-ray areuniformity spectroscopy examples and of repeatability.anode(EDX) samplesmeasurement. These with findings excellent Figure lay3 uniformity (right) a solid are foundation andexamples repeatability. forof anode the next These samples LAM findings processwith layexcellent of a solidYSZ, uniformity foundationthe electrolyte and for layer.therepeatability. next LAM These process findings of YSZ, lay the a electrolyte solid foundation layer. for the next LAM process of YSZ, the electrolyte layer.

Figure 2. (Left) Density as a function of laser power at a scan speed of 300 mm/s. (Right) Microscopic imagesFigureFigure 2.2.of (meltedLeft) Density layer with as a a function target power of laser of power 130 W, atat hatching aa scan speed space ofof of 300300 30 mmmm/s.μm,/ s.scanning (Right) speed Microscopic at 300 µ mm/simagesimages (a of ,b ),melted melted 400 mm/s layer layer ( cwith, withd), and a a target target 500 mm/spower power ( eof,f of ).130 ( 130Right W, W, hatching) hatchingcolumn space images space of 30 of(b , 30μdm,,f) m,arescanning scanningmagnified speed speed images at 300 at a b c d e f Right b d f of300mm/s the mm left(a/s,b (column),, 400), 400 mm/s (a mm,c, e(/).cs, d ( ),, and), and 500 500 mm/s mm /(se (,f)., ).(Right ( ) column) column images images (b ( ,d, ,f,)) are are magnified magniﬁed images of the left column (a,c,e). of the left column (a,c,e).

Support anode layer Support anode layer

Functional anode layer Functional anode layer

Figure 3. Left Figure 3. ((Left)) Cross-section Cross-section scanning scanning electromagnetic electromagnetic (SEM) (SEM) image image of of the interface between the functional layer and support layer of the anode. (Right) Ni–YSZ anode samples (the largest size was functionalFigure 3. (layerLeft) andCross-section support layer scanning of the electromagneticanode. (Right) Ni–YSZ (SEM) anodeimage samplesof the interface (the largest between size was the 30 30 mm). 30functional × 30 mm). layer and support layer of the anode. (Right) Ni–YSZ anode samples (the largest size was 30 × 30 mm). Appl. Sci. 2020, 10, 979 4 of 8

Appl. Sci. 2020, 9, x FOR PEER REVIEW 4 of 8 Second, a thin layer of YSZ, either in powder or ink, was spread on top of the Ni–YSZ anode supportSecond, substrate a thin and layer processed of YSZ, witheither the in laser.powder Both or can ink, be was melted spread and on bonded top of onthe top Ni–YSZ of the anode supportsurface, butsubstrate the ink and type processed of YSZ provides with the betterlaser. Both uniformity. can be Themelted scan and patterning bonded on and top hatching of the anode space playsurface, critical but rolesthe ink in thetype melting of YSZ of provides YSZ on the better Ni–YSZ uniformity. anode. TwoThe patterningscan patterning types and were hatching systematically space playstudied critical as shown roles inin Figure the melting4, and the of staggering YSZ on the pattern Ni–YSZ was betteranode. than Two the patterning line pattern, types with were less systematicallythermally induced studied peeling as shown oﬀ. YSZ in wasFigure strongly 4, and bonded the staggering on Ni–YSZ. pattern The YSZwas layerbetter was than fabricated the line pattern,from 10 towith 50 lessµm thermally in thickness. induced peeling off. YSZ was strongly bonded on Ni–YSZ. The YSZ layer was fabricated from 10 to 50 μm in thickness.

Staggered

Figure 4. Types of scan patterning of LAM process for YSZ on Ni–YSZ. Staggered pattern (Right) Figure 4. Types of scan patterning of LAM process for YSZ on Ni–YSZ. Staggered pattern (right) shows less peel-off issues during the AM process than the line pattern (Left). shows less peel-off issues during the AM process than the line pattern (left). The melting effect of YSZ on Ni–YSZ as a function of hatching space was investigated and The melting effect of YSZ on Ni–YSZ as a function of hatching space was investigated and summarized in Figure5. Figure5a,b show the YSZ layers printed on the Ni–YSZ anode substrates summarized in Figure 5. Figure 5a,b show the YSZ layers printed on the Ni–YSZ anode substrates with a hatching space of 15 µm, scanning speed of 80 mm/s, and on-target power of 78 W. Figure5c,d with a hatching space of 15 μm, scanning speed of 80 mm/s, and on-target power of 78 W. Figure 5c,d show the results with hatching space of 20 µm, scanning speed of 100 mm/s and on-target power show the results with hatching space of 20 μm, scanning speed of 100 mm/s and on-target power of of 78 W. Figure5e,f show the results with a hatching space of 20 µm, scanning speed of 80 mm/s, 78 W. Figure 5e,f show the results with a hatching space of 20 μm, scanning speed of 80 mm/s, and and target power of 78 W. Figure5g,h show the results with a hatching space of 25 µm, scanning target power of 78 W. Figure 5g,h show the results with a hatching space of 25 μm, scanning speed speed of 100 mm/s, and target power of 78 W. Images on the right column (b, d, f, h) differ only in the of 100 mm/s, and target power of 78 W. Images on the right column (b, d, f, h) differ only in the magnifying scale from their counterparts on the left column (a, c, e, g). magnifying scale from their counterparts on the left column (a, c, e, g). When the hatching space is too small, continuous ripple lines in the narrow space can be found. When the hatching space is too small, continuous ripple lines in the narrow space can be found. This results in more cracks propagating in the cross direction. Consequently, the porosity is large This results in more cracks propagating in the cross direction. Consequently, the porosity is large (as (as shown in Figure5a,b). When the hatching space is too large, melted lines cannot be formed shown in Figure 5a,b). When the hatching space is too large, melted lines cannot be formed continuously, resulting in unmelt regions and fractured structure with large porosity (as shown in continuously, resulting in unmelt regions and fractured structure with large porosity (as shown in Figure5g,h). At the hatching space of 20 µm, large uniformly melted regions with less cracks were Figure 5g,h). At the hatching space of 20 μm, large uniformly melted regions with less cracks were found, as shown in Figure5c–f. After the YSZ electrolyte layer was printed on the anode substrate, found, as shown in Figure 5c–f. After the YSZ electrolyte layer was printed on the anode substrate, a a thin layer of LSM (ink type) was spread on the YSZ layer and melted to serve as the cathode of thin layer of LSM (ink type) was spread on the YSZ layer and melted to serve as the cathode of the the SOFC. SOFC.

Appl. Sci. 2020, 10, 979 5 of 8 Appl. Sci. 2020, 9, x FOR PEER REVIEW 5 of 8 a b

c d

e f

g h

FigureFigure 5. 5.MicroscopicMicroscopic images images (di(differentﬀerent scales) scales) of of the the YSZ YSZ layer layer printed printed on the on Ni–YSZ the Ni–YSZ anode substrates. anode substrates.(a,b) Hatching (a,b) Hatching space of space 15 µm, of scanning 15 μm, scanning speed of speed 80 mm of/ s,80 and mm/s, target and power target ofpower 78 W; of ( c78,d )W; hatching (c,d) hatchingspace of space 20 µ m,of 20 scanning μm, scanning speed ofspeed 100 mmof 100/s, andmm/s, target and powertarget ofpower 78 W. of ( e78,f) W. Hatching (e,f) Hatching space of space 20 µm, of scanning20 μm, scanning speed of speed 80 mm of/s, 80 and mm/s, target and power target of 78 power W. (g ,ofh) Hatching78 W. (g,h space) Hatching of 25 µ spacem, scanning of 25 μ speedm, scanningof 100 mm speed/s, andof 100 target mm/s, power and oftarget 78 W. power of 78 W.

Appl. Sci. 2020, 10, 979 6 of 8 Appl. Sci. 2020, 9, x FOR PEER REVIEW 6 of 8

3.2.3.2. Three-Layered Three-Layered Cell Cell Testing Testing and and Analysis Analysis FigureFigure 6a6 isa isa photo a photo of ofthe the SOFC SOFC cells cells made made from from the the same same batch. batch. Consistent Consistent and and uniform uniform three three layered cells were made in our laboratories with up to a 30 30 mm active area. Figure6b shows that layered cells were made in our laboratories with up to a 30 × 30× mm active area. Figure 6b shows that thethe high high strength strength of Ni–YSZ of Ni–YSZ supporting supporting anode anode can can be beachieved achieved with with controlled controlled porosity porosity at properly at properly adjustedadjusted laser laser power power and and scan scan speed. speed. WeWe have have developed developed a aprocess process to to control thethe laserlaser power, power, scan scan speed, speed, and and scan scan patterning patterning to obtain to obtaincomplete complete cell unitcell unit fabrication fabrication (anode, (anode, electrolyte, electrolyte, and and cathode). cathode). Over Over 50% 50% yield yield of complete of complete SOFC SOFCcells cells was was achieved achieved from from the the anode anode substrates. substrates. The The boundary boundary and and porosityporosity werewere wellwell controlled to to achieve desireddesired valuesvalues as as described described in in Table Table1. A1. completeA complete SOFC SOFC cell wascell was sent tosent a thirdto a third party party (Fiaxell, (Fiaxell,Lausanne, Lausanne, Switzerland) Switzerland) for performance for performance tests including tests including polarization polarization (I-V) curve (I-V) and curve electro and impedance electro impedancespectroscopy spectroscopy (EIS). Figure (EIS).6c,d Figure show 6c,d the show cell before the cell and before after and the performanceafter the performance tests. Of note,tests. sinceOf note,there since is nothere interconnect is no interconnect integrated integrated with the with cell sample,the cell itsample, is a challenge it is a challenge for the Au for grid the attachedAu grid to attachedthe cell to to the have cell ato perfecthave a contact.perfect contact. This may This impact may impact the ﬁnal the results.final results. However, However, the purpose the purpose of this of publicationthis publication is to is demonstrate to demonstrate that thethat basic the ba three-layeredsic three-layered structure structure (anode, (anode, electrolyte, electrolyte, and cathode) and cathode)made withmade LAM with canLAM generate can generate a current. a current.

a b

c d

FigureFigure 6. ( 6.a) (Completea) Complete three three layered layered cell cell samples samples (30 (30× 30 mm);30 mm); (b) (cross-sectionb) cross-section SEM SEM image image of the of the × interfacesinterfaces for fora complete a complete cell cell sample; sample; (c) (ac cell) a cell sample sample is mounted is mounted with with a Au a Augrid grid for fortesting testing I-V I-Vcurve curve andand EIS; EIS; (d) (thed) thecell cell sample sample after after testing. testing.

TheThe testing testing results results are are shown shown in Figure in Figure 7. 7The. The cell cell sample sample was was functional functional and and an anelectrical electrical current current waswas generated. generated. Two Two measurements measurements were were implemen implementedted to reflect to reﬂect thethe time time dependent dependent performance. performance. DataData with with solid solid markers markers were were taken taken right right after after inject injectinging the the H2 H2 into into the the cell. cell. The The cell cell was was then then kept kept in in aa protective protective atmosphere atmosphere (200 (200 mL/min mL/min of of 92%N2 92%N2 ++ 8%H2)8%H2) overnight. overnight. Data Data with with open open markers markers were were takentaken after after 18 h 18 in h 92%N2 in 92%N2 + 8%H2,+ 8%H2, followed followed by the by injection the injection of H2. of Three H2. Threetemperatures temperatures (700 °C, (700 800◦ C, °C,800 and◦ C,900 and °C) 900 were◦C) sequentially were sequentially tested. tested.The resist Theance resistance of cells was of cells found was to found be much to be higher much than higher thethan SOFC the made SOFC by made conventional by conventional approaches approaches such as cold such spray. as cold The spray. higher The resistance higher causes resistance a lower causes voltagea lower and voltage lower current and lower at the current same attime. the Based same time.on our Based current on LAM our current process, LAM the high process, resistance the high mayresistance be caused may by bethe caused thickness by thecontrol thickness of the control YSZ electrolyte, of the YSZ with electrolyte, the thickness with thebeing thickness either too being largeeither or non-uniform. too large or non-uniform. The second possibility The second is possibilitythe YSZ electrolyte is the YSZ layer electrolyte does not layer connect does well not connectwith bothwell the with Ni–YSZ both anode the Ni–YSZ and LSM anode cathode and LSM layer cathode (disconnection). layer (disconnection). The third issue The may third come issue from may the come Aufrom grid theattachment Au grid attachmentin testing. inThe testing. Au grid The migh Au gridt not might be in notphysical be in physicalcontact contactwith the with testing the testingcell duringcell duringthe 18 theh testing 18 h testing period. period. All these All theseissues issues are related are related to the to theuniformity uniformity control control of the of theLAM LAM processprocess over over the the entire entire SOFC SOFC area, area, especially especially th thee surface roughnessroughnessof of the the Ni–YSZ Ni–YSZ anode anode and and thin thin layer layerprinting printing uniformity uniformity of theof the YSZ YSZ electrolyte. electrolyte. We areWe currentlyare currently working working to improve to improve the control the control process. process. In future activities, a full temperature range, temperature cycling, and damp heat will be Appl. Sci. 2020, 10, 979 7 of 8

Appl. Sci. 2020, 9, x FOR PEER REVIEW 7 of 8 In future activities, a full temperature range, temperature cycling, and damp heat will be used to test usedits to reliabilitytest its reliability and endurance. and endurance. A model related A model to the related LAM to process the LAM will process be built towill analyze be built its failureto analyze its failuremode mode and mechanism. and mechanism.

Figure 7. Polarization (I-V) curve and electro impedance spectroscopy (EIS) test curve for a cell sample. Figure 7. Polarization (I-V) curve and electro impedance spectroscopy (EIS) test curve for a cell 4.sample. Conclusions For the first time, a three layered structure, as a basic cell unit of SOFC, was built and demonstrated 4. Conclusions via laser additive manufacturing. The processed multi-layer structure was examined and characterized, andFor athe basic first cell time, sample a wasthree tested layered on polarization structure, I-V as and a EISbasic curves. cell Itunit shows of thatSOFC, the densitywas built and and demonstratedporosity can via be laser well controlled additive duringmanufacturing. the LAM process, The processed and that the mu basiclti-layer SOFC structure cell sample was functions. examined and Wecharacterized, believe that onceand thea basic interface cell strengthsample andwas uniformity tested on ofpolarization the YSZ electrolyte I-V and thickness, EIS curves. reduction It shows of the Ni–YSZ surface roughness, and the LAM integration of the interconnect are improved, there will that the density and porosity can be well controlled during the LAM process, and that the basic SOFC be great potential for many applications. cell sample functions. We believe that once the interface strength and uniformity of the YSZ electrolyteAuthor thickness, Contributions: reductionConceptualization, of the Ni–YSZ J.L. and surface S.B.; methodology, roughness, S.B.; and formal the LAM analysis, integration S.B. and J.L.; of the investigation, S.B. and J.L.; writing—original draft preparation, S.B.; writing—review and editing, J.L.; supervision, interconnectJ.L.; project are administration, improved, J.L.; there funding will acquisition,be great potential J.L. All authors for many have read applications. and agree to the published version of the manuscript. AuthorFunding: Contributions:This research Conceptualization, was funded by the J.L. Department and S.B.; of Energymethodology, (DOE) Small S.B.; Business formal Innovationanalysis, S.B. Research and J.L.; investigation,(SBIR) SBIR S.B. DE-SC0015199. and J.L.; writing—original draft preparation, S.B.; writing—review and editing, J.L.; supervision,Acknowledgments: J.L.; projectThis administration, research was J.L.; funded funding by the acquisition, Department J.L. of Energy (DOE) Small Business Innovation Research (SBIR) SBIR DE-SC0015199. Special thanks go to Patcharin Burke for her guidance and support. Funding: This research was funded by the Department of Energy (DOE) Small Business Innovation Research (SBIR)Conflicts SBIR DE-SC0015199. of Interest: The authors declare no conflicts of interest.

Acknowledgments:References This research was funded by the Department of Energy (DOE) Small Business Innovation Research (SBIR) SBIR DE-SC0015199. Special thanks go to Patcharin Burke for her guidance and support. 1. Kamath, C.; El-dasher, B.; Gallegos, G.F.; Gallegos, W.; King, W.E.; Sisto, A. Density of Conflicts additively-manufactured,of Interest: The authors declare 316L SS no parts conflicts using laserof interest. powder-bed fusion at powers up to 400 W. Int. J. Adv. Manuf. Technol. 2014, 74, 65–78. [CrossRef] References2. Abe, F.; Osakada, K.; Shiomi, M.; Uematsu, K.; Matsumoto, M. The manufacturing of hard tools from metallic powders by selective laser melting. J. Mater. Process. Technol. 2001, 111, 210–213. [CrossRef] 1. 3.Kamath,Bai, S.;C.; Yang, El-dasher, L.; Liu, J. ManipulationB.; Gallegos, of G.F.; Microstructure Gallegos, in W.; Laser King, Additive W.E.; Manufacturing. Sisto, A. DensityAppl. Phys. of Aadditively-2016, manufactured,122, 495. [CrossRef 316L SS] parts using laser powder-bed fusion at powers up to 400 W. Int. J. Adv. Manuf. 4.Technol.Nie, 2014 B.; Huang,, 74, 65–78. H.; Bai, S.; Liu, J. Femtosecond laser melting and resolidifying of high-temperature powder 2. Abe,materials. F.; Osakada,Appl. K.; Phys. Shiomi, A 2015 ,M.;118 ,Uematsu, 37–41. [CrossRef K.; Matsumoto,] M. The manufacturing of hard tools from 5.metallicNie, powders B.; Yang, L.;by Huang,selective H.; laser Bai, S.;melting. Wan, P.; J. Liu, Mater. J. Femtosecond Process. Technol laser.additive 2001, 111 manufacturing, 210–213. of iron and 3. Bai, S.;tungsten Yang, parts. L.; Liu,Appl. J. Phys.Manipulation A 2015, 119 of, 1075–1080.Microstructure [CrossRef in ]Laser Additive Manufacturing. Appl. Phys. A 6.2016,Fulcher, 122, 495–460. B. Eﬀects of Laser Parameter Variation and Thermal Processing on Mechanical Properties and Microstructure of Laser-Melted Inconel 718. In Proceedings of the Additive Manufacturing with Powder 4. Nie, B.; Huang, H.; Bai, S.; Liu, J. Femtosecond laser melting and resolidifying of high-temperature powder Metallurgy Conference, Orlando, FL, USA, 19–20 May 2014; p. 146. materials. Appl. Phys. A 2015, 118, 37–41. 5. Nie, B.; Yang, L.; Huang, H.; Bai, S.; Wan, P.; Liu, J. Femtosecond laser additive manufacturing of iron and tungsten parts. Appl. Phys. A 2015, 119, 1075–1080. 6. Fulcher, B. Effects of Laser Parameter Variation and Thermal Processing on Mechanical Properties and Microstructure of Laser-Melted Inconel 718. In Proceedings of the Additive Manufacturing with Powder Metallurgy Conference, Orlando, FL, USA, 19–20 May 2014; p. 146. 7. Huang, W.L.; Zhu, Q.; Xie, Z. Gel-cast anode substrates for solid oxide fuel cells. J. Power Sour. 2006, 162, 464–468. Appl. Sci. 2020, 10, 979 8 of 8

7. Huang, W.L.; Zhu, Q.; Xie, Z. Gel-cast anode substrates for solid oxide fuel cells. J. Power Sour. 2006, 162, 464–468. [CrossRef] 8. Sammes, N.M.; Du, Y.; Bove, R. Design and fabrication of a100W anode supported micro-tubular SOFC stack. J. Power Sour. 2005, 145, 428–434. [CrossRef] 9. Lawlor, V. Review of the micro-tubular solid oxide fuel cell (PartII: Cell design issues and research activities). J. Power Sour. 2013, 240, 421–441. [CrossRef] 10. Ding, J.; Liu, J.; Yuan, W.; Zhang, Y. Slip casting combined with colloidal spray coating in fabrication of tubular anode-supported solid oxide fuel cells. J Eur. Ceram. Soc. 2008, 28, 3113–3117. [CrossRef] 11. Liu, J.; Bai, S. Femtosecond laser additive manufacturing of YSZ. Appl. Phys. A 2017, 123, 293. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).