materials Article

Fabrication and Characterization of 3D-Printed Highly-Porous 3D LiFePO4 Electrodes by Low Temperature Direct Writing Process Changyong Liu 1 ID , Xingxing Cheng 1 , Bohan Li 2 , Zhangwei Chen 1 , Shengli Mi 2, * and Changshi Lao 1, * 1

2

*

Additive Manufacturing Research Institute, College of Mechatronics and Control Engineering, Shenzhen University, Shenzhen 518060, China; [email protected] (C.Li.); [email protected] (X.C.); [email protected] (Z.C.) Division of Advanced Manufacturing, Graduate School at Shenzhen, Tsinghua University, Beijing 518000, China; [email protected] Correspondence: [email protected] (S.M.); [email protected] (C.La.)

Received: 17 July 2017; Accepted: 8 August 2017; Published: 10 August 2017

Abstract: LiFePO4 (LFP) is a promising cathode material for lithium-ion batteries. In this study, low temperature direct writing (LTDW)-based 3D printing was used to fabricate three-dimensional (3D) LFP electrodes for the first time. LFP inks were deposited into a low temperature chamber and solidified to maintain the shape and mechanical integrity of the printed features. The printed LFP electrodes were then freeze-dried to remove the solvents so that highly-porous architectures in the electrodes were obtained. LFP inks capable of freezing at low temperature was developed by adding 1,4 dioxane as a freezing agent. The rheological behavior of the prepared LFP inks was measured and appropriate compositions and ratios were selected. A LTDW machine was developed to print the electrodes. The printing parameters were optimized and the printing accuracy was characterized. Results showed that LTDW can effectively maintain the shape and mechanical integrity during the printing process. The microstructure, pore size and distribution of the printed LFP electrodes was characterized. In comparison with conventional room temperature direct ink writing process, improved pore volume and porosity can be obtained using the LTDW process. The electrochemical performance of LTDW-fabricated LFP electrodes and conventional roller-coated electrodes were conducted and compared. Results showed that the porous structure that existed in the printed electrodes can greatly improve the rate performance of LFP electrodes. Keywords: low temperature direct writing; 3D printing; LiFePO4 ; lithium-ion battery

1. Introduction In recent years, as 3D printing technology advances, its application has extended from mechanical components to functional devices including energy storage devices [1], biomedical devices [2–4], electrical/electronic devices [5,6], and so forth. These emerging applications may have great impacts on the current design methodologies of functional devices. With the aid of 3D printing, 3D architectures that were impossible to fabricate in the past have now become very simple and convenient to prepare. Device performance can be greatly enhanced based on novel and unconventional 3D-printed architectures. For energy storage devices, there have been several demonstrations of 3D-printed devices including lithium-ion batteries [7–10], super capacitors [11] and redox flow cells [12], among others. Among all these applications, 3D-printed lithium-ion batteries have drawn much attention due to its wide applications and enormous impacts on modern society [13–15].

Materials 2017, 10, 934; doi:10.3390/ma10080934

www.mdpi.com/journal/materials

Materials 2017, 10, 934

2 of 13

In a conventional lithium-ion battery, cathode and anode plates are fabricated by roller coating of active materials on current collectors, and then parallel layer-stacking or winding with each other [16]. Therefore, conventional batteries are fundamentally 2D designs. However, 2D design has unavoidable challenges due to the trade-off between energy density and power density [17]. Three-dimensional (3D) lithium-ion battery architectures have been proposed to tackle this issue [18]. In a 3D lithium-ion battery, energy density can be improved by increasing the electrode height to accommodate more active materials while maintaining the Li-ion diffusion length constant. Therefore, the energy density can be enhanced without sacrificing power density [19–21]. To date, 3D lithium-ion batteries have been produced using many technologies including lithographic methods, electro-deposition, atomic layer deposition and various template synthesis methods [17]. Although significant progress has been achieved, the fabrication of 3D lithium-ion battery in a simple and low-cost way remains a challenge. In recent years, some research groups have tried to fabricate 3D lithium-ion batteries using 3D printing due to its convenience and declining cost. J. A. Lewis’s group demonstrated the first 3D-printed 3D lithium-ion battery [10]. 3D interdigitated lithium-ion batteries composed of LiFePO4 (LFP) cathode and Li4 Ti5 O12 (LTO) anode were manufactured. Liangbing Hu’s group developed graphene oxide-based LFP/LTO inks to improve the electrical conductivity of LFP and LTO electrodes [7]. They also developed solid-state electrolyte inks to achieve all-component 3D-printed lithium-ion batteries. Feng Pan’s group developed a novel ink LiMn0.21 Fe0.79 PO4 (LMFP) with doping of Mn and carbon coating [8]. 3D-printed LMFP cathode was fabricated and showed impressive electrical performance such as high capacity and rate performance. Ryan R. Kohlmeyer et al. developed a universal approach to fabricate 3D-printable lithium-ion battery electrodes [9]. Well-dispersed mixture of active materials, carbon nanofibers and polymer were used to make printable electrode inks. By tuning the ratios of these components, LFP and LTO electrodes can be fabricated using direct ink writing process. Previous research mainly focused on the development of novel inks based on the modification of active materials to improve the electrical performance of 3D-printed lithium-ion batteries. Besides inks, 3D-printing technologies also play an important role. To date, room temperature extrusion-based direct ink writing process has been used to fabricate 3D electrodes. For successful deposition of electrode ink layer upon layer, it is crucial to maintain the structural integrity and strength of printed features during the printing process. However, the electrode ink flows easily and it is difficult to maintain the shape. To solve the problem, J. A. Lewis’s group developed a graded volatility solvent system in which water evaporation induces partial solidification to ensure the shape and structural integrity of printed features [10]. In our study, low temperature direct writing (LTDW)-based 3D printing was used to fabricate 3D LFP electrodes for the first time. LFP ink was deposited into a low temperature chamber (−40 degrees Celsius) and froze as the ink was printed layer by layer. Then the frozen electrode was further freeze-dried to obtain a porous 3D electrode. Unlike the previously mentioned studies that used the evaporation of solvent to induce ink solidification, low temperature was used to freeze the electrode and maintain its shape and structural integrity in the present study. LTDW was usually used to fabricate biomaterial scaffolds with interconnected pores and graded porosity [22–24]. In this study, 3D-printed LFP electrode was fabricated by LTDW. LFP inks capable of freezing at low temperature were prepared and the rheological behavior were measured. The printing process was optimized and printing accuracy variation with processing parameters was obtained. Printing results were compared between LTDW and conventional room temperature extrusion-based printing process. Results showed that LTDW can significantly improve the structural integrity while serious collapse of printed features was observed with conventional method. The microstructure, pore size and distribution were characterized using -scanning electron microscope (SU-70, Hitachi, Tokyo, Japan), N2 adsorption isotherms (ASAP2020, Micromeritics, Norcross, GA, USA) and mercury porosimetry (AutoPore IV, Micromeritics, Norcross, GA, USA). Results showed that highly-porous 3D LFP electrodes with improved porosity can be obtained using LTDW. The improved porosity can promote

Materials 2017, 10, 934

3 of 13

the infiltration and contact of electrolyte with electrode active materials and, therefore, is beneficial to the lithium-ion diffusion and intercalation process which greatly influence the rate performance of Materials 2017, 10, 934 3 of 13 lithium-ion batteries. to the lithium-ion diffusion and intercalation process which greatly influence the rate performance

2. Experimental of lithium-ion batteries. 2.1. Materials Preparation 2. Experimental

To obtain a printable LiFePO4 ink capable of freezing at low temperature, the ink composition, ratio and preparation process need to be carefully designed. The ink composition can be divided obtainactive a printable LiFePO 4 ink capable4 ;ofadditives freezing at low temperature, inkconductive composition, into threeToparts: materials like LiFePO including bindersthe and agents, and preparation process need to be carefully designed. The inksolution. composition can be divided intoactive and;ratio solvents that dissolve polymer binders to make a polymer Powders including three parts: active materials like LiFePO4; additives including binders and conductive agents, and; materials and conductive agents were evenly dispersed into the binder polymer solution to prepare solvents that dissolve polymer binders to make a polymer solution. Powders including active the ink. In the composite materials, the only freezable agent is the solvent. Usually, for conventional materials and conductive agents were evenly dispersed into the binder polymer solution to prepare LiFePO slurry, the solvent is NMP (N-methyl-2-pyrrolidone) and the binder is PVDF (poly(vinylidene the4ink. In the composite materials, the only freezable agent is the solvent. Usually, for conventional fluoride)). LTDW, freezing rate plays an important part in the freezing of the ink during the printing LiFePOIn 4 slurry, the solvent is NMP (N-methyl-2-pyrrolidone) and the binder is PVDF process. To achieve high freezing rate,freezing a large undercooling degree ispart required but is very difficult to (poly(vinylideneafluoride)). In LTDW, rate plays an important in the freezing of the ink execute for the NMP with process. a freezing ofa− 24 degrees Therefore, ink composition and ratios during printing To point achieve high freezing Celsius. rate, a large undercooling degree is required is very difficult to execute for NMP with a freezing pointinofLTDW −24 degrees Celsius. Therefore, ink need usedbut in conventional lithium-ion batteries are not applicable process and new materials to becomposition developed.and ratios used in conventional lithium-ion batteries are not applicable in LTDW process and newstudy, materials need to be developed. In our 1,4 dioxane was added into the solvent system and functioned as the freezing agent In our study, 1,4 dioxane was added into the solvent system and functioned as the freezing agent due to its high freezing point (11.8 degrees Celsius) and low specific heat capacity. 1,4 dioxane and due to its high freezing point (11.8 degrees Celsius) and low specific heat capacity. 1,4 dioxane and deionized water were mixed to obtain a mixed solution. CMC (carboxy methyl cellulose) was used as deionized water were mixed to obtain a mixed solution. CMC (carboxy methyl cellulose) was used the binder to be dissolved in the mixed solution. Then, LiFePO powder was added and stirred by a as the binder to be dissolved in the mixed solution. Then, LiFePO44 powder was added and stirred by vacuum mixer to obtain thethe prepared ink. possibleclogging clogging due to agglomeration, the ink a vacuum mixer to obtain prepared ink.To Toavoid avoid possible due to agglomeration, the ink was filtered by abymesh screen totoremove was filtered a mesh screen removelarge large clusters. clusters. 2.1. Materials Preparation

2.2. LTDW Process 2.2. LTDW Process schematic LTDWprocess processis is shown shown in first step is CAD modeling of 3Dof 3D The The schematic of of thethe LTDW inFigure Figure1.1.The The first step is CAD modeling electrodes using CAD software (Solidworks, Dassault Systemes, Paris, France). Then the CAD model electrodes using CAD software (Solidworks, Dassault Systemes, Paris, France). Then the CAD model digitally sliced into layerstotoobtain obtain the the slice slice file. loaded intointo a syringe and and extruded was was digitally sliced into layers file.Ink Inkwas was loaded a syringe extruded through a micro-nozzle and deposited layer by layer according to the CNC (computer numerical through a micro-nozzle and deposited layer by layer according to the CNC (computer numerical control) paths of the slice file. Then the printed electrode was transferred to a freeze-drying machine control) paths of the slice file. Then the printed electrode was transferred to a freeze-drying machine to to remove the solvents and a porous electrode was obtained.

remove the solvents and a porous electrode was obtained.

Figure 1. Fabrication process (LFP)electrode electrodebyby low temperature direct writing (LTDW). Figure 1. Fabrication processofofLiFePO LiFePO44 (LFP) low temperature direct writing (LTDW). (a) CAD modeling of 3D electrodes, preparation,(c)(c) low temperature direct writing process, (a) CAD modeling of 3D electrodes,(b) (b) slurry slurry preparation, low temperature direct writing process, (d) freeze-drying process, (e)(e) porous (d) freeze-drying process, porouselectrodes. electrodes.

Materials 2017,2017, 10, 934 Materials 10, 934

4 of 134 of 13

Toward this goal, a LTDW machine (Shenzhen University, Shenzhen, Guangdong, China) Toward this goal, a LTDW machine (Shenzhen University, Shenzhen, Guangdong, China) (shown (shown in Figure 2) composed of a low temperature chamber, an extrusion nozzle, a XYZ motion in Figure 2) composed of a low temperature chamber, an extrusion nozzle, a XYZ motion stage, a CNC stage, a CNC control system and GUI software was developed. The printing head was equipped with control system and GUI software was developed. The printing head was equipped with a heating a heating to the avoid the solidification of LFP The LFP ink loaded in was the syringe unit tounit avoid solidification of LFP inks. The inks. LFP ink loaded in the syringe extrudedwas by aextruded step by amotor-driven step motor-driven extrusion rod and deposited the low chamber. temperature With extrusion rod and deposited into the lowinto temperature With chamber. the scanning of the scanning of printing head the XY printed direction, LFP materials lines are deposited printing head in the XY in direction, LFPprinted materials lines are deposited layer by layer tolayer formby 3Dlayer porous to form 3D electrodes. porous electrodes.

Figure 2. Schematic andand image ofofthe (a)Schematic Schematicofof LTDW machine, (b) the Figure 2. Schematic image theLTDW LTDW machine. machine. (a) thethe LTDW machine, (b) the LTDW machine. LTDW machine.

2.3. Characterization 2.3. Characterization Rheological behavior of of the LFPLFP inks was using rotational Rheological behavior theprepared prepared inksmeasured was measured using rheometer rotational(AR1000, rheometer TA Instruments, New Castle, DE, USA). Appropriate compositions and ratios of LFP inks (AR1000, TA Instruments, New Castle, DE, USA). Appropriate compositions and ratios of were LFP inks selected according to the rheological results. The accuracy of the printed electrodes was characterized were selected according to the rheological results. The accuracy of the printed electrodes was using an advanced optical 3D microscope (VHX-5000, Keygence, Osaka, Japan). The printing characterized using an advanced optical 3D microscope (VHX-5000, Keygence, Osaka, Japan). The parameters influencing the printing accuracy were examined and optimal printing parameters were printing parameters influencing the printing accuracy were examined and optimal printing obtained. Detailed microstructures of the printed electrodes was characterized using a scanning parameters were obtained. Detailed microstructures of the printed electrodes was characterized electron microscope (SU-70, Hitachi, Tokyo, Japan). Nitrogen adsorption isotherms were measured usingonaASAP scanning electron microscope (SU-70, Hitachi, Nitrogen adsorption isotherms 2020(Micromeritics, Norcross, GA, USA). TheTokyo, specific Japan). surface area was computed using the weremultipoint measuredBrunauer-Emmett-Teller on ASAP 2020(Micromeritics, Norcross, GA, USA). The specific surface area was (BET) method. Pore size and distribution was computed using computed using the multipoint Brunauer-Emmett-Teller method. Pore size and distribution the Barrett-Joyner-Halenda (BJH) method. Since nitrogen (BET) adsorption method was only capable of pores the diameter rang of 2 nm~200 Mercury porosimetry was performed was characterizing computed using theinBarrett-Joyner-Halenda (BJH) nm. method. Since nitrogen adsorption method to measure the pore size, distribution and porosity on mercury intrusion machine (AutoPore IV, was only capable of characterizing pores in the diameter rang of 2 nm~200 nm. Mercury porosimetry was Micromeritics, performed toNorcross, measureGA, theUSA). pore size, distribution and porosity on mercury intrusion machine To verify the impact of the porousGA, structure on the rate performance of LiFePO4 electrodes, the (AutoPore IV, Micromeritics, Norcross, USA). rate performance of LiFePO4 electrodes fabricated by LTDW process and conventional roller coating To verify the impact of the porous structure on the rate performance of LiFePO4 electrodes, the method was compared. The electrodes were assembled into coin-type half-cells using Li-metal as the rate performance of LiFePO4 electrodes fabricated by LTDW process and conventional roller coating counter electrode. The charge/discharge performance was conducted at cutoff voltage between 2.5 method was compared. The electrodes were assembled into coin-type half-cells using Li-metal as the and 4.2 V. The current rate was 0.1, 0.2, 0.5, 1, 5, and 10 C. Cycle performance was also studied at 0.5 C counter electrode. for 100 cycles. The charge/discharge performance was conducted at cutoff voltage between 2.5

and 4.2 V. The current rate was 0.1, 0.2, 0.5, 1, 5, and 10 C. Cycle performance was also studied at 0.5 C for 100 cycles.

Materials 2017, 10, 934 Materials 2017, 10, 934

5 of 13 5 of 13

3. Results 3. Results

3.1. Rheological Properties of LFP Slurries 3.1. Rheological Properties of LFP Slurries

CMC is a commonly used binder for lithium-ion batteries. In addition to binding the LFP particles, CMC also beused used to adjust the viscosity of theInLFP ink that is very to its CMC is a can commonly binder for lithium-ion batteries. addition to binding theimportant LFP particles, CMC can also be used to adjust the viscosity of the LFP ink is very important printability. printability. To investigate the influence of CMC content on that the apparent viscositytoofitsLFP ink, CMC To investigate the influence of CMC on the viscosity of % LFP ink,prepared. CMC aqueous aqueous solutions with the loading of 1content wt %, 1.33 wtapparent %, 2 wt %, and 4 wt were Results solutions ofof 1 wt %, 1.33 wt %,the 2 wt %, and increased 4 wt % were prepared. Results showed thatwith withthe theloading increase CMC content, viscosity dramatically (shownshowed in Figure with increase of CMC content, theapparent viscosityviscosity increasedvaried dramatically Figure 3).that When thethe CMC content reached 2%, the between(shown 10 Pa·sinand 100 3). Pa·s When the1~50 CMCS−1 content reached of 2%, the apparent varied between ·s and 100impact Pa·s (shear (shear rate ). The adding 1,4 dioxane to viscosity the solvent system had 10 an Pa important on the rate 1~50 S−1 ). The adding of 1,4 dioxaneoftoCMC-water-1,4 the solvent system had an important the solution properties. The apparent viscosity dioxane solution was impact greatly on reduced solution properties. The apparent viscosity of CMC-water-1,4 dioxane solution was greatly reduced (shown in Figure 3). The apparent viscosity increased with the increase of 1,4 dioxane proportion in Figure 3). The apparent viscosity increased with the increase of 1,4 dioxane proportion in the(shown solventinsystem. When the 1,4 dioxane proportion exceeded 50%, white flocs were observed as the solvent system. When the 1,4 dioxane proportion exceeded 50%, white flocs were observed as CMC could not completely dissolve in the solvent. As a result, 50% proportion was selected. Then CMC could not completely dissolve in the solvent. As a result, 50% proportion was selected. Then LFP LFP particles were mixed into the CMC-water-1,4 dioxane solution to obtain the LFP ink. The particles were mixed into the CMC-water-1,4 dioxane solution to obtain the LFP ink. The apparent apparent viscosity increased remarkably with the increase of LFP amount. When the amount of LFP viscosity increased remarkably with the increase of LFP amount. When the amount of LFP powder powder reached 25 g/50 mL, the apparent viscosity of LFP ink varied between 50 Pa·s and 500 Pa·s reached 25 g/50 mL, the apparent viscosity of LFP ink varied between 50 Pa·s and 500 Pa·s (shear (shear rate 1~50 S−1). If the amount of LFP exceeded 25 g/50 mL, the ink tended to dry very quickly rate 1~50 S−1 ). If the amount of LFP exceeded 25 g/50 mL, the ink tended to dry very quickly and and dried particles greatly affected the printing fluency and continuity. evaluation, thethe dried particles greatly affected the printing fluency and continuity. After After carefulcareful evaluation, LFP LFP ink with 25 g LFP powder dispersed in 50 mL CMC solution (25 mL water and 25 mL 1,4 dioxane) ink with 25 g LFP powder dispersed in 50 mL CMC solution (25 mL water and 25 mL 1,4 dioxane) was selected forfor the the printable printablemaximum maximumand andminimum minimum mass was selected theLTDW LTDWprocess. process.In Inour our experiments, experiments, the mass loading of of active 30 g/50 g/50 mL 15 g/50 g/50 mL, respectively. loading activematerials materialsininthe theLFP LFPslurry slurry was was 30 mL and and 15 mL, respectively.

Figure Rheological properties LFP inks: carboxy methyl cellulose (CMC) watersolution; solution;(b) Figure 3. 3. Rheological properties of of LFP inks: (a)(a) carboxy methyl cellulose (CMC) water (b) CMC-water-1,4 dioxane solution; (c) LFP-CMC-water-1,4 dioxane ink. CMC-water-1,4 dioxane solution; (c) LFP-CMC-water-1,4 dioxane ink.

Printing ProcessOptimization Optimizationand andAccuracy Accuracy Characterization Characterization 3.2.3.2. Printing Process LTDWprocess, process,main mainprinting printing parameters parameters include: vj ,vscanning speed vxyvxy ForFor thethe LTDW include:extrusion extrusionspeed speed j, scanning speed and layer thicknesshsh. sTo . Toinvestigate investigatethe theinfluence influence of of printing results, anan and layer thickness printing parameters parameterson onthe theprinting printing results, 3 ) was designed (shown in Table 1). Printing tests were carried out using a orthogonal test scheme L (3 3 9 orthogonal test scheme L9(3 ) was designed (shown in Table 1). Printing tests were carried out using gridpattern pattern with with the printing a grid the dimensions dimensionsofof2020mm mm××20 20mm mm×× 22 mm mm with withline linespacing spacingofof1 1mm. mm.The The printing results are shown in Figure 4. results are shown in Figure 4.

Materials 2017, 10, 934 Materials 2017, 10, 934

6 of 13 6 of 13

Table1.1.Printing Printingparameters. parameters. Table No. 1 2 3 4 5 6 7 8 9

No. 1 2 3 4 5 6 7 8 9

Extrusion Speed vj Extrusion (mm/s)Speed vj (mm/s) 0.003 0.003 0.003 0.003 0.003 0.003 0.006 0.006 0.006 0.006 0.006 0.006 0.009 0.009 0.009 0.009 0.009 0.009

Scanning Speed vxy Layer Thickness hs Printed Structures Scanning Speed Layer Thickness Printed Structures (mm/s) (mm) (Figure 4) vxy (mm/s) hs (mm) (Figure 4) 2 0.1 (i) 2 4 0.1 0.15 (i) (b) 4 6 0.15 0.2 (b) (a) 6 2 0.2 0.15 (a) (h) 2 0.15 (h) 4 0.2 (d) 4 0.2 (d) 6 0.1 (e) 6 0.1 (e) 2 0.2 (g) 2 0.2 (g) (f) 4 4 0.1 0.1 (f) 6 0.15 (c) 6 0.15 (c)

Figure 0.003 mm/s, Figure4. 4. Images Imagesof ofprinted printedLFP LFPgrid gridpatterns patternsusing usingdifferent differentprinting printingparameters. parameters.(a) (a)vjv= j = 0.003 mm/s, vvxy = 6 mm/s, h = 0.2 mm; (b) v = 0.003 mm/s, v 4 = 4 mm/s, hs = mm; 0.15 mm; vj = 0.009 xy = 6 mm/s, hs s= 0.2 mm; (b) vj =j 0.003 mm/s, vxy =xy mm/s, hs = 0.15 c) vj =c)0.009 mm/s,mm/s, vxy = 6 vmm/s, mm/s, h = 0.15 mm; (d) v = 0.006 mm/s, v = 4 mm/s, h = 0.2 mm; (e) v = 0.006 mm/s, xy = 6 h s xy s j j s = 0.15 mm; (d) vj = 0.006 mm/s, vxy = 4 mm/s, hs = 0.2 mm; (e) vj = 0.006 mm/s, vxy = 6 mm/s, hs v=xy0.1 = mm; 6 mm/s, = 0.1 mm/s, mm; (f) 0.009 mm/s, 4 mm/s, hs = 0.1 mm; vj = 0.009 mm/s, xy = (g) vxyv=j = 4 mm/s, hs = 0.1 vmm; vj = 0.009 mm/s, vxy =(g) 2 mm/s, hs = 0.2 mm; (f) vj h=s 0.009 vxy = 2 mm/s, hs = 0.2 mm; (h) vj = 0.006 mm/s, vxy = 2 mm/s, hs = 0.15 mm; (i) vj = 0.003 mm/s, (h) vj = 0.006 mm/s, vxy = 2 mm/s, hs = 0.15 mm; (i) vj = 0.003 mm/s, vxy = 2 mm/s, hs = 0.1 mm. vxy = 2 mm/s, hs = 0.1 mm.

It can be seen that the printing parameters have important impacts on the fabrication results. canprinted be seenpatterns that theshown printing parameters have important impacts on the fabrication results. FromIt the in Figure 4a–e, disconnection and discontinuity of the printed lines From printed as patterns in Figure 4a–e, disconnection and discontinuity of the printed lines can bethe observed a resultshown of non-uniform materials supply caused by inappropriate combination of can be observed as a result of non-uniform materials supply caused by inappropriate combination of printing parameters. In Figure 4f–i, by contrast, clear and continuous printed patterns can be seen. printing parameters. In Figure 4f–i, by contrast, and continuous printed patterns can be seen. However, the printed line width varied withclear different combinations of printing parameters. However, the printed varied with different combinations of printing parameters. Figure 4i Figure 4i exhibits the line bestwidth printing accuracy with the finest line width. exhibits best printing the finest line width. To the characterize the accuracy printing with accuracy, the values of line width and diagonal distance of the To characterize printing accuracy, the values width and diagonal of the1 intersection point arethe shown in Figure 5. Three groupsofofline parameters displayed in distance bold in Table intersection point are shown in Figure 5. Three groups of parameters displayed in bold in Table 1 with good printing quality are compared. A unit flow equivalent number was defined as: vj/(vxy with × hs). good printingdenotes qualitythat are unit compared. A unit flow equivalent number wasand defined as: vproportional The number flow is proportional to the extrusion speed inversely j /(vxy × hs ). The number denotes that unit flowthickness. is proportional to the extrusion speed and and diagonal inversely distance proportional to the scanning speed and layer The variation of line width with to the scanning speed and layer thickness. of line width and diagonal distance with equivalent number is shown in Figure 5d. ItThe canvariation be seen that both line width and diagonal distance equivalent number is shown Figure 5d. It number. can be seen line width and distance increased with the increase ofin the equivalent Thethat bestboth line width of 250 μmdiagonal using parameters increased the of hthe The best line width of 250 µm using parameters vj = 0.003 with mm/s, vxyincrease =2 mm/s, s = equivalent 0.1 mm cannumber. be obtained. vj = 0.003 mm/s, vxy = 2 mm/s, hs = 0.1 mm can be obtained.

Materials 2017, 10, 934 Materials 2017, 10, 934 Materials 2017, 10, 934

7 of 13 7 of 13

7 of 13

Figure 5.5.Comparison Comparison on the printing accuracy accuracyof ofLTDW LTDWprinted printedgrid gridpatterns patterns using different Figure5. Comparison on on the the printing printing accuracy using different Figure of LTDW printed grid patterns using different parameters: (a) v 0.003 mm/s, = 2 mm/s, h = 0.1 mm; v = 0.006 mm/s, v = mm/s, j = 0.003 mm/s, v xyv= 2 mm/s, h s = 0.1 mm; (b) v j =(b) 0.006 mm/s, v xy = 2 mm/s, h s 2 parameters: (a) v xy s xy j j = 0.15 parameters: (a) vj = 0.003 mm/s, vxy = 2 mm/s, hs = 0.1 mm; (b) vj = 0.006 mm/s, vxy = 2 mm/s, =hs0.15 h = 0.15 mm; (c) v = 0.009 mm/s, v = 4 h = 0.1 mm; (d) variation of line width and diagonal j = 0.009 mm/s, v xy = 4 mm/s, h s =mm/s, 0.1 mm; (d) variation of line width and diagonal distance mm; (c) v s xy s j mm; (c) vj = 0.009 mm/s, vxy = 4 mm/s, hs = 0.1 mm; (d) variation of line width and diagonal distance distance with unit flow equivalent with unit flow equivalent number.number. with unit flow equivalent number.

To further characterize the printing accuracy on the height direction, a comb-shaped LFP thethe printing accuracy on the direction, a comb-shaped LFP electrode To further furthercharacterize characterize printing accuracy onheight the height direction, a comb-shaped LFP electrode was printed and shown in Figure 6. The LFP electrodes were printed with 8 layers, 12 layers was printed shown Figurein6.Figure The LFP electrodes were printed with 8with layers, 12 layers and electrode wasand printed andinshown 6. The LFP electrodes were printed 8 layers, 12 layers and 16 layers, respectively. Then the height of LFP electrodes was measured, and the variation of 16 layers, respectively. Then the height of LFP of electrodes was measured, and the variation of electrode and 16 layers, respectively. Then the height LFP electrodes was measured, and the variation of electrode height with the number of layers is shown in Figure 6d. It can be seen that the actual height height with the with number of layersofislayers shown in Figure 6d. It6d. canIt be thatthat the the actual height of electrode height the number is shown in Figure canseen be seen actual height of printed electrodes was about 85% of the theoretical values, indicating the existence of possible printed electrodes waswas about 85% 85% of theoftheoretical values,values, indicating the existence of possible material of printed electrodes about the theoretical indicating the existence of possible material spreading. We guess there are some factors in the whole process that may lead to the spreading. We guessWe there arethere some factors the whole process that mayand lead thelead spreading material spreading. guess are somein in to the whole process that may to the spreading of material, including the exposure offactors materials the atmosphere theto freeze-drying ofprocess. material, including the exposure of materials to the atmosphere and the freeze-drying process. spreading of material, including the exposure of materials to the atmosphere and the freeze-drying Although the spreading of materials may be difficult to eliminate, the height reduction did Although the spreading of materials bemay difficult to eliminate, height reduction did did not process. Although the spreading of materials be difficult eliminate, the height reduction not influence the printed shape and themay mechanical integrity cantobe wellthe maintained. influence the printed shape andand the mechanical integrity can can be well maintained. not influence the printed shape the mechanical integrity be well maintained.

Figure 6. Printed comb-shaped LFP electrode: (a) 8 layers; (b) 12 layers; (c) 16 layers; (d) electrode height variation with number of layers.

Figure 6. Printed comb-shaped LFP electrode: (a) 8 layers; (b) 12 layers; (c) 16 layers; (d) electrode Figure 6. Printed comb-shaped LFP electrode: (a) 8 layers; (b) 12 layers; (c) 16 layers; (d) electrode height variation with number of layers. height variation with number of layers.

Materials 2017, 10, 934 Materials 2017, 10, 934

8 of 13 8 of 13

To further verify the effectiveness of LTDW in maintaining the shape of printed LFP electrodes, Materials 934 8 of 13 To 2017, further verifybetween the effectiveness of LTDW in maintaining the shape of printed electrodes, comparison was10,made the patterns printed at low temperature and at LFP room temperature comparison was made between the patterns printed at low temperature and at room temperature respectively using exactly the same printing parameters. The printed results are shown in Figure 7. To further verify the effectiveness of LTDW in maintaining the shape of printed LFPinelectrodes, using exactly the same printing parameters. printed results are shown Figureat7.room It canrespectively be seen that serious spreading of materials can The be observed for patterns printed comparison was made between the patterns printed at low temperature and at room temperature It can be seen that serious spreading of materials can be observed for patterns printed at room temperature. Both theexactly line width and printing diagonal distance were much results larger than LTDW respectively using same parameters. Themuch printed are shown in printed Figure 7.ones. temperature. Both the line the width and diagonal distance were larger than LTDW printed ones. The comparison between the two printedofpatterns indicates that LTDW can effectively improve the It can be seen that serious materialsindicates can be observed forcan patterns printed at room The comparison between thespreading two printed patterns that LTDW effectively improve the capacity for maintain the shape and mechanical integrity in comparison with conventional room temperature. Both the line width and and diagonal distance were larger than printed room ones. capacity for maintain the shape mechanical integrity inmuch comparison withLTDW conventional temperature methods. The comparison between the two printed patterns indicates that LTDW can effectively improve the temperature methods. capacity for maintain the shape and mechanical integrity in comparison with conventional room temperature methods.

Figure 7. Comparison of LTDW printed patterns and room temperature printed patterns: (a) LTDW

Figure 7. Comparison of LTDW printed patterns and room temperature printed patterns: (a) LTDW printed patterns; (b) the printing accuracy of LTDW printed patterns; (c) room temperature printed printed patterns; (b) the printing of LTDW patterns; (c) room temperature printed Figure 7. (d) Comparison of accuracy LTDWaccuracy printed patterns andprinted room temperature patterns; the printing of room temperature printed patterns. printed patterns: (a) LTDW patterns; (d) the printing accuracy of room temperature printed patterns. printed patterns; (b) the printing accuracy of LTDW printed patterns; (c) room temperature printed patterns; (d) theCharacterization printing accuracy of room temperature printed patterns. 3.3. Microstructure

3.3. Microstructure Characterization

To characterize the microstructure of the 3D-printed LFP electrodes, SEM images were taken on 3.3. Microstructure Characterization the surface and cross section of bothofelectrodes, as shown Figures 8 and respectively. Porouson the To characterize the microstructure the 3D-printed LFPinelectrodes, SEM9,images were taken To characterize the microstructure of theand 3D-printed LFP electrodes, SEM images were taken by on features can be observed both on the surface the cross section. The LFP particles were bonded surface and cross section of both electrodes, as shown in Figures 8 and 9, respectively. Porous features the surface and cross section of both electrodes, asfrom shown intoFigures 8The and 9, respectively. Porous CMC binders. The diameter of LFP particles ranges 1μm 10μm. pores and voids between can be observed both on the surface and the cross section. The LFP particles were bonded by CMC features can be observed onfor theelectrolyte surface and thebecross section. The LFP particles wereimpacts bonded on by LFP particles that provideboth space can observed and have very important binders. The diameter of LFP particles rangesranges from from 1µm1μm to 10µm. TheThe pores and CMC binders. The diameter of LFP particles to 10μm. pores andvoids voidsbetween between LFP the electrochemical performance. particles provide space for electrolyte can be andand have very on the LFP that particles that provide space for electrolyte canobserved be observed have veryimportant important impacts impacts on electrochemical performance. the electrochemical performance.

Figure 8. SEM images of the surface of printed LFP electrodes: (a,b) room temperature printed LFP electrodes; (c,d) LTDW printed LFP electrodes. Figure 8. SEM images of the surfaceofofprinted printedLFP LFP electrodes: electrodes: (a,b) temperature printed LFP LFP Figure 8. SEM images of the surface (a,b)room room temperature printed electrodes; (c,d) LTDW printed LFP electrodes.

electrodes; (c,d) LTDW printed LFP electrodes.

Materials 2017, 10, 934 Materials 2017, 10, 934 Materials 2017, 10, 934

9 of 13 9 of 13 9 of 13

Figure 9. 9. SEM SEM images images of of the the cross cross section section of of printed printed LFP LFP electrodes: electrodes: (a,b) (a,b) room room temperature temperature printed printed Figure Figure 9. SEM images of the cross section of printed LFP electrodes: (a,b) room temperature printed LFP electrodes; (c,d) LTDW printed LFP electrodes. LFP electrodes; electrodes; (c,d) (c,d) LTDW LTDW printed printed LFP LFP electrodes. electrodes. LFP

3.4. Characterization Characterization of Pore Pore Volume, Volume, Size Size and and Distribution Distribution 3.4. 3.4. Characterization of of Pore Volume, Size and Distribution To further further characterize characterize the the pore pore size size and and distribution, distribution, nitrogen nitrogen adsorption adsorption isotherms isotherms were were To To further characterizepore the volume pore size and distribution, nitrogen and adsorption isotherms were measured. The incremental variation with pore diameter cumulative pore volume measured. The incremental pore volume variation with pore diameter and and cumulative cumulative pore pore volume volume measured. The incremental poreshown volume variation with pore diameter varied with pore diameter are in Figure 10. It can be seen that the incremental pore volume varied with with pore pore diameter diameter are are shown shown in in Figure Figure 10. 10. It can be be seen seen that that the the incremental incremental pore pore volume volume varied It can for pores of diameter less than 30 nm is approximately the same for LTDW and room temperature for pores pores of of diameter diameter less less than than 30 30 nm nm is is approximately approximately the same for LTDW and and room room temperature temperature for same for LTDW printed electrodes. electrodes. For pores pores of of diameter diameter over 50 50 nm, nm,the LTDW printed electrodes had larger larger pore pore printed For over LTDW printed electrodes had printed electrodes. For pores of diameter over 50 nm, LTDW printed electrodes had larger pore volume than than room room temperature temperature printed printed electrodes. electrodes. From From the the cumulative cumulative pore pore volume volume curves, curves, itit can can volume volume than room temperature printed electrodes. From the cumulative pore volume curves, it can be seen that the total pore volume of LTDW printed electrodes was larger than room temperature be seen seen that that the the total total pore pore volume volume of of LTDW LTDW printed printed electrodes electrodes was was larger larger than than room room temperature temperature be printed electrodes. The results indicate that LTDW can be used to print LFP electrodes with improved printed electrodes. electrodes. The The results results indicate indicate that that LTDW can be be used electrodes with improved printed LTDW can used to to print print LFP LFP electrodes with improved pore volume. Besides pore volume, the BET (Brunauer-Emmett-Teller) specific surface area was also also pore volume. volume. Besides pore volume, the (Brunauer-Emmett-Teller) specific surface surface area area was was pore Besides pore volume, the BET BET (Brunauer-Emmett-Teller) specific also 2/g for room temperature obtained. The specific surface area value was 28.03 m printed electrodes and obtained. The specific surface area value was 28.03 m2m /g2for room temperature printed electrodes and obtained. The specific surface area value was 28.03 /g for room temperature printed electrodes 2 28.13 m m2/g /g for for LTDW printed printed electrodes, electrodes, which which are are very very close. close. Since Since the the specific specific surface surface area area was 28.13 and 28.13 m2 /gLTDW for LTDW printed electrodes, which are very close. Since the specific surface area was was mainly determined by the diameter of LFP particles, the fabrication process had little impact. mainly determined determined by by the the diameter diameter of of LFP LFP particles, particles, the the fabrication fabrication process process had had little little impact. impact. mainly

Figure Pore (a) incremental Figure10. 10. Pore Pore size size and and distribution distribution measured measured by bynitrogen nitrogenadsorption: adsorption: (a) (a) incremental incremental pore pore volume volume Figure 10. size and distribution measured by nitrogen adsorption: pore volume variation with pore diameter; (b) cumulative pore volume variation with pore diameter. variation with pore diameter; (b) cumulative pore volume variation with pore diameter. variation with pore diameter; (b) cumulative pore volume variation with pore diameter.

The SEM SEM images images shown shown in in Figures Figures 888 and and 999 revealed revealed that that aaa large large number number of of pores pores with with diameter diameter The shown in Figures and revealed that large number of pores The SEM images with diameter over111µm μmexisted existedin inthe theLFP LFPelectrodes. electrodes.Since Sincethe the nitrogen adsorption isotherms were only capable over nitrogen adsorption isotherms were only capable to over μm existed in the LFP electrodes. Since the nitrogen adsorption isotherms were only capable to characterize the pores in the diameter range of 2~200 nm. Therefore, in order to better characterize characterize the pores in the diameter range of 2~200 nm. Therefore, in order to better characterize the to characterize the pores in the diameter range of 2~200 nm. Therefore, in order to better characterize the pore pore size size and and distribution, distribution, mercury mercury porosimetry porosimetry capable capable of of measuring measuring pores pores in in the the diameter diameter the range from from tens tens of of nanometers nanometers to to hundreds hundreds of of micrometers micrometers was was performed. performed. Pore Pore volume volume variation variation range

Materials 2017, 10, 934

10 of 13

Materials 2017, 10, distribution, 934 pore size and

of 13 mercury porosimetry capable of measuring pores in the diameter10range from tens of nanometers to hundreds of micrometers was performed. Pore volume variation with pore with pore is diameter is shown 11a and percentage distribution of pores variation pore diameter shown in Figure in 11aFigure and percentage distribution of pores variation with porewith diameter diameter Figure It canthat be seen that theofvolume pores of below diameter nm is shown is inshown Figure in 11b. It can11b. be seen the volume pores ofofdiameter 200below nm for200 LTDW for LTDW printed electrodes is slightly larger than room temperature printed electrodes. The volume printed electrodes is slightly larger than room temperature printed electrodes. The volume of pores of of diameter overis 50 μm is approximately identical for electrodes both of pores diameter over 50 µm approximately identical for electrodes fabricatedfabricated using bothusing processes. processes. In the diameter range of 200~50 μm, the pore volume of LTDW printed electrodes is much In the diameter range of 200~50 µm, the pore volume of LTDW printed electrodes is much larger than larger than room temperature printed ones. The porosity of printed electrodes was also obtained by room temperature printed ones. The porosity of printed electrodes was also obtained by the mercury the mercury intrusion process. Results showed that the porosity of the LTDW printed electrodes was intrusion process. Results showed that the porosity of the LTDW printed electrodes was 71.8%, while 71.8%, while the porosity was 61.4% for room temperature printed ones. the porosity was 61.4% for room temperature printed ones.

Figure 11. 11. Pore Pore size size and and distribution distribution measured measured by by mercury mercury porosimetry: porosimetry: (a) (a)pore pore volume volume variation variation Figure with pore diameter; (b) percentage distribution of pores variation with pore diameter. with pore diameter; (b) percentage distribution of pores variation with pore diameter.

3.5.Electrochemical ElectrochemicalPerformance Performance 3.5. To verify verify the the effectiveness effectiveness of of the the porous porous structure structure in in improving improving the the rate rate performance performance of of LiFePO LiFePO44 To electrodes,the the electrochemical electrochemicalperformance performanceof ofthe theelectrodes electrodesfabricated fabricatedby byLTDW LTDWprocess process and and the the electrodes, conventionalroller rollercoating coatingprocess processwas wascompared. compared.To Tomake makethe theresults results comparable, comparable,the themass massof ofthe the conventional activematerials materialscoated coatedon onthe thealuminum aluminumfoil foilfor forboth bothof ofthe thetwo twotypes typesof ofelectrodes electrodeswas wasadjusted adjustedto to active around 4.0 12 shows the charge/discharge curves, rate performance and cycle performance around 4.0mg. mg.Figure Figure 12 shows the charge/discharge curves, rate performance and cycle of the two types of two electrodes. performance of the types of electrodes. canbebeseen seen that discharge capacities at various current rates (0.1~10 C)LTDWof the ItIt can that thethe discharge capacities at various current rates (0.1~10 C) of the LTDW-fabricated electrodes have beenenhanced greatly enhanced in comparison with conventional ones fabricated electrodes have been greatly in comparison with conventional ones fabricated 1 forconventional fabricated by rollerWhen coating. When is 10 the discharge capacity mAhg−1−for conventional by roller coating. the ratethe is rate 10 C, theC,discharge capacity is is 6161mAhg 1 for −1− electrodeswhile while82 82mAhg mAhg LTDW-fabricated electrodes. Cycle performance measured at electrodes for LTDW-fabricated electrodes. Cycle performance waswas measured at 0.5 ◦ 0.5forC 100 for 100 cycles. capacity retention is around 85% both types electrodesafter after100 100cycles. cycles. °C cycles. TheThe capacity retention is around 85% forfor both types ofof electrodes Electrochemical results results showed showed that that the the porous porous structure structure achieved achieved using using the the LTDW LTDW process process can can Electrochemical improve the the rate rate performance performance of of LiFePO LiFePO44electrodes. electrodes. improve In comparison comparison with with other other studies studies that that tried tried to to improve improve the the rate rate performance performance of of LiFePO LiFePO44 by by In materials modification and which cancan improve thethe discharge capacity to over materials andnovel novelsynthesis synthesisapproach approach which improve discharge capacity to −1 (10 −1 (10 100 mAhg C) [25,26], the the raterate performance presented in this study is not charming. There are over 100 mAhg C) [25,26], performance presented in this study is not charming. There twotwo reasons limiting thethe improvement of rate performance. One is that thethe average particle size of are reasons limiting improvement of rate performance. One is that average particle size commercial LiFePO powder used in this study was ~2 µm and larger diameter of particle would limit of commercial LiFePO 4 powder used in this study was ~2 μm and larger diameter of particle would 4 the transport of Li-ions in solid activeactive materials. Another important factorfactor is theisbinder CMCCMC used limit the transport of Li-ions in solid materials. Another important the binder in our Usually, PVDF is theismost commonly usedused binder for LiFePO . We speculate that the used instudy. our study. Usually, PVDF the most commonly binder for LiFePO 4 . We speculate that 4 the combination of active materials and binders relatively poor rate performance obtained combination of active materials and binders ledled to to thethe relatively poor rate performance obtained in in this study. However, LTDW process which can produce highly porous electrodes may used this study. However, thethe LTDW process which can produce highly porous electrodes may bebe used to to improve rate performance lithium-ion batteries. improve thethe rate performance of of lithium-ion batteries.

Materials 2017, 10, 934 Materials 2017, 10, 934

11 of 13 11 of 13

Figure 12.Electrochemical Electrochemical performance of LTDW-fabricated LiFePO4 and electrodes and Figure 12. performance of LTDW-fabricated LiFePO4 electrodes conventional conventional roller-coated electrodes: (a) charge/discharge curves of LTDW-fabricated electrodes; roller-coated electrodes: (a) charge/discharge curves of LTDW-fabricated electrodes; (b) (b) charge/discharge curves of roller-coated electrodes; (c) rate performance; (d) cycle performance. charge/discharge curves of roller-coated electrodes; (c) rate performance; (d) cycle performance.

4. Discussion 4. Discussion From the quality qualityof ofthe theprinted printedresults resultsand and characterization of the printed electrodes, it From the thethe characterization of the printed electrodes, it can can be concluded that LTDW is an effective process for fabricating LFP electrodes. idea of be concluded that LTDW is an effective process for fabricating 3D LFP3D electrodes. The ideaThe of freezing freezing LFP ink during theprocess printing tothe maintain the mechanical integrity andofthe of LFP ink during the printing toprocess maintain mechanical integrity and the shape theshape printed the printed features is viable and practical. In comparison with previously reported printing strategy features is viable and practical. In comparison with previously reported printing strategy by rapid by rapidpartial solventevaporation partial evaporation by a specially developed the solvent enabled enabled by a specially developed graded graded solvent solvent system,system, the height height reduction (around 15%) in this study is slightly larger, which is considered to be mainly due to reduction (around 15%) in this study is slightly larger, which is considered to be mainly due to the the exposure of frozen electrode to the atmosphere and spread of materials during the freeze-drying exposure of frozen electrode to the atmosphere and spread of materials during the freeze-drying process. However, the process. However, the printed printed results results have have demonstrated demonstrated the the capacity capacity of of maintaining maintaining the the shape shape and and mechanical integrity. From the characterization of pore size and distribution using nitrogen adsorption mechanical integrity. From the characterization of pore size and distribution using nitrogen isotherms mercuryand intrusion process, it canprocess, be seen it that can that be used to fabricate 3D LFP adsorptionand isotherms mercury intrusion canLTDW be seen LTDW can be used to electrodes with and poreporosity volume.and Thepore enhanced pore volume and pore porosity can fabricate 3D LFPimproved electrodesporosity with improved volume. The enhanced volume provide more can space for electrolyte accommodation andaccommodation infiltration, which beneficial towhich the rate and porosity provide more space for electrolyte andis infiltration, is performance lithium-ion batteries. Electrochemical performance the effectiveness beneficial toofthe rate performance of lithium-ion batteries. demonstrated Electrochemical performance of porous structure on improving discharge capacity at high current rates. Gencapacity Inoue etat al.high [27] demonstrated the effectiveness ofthe porous structure on improving the discharge conducted an numerical and experimental evaluation of the relationship between porous electrode current rates. Gen Inoue et al. [27] conducted an numerical and experimental evaluation of the structure andbetween effectiveporous conductivity of ions in lithium-ion batteries. It showedof that in ainmacroscopic relationship electrode structure and effective conductivity ions lithium-ion porous electrode, the effective ionic conductivity is proportional to the porosity of electrodes and batteries. It showed that in a macroscopic porous electrode, the effective ionic conductivity is the increase oftoporosity can greatly enhance and the transport of Li-ions. This coincides with the results proportional the porosity of electrodes the increase of porosity can greatly enhance the presented in Li-ions. this study. transport of This coincides with the results presented in this study.

Materials 2017, 10, 934

12 of 13

5. Conclusions A LTDW-based 3D printing process was used to fabricate 3D LFP electrodes for the first time. A LTDW machine was developed and LFP inks capable of freezing at low temperature were developed. 1,4 dioxane was added as a freezing agent to meet the LTDW process requirements. The rheological behavior of the LFP ink was measured and appropriate viscosity range was obtained. Printing parameters were optimized by an orthogonal experiment and the optimal printing parameters were obtained. The printing accuracy was characterized and the printed line width and diagonal distance variation with unit flow equivalent number was obtained. Printed results showed that LTDW process can effectively maintain the shape and mechanical integrity of the printed structures, while room temperature printed electrodes cannot. The comparison demonstrates the feasibility and capacity of LTDW in fabricating 3D LFP electrodes. SEM characterization on the microstructures of printed electrodes using LTDW process and room temperature printing process was performed. Results showed that porous electrodes can be fabricated using both processes. Nitrogen adsorption isotherms and mercury porosimetry were used to characterize the pore size and distribution of the printed 3D LFP electrodes. Results showed that improved pore volume and porosity were obtained by the LTDW process. The process can be used to fabricate highly porous 3D LFP electrode and the improved electrode porosity can enhance the rate performance, which was verified by the electrochemical results. Acknowledgments: This study was granted financial support by Shenzhen Science & Technology Innovation Agency (Grant No. JCYJ20150626090430369) and Natural Science Foundation of Shenzhen University (Grant No. 2016034). Author Contributions: Changyong Liu developed the LTDW machine, conducted the data analysis and wrote the manuscript. Xingxing Cheng printed all the electrodes and optimized the printing process. Bohan Li performed the mercury intrusion test and analyzed the data. Zhangwei Chen conducted analysis on the N2 adsorption isotherms and revised the manuscript. Shengli Mi and Changshi Lao conceived the idea and designed the experiments. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4. 5. 6. 7.

8.

9.

10.

Ambrosi, A.; Pumera, M. 3D-printing technologies for electrochemical applications. Chem. Soc. Rev. 2016, 45, 2740–2755. [CrossRef] [PubMed] Calvert, P. Printing cells. Science 2007, 318, 208–209. [CrossRef] [PubMed] Durmus, N.G.; Tasoglu, S.; Demirci, U. Bioprinting functional droplet networks. Nat. Mater. 2013, 12, 478–479. [CrossRef] [PubMed] Norotte, C.; Marga, F.S.; Niklason, L.E.; Forgacs, G. Scaffold-free vascular tissue engineering using bioprinting. Biomaterials 2009, 30, 5910–5917. [CrossRef] [PubMed] Wang, Q.; Yu, Y.; Yang, J.; Liu, J. Fast Fabrication of Flexible Functional Circuits Based on Liquid Metal Dual-Trans Printing. Adv. Mater. 2015, 27, 7109–7116. [CrossRef] [PubMed] Wu, S.-Y.; Yang, C.; Hsu, W.; Lin, L. 3D-printed microelectronics for integrated circuitry and passive wireless sensors. Microsyst. Nanoeng. 2015, 1, 15013–15021. [CrossRef] Fu, K.; Wang, Y.B.; Yan, C.Y.; Yao, Y.G.; Chen, Y.A.; Dai, J.Q.; Lacey, S.; Wang, Y.B.; Wan, J.Y.; Li, T.; et al. Graphene Oxide-Based Electrode Inks for 3D-Printed Lithium-Ion Batteries. Adv. Mater. 2016, 28, 2587–2594. [CrossRef] [PubMed] Hu, J.T.; Jiang, Y.; Cui, S.H.; Duan, Y.D.; Liu, T.C.; Guo, H.; Lin, L.P.; Lin, Y.; Zheng, J.X.; Amine, K.; et al. 3D-Printed Cathodes of LiMn1-xFexPO4 Nanocrystals Achieve Both Ultrahigh Rate and High Capacity for Advanced Lithium-Ion Battery. Adv. Energy Mater. 2016, 6, 1600856. [CrossRef] Kohlmeyer, R.R.; Blake, A.J.; Hardin, J.O.; Carmona, E.A.; Carpena-Nunez, J.; Maruyama, B.; Berrigan, J.D.; Huang, H.; Durstock, M.F. Composite batteries: a simple yet universal approach to 3D printable lithium-ion battery electrodes. J. Mater. Chem. A 2016, 4, 16856–16864. [CrossRef] Sun, K.; Wei, T.S.; Ahn, B.Y.; Seo, J.Y.; Dillon, S.J.; Lewis, J.A. 3D printing of interdigitated Li-ion microbattery architectures. Adv. Mater. 2013, 25, 4539–4543. [CrossRef] [PubMed]

Materials 2017, 10, 934

11. 12. 13. 14. 15. 16. 17. 18. 19.

20. 21.

22.

23.

24. 25. 26.

27.

13 of 13

Zhao, C.; Wang, C.; Gorkin, R.; Beirne, S.; Shu, K.; Wallace, G.G. Three dimensional (3D) printed electrodes for interdigitated supercapacitors. Electrochem. Commun. 2014, 41, 20–23. [CrossRef] Arenas, L.F.; Walsh, F.C.; de Leon, C.P. 3D-Printing of Redox Flow Batteries for Energy Storage: A Rapid Prototype Laboratory Cell. ECS J. Solid State Sci. Technol. 2015, 4, P3080–P3085. [CrossRef] Chiang, Y.M. Building a Better Battery. Science 2010, 330, 1485–1486. [CrossRef] [PubMed] Goodenough, J.B. Electrochemical energy storage in a sustainable modern society. Energy Environ. Sci. 2014, 7, 14–18. [CrossRef] Tarascon, J.M.; Armand, M. Issues and challenges facing rechargeable lithium batteries. Nature 2001, 414, 359–367. [CrossRef] [PubMed] Armand, M.; Tarascon, J.M. Building better batteries. Nature 2008, 451, 652–657. [CrossRef] [PubMed] Ferrari, S.; Loveridge, M.; Beattie, S.D.; Jahn, M.; Dashwood, R.J.; Bhagat, R. Latest advances in the manufacturing of 3D rechargeable lithium microbatteries. J. Power Sources 2015, 286, 25–46. [CrossRef] Long, J.W.; Dunn, B.; Rolison, D.R.; White, H.S. Three-dimensional battery architectures. Chem. Rev. 2004, 104, 4463–4492. [CrossRef] [PubMed] Arthur, T.S.; Bates, D.J.; Cirigliano, N.; Johnson, D.C.; Malati, P.; Mosby, J.M.; Perre, E.; Rawls, M.T.; Prieto, A.L.; Dunn, B. Three-dimensional electrodes and battery architectures. MRS Bull. 2011, 36, 523–531. [CrossRef] Cobb, C.L.; Ho, C.C. Additive Manufacturing: Rethinking Battery Design. Electrochem. Soc. Interface 2016, 25, 75–78. [CrossRef] Roberts, M.; Johns, P.; Owen, J.; Brandell, D.; Edstrom, K.; el Enany, G.; Guery, C.; Golodnitsky, D.; Lacey, M.; Lecoeur, C.; et al. 3D lithium ion batteries—From fundamentals to fabrication. J. Mater. Chem. 2011, 21, 9876. [CrossRef] Liu, L.; Xiong, Z.; Yan, Y.N.; Hu, Y.Y.; Zhang, R.J.; Wang, S.G. Porous morphology, porosity, mechanical properties of poly(alpha-hydroxy acid)-tricalcium phosphate composite scaffolds fabricated by low-temperature deposition. J. Biomed. Mater. Res. A 2007, 82, 618–629. [CrossRef] [PubMed] Liu, L.; Xiong, Z.; Yan, Y.N.; Zhang, R.J.; Wang, X.H.; Jin, L. Multinozzle Low-Temperature Deposition System for Construction of Gradient Tissue Engineering Scaffolds. J. Biomed. Mater. Res. B 2009, 88, 254–263. [CrossRef] [PubMed] Xiong, Z.; Yan, Y.N.; Wang, S.G.; Zhang, R.J.; Zhang, C. Fabrication of porous scaffolds for bone tissue engineering via low-temperature deposition. Scri. Mater. 2002, 46, 771–776. [CrossRef] Eftekhari, A. LiFePO4 /C nanocomposites for lithium-ion batteries. J. Power Sources 2017, 343, 395–411. [CrossRef] Wang, P.; Zhang, G.; Li, Z.C.; Sheng, W.J.; Zhang, Y.C.; Gu, J.J.; Zheng, X.S.; Cao, F.F. Improved Electrochemical Performance of LiFePO4 @N-Doped Carbon Nanocomposites Using Polybenzoxazine as Nitrogen and Carbon Sources. ACS Appl. Mater. Interfaces 2016, 8, 26908–26915. [CrossRef] [PubMed] Inoue, G.; Kawase, M. Numerical and experimental evaluation of the relationship between porous electrode structure and effective conductivity of ions and electrons in lithium-ion batteries. J. Power Sources 2017, 342, 476–488. [CrossRef] © 2017 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/).