CHAPTER 2 Literature Review
2.5. Motivation
2.5.2. Dispersion of CNTs in LiCoO 2 electrode of LIBs
In 1990, the LIBs were commercialized by Sony Energytech. Owing to the need of small power source with high energy density and long cycle life for portable devices, the lithium-ion batteries have been used popularly for 3C products. The lithium metal oxides LiMO2 (M = Co, Ni and NiCo) have been used as cathode materials for lithium-ion batteries. However, the powder conductivities are poor for the pure lithium metal oxides. Various conductive materials such as carbon black, graphite, carbon fiber and CNT are used to improve the conductivity of positive electrode.
CNT is an alternative of conductive agent in LIB electrode due to its relatively high mechanical strength and electrical conductivity as well as the nature of forming a network-like morphology. However, in most cases, the CNTs aggregates seriously like tied hair and its aggregates do not break up during the ball mixing process. Up to now, study relating to dispersion of CNT is still missing. This work demonstrates the dispersion method of CNTs and its applicability to the performance improvement of LIBs.
CHAPTER 3
Experimental Methods
3.1. Dispersion and Preparation of Nano-ceramic Powders 3.1.1. Raw Materials
(1) BTO samples: High purity, amorphous BTO powder (Seedchem, Australia, Purity
≈ 99.95) with Ba/Ti = 0.915, average size = 100 nm, and specific area = 9.18 m2/g was utilized as the raw material of experiment.
(2) BST samples: The BTO (Ba0.7Sr0.3TiO3) specimens were synthesized based on the commercial BTOand SrTiO3 (STO) powders supplied by Ferro Corp., Penn Yan, NY, USA, and fabricated via the solid-state reaction process.54 According to the supplier’s data, the purities and average sizes of BTO and STO were both 99.9%
and about 100 nm. The crystal structures of BTO and STO are both cubic. The chemical dispersant for nano-ceramic powders, PDAAE, was prepared from acrylamide and (α-N, N-dimethyl-N-acryloyloxyethyl) ammonium ethanate (DAAE) via the free-radical polymerization and the detailed preparation procedures of PDAAE could be found elsewhere.55 Powder slurries were prepared by mixing the BTO and STO powders at fixed molar ratio of 7:3 and then grinding with appropriate amount of PDAAE chemical dispersant in a laboratory
appropriate time spans.
Figure 3-1. The structure of MiniZeta 03 laboratory mill.
(3) La-doped BST samples: Commercially available BTO, STO (Ferro Co., Penn Yan, NY, USA) and La2O3 (99.99%, Genesis Nanotech Co., Taiwan) powders were adopted as the starting materials. Powder slurries were prepared by mixing
the BTO and STO powders at fixed molar ratio of 7:3 and then grinding with 5 wt.% of PDAAE chemical dispersant in the MiniZeta 03 laboratory mill at the
speed of 1800 rpm for 30 min. Afterward, 0.5, 1.0 and 1.5 mol.% (0.74, 1.48 or 2.22 wt.%) of La2O3 was added in to the BTO/STO mixture with PMAA chemical dispersant further grounded in the laboratory mill at 2400 rpm for 20 min to form the slurry containing the ceramic powders of average size = 70 nm. Hereafter, the capacitor samples fabricated based on such a powder sulrry are termed as xLBST in which x denotes the dopant concentration in mol.% as cited above.
3.1.2. Chemical Dispersants
(1) PMAA-Na: PMAA-Na is a kind of anionic dispersant used to disperse ceramic powders. The anionic function group is COO−. The detail data of PMAA-Na is listed in Table 3-1.
(2) PDAAE: An amphibious dispersant was adapted. There are two kinds of function groups. The anionic function group is COO−, same as PMAA-Na. The cationic function group is N+. The detail data of PDAAE is also listed in Table 3-1.
Table 3-1. Properties of chemical dispersants.
Formula Show as below
Function group −COONa and N+ −COONa
Outlook Citrine liquid Yellow liquid
Molecular weight
3.1.3. The Experimental Flow
The experimental flows for this study are shown in Figs. 3-2 to 3.4. The experimental procedures are adjusted according to the difference of starting materials.
Figure 3-2. Experimental flow of BTO specimens.
Dispersants BTO
Laboratory mill 30 min/3600 rpm
Slurry
Dry, sieving and die pressing
Sintering
1100 ~ 1300°C/1 ~ 12 hrs
Microstructure observation Phase identification
Density measurement Electric measurement
Particle size distribution
Viscosity
Adsorption of dispersant Zeta potential
Figure 3-3. Experimental flow of BST specimens.
BTO STO Dispersants
Laboratory mill 30 min/3600 rpm
Slurry
Dry, sieving and die pressing
Sintering 800 ~1400°C/1 ~ 12 hrs
Microstructure observation Phase identification
Density measurement Electric measurement
Particle size distribution
Viscosity
BTO STO Dispersant
Laboratory mill 30 min/3600 rpm
Slurry
La2O3
Dispersant Laboratory mill
20 min/3600 rpm
Slurry
Dry, sieving and die pressing
Sintering
1000 ~1400°C/1 ~ 12 hrs
Microstructure observation Phase identification
Density measurement Electric measurement
Particle size distribution
Viscosity
3.2. Characterizations of Aqueous Slurries
3.2.1. Particle Size Distribution
Different amounts of PMAA-Na and PDAAE dispersants based on the weight of ceramic powder were added into deionized water and the pH value of solutions were adjusted to 9.5 by adding appropriate amounts of NH4OH(aq). The purpose of this procedure is to ensure the complete dissociation of chemical dispersants.55 The nano BTO powder, BST or xLBST mixtures was added. After mechanical grinding/mixing by the MiniZeta 03 laboratory mill, a small amount of slurry was taken out and sent to a particle size analyzer (Malvern Mastersizer 2000, UK) for the examination of the size distribution of powders in the slurry.
3.2.2. Zeta potential
After mechanical milling, a small amount of supernatant was removed from slurries by centrifugation, and zeta potential of the remaining powders in the supernatant was then measured by using a zeta meter (PEN KEN Inc. 501, France).
3.2.3. Viscosity
The viscosity of 60 wt.% ceramics suspensions with or without chemical dispersants subjected to mechanical grinding/mixing process via a laboratory mill was
evaluated by a viscometer (Brookfield DV-II, USA) at a rotation speed of 0- 60 rpm.
3.2.4. Adsorption of Dispersant
BTO aqueous suspensions of 30 wt.% with different concentrations of either PDAAE or PMAA-Na were prepared at pH=9.5. These suspensions were deagglomerated by a laboratory mill. After laboratory milled, the suspensions were centrifuged at a speed of 6 ×103 rpm for 30 min to obtain supernatants. The residual dispersant concentration in the supernatants was analyzed and determined by a titration procedure mentioned above. The amount of dispersant adsorbed on ceramics was calculated from the difference in dispersant concentration before and after adsorption.
3.3. Characterizations of Sintered Samples
3.3.1. Density Measurement
The density of the samples was measured by the Archimedes method. At first, the samples were bake at 150°C for 1 hr. The dry-weight W1 of it is weighted out with an electronic balance. Sample was then immersed in de-ionized water and heated at 100°C for 30 min with a hot plate. The samples were taken off and laid on a dry tissue to remove the moisture on the samples. The wet-weight of the sample W3 was
obtained. Then, the samples was hung with a small net and put in a glass of DI water in order to measure the floating weight W2. The density D of the sample is obtained by
Grain sizes, morphology and element distribution of sintered bodies were examined by scanning electron microscopy (SEM, Hitachi 4700) and transmission electron microscope (TEM, Philips, Tecnai 20, G-2) attached with energy dispersive spectroscopy (EDS, STUW 3.3). The average grain size was obtained by SEM and the intercept method.43
3.3.3. The Phase Identification of Materials
X-ray diffraction (XRD) was performed by using a diffractometer (Siemens D5000) within Cu-Kα radiation (λ = 0.1542 nm) at the scanning rate of 1°/min in order to identify the phase constitution of samples.
3.3.4. Dielectrical Properties
The indium-contained silver paste was adopted as the electrode material to provide Ohmic contact. The dielectric properties of sintered body were measured by a HP 4194A impedance analyzer at 1 kHz. Before the measurement was carried out, calibration must be done. With the capacitance of the sample measured, the real part of the relative dielectric constant εr was calculated as follows:
A
(1) Materials: NMP (Aldrich, 99.99%) and N,N-dimethyl -formamide (DMF) (Aldrich, 99.99%) were distilled over CaH2 (Aldrich, 99.95%) under reduced pressure. Pyridine (Aldrich, 99.99%) was dried by distillation after being refluxed with KOH (Aldrich, 99.99%). 4-[Di(4-aminophenyl)-methyl]phenol (Lancaster, 99.99%) was recrystallized from MeOH (Aldrich, 99.99%). Triphenyl phosphate
(TPP) (Aldrich, 99.99%) was purified by distillation under reduced pressure. All other reagents and solvents were used as received from venders unless otherwise stated.
(2) 4-{4-[Di(-aminophenyl)methyl]phenoxy}phthalonitrile (1): A mixture of 4-[Di(4-aminophenyl)-methyl]phenol (5 g, 26.96 mmol), 4-nitrophthalonitrile (4.65 g, 26.9 mmol), K2CO3 (5 g, 36.2 mmol) (Aldrich, 99.99%), and DMF (20 ml) was heated at 130°C for 4 hrs. After the heating period, the reaction mixture was cooled and poured into by filtration, dried under vacuum and purified via recrystallization from ethyl acetate to yield compound 1 (8.2 g, yield ratio ≈ 85 %).
(1)
(3) 4-{4-[Di(-aminophenyl)methyl]phenoxy}phthalic acid (2): Solution of KOH (14 g, 0.25 mmol) in a mixture of water-ethanol (100 ml/ 80 ml) was added into
compound 1 (8 g, 28.3 mmol). The mixture was then refluxed for 4 hrs. The resulted solution was diluted with water (250 ml) and acidified with 6 N HCl(aq) to pH ≈ 1.
The precipitate was filtered, washed thoroughly with water, and dried to give compound 2 (7 g, yield ratio ≈ 88 %).
(2)
(4) 4-{4-[Di(-aminophenyl)methyl]phenoxy}phthalic Anhydride (3): A mixture of compound 2 (6 g, 17.3 mmol), acetic anhydride (6 ml), and acetic acid (6 ml) was stirred under reflux for 6 hrs. After cooling, the precipitate was collected by filtration, washed with a small amount of glacial acetic acid, and then dried in vacuum at 130°C for 12 hrs to yield compound 3 (4.85 g, yield ratio ≈ 80 %).
(3)
(5) PEI: A solution of compound 3 (3 g, 2.61 mmol), TPP (1 ml, 5.2 mmol), and LiCl (160 mg, 5.77 mmol) in NMP/pyridine (50 ml, 4:1 v/v) was heated at 120°C in nitrogen ambient for 4 hrs. After cooling, the resulted polymer was precipitated in methanol. The polymer was collected, washed with hot water, purified by reprecipitation from DMF into methanol twice and dried in vacuum to produce PEI (2.32 g, yield ratio ≈ 70 %). The chemical reaction was showed in Scheme 3-1 and 3-2. The chemical structure of PEI shown in Fig. 3-5 was identified by 1H-NMR (Varian Unity-300 MHz) and FTIR (Nicolet 360).
Figure 3-5. Chemical Structure of PEI.
Scheme 3-1
Scheme 3-2
3.4.2. Experimental Flow of LIB Experiment
The experimental flow for this part of study is shown in Fig. 3-6. The LiCoO2
electrode contained 90 wt.% LiCoO2 (LICO, Taiwan), 3 wt.% polyvinylidene fluoride (PVDF, Kuraha Chemical, Japan) binder and 7 wt.% conducting additives. The conducting additives were synthetic graphite (Timrex KS6, Timcal, Switzerland) and CNTs dispersed by various amounts of PEI. The mixture containing above ingredients were mixed by ball milling for 4 hrs and then coated onto 20-µm thick Al foil, dried at 140°C and pressed to form the LiCoO2 electrode with final thickness of 120 µm. The coin cell assembly of LIB test cell is showed in Fig. 3-7.
.
Figure 3-6. Experimental flow for LIB experiment.
Resistivity measurement Electric measurement
Viscosity
Figure 3-7. The coin cell assembly of LIB test cell.
3.4.3. Optical Microscopy
The CNTs slurry was first prepared with dispersant to confirm that the CNTs was dispersed. The dispersive effect was determined to Optical Microscope (OM, OLYMPUS-Mx40, Japan).
3.4.4. Viscosity Measurement
The rheology of CNTs slurry without or with chemical dispersants subjected to ball mixing was evaluated by a viscometer (ARES-LS1, Germany).
3.4.5. Adsorption of Dispersant
The CNTs suspensions with different concentrations of PEI were prepared. These suspensions were deagglomerated by a 3D ball mixing. After ball mixing, the suspensions were centrifuged at a speed of 1.5 ×104 rpm for 2 hrs to obtain supernatants. The residual dispersant concentration in the supernatants was analyzed and determined by a UV-Visible (V-600, JASCO/Japan). The amount of dispersant adsorbed on CNTs was calculated from the difference in dispersant concentration before and after adsorption.
3.4.6. Microstructure Observation
The morphology and CNTs dispersive effect in LiCoO2 electrode was obtained by scanning electron microscopy (SEM, Hitachi 4700).
3.4.7. Cyclic Voltammograms
The test cells without and with 2 wt.% PEI in LiCoO2 electrode were evaluated
by a cyclic voltammograms (Biologic, Mac Pile II, USA) operating in between 2.9 and 4.2 V. The purpose of cyclic voltammogram test is to explore the oxidation reduction potential of the test cell and the electrochemistry change in the test cells.
3.4.8. Electrical Characterizations
Electrochemical impedance spectroscopy (EIS, Schlumberger, SI 1286 and SI
1255, England) operating at 3.5 V was adopted to characterize the impedance profile of half cell containing modified LiCoO2 electrode. The measurement was carried out at the scanning frequencies ranged from 0.01 Hz to 50 kHz and with the perturbation amplitude of 10 mV. The cycling performance and charge-discharge capability of LIB test cells were carried out with the aid of charge-discharge apparatus (Arbin, model BT2042, USA) which enables a control of 5 cells synchronously at most. A charge-discharge test consisted of five cycles. The procedure was composed of a constant-current of 2 mA followed by constant-voltage at 4.2 V until the current tapered down to 0.2 mA. Both test cells were first charged at 0.2-C current density.
CHAPTER 4
Results and Discussion
4.1. Dispersion and Preparation of Nano-scale BTO materials 4.1.1. Rheological Behaviors
Viscosity measurement is commonly adopted to evaluate the dispersed extent of concentrated suspensions. Figure 4-1 shows the effect of dispersant concentration on the viscosity of 60 wt.% BTO suspensions subjected to conventional ball milling or nano grinding/mixing via MiniZeta mill. The viscosity of suspensions without dispersion treatment is about 3200 cps. For the suspensions subjected to different mechanical milling processes, their viscosities decrease when PMAA-Na or PDAAE is incorporated. The viscosity of suspensions first decreases with the increase of dispersant concentration then reaches a plateau. At same of dispersants concentration, the viscosity of suspensions subjected to nano grinding/milling exhibits a lower value in comparison with that of suspensions subjected to conventional ball milling. For the suspensions subjected to nano grinding/mixing, the optimum chemical dispersant concentrations in BTO suspensions are about 2 wt.%
for PMAA-Na and 3 wt.% for PDAAE, which respectively result in the lowest viscosities of 15 and 76 cps.
sharpest descend of viscosity and the amount of PMAA-Na to achieve the lowest value of viscosity is less in comparison with that of PDAAE. As the PMAA-Na and PDAAE both have similar molecular weight, the difference in the rheological behaviors is attributed to the difference of chemical structures. The PMAA-Na is anionic polyelectrolyte, which causes mainly electrostatic repulsion when adsorbed on BTO particles. Though PDAAE also causes electrostatic repulsion when adsorbed on BTO surface, the effect is relatively less because PDAAE is amphibious polyelectrolyte.
Figure 4-1. The effect of dispersant concentration on the viscosity of 60 wt.% BTO suspensions subjected to various grinding/mixing processes.
4.1.2. Particle Size Distribution
In conventional ball milling, slurry and grinding media are sealed in a bottle and
the grinding/mixing is achieved by the rotation motion of grinding balls. Such a milling process provides a relatively low shear force that it is insufficient to achieve nano-scale dispersion. In this work, the MiniZeta mill using the 2-mm-diameter ZrO2
balls as the grinding media was able to grind the BTO powder down to the sizes smaller then 100 nm. By using the principle of agitator bead mills, a special agitator shaft rotating at a speed as high as 3600 rpm accelerated the grinding media in the MiniZeta mill chamber. It not only provided superfine grinding, but also generated a rather large shear force to disperse the slurry. In addition, the multi-passage grinding/mixing path adopted by MiniZeta mill is able to reduce the sizes of BTO powders in a relatively short time span.
Figure 4-2 presents the mean particle size (d50) of powder as a function of dispersant concentration in 30 wt.% BTO suspensions subjected to nano or conventional milling processes. The initial value of d50 of powder without chemical dispersants in suspensions is 1 µm. With the increasing amount of chemical dispersant, the values of d50 of powders in all types of suspensions decrease. As shown in Figure 4-2, the variation of d50 with the dispersant concentration is very much the same as that of viscosity presented in Fig. 4-2. The smallest d50 of BTO suspensions occurs at
about 2 wt.% for PMAA-Na and 3 wt.% for PDAAE; it equals to 80 nm and 83 nm, respectively. Apparently, more PDAAE are required to achieve the smallest d50 in suspensions and the minimum particle size in suspension containing PDAAE is slightly larger than that in suspension containing PMAA-Na.
0 1 2 3 4 5 6
Figure 4-2. The effect of dispersant concentration on the mean particle size (d50) of powder in 30 wt.% BTO suspensions subjected to different grinding/mixing processes.
Figures 4-3 and 4-4 respectively show the particle size distributions of 30 wt.%
BTO slurries containing 2 wt.% PMAA-Na and 3 wt.% PDAAE subjected to conventional and nano milling processes. The d50’s of suspensions subjected to the
nano grinding/mixing are 80 nm for PMAA-Na and 83 nm for PDAAE, while that of suspensions subjected to conventional ball milling are 180 nm for PMAA-Na and 181 nm for PDAAE. Furthermore, narrower shape of distribution curve implies a more uniform size of BTO particles in the suspensions subjected to nano grinding/mixing process. It was also found that average size of BTO in suspension without any mechanical milling is in the range of 200 nm. This indicated that the high shear force provided by an appropriate mechanical milling process is also beneficial to the stability of suspensions. Owing to the larger SSA and van der Waals force, the nano-scale particles in slurry tend to agglomerate together. In addition to the chemical dispersant, a mechanical milling process that is able to provide sufficient shear force is required in order to achieve satisfactory dispersion of nano-scale BTO particles in aqueous suspension.
Experimental results of viscosity and particle size distribution also indicated that a sole mechanical milling process is insufficient to disperse nano-sized BTO particles in suspensions. Though nano grinding/mixing via the MiniZeta mill provides a large shear force to disperse the particles, it cannot stabilize the suspensions. Owing to the large SSA and van der Waals force, the particles subjected to nano grinding/mixing still tend to agglomerate together if the suspensions were free of chemical dispersant.
Hence, in addition to mechanical milling, appropriate amount of chemical dispersant
has to be added to provide sufficient electrostatic repulsions and steric effect in suspensions so that a stable, uniform dispersion in nanometer scale could be obtained.
0.0 0.4 0.8 1.2
Figure 4-3. Particle size distribution of BTO powder containing 2 wt.% PMAA-Na subjected to conventional ball milling and nano grinding/mixing processes.
4.1.3. Zeta Potentials
Electrokinetic measurements are widely used to assess the magnitude and sign of
particle surface charge as a function of pH value in suspension.21,55 Figure 4-5 presents the zeta potentials of 30 wt.% BTO powder as function of pH values of suspension. Isoelectric points for 80-nm-BTO was at pH = 3.8; for 180-nm-BTOwas at pH = 4.5; and for 1-µm-BTO was at pH = 5.2. This indicates the BTO particles tend
to stabilize in aqueous solutions of high pH values.
Figure 4-4. Particle size distribution of BTO powder containing 3 wt.% PDAAE subjected to conventional ball milling and nano grinding/mixing processes.
The hydrolysis of BTO in acidic environment is as follows18:
BaTiO3(s) + H2O(l) = Ba2+(aq) + TiO2(rutile)+ 2OH−(aq) (4-1) BaTiO3(s) + 2H+(aq) = Ba2+(aq) + TiO2(rutile)+ H2O(l) (4-2)
Nano-scale BTO particles possess large SSA. When HNO3(aq) was added into the
surface of BTO. The Ba2+ ions could nullify some negative charges on the surface of the BTO, so the zeta potential shows a more positive value as revealed by Eq.(4-2).
On the other hand, when NH4OH was added into the suspension, the pH increased and the resulted OH− ions could neutralize the positive charges on the surface of BTO so that the negative trend of the zeta potential increased.
2 4 6 8 10 12
Figure 4-5. Zeta potentials of BTO suspensions at different particle sizes.
Figure 4-6 shows that the zeta potentials for BTO suspensions containing PMAA-Na or PDAAE at pH = 9.5. It can be seen that at the same particle size, the zeta potential for the suspension containing PMAA-Na is more negative than that
negative charged so that the adsorption of dispersant on particle surface is more difficult in comparison with the micro-scale particles. This suggests that the rejection caused by the net negative electric charge of the particles and, for PMAA molecules, rejection becomes more severe when the particle size is down to nanometer scale.
However, the PDAAE molecules become amphibious in a basic solution (i.e., at the condition of pH = 9.5). The molecules contain cationic functional groups (−N+R3) would favor themselves to be adsorbed on the surface of BTO particles.
0 200 400 600 800 1000
Figure 4-6. Zeta potentials of BTO suspensions containing different chemical
Figure 4-6. Zeta potentials of BTO suspensions containing different chemical