A unified approach to understanding conductivity enhancement in nanoparticle-filled solid polymer electrolytes for lithium-ion batteries
Open Access
- Author:
- Fullerton Shirey, Susan Karen
- Graduate Program:
- Chemical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- June 19, 2009
- Committee Members:
- Janna Kay Maranas, Dissertation Advisor/Co-Advisor
Janna Kay Maranas, Committee Chair/Co-Chair
Ralph H Colby, Committee Member
James Patrick Runt, Committee Member
Michael John Janik, Committee Member - Keywords:
- dielectric spectroscopy
nanocomposite
QENS
SANS
neutron scattering
Al2O3
LiClO4
battery
conductivity
nanoparticle
polyethylene oxide
nanoparticle aggregation
concentration fluctuations - Abstract:
- The relationship between structure, polyethylene oxide [PEO] mobility, and ionic conductivity is investigated for the solid polymer electrolyte, PEO/LiClO4, filled and unfilled with Al2O3 nanoparticles. Oxide nanoparticles are known to improve conductivity in solid polymer electrolytes; however, the mechanism is not well understood. We measure semi-crystalline and amorphous samples over a range of lithium and nanoparticle concentrations. Unfilled samples are prepared with ether oxygen to lithium ratios of 4:1, 8:1, 10:1, 14:1, 30:1 and 100:1, where 10:1 is the eutectic concentration. Filled samples are prepared with Al2O3 nanoparticle concentrations of 5, 10 and 25wt%, at LiClO4 concentrations of 8:1, 10:1 and 14:1. Previous X-ray diffraction results show that three crystalline phases can form in this system depending on the LiClO4 concentration and temperature: (PEO)3:LiClO4, pure PEO and (PEO)6:LiClO4. We use small-angle neutron scattering [SANS] to determine that the (PEO)3:LiClO4 phase forms cylinders with radius 125Ang., and length 700Ang. We measure the amount and size of pure PEO lamellae by exploiting the neutron scattering contrast that arises because of crystallization, and we learn that nanoparticles do not affect the extent or crystal structure of pure PEO. We also learn that crystalline (PEO)6:LiClO4 does not form immediately, but requires several days if the sample is dry, or weeks if the sample is exposed to moisture. It is generally accepted that ion mobility is maximized in amorphous polymer electrolytes, because polymer mobility (and therefore ion mobility) is faster in amorphous regions. However, conductivity through the (PEO)6:LiX crystal structure, where X is the anion, has been reported to exceed that through the amorphous equivalent. Improved conductivity is attributed to the formation of cylindrical PEO channels that direct ion transport. The channels are formed by two PEO chains wrapped around a column of lithium ions at an ether oxygen to lithium ratio of 6:1. The channels are highly conductive at low PEO molecular weight where the SPE is a powder. When the molecular weight is increased to create a flexible solid, the crystalline channels misalign and the conductivity plummets. We measure structure and mobility in dry samples at high molecular weight where the (PEO)6:LiClO4 crystal phase has not formed; however, remnants of this structure are known to persist in the liquid phase. In fact, our SANS results yield scattering consistent with concentration fluctuations that may represent the remnants of (PEO)6:LiClO4 in the liquid phase. We use quasi-elastic neutron scattering [QENS] to learn more about the molecular-level mobility of the (PEO)6:LiClO4 remnants, and our data reveals two dynamic processes. The first process at short times, is attributed to the segmental mobility of PEO, and the second process at longer times, is attributed to the restricted rotation of protons around the lithium ions. The type of motion and the radius of rotation are consistent with the cylindrical remnants of (PEO)6:LiClO4. By comparing structure, mobility and conductivity results on all the unfilled samples, we determine that a semi-crystalline sample (concentration of 14:1) has the highest conductivity at 50C, despite being less mobile, partially crystalline, and having less charge carriers than amorphous samples at the same temperature. This result suggests a decoupling of ionic conductivity and polymer mobility. It is possible that the pure crystalline PEO in the 14:1 sample stabilizes the conductive (PEO)6:LiClO4 remnants, allowing them to persist long enough for conduction to occur. When nanoparticles are added, the rotation of the (PEO)6:LiClO4 remnant becomes more restricted at an ether oxygen to lithium ratio of 8:1, suggesting a direct interaction between the ether oxygen atoms in (PEO)6:LiClO4 and acidic sites on the nanoparticle surface. The rotation is unaffected by nanoparticles at a concentration of 10:1, where crystalline nuclei of pure PEO and (PEO)6:LiClO4 are present at the eutectic temperature of 50C. We suggest that pure PEO and (PEO)6:LiClO4 form alternating layers extending away from the nanoparticle surface - consistent with the structure expected at a eutectic. This could provide a conductive pathway for lithium ions, accounting for the improved conductivity at this concentration. Above the eutectic temperature, the layers can fluctuate and rearrange easily, and are likely stabilized by the nanoparticle surface. These results suggests a new mechanism for increased lithium-ion transport in nanoparticle-filled solid polymer electrolytes. When water is introduced to the electrolyte (30 - 50% relative humidity), the crystallization of (PEO)6:LiClO4 is delayed by an additional two weeks, and pure PEO does not crystallize. It is known that multiple water molecules bind to the lithium ions, and the delayed crystallization likely results from the formation of Li+/H2O complexes that hinder the chain folding required for crystallization. An additional high-temperature melting feature also arises in the presence of water, and could possibly represent the formation of the lithium hydrate, LiClO4*3H2O. Water boosts the conductivity in both filled and unfilled samples. This is attributed to the fact that water increases the segmental motion of the polymer, and therefore the ion mobility. When nanoparticles are added, the conductivity boost is unaffected at the 8:1 concentration, whereas nanoparticles decrease the conductivity boost at a concentration of 10:1. While we do not know for certain, it is possible that the 8:1 sample undergoes phase-separation into regions rich and poor in Li+/H2O, with nanoparticles located in the Li+/H2O-poor regions. Conduction will occur in the Li+/H2O-rich regions, meaning that nanoparticles will have no influence on the conductivity-boost with water. In contrast, it is likely that the 10:1 sample will not phase separate, owing to that fact that it is at the eutectic concentration where the energies of all phases are equivalent. In this case, Li+/H2O will be dispersed throughout the sample, including those regions were nanoparticles are located. Thus, ion transport could be affected by the fact that nanoparticles will absorb water at their surfaces, decreasing the conductivity-boost with increasing nanoparticle concentration. The results of this study suggest that structure could play an important role for improving ionic conductivity in solid polymer electrolytes, despite the fact that ion transport through structure is often dismissed in favor of transport through purely amorphous regions. We suggest that nanoparticles improve conductivity by stabilizing and aligning the conductive (PEO)6:LiClO4 remnants. Understanding transport through the (PEO)6:LiX structure is important for designing a solid polymer electrolyte with adequate conductivity to operate a portable device at room temperature.