PC BASED ELECTROLYTES WITH LIDFOB AS AN ALTERNATIVE SALT FOR LITHIUM-ION BATTERIES

Lithium-ion batteries (LIBs) have been greatly sought after as a source of renewable energy storage. LIBs have a wide range of applications including but not limited portable electronic devices, electric vehicles, and power tools. As a direct result of their commercial viability an insatiable hunger for knowledge, advancement within the field of LIBs has been omnipresent for the last two decades. However, there are set backs evident within the LIB field; most notably the limitations of standard electrolyte formulations and LiPF6 lithium salt. The standard primary carbonate of ethylene carbonate (EC) has a very limited operating range due to its innate physical properties, and the LiPF6 salt is known to readily decompose to form HF which can further degrade LIB longevity. The goal of our research is to explore the use of a new primary salt LiDFOB in conjunction with a propylene carbonate based electrolyte to establish a more flexible electrolyte formulation by constructing coin cells and cycling them under various conditions to give a clear understanding of each formulation inherent performance capabilities. Our studies show that 1.2M LiDFOB in 3:7 PC/EMC + 1.5% VC is capable of performing comparably to the standard 1.2M LiPF6 in 3:7 EC/EMC at 25°C and the PC electrolyte also illustrates performance superior to the standard at 55°C. The degradation of lithium manganese spinel electrodes, including LiNi0.5Mn1.5O4, is an area of great concern within the field of lithium ion batteries (LIBs). Manganese containing cathode materials frequently have problems associated with Mn dissolution which significantly reduces the cycle life of LIB. Thus the stability of the cathode material is paramount to the performance of Mn spinel cathode materials in LIBs. In an effort to gain a better understanding of the stability of LiNi0.5Mn1.5O4 in common LiPF6/carbonate electrolytes, samples were stored at elevated temperature in the presence of electrolyte. Then after storage both the electrolyte solution and uncharged cathode particles were analyzed. The solid cathode particles were analyzed via scanning electron microscopy (SEM) whereas the electrolyte solution was analyzed using inductively coupled plasma mass spectroscopy (ICP-MS). The SEM analysis assists with elucidation of changes to the surfaces of the cathode particles. The ICP-MS of the electrolyte allows the determination of the extent of Mn and Ni dissolution. Samples of LiNi0.5Mn1.5O4 with different crystal surface facets were prepared to investigate the role of particle morphology in Mn and Ni dissolution. The factors affecting Mn and Ni dissolution and methods to inhibit dissolution will be discussed


INTRODUCTION Background
The search for renewable energy sources has been motivated by the need for an alternative to our inherent reliance on fossil fuels which began in the wake of the Industrial Revolution. 1 Due to major complications with lithium metal batteries 2 , lithium-ion batteries were sought after to fill the void left by their lithium metal predecessors due to their high gravimetric and volumetric energy densities. 3 Today, lithium-ion batteries are the main choice for many applications including (but not limited to) portable electronics (such as cellular phones), hybrid electric vehicles, electric vehicles, and aerospace applications. [4][5][6] The wide array of applications in conjunction with the omnipresent energy crisis facing the world furthers the demand for deeper understanding of mechanics involved in battery systems; understanding and how these systems affect all the major components of the lithium-ion battery will greatly augment the knowledge base of the within the field of renewable energy sources as a whole.
Lithium-ion batteries use lithium-ion intercalation mechanics in order to store charge within a particular battery cell. 3 In order to achieve proper intercalation mechanics very specific types of electrodes must be used. The cathode generally consists of a lithium metal oxide which functions as a source for lithium-ions. The anode is usually a graphitic type carbon which in turn can act as an acceptor for lithium-ions. During a charge cycle, lithium-ions are removed from the cathode and intercalate in-between the graphene sheets of the anode. Conversely, during discharge the lithium-ions de-intercalate and traverse back to the cathode. Thus, the theoretical capacities of every lithium-ion cell are largely dependent on the physical properties of the materials used; ultimately the cathode's material will determine the maximum capacities that can be achieved. However, other primary components of the lithiumion battery system also play a large role in the overall capacities that can be seen. 3 The separator is a crucial structure within the lithium-ion battery cell. The separator is generally composed of a porous polyolefin that allows for ion transfer to and from the electrodes. The major purpose of the separator is to isolate both the cathode and the anode from each other in order to prevent an electrical short during cycling. Therefore it is important that the separator be thin and porous enough to limit ionic internal resistance (allowing for facile lithium-ion flow) but also be durable enough to prevent any deformation that might lead to the electrodes coming in contact with each other. Additionally, the separator must allow for the electrolyte and its' components to flow freely through the separator. Therefore, the assertion that the separator is a crucial component of the battery cell construction is irrefutable.
However, as chemists, understanding the electrolyte itself is easily the most compelling aspect of the entire lithium-ion battery system.
The electrolyte is indubitably one of the most important components of the lithium-ion battery systems as its characteristics are multifaceted. The electrolyte facilitates lithium-ion flow through the entire system; without the electrolyte no ion transfer would occur. The electrolyte is usually composed ethylene carbonate and a blend linear carbonates which in turn solvate the transferable lithium-ions from one electrode to the other. 3 In conjunction with the carbonates, an appropriate lithium metal salt (traditionally lithium hexafluorophosphate) is added to the carbonate blend in order to assist with the ion conductivity of the system in its entirety. The carbonates work to create a solvent sheath around transferable lithium-ions allowing them to flow freely from one electrode to the other. 2 Therefore, optimizing the conductivity and solubility (as well as other colligative properties) of the electrolyte are crucial to the overall performance and inherent kinetics of any lithium-ion battery system.
Furthermore, the electrolyte also contributes to another inextricable component of the systems, the solid electrolyte interface.
The solid electrolyte interface (SEI) is pivotal to the overall functioning of any lithium-ion battery system. [7][8] During cycling, the electrolyte inevitably decomposes and the reduction products from the electrolyte composition are utilized in the formation of the SEI. Logically, due to the fact the SEI is composed of reduction products, the SEI forms primarily on the surface of the anode. The SEI acts as a passivation layer that allows for ion transfer through the SEI to the electrode as well as working to protect the electrolyte from continual decomposition. [7][8] The SEI acting as a passivation layer protects against further electrolyte reduction on the surface which in turn helps to extend the overall life of the electrolyte and by association the life of the battery itself. Therefore, understanding the SEI and especially what helps to create a stable SEI is an incredibly powerful force to wield within the realm of lithium-ion batteries and thus should not be ignore but rather embraced. Due to the enticing possibilities that exist in understanding the SEI, it is the major aspect of our research. In addition to the desirability of PC, the use of a different primary salt within the electrolyte matrix also has an undeniable allure. While LiPF 6 is a very conductive salt, it is not a very stable lithium salt. 2,11 Due to its' inherent thermal instability,

PC and LiDFOB Incorporation into the Electrolyte
LiDFOB is an attractive alternative to investigate as it is demonstrates resistance to thermal decomposition which could help with increasing the lifetime of the electrolyte which in turn will increase the life of any lithium-ion battery. Studies have been done showing that PC did work well with LiFOP as a primary salt thus it seemed prudent to try a PC based electrolyte with LiDFOB as the primary salt with the hope that comparable cycling performance could be obtain. The utilization of PC and LiDFOB is the primary focus of Chapter 2 of this dissertation and is discussed therein.

Transition Metal Dissolution Analysis from High Voltage Spinel Cathode Powder
High voltage spinel cathode material has been highly sought after within the lithium-ion battery community due to its larger operating potential. 3 However, high voltage materials also support additional concerning the materials themselves. When charged to high potentials, high voltage spinel materials are known to experience dissolution of major transition metals (specifically Mn) which ultimately causes greater problems within the lithium-ion battery cell itself. Researchers have implemented a variety of methods to overcome transition metal dissolution and ultimately improve the performance of high voltage lithium-ion battery cells with varying degrees of success reported. 14-16 Chapter 3 of this dissertation discusses our research with high voltage spinel cathode powder primarily motivated by investigating the effect crystal structure of the spinel has on transition metal dissolution.

Lithium-ion Capacitors
Lithium-ion capacitors share some similarity and stark contrast to their lithium-ion battery counterparts. In terms of construction, lithium-ion capacitors also incorporate a graphitic type anode used for lithium ion intercalation however this is where all similarities end. While the anode does experience initial lithiation, the lithiation process is solely to achieve a greater potential difference between the cathode and anode of the cell; effectively the lithiated graphite does not cycle lithiumion in and out of the material. 17  Electrodes used for coin cell construction were received from MTI and consisted of natural graphite anodes and LiCoO 2 cathodes. Anodes were composed of 85% natural graphite (NG), 10% conductive carbon, and 5% styrene-butadiene rubber (SBR) and carboxymethyl cellulose (CMC) pre-coated on a copper foil current collector. The cathodes were comprised of 85% LiCoO 2 , 10% conductive carbon, and 5% PVDF pre- were obtained using a.

Results and Discussion
Cycling Performance of Coin Cells at Room Temperature (25°C) As illustrated by Figure

Room Temperature Resistance Measurements (25°C)
The resistance measurements shown in Figure

Cycling Performance of Coin Cells at Elevated Temperature (55°C)
When observing the elevated temperature cycling performance in Figure 2-

Elevated Temperature Resistance Measurements (55°C)
As can be seen in Figure 2-4, the resistance measurement obtained at 55°C are also consistent with the cycling data seen previously.

Coin Cell Efficiencies (25°C & 55°C)
The efficiencies obtained for all three formulations can be seen in efficiency. This seems to indicate that all the cells performed well within each formulation and any discrepancies seen can therefore be attributed to electrochemical differences between the cells with minimal cell to cell variance.

SEM Images of Anodes after Room Temperature Cycling (25°C)
The SEM images seen in Figure  The observed result is not surprising as most of the damage incurred to the surface morphologies of electrodes is traditionally seen on the anode and not the cathode. As there was no damage seen to the anodes it is also unlikely (not impossible however) that surface particle damage would be seen on the cathodes thus, this is an anticipate result.

SEM Images of Anodes after Elevated Temperature Cycling (55°C)
The SEM images of the anodes acquired in Figure

SEM Images of Cathodes after Elevated Temperature Cycling (55°C)
The SEM Images in Figure 2-9 clearly illustrates that no significant morphological changes occurred to any of the species investigated. As previously reflected upon, it is not unusual to see no substantial changes to the cathode material under these conditions; most of the changes to the surface morphology will be observed on the anode side (if any changes occur at all). Despite the blatantly harsh cycling environment, the cathode remains unchanged.

XPS Spectra of Anodes after Room Temperature Cycling (25°C)
The XPS spectra of the anodes at room temperature give significant insight into the chemicals found on the surface of the electrodes. Figure 2-10 illustrates that there are almost certainly oxalate species from the LiDFOB present on the surface of the anode based on the C1s spectrum observed. As the C1s spectrum of the LiDFOB with PC clearly illustrates, the oxalate peak seen between 288eV and 287.5eV.
Interestingly, with the LiDFOB with EC formulation the observed peak is much larger. The discrepancy seen with the peak intensity might be correlated to electrolyte stability. The EC/LiDFOB formulation did not perform nearly as well as the PC/LiDFOB formulation as illustrated previously by the cycling performance. Due to the large amount of oxalate believed to be on the surface of the LiDFOB/PC, this may indicated that the EC based system decomposed more readily than that of the PC based electrolyte tested; this would account for the performance differences observed between the two matrices.
The F1s and O1s spectra provide minimal insight into the differences in performances observed between cells. The fluorine spectra seen of the PC and EC Aside from the EC/LiDFOB spectra observed the remaining cathode spectra leave much to be desired in terms of elucidating potential surface films present on the cathode. Figure 2-11 shows that there does appear to be a minimal surface film being formed on the surface of the cathode within the O1s spectrum of the LiDFOB/PC cathode. However, aside from noticeable diminishment of the metal oxide peak seen within the LiDFOB/PC cathode spectrum and the and increase of the peaks seen at ~534eV (most likely due to LiDFOB decomposition based on the similarity to the EC/LIDFOB spectrum seen), the PC/LiDFOB spectra vary minimally from that of the standard 1.2M LiPF 6 electrolyte As previously stated, there does appear to be a surface film formed on the cathode of the PC based electrolyte however the layer appears to be extremely thin as the metal oxide is still clearly present on the surface.

XPS Spectra of Anodes after Elevated Temperature Cycling (55°C)
The XPS spectra seen in Figure 2- The O1s spectra obtained to relay much information about the possible reduction products present on the surface of the anodes. This is mostly due to the fact that there was an abundance of oxygen containing species seen in the fresh. However, it can be inferred that any new oxygen containing species that have arisen are most likely due to the reduction of the EC and PC on the surface of the anode.

XPS Spectra of Cathodes after Elevated Temperature Cycling (55°C)
The major area of interest with the cathode spectra obtained can be found within that of the O1s spectra, specifically the spectrum of the PC based electrolyte.
As can be seen in the fresh and LiPF 6 spectra, the metal oxide peak s seen at 529eV is clearly present in both spectra. However, there is no metal oxide peak present within the spectrum of the cathode with the PC based electrolyte. This is clear evidence that when subjected to high temperatures the LIDFOB must be capable of creating oxidation products that form a thin film on the cathode. This is further supported by the fact the C1s spectrum for the anode of the PC based electrolyte also clearly shows the presence of oxalate which also suggests the LiDFOB might form an oxidative film on the surface of the cathode.

Elemental Concentrations on the Surface of the Electrodes Obtained via XPS
The elemental concentrations for the anodes seen in Table 2-1further reinforce the results seen within the XPS spectra themselves. As to be expected, the carbon elemental percentage ratio decreases significantly in all cycled test samples. This is an expected result as all formulations should form an SEI on all of the anodes tested in order show any type of appreciable cycling performance. The electrolyte decomposition products contain carbon but also contain a plethora of oxygen and fluorine thus the relative ration of carbon (due to graphite) should decrease considerably upon cycling which all of the samples illustrate.
Another interesting aspect of the carbon data obtained lies within the standard LiPF 6 electrolyte at both 25°C and 55°C. The concentration of carbon decreases when cell is exposed to elevated temperatures. This result reveals much about electrolyte performance at elevated temperatures. Electrolyte decomposition increases as temperature increases. However, the thermal decomposition seems to include the lithium salt in addition to the normal carbonate reduction. Therefore, we see an increase in the fluorine and oxygen concentrations at the surface due to the increased reduction seen at elevated temperature this lowering the relative carbon concentration.

FT-IR Analysis of Anodes at Room Temperature (25°C)
The FT-IR spectra obtained of the anodes do support some of the findings seen within the XPS spectra seen previously. As can be seen by Figure 2-14, as previously speculated, the binder used is most definitely not the standard PVDF as none the peaks observed. This is consistent with the data seen from the XPS. Within the standard 1.2M LiPF 6 electrolyte there is very little change in the spectrum within regard to the fresh aside from a rather large and broad peak at ~1500cm -1 . However, since the XPS spectra obtained indicated a relatively thin SEI, the peak seen is most likely a combination of surface reduction products (lithium alkyl carbonates) and the residual binder itself. This is mostly attributed to the fact that the overall shape of the 1.2M LiPF 6 anode spectrum does not vary much from what can be observed from the fresh anode.
The EC/LiDFOB and PC/LiDFOB spectra clearly indicate that the SEI formed is very similar in both cases based on the spectra presented in Figure 2-14. This is not unusual as the date collected from the XPS seemed to indicate that surface of both electrodes were in fact quite similar. The intensity of the peak observed that can be attributed to the binder is greatly diminished indicating there is a thicker SEI formed for both LiDFOB derivatives. Two almost identical peaks arise at ~1650cm -1 and 1100cm -1 as well as two smaller peaks at 1450cm -1 and 1350cm -1 . Clearly these peaks seem to represent the decomposition of the LiDFOB on the surface of the electrodes.
This can be inferred due to the overwhelming similarity of both the PC and EC based with LiDFOB electrolytes in contrast to the large disparity seen within regard to the standard LiPF 6 based electrolyte. Also, the EC/LiDFOB and EC/LiPF6 electrolytes clearly do not form a similar SEI based on the spectra obtained here as well as the spectra obtained from the XPS. These results suggest that the primary component responsible for SEI formation is not lithium ethylene dicarbonate (LEDC) within these systems which contrasts directly to that of LiPF 6 based electrolytes.

FT-IR Analysis of Anodes at Elevated Temperature (55°C)
The results found in Figure 2-15 are somewhat unexpected. Based on the spectra obtained, the electrolyte decomposition at elevated temperatures seems to align more closely than at room temperature. The spectrum of the PC based electrolyte looks very similar to that of the standard. The only errant peak observed within the PC FT-IR spectrum can be seen at approximately 1650cm -1 . This peak is most likely due to the oxalate which in turn reinforces that there is almost definitely oxalate formed on the surface of the anode. The XPS C1s spectrum obtained for the PC based electrolyte was somewhat inconclusive as there was no noticeable oxalate peak seen which was very anomalous. However, the other spectra seemed to indicate there was in fact oxalate present and the FT-IR spectrum of the PC/LiDFOB electrolyte even also indicates there is oxalate present on the surface of the anode. This result falls in line with what was to be expected and clarifies previous discrepancies seen within the results PC anode results.

Conclusions
Based on the results acquired, 1.2M LiDFOB in 3:7 PC/EMC + 1.5% seems like an acceptable substitute for the standard 1.2M LiPF 6 electrolyte. The cells constructed formed indistinguishably from the standard LiPF 6 electrolyte and retained greater stability after cycling at elevated temperature. This result is somewhat miraculous considering PC/EMC without VC will not cycle in the slightest. The addition of a minimal amount of VC seems to indicate that it must be conducive to stable SEI formation within PC based electrolytes. The problems encountered with LiDFOB and EC may in fact be very similar to the problems originally encountered with PC when not using VC. Thus, the next logical step would be to compare the performance of the EC electrolyte with VC to see if it in fact surpasses the PC based electrolyte in performance.
The surface analysis results seem to indicate that oxalate is a prominent player in the SEI formation as well as boron. The major complications facing the EC based electrolyte seem to be related to electrolyte stability with EC thus the VC addition in the PC based electrolyte almost certainly assists in SEI stability and allows good electrolyte passivation to occur.      facets were prepared to investigate the role of particle morphology in Mn and Ni dissolution. The factors affecting Mn and Ni dissolution and methods to inhibit dissolution will be discussed.

Introduction
High voltage spinel type cathodes have been a greatly sought resource for its inherent higher potential window and thus greater achievable capacities than the standard lithium metal oxide cathodes traditionally used. 1 While high voltage cathodes may superficially look like the obvious choice for an electrode, further investigation reveals that high voltage materials are not without their own limitations.
Transition metals from the crystal structure of the spinel have been known to dissolute from the crystal structure 2 and potentially cause damage to the anode side of the cathode decreased cell performance or failure. 3 In order to overcome these shortcomings, researchers have investigated ways to limit transition metal dissolution from the bulk material [4][5] , specifically in regard to the altering the crystal structure spinel material. 6 High voltage spinel is host to a plethora of LiMnO 4 type derivatives that make up the material. 1 Thus, due to the variability available to constructing spinellike structures implies and altered crystal structure can be implemented. By changing the crystal structure researchers hope to help end the tyranny of transition metal dissolution and preserve a more stable lithium-ion battery system that can sustain higher potentials.
Our research interests lie with the investigation of the octahedral and plate crystal structures of LiNi 0.5 Mn 1.5 O 4 high voltage spinel. In order to determine whether or not crystal structure and/or ordering within that crystal structure has an overall effect on the amount of transition metal dissolution observed, ICP-MS will be used.
In addition the ICP-MS, SEM will be used in an attempt to further elucidate changes within the morphology that may help to confirm the superiority of one crystal structure over the other.

Experimental
High voltage spinel powder samples were obtained from LBL and brought into an Ar filled glove box for storage. 1.2M LiPF 6 in 3:7 EC/EMC electrolyte was received from BASF, stored under argon, and used as received. 30mg of spinel powder was put into an ampoule and then had 2ml of 1.2M LiPF 6 in3:7 EC/EMC and subsequently the ampoules were sealed with septa. Ampoules were removed from the argon glove box and taken to a schlenk line for flame sealing. The inside of the ampoules were purged with nitrogen gas and, while on the schlenk line, partial vacuum was pulled within the ampoule. A torch was then used to melt the glass neck of the ampoules in order to ensure no air or water contamination could occur. The ampoules were then place in an oil bath of 85°C for 6 days for thermal storage. Upon thermal storage completion, the ampoules were taken into the glove box, cracked, and 1mL of electrolyte was extracted and placed into a recently dried vial for further analysis. The electrolyte was removed from the glove box and heated on a hot plate within a hood in order to boil off the remaining organic solutions. Once dry, 2mL of a 3% nitric acid solution were added to the remaining polymer within each vial. Once completely dissolved, 1mL of the remaining solution was obtained from each vial and diluted to 10mL. These 10mL samples were then taken to the ICP-MS for analysis.
In addition to the electrolyte the powder within the ampoules was also taken, rinsed 3 times with DMC and dried overnight within the antechamber under vacuum in preparation for SEM analysis.

ICP-MS Analysis of Electrolyte
The powder samples with the ordered plate crystals 112 facets and the samples with the octahedral crystals with 111 facets were taken to ICP-MS for analysis. The results of the ICP-MS analysis can be seen In Table 3-1. Based on the percent of transition metal dissolution the less ordered octahedral crystal structure experienced the least amount of transition metal dissolution. Interestingly, in both cases the less ordered crystal structure for both the plate and octahedral crystals experienced less transition metal dissolution which seems to indicate that ales ordered spinel crystal structure is more appreciable to being used when cycling at high voltages due solely to the face that it is more resistance to transition metal dissolution. any damage were to be one would need a more powerful SEM in order to increase the magnification sufficiently to see small details.

Conclusions
Based on the data found herein there is not much to reinforce the assertions made by the ICP-MS results. While the results do suggest that a less ordered structure is conducive to minimizing transition metal dissolution from the cathode. However, since these results are not reflected in the SEM data, they should be regarded with the appropriate level of skepticism. .