The Nature of the Thermal and Electrochemical Degradants in Lithium Battery Electrolytes.

[NNadir]

The two papers from the primary scientific literature that I'll discuss in this post are these:

Reaction Product Analyses of the Most Active “Inactive” Material in Lithium-Ion Batteries—The Electrolyte. I: Themal Stress and Marker Molecules (Sascha Nowak et al, Chem. Mater. 2019, 31, 24, 9970?9976)

...and...

Reaction Product Analysis of the Most Active “Inactive” Material in Lithium-Ion Batteries—The Electrolyte. II: Battery Operation and Additive Impact (Sascha Nowak et al, Chem. Mater. 2019, 31, 24, 9977?9983)

Chemistry of Materials is a publication of the American Chemical Society, the world's largest professional scientific organization.

The cartoon graphics for introducing the papers are these:



...and...



There is a widespread belief, particularly in the Western World, that somehow, batteries, with which we are all familiar, and which we all use, will save the world. A belief in a concept, of course, is very different than facts related to a concept. Batteries will not save the world. A battery is a device that wastes primary energy. The fastest growing source of primary energy on this planet has been, in this century as in the last, dangerous fossil fuels. The world was utilizing, as of 2018, the last year for which we have complete compiled data, 599.34 exajoules of primary energy. This compares to 420.19 exajoules of primary energy utilized in the year 2000. (So much for the widespread belief that energy conservation would save the world.) Of the increase of 179.15 exajoules, 148.34 exajoules has been provided by increases in the use of the three dangerous fossil fuels, lead by dangerous coal, (63.22 exajoules) and dangerous natural gas, (50.33 exajoules).

The belief that batteries would save the world, is connected with the associated belief that so called "renewable energy" would save the world, a belief that has not been borne out by experiment: Despite an investment of over 2.1 trillion dollars in the last ten years alone, so called renewable energy has not saved the world; it isn't saving the world; and - I keep trying to burst this toxic fantasy with only limited success - it won't save the world. In this century, all the primary energy produced by wind, solar, geothermal and tidal energy on this planet grew by 9.76 exajoules. In the "percent talk" so favored by advocates of vast expenditures (and destruction) associated with so called "renewable energy," all the world's solar, wind, geothermal, and tidal output grew by 15.4% as fast as the most lethal and dangerous of all the dangerous fossil fuels, coal.

That's a fact. Facts matter.

Here's another fact: On this planet, when one charges a battery, whether its for your computer, your phone, or for Elon Musk's silly electric car for millionaires and billionaires, there is an overwhelming probability that one is storing, and wasting, almost in its entirety, dangerous fossil fuel generated primary energy.

Here's another fact: The existence of the most efficient batteries known, lithium batteries, for the discovery of which Nobel Prizes were awarded this year, is very much dependent on the use of energy generated by dangerous fossil fuels and materials dependent, for their manufacture, by dangerous fossil fuels. Further, there is not, on this planet, enough lithium, nor for that matter and considerably more important and dire, not enough cobalt on this planet to store more than 600 exajoules of primary energy.

Before going into more detail about the two papers cited at the outset, let's look at the graphic of figure 1 of the first of the two papers, which looks at the common formulation of electrolytes in most of the world's lithium batteries.



The caption:

Figure 1. Common electrolyte formulations used in state-of-the art LIB cells. In this study, ternary mixtures of varying linear carbonates (DEC, DMC, EMC) in combination with EC and LiPF6 were investigated by means of LC–MS.

DEC here is diethyl carbonate. DMC is dimethyl carbonate. EMC is ethyl methyl carbonate. EC is ethylene carbonate. Historically, and in many cases, still these organic carbonates are obtained by the use of phosgene, a war gas, made from dangerous natural gas and chlorine gas, that killed tens of thousands of soldiers in the First World War and organic alcohols. Most of the world's methanol is produced from dangerous natural gas. Ethanol for DMC is obtained from corn, a "renewable fuel" that has been responsible for the destruction of the Mississippi Delta's ecosystem. Ethylene glycol, the starting material for EC, the cyclic carbonate in the paper is made via the thermal cracking, at temperatures of around 700°C , of dangerous petroleum to give ethylene (aka "ethene&quot , and epoxidation of this highly flammable gas with oxygen under controlled conditions. The most common use of ethylene glycol is for use as antifreeze in dangerous automobiles, combusting largely dangerous petroleum products.

Other abbreviations that come up in these papers.

LC-MS often written as LC/MS/MS and written in this case as LC-IT-MS or (as the case actually involved here, LC-HRMS or LC-IT-TOF-MS) refers to liquid chromatography (LC) coupled to various permutations of mass spectrometry, ion trap (IT), high resolution (HR) and Time of Flight (TOF).

The authors here, all of whom work in Germany, have studied what happens to these electrolyes in use, storage, and upon exposure to heat. Lithium batteries, as most of us know, do generate heat on charge and discharge, this heat representing the wasted primary energy that their use involves.

From the introduction of the first paper:

Today’s applications of lithium-ion batteries (LIBs) ranging from small consumer electronics (e.g., smartphones, watches) to large battery systems in electric vehicles (xEVs) are accompanied by new challenges regarding heat generation at the cell level.(1) On the one hand, various novel, for example, inductive charging, techniques are rising in popularity; on the other hand, fast-charging capabilities for xEVs are the considered key for consumer acceptance and will be affected by overpotentials.(2) Both requirements, combined with ambient temperature fluctuations, and in particular direct insolation, will result in an increased thermal stress for LIBs.(3) As a consequence, the most susceptible cell component in terms of thermal stability, the LIB electrolyte, is prone to decomposition at elevated temperatures.(4?8)

Manifold aging phenomena have been described for LIB cell components in recent years.(9,10) In detail, the thermal decomposition represents one possible electrolyte degradation pathway. Part 1 of this study focuses solely on the thermal influence on the electrolyte decomposition; Part 2 involves electrochemical influences such as interphase formation reactions and varying the applied voltage at room temperature. The most commonly applied electrolyte formulations consist of lithium hexafluorophosphate (LiPF6) as conducting salt dissolved in a mixture of cyclic ethylene carbonate (EC) and at least one linear carbonate (e.g., dimethyl carbonate, DMC; diethyl carbonate, DEC; ethyl methyl carbonate, EMC) (Figure 1).(5,11) Numerous analytical techniques were applied for the identification and quantification of electrolyte degradation products.(12,13) Particularly, the LiPF6 decomposition route was studied intensively, leading to the identification of potentially toxic organo(fluoro)phosphates (O(F)Ps) in a high structural variety.(14?22) Electrolyte degradation products originating from carbonate solvents have been reported in literature(23,24) as well as their impairing effect on LIB performance.(25) According to Lee et al. and Ariga et al., a ring-opening reaction of EC (via nucleophilic attack or cationic activation) results in polymerization with occasional decarboxylation.(26,27) Consequently, reliable structural elucidation of the decomposition product variety enables the understanding of reaction pathways and represents the initial step for the development of prevention strategies. Eventually, deciphering the pathway of decomposition products might enable the definition of marker molecules in the electrolyte to identify prior thermal strain for the LIB, enhancing the capabilities of post-mortem analysis and the validity of analytical investigations in terms.

Since I regard mass spectrometry as the most important of all analytical chemistry methods, please indulge me as I describe the experimental procedure:

For LC–MS investigations, a Nexera X2 UHPLC system (Shimadzu, Kyoto, Japan) hyphenated to a LCMS-IT-TOF (Shimadzu) was used. Reversed-phase chromatography was conducted on a ZORBAX SB-C18 column (100 × 2.1 mm, 1.8 ?m; Agilent Technologies, Inc., Santa Clara, CA, USA) at 40 °C and a flow rate of 0.5 mL min–1. The mobile phase consisted of water (A) and acetonitrile (B). The gradient started with 5% B from 0 to 1.8 min and increased to 60% B within 12.2 min. Subsequently, the mobile phase was kept constant at 60% B for 2 min. Finally, the column was equilibrated at 5% B for 4 min. The injection volume was 5 ?L. To protect the IT-TOF mass spectrometer from high concentrations of conducting salt LiPF6 and thermally induced acidic decomposition products, the flow line switched to the MS after 1.8 min. Ionization was performed in the ESI(+) mode at 4.5 kV. The curved desolvation line and heat block temperature were 230 °C. The drying gas pressure was set to 100 kPa, and the nebulizer gas flow was 1.5 L min–1. The ion trap was operated in an automatic MS2 mode with an ion accumulation time of 10 ms in MS1 and 40 ms in MS2, leading to a loop time of 260 ms. The mass range was set to a mass-to-charge ratio (m/z) of 100–400 in MS1 and 50–400 in MS2 for the low-mass window, an m/z of 200–1000 in MS1 and 150–1000 in MS2 for the high-mass window.

Trust me, it's cool.

Figures 2 and 3 show the mass chromatograms of the products of the forced degradation studies, conducted at at 80 °C (accelerated conditions) on the electrolytes:




The caption:

Figure 2. Identification of decomposition product signals via the formed adduct pattern of species (H+, NH4+, and Na+) with RPLC-HRMS in ELDMC. The excerpt shows the average mass spectrum between 8.0 and 11.0 min. Coeluting compounds resulting in chromatographic interferences with the same nominal mass (*) and spectral interferences (**) are indicated.

The caption:

Figure 3. Chromatographic separation (c) of five compounds with the same nominal mass in ELDMC electrolyte. MS2 experiments (a, b, d, e, f) show differences in fragments formed and intensity pattern. An m/z shift of 0.036 was observed between chromatographic peaks (c) and within almost all fragment signals exemplarily shown in (d).

In mass spectrometry, mass transitions for different compounds can result in similar or identical "first pass" signals, a factor that is subject to clarification by the LC component's orthogonality (differences in the retention time in the chromatographic column). An ion trap, allows for further orthogonality, resulting in a deeper understanding of the precise structures, further elucidated by the high resolution achievable by the use of time of flight techniques.

Table 1 in the paper reflects differentiation of these similar (isobaric) fragments to give the structures of different molecules using these powerful techniques:



The next figure shows the thermal degradants putative structures:



The caption:

Figure 4. MS2 fragmentation of the precursor m/z 416.2126 (Figure 3d). Obtained information was used for fragment and precursor structure predictions. Cleavage positions are indicated in the precursor structure suggestion, and incongruous fragments are highlighted in red.

The final graphic suggests a nomenclature for the decomposition products:



The caption:

Figure 5. Copolymer scheme based on carboxylate and ethylene oxide linkages. Different terminal groups formed depending on the thermally degraded electrolyte formulation are shown (R1, R2). The copolymer scheme for m/z 416.2126 is shown exemplarily.(sic)

From the conclusion to the paper:

LIB electrolyte decomposition products formed during exposure to elevated temperatures were studied in this Part 1 of a two-part study. The identification of oligomeric compounds led to the generation of a target list of 206 unique species with an upper molecule mass border of 602 u. Optimal confidence for structure elucidation was achieved via consideration of HRMS, MS2, chromatographic correlation, as well as the interrelation of findings. In electrolytes containing either DMC, DEC, or EMC, 140 different species were identified and structures suggested. In this regard, a classification of solvent-based degradation products was developed, overcoming the structural uncertainties of possible isomers. Terminal hydroxy groups were scarcely detected, which is good agreement with the postulated correlation to water contamination. Furthermore, a statistical distribution of the decomposition product formation process was postulated. Overall, thermal aging revealed a conducting salt degradation to O(F)Ps, distinctly separated from solvent decompositions to oligomeric carbonates, ethylene glycols, and co-oligomers. Consequently, no intermixtures of O(F)Ps and oligomeric compounds, or in other words no mixed compounds of conducting salt and electrolyte solvent entities, were detected in these thermal aging experiments.

Part II gives rise to some products that are of a little more concern to me, since there are clear toxicological implications. It involves the degradation related to charge and discharge, which is, of course, what batteries are all about.

From the introduction:

Since their market introduction in the early 1990s, the requirements for lithium-ion batteries (LIBs) extended with its fields of application.(1,2) The competition for enhanced energy densities(3?5) yields, high-voltage materials,(6?8) large-capacity negative-electrode materials,(9?12) and high-energy/capacity battery cells drift further apart. In contrast, the electrolyte remained almost unchanged for three decades.(13)

Lithium hexafluorophosphate (LiPF6) dissolved in linear and cyclic organic carbonates still represents the state-of-the-art electrolyte. The limited redox stability of the electrolyte toward electrode potentials results in parasitic (aging) and beneficial (interphase formation) decomposition reactions.(14,15) With regard to electrode materials, carbonaceous negative electrodes in combination with a lithium metal oxide (LiMO2; M: Mn, Co, Ni, Al) as positive electrode, spatially separated by polyolefin layers, are commonly used in commercial LIB cells. One crucial step for LIB cell performance is the generation of a solid electrolyte interphase (SEI) on the negative electrode surface during the first cycles.(16?21) On the positive electrode surface, the cathode electrolyte interphase (CEI) is formed. The interphase composition and variability during charge/discharge are still under discussion in literature.(16,22,23) The reductive degradation at the negative electrode leads to organic and inorganic parts within the SEI.(24) The SEI is a first-order electronic insulator while being lithium-ionconductor. This represents the starting point of the film-forming electrolyte additives, which decompose reductively at higher potentials than ethylene carbonate (EC); thus, the products of the additive (e.g., vinylene carbonate; VC) integrate into the SEI. Consequently, the properties of the SEI can be modified by electrolyte additives and its passivation capabilities toward electrolyte components.(25) The structure elucidation of decomposition products soluble in the electrolyte after parasitic reactions are the focus of this Part 2 of the study.

Approaches for the characterization and structure elucidation of the electrolyte decomposition in LIB cells started in the late 1990s(26) and culminated in structure predictions in combination with reaction pathway suggestions mainly by the working group of Laruelle and others.(27?34) Gachot et al. described the formation of compounds incorporating carbonates and phosphates,(32) which was extended by more complex structures and oligo phosphates in 2015.(35) The picture of electrolyte decomposition products with molecular weights exceeding those of the solvents became increasingly clear via the application of liquid chromatography-mass spectrometry (LC-MS) with high-resolution mass spectrometry (HRMS), fragmentation (MS2) information and hydrophobic retention on reversed-phase (RP) chromatography.(35?38) To generate optimum structural confidence with LC-MS, these features were considered in highly complex LIB electrolyte samples. Nonetheless, quadrupole-based low-resolution studies can lead to less dependable results and misleading structure suggestions.(39)

The authors constructed some typical types of cells (of a type making components available for examination using analytical chemistry, again LC/MS. They ran the cells through 1000 charge and discharge cycles.

Some graphics:




The caption:

Figure 1. Overview of products identified after thermal (gray) and electrochemical decomposition. Compounds with a maximum phosphate (P) content of 3 and carbonate (C) count of 3 were identified and are shown in the Supporting Information

The caption:

Figure 2. RPLC-IT-TOF-MS chromatogram of diphosphates (top) with varied alkylation in ELEMC+VC after >1000 cycles. Isobaric interferences are indicated in gray. Fragmentation spectra of the proton (m/z 307.0706) and lithium (m/z 313.0756) adducts of double methylated and ethylated diphosphate are allocated to specific fragments. The fragmented peak is highlighted with (I).

The caption:

Figure 3. RPLC-IT-TOF-MS chromatogram of a phosphate–carbonate compound (top) with different alkylations in ELEMC+VC after >1000 cycles. Chromatographic peak splitting of double-methylated or -ethylated compounds can be deciphered via MS2 experiments of the [M + H]+ adduct (bottom), forming carbonate fragments. The specific fragmented peaks are highlighted with (I) and (II).

The caption:

Figure 4. RPLC-IT-TOF-MS chromatogram of isomers of phosphate with three carbonate groups (top) in ELEMC+VC after >1000 cycles. Three different chromatographic peaks were obtained and allocated to the shown configuration via MS2 experiments, identifying characteristic fragments (bottom, green). The fragmented peaks are highlighted with (I), (II), and (III).

The caption:

Figure 5. RPLC-IT-TOF-MS chromatogram of three cyclic ether carbonate oligomers in ELEMC after >1000 cycles (top). MS2 experiments were applied for structure confirmation (bottom), including structure suggestions of formed fragments. The fragmentated peak is highlighted with (I).

This table from the paper compares the results of the number and types of compounds found in each of the two types of experiments from Part 1 and Part 2 of the series.



Organophosphate esters are widely utilized compounds. They have been utilized as flame retardants, replacing the halogenated aromatic ethers that have resulted in a huge environmental contamination problem, particularly for people recycling electronic waste, most of whom are in the third world where we do not have to pay attention to the consequences on their lives. It is not entirely clear that the replacement organophosphates are totally benign. Organophosphates are also utilized as chemical warfare agents, in particular, sarin, which probably has a precursor similar the anion to in the lithium batteries, the perfluorophosphium ion. (There is no evidence of sarin or closely related compounds in this paper.) Sarin is an acetylcholine esterase inhibitor, and thus a nerve agent. Similar nerve agents are widely used as insecticides. The closest isosteric compound to an insecticide found in this paper is the compound having a mass of 167.0472 amu in figure 2. It is nearly isosteric with dichlorovos, a halogenated analogue, which was an insecticide banned in Europe in 1998, also a neurotoxic compound.

I have written in the past in this space on the subject of recycling of lithium batteries, and some of the issues connected with doing so.

Trust me, this kind of recycling may not fall under the rubric of "green." Things that sound good in the abstract have real consequences. The most destructive distributed energy device ever invented, the automobile, was originally promoted as an alternative to the very serious historical problem of horse manure accumulations in cities, a health, environmental and aesthetic problem of consequence. If we take our hippie rose colored glasses off, arguably the cure was far worse than the problem.

I wish you a successful, happy and healthy New Year.