The gelation of the slurry during electrode fabrication is a known issue, especially for the upcoming nickel-rich cathode materials such as LiNi0.8Mn0.1Co0.1O2 (NMC811).[1–3] The gelation originates from the reaction between hydroxides, which form when the NMC comes in contact with moisture, and the polyvinylidene difluoride (PVdF) binder.[4,5] To avoid potential negative effects on the processability and performance, nickel-rich NMC cathodes are normally prepared under dry room conditions. However, these working conditions are a significant cost factor which lie roughly 130 % above the production costs of previous NMC generations like NMC111 or NMC532.[6] To potentially avoid the need for such costly processing environment, an in-depth understanding of the reactions[4] and their eventual impact on the electrode performance is essential. On one side, the contact to moisture influences the processability since the slurry shows different viscosity at different aging stages. On the other side, the physical properties of the resulting electrodes change. For longer exposure to a moist atmosphere, the adhesion to the current collector collapses and the distribution of the active material, the PVdF binder and the conductive carbon becomes more and more inhomogeneous. As a final result, also the electrochemical performance of the electrodes is affected and we can observe an increased voltage hysteresis, lower initial capacity and stronger capacity fading upon long-term cycling.
In more detail, we herein report the first study of the correlation between the relative humidity and the gelation time. We made use of a two chamber setup to allow sample transfer without significant distortion of the present relative humidity (RH). The respective RH value was set employing saturated solutions of different salts.[7] Even though, a visible change of the slurry’s stored in moist atmosphere occurs, as can be seen in the pictures, for a more accurate characterization of the gradual degradation, viscosity measurements were performed. An example for the typical development of the viscosity over time is shown in panel 3 for 75 % RH over a total storage time of 8 hours. In a first step, the viscosity decreases by about 40 % during the first hour. This value stays constant under dry room conditions, but also at such high humidity values no change occurs for roughly 2-3 hours. Here, the first signs of PVdF decomposition and the connected gelation process occur after 4 hours (change of slope), followed by a strong increase after 6 and 8 hours. Based on that, the gelation point for 75 % RH should be identified as 6h of storage, however, the physical properties of the resulting electrode show a drastic change already after 4 hours. Firstly, the adhesion between the coating layer and the current collector breaks down to about 25 % of its original value. Further, the SEM/EDX micrographs show clear agglomeration of the PVdF decomposition products and the conductive carbon. As a consequence, the very first change in slope during the viscosity measurement should be considered as turning point in the slurry’s quality. This correlates with a visible change in the appearance, i.e. the slurry becoming grainy. In order to put these findings into context with the electrochemical performance of such altered NMC811 electrode slurries, Li metal and Li-ion cells versus a graphite anode were studied. Just as for the physical properties, the slurry with 2h exposure time shows rather identical electrochemical performance in both cell configurations. Interestingly, the 2h exposed electrodes even had lower voltage hysteresis in Li-ion cells, but anyhow, both (the pristine and the 2h) have a stable internal resistance.[8] Starting from 4h and significantly more pronounced for the 6h exposure time, a decay in the specific discharge capacity and the capacity retention is observed. In the Li-metal cell, the same capacity is reached during slow dis-/charge cycles but a drop for the 1C capacity is present. Looking at the ΔV this can be assigned to a higher overpotential, originating from the agglomerates visible in the SEM. In difference, the initial charge and discharge capacity for the 6h exposed sample is lower in graphite‖NMC cells, revealing that there is not just stronger polarization but lithium consuming side reactions as well. This fact becomes particularly visible if the capacity retention upon subsequent cycling at 1C is compared to the capacity retention of the C/3 control cycles. Here the 4h exposed sample is capable of restoring 88 % of its initial capacity, comparable to the 89 % (2h) and 91 % (fresh) of the less aged material. Hence, the 1C fading is mostly based on cell polarization. For the 6h exposed sample, however, the significantly stronger cell polarization during 1C cycling is accompanied by accelerated side reactions as well, leaving only 71 % of its initial capacity at C/3.
As a summary, the influence of the contact to moisture on the overall quality of the cathode coating and further on the electrochemical performance in lithium-metal and lithium-ion battery cells was characterized. Thereby, different threshold times could be identified which mark important steps during the aging and degradation of the slurry. While short exposure times (e.g. 2h at 75 % RH) have virtually no influence on processability, homogeneity and the electrochemistry – different behavior is observed for longer exposure to moisture. The next aging stage (e.g. 4h at 75 % RH) already shows visible grains in the slurry and a breakdown of the adhesion along with agglomerates in the electrode coating can be observed. Such electrodes lead to stronger cell polarization, however, we can still access the full lithium inventory and no sever side reactions occur during long-term cycling in the Li-ion cell. In the final stage, the gelation arises in form of high viscosity and consequently poor processability. The resulting electrodes reveal even denser agglomerates and higher polarization. More importantly though, this kind of aging leads to side reactions during electrochemical cycling and thereby loss of lithium inventory. Those main four stages identified during the extreme case of 75 % RH, i.e. the fresh slurry, 2h, 4h and 6h exposure, are present for lower humidity values as well and their timing can be extrapolated. As we found, the gelation time is increasing significantly for lower RH values, e.g. up to over the 8-fold to 50 hours for 20 % RH. Accordingly, this study may serve as a model case for the electrode preparation process, including the important insight that the viscosity of the slurry can be used as a general quality indicator. For process monitoring on larger scale a complementary approach combining viscosity measurements with optical methods could be applied, since the different stages are accompanied by visible changes of the slurry’s appearance. Importantly, this study shows that short exposure times do not affect the electrode quality and electrochemical performance compared to electrodes processed under dry room conditions, thus, enabling a significant potential cost reduction as standard air conditioning can reach RH values around 20-30 % with ease.

References
[1] J. Zheng, W. H. Kan, A. Manthiram, ACS Appl. Mater. Interfaces 2015, 7, 6926–6934.
[2] J. Xu, F. Lin, M. M. Doeff, W. Tong, J. Mater. Chem. A 2017, 5, 874–901.
[3] F. Y. Tsai, J. H. Jhang, H. W. Hsieh, C. C. Li, J. Power Sources 2016, 310, 47–53.
[4] A. R. Schuer, M. Kuenzel, S. Yang, M. Kosfeld, F. Mueller, S. Passerini, D. Bresser, J. Power Sources 2022, 525, 231111.
[5] G. J. Ross, J. F. Watts, M. P. Hill, P. Morrissey, Polymer (Guildf). 2000, 41, 1685–1696.
[6] M. Greenwood, M. Wentker, J. Leker, J. Power Sources Adv. 2021, 9, 100055.
[7] D. S. Carr, B. L. Harris, Ind. Eng. Chem. 1949, 8–9.
[8] J. E. Harlow, X. Ma, J. Li, E. Logan, Y. Liu, N. Zhang, L. Ma, S. L. Glazier, M. M. E. Cormier, M. Genovese, S. Buteau, A. Cameron, J. E. Stark, J. R. Dahn, J. Electrochem. Soc. 2019, 166, A3031–A3044.