Factors That Influence Formation of AlF 3 Passive Film on Aluminum in Li-Ion Battery Electrolytes with LiPF 6 Xueyuan Zhang a and T. M. Devine a,b, * a Lawrence Berkeley National Laboratory, Energy, Environment and Technology Division, Berkeley, California 94720, USA b Department of Materials Science and Engineering, University of California, Berkeley, Berkeley, California 94720, USA The effects of ipotential, iithe electrolyte’s concentrations of LiPF 6 and H 2 O, iiithe aluminum’s air-formed oxide, ivan anodized layer of Al 2 O 3 , and vthe purity of aluminum, on the formation of a protective layer of AlF 3 during anodic polarization of aluminum were investigated. Most tests were conducted in lithium-ion battery electrolytes with 1.0 M LiPF 6 . At potentials above a critical value, a film of AlF 3 forms on top of the air-formed oxide, creating a duplex, or two-layered film. The thickness of the AlF 3 increases with applied potential. The critical value of potential for formation of AlF 3 is independent of the composition of the aluminum alloys investigated, and increases with the thickness of pre-existing surface films, such as the native air-formed oxide. There is a threshold value of concentration of LiPF 6 =0.05 M LiPF 6 for the formation of a protective layer of AlF 3 . Protective films of AlF 3 are formed in 1 M LiPF 6 electrolytes with concentrations of H 2 O in the range of 2 to 4000 ppm. LiTFSI causes severe pitting corrosion of aluminum that is covered by only its air-formed oxide. A thin film 2 nmof AlF 3 significantly improves the resistance of aluminum to pitting corrosion in 1:1 ethylene carbonate + dimethyl carbonate with 1 M LiTFSI. © 2006 The Electrochemical Society. DOI: 10.1149/1.2218816All rights reserved. Manuscript submitted July 5, 2005; revised manuscript received May 9, 2006. Available electronically July 19, 2006. This is the second of a pair of papers that investigates the pro- tective film that forms on aluminum in lithium-ion battery electro- lytes with LiPF 6 . In the preceding paper, the protective film, which forms on top of aluminum’s air-formed oxide, was identified as AlF 3 . 1 The present paper is focused on determining some of the factors that affect the formation of AlF 3 . The outer layer of AlF 3 is responsible for protecting aluminum current collectors against corrosion at high potentials in battery elec- trolytes. The air-formed oxide can protect aluminum against corro- sion at potentials below approximately 4.0 V. 1-4 Above 4.0 V, a layer of AlF 3 forms on aluminum. 1-9 Aluminum exposed to the elec- trolyte at pores in the cathode can experience potentials greater than 4.0 V during charging of the battery. On occasion, aluminum current collectors are severely corroded at pores in the cathode. The corro- sion is not simply the result of the potential exceeding 4.0 V, be- cause in the absence of the cathode, aluminum can be polarized in the laboratory to potentials close to 30 V without experiencing lo- calized corrosion. 10 The current collectors corrode by a type of crev- ice corrosion called underdeposit corrosion. 10 The occurrence of underdeposit corrosion is the result of either breakdown of the film of AlF 3 or the failure to form AlF 3 on alu- minum exposed to the pore’s electrolyte. The current paper investi- gates the influence of a number of factors on the formation of AlF 3 . The variables considered are applied potential, time at constant po- tential, concentration of LiPF 6 , concentration of H 2 O, the presence of an air-formed oxide, the presence of a thick, anodized oxide, and the composition purityof aluminum. Experimental Potentiodynamic, potentiostatic, and cyclic polarization tests, along with electrochemical impedance spectroscopy EISand elec- trochemical quartz crystal microbalance EQCMmeasurements were used to investigate the formation of passive films of AlF 3 on aluminum samples. Electrolytes.— Electrochemical tests were conducted in two commercially prepared, lithium-ion battery electrolytes, Gen 2 pur- chased from Merck and purified at Quallionand Gen 3. The nomi- nal composition of Gen 2 is 1:1 ethylene carbonate EC + ethylmethyl carbonate EMCwith 1.2 M LiPF 6 , and the nominal composition of Gen 3 is 1:1:1 propylene carbonate PC+ EC + dimethyl carbonate DMCwith 1.2 M LiPF 6 . A number of electrochemical tests were also conducted in elec- trolytes prepared in the laboratory. The laboratory-prepared electro- lytes were mixed from their components in a dry and deoxygenated glove box. Three different types of laboratory electrolytes were em- ployed. The first type of laboratory electrolyte consisted of 1:1 EC + DMC with concentrations of LiPF 6 that ranged from 0.001 to 1.0 M. The second type of laboratory electrolyte was com- posed of 1:1 EC + DMC with 0.38 M LiPF 6 and two different water concentrations: 2 ppm no deliberate addition of waterand 2469 ppm. The third type of laboratory electrolyte consisted of 0.35 M LiPF 6 in PC with three different water concentrations: 2, 874, and 4000 ppm. Water concentrations were measured by Karl Fischer titration. Aluminum.— Electrochemical tests were conducted on two alu- minum alloys A1100 rod, 0.5 cm diameterand A2024 sheet, 25.4 cm 25.4 cm 2 mmand three aluminum foils. The chemical compositions of the three foils were i98.5% Al iron and silicon were the main alloying elements; ii99.8% Al; and iii 99.999% Al. In addition, sputter deposited films of aluminum were used for EQCM. The effect of potential on the formation of a passive film of AlF 3 was investigated by cyclic voltammetry described in the next sub- sectionof 25 m thick Reynold’s aluminum foil 98.5% aluminum, with iron and silicon the main alloying elementsimmersed in the battery electrolyte Gen 2. The tested area of samples of aluminum foil was 0.5 cm 2 and was circular in shape. The effect of time at potential was examined by potentiostatic measurements of oxidation current and mass change of 1 m thick film of aluminum deposited ontoAT-cut, 5 MHz quartz crystal. The test was conducted in Gen 2 at a potential of 4.5 V vs Li/Li + , using an electrochemical quartz crystal microbalance model no. RQCM purchased from Maxtek, Inc. Aluminum alloys 2024 and 1100 were used to investigate the influence of the surface condition of aluminum on the formation of a passive film of AlF 3 in Gen 3. Samples of Alloy 2024 were cut with a test area of 0.5 cm 2 . Cylindrical samples of alloy 1100 were cut from a 0.5 cm diameter rod. Tests that examined the influence of aluminum’s surface condi- tion consisted of potentiodynamic anodic polarization and EIS in Gen 3. Six different surface conditions were studied: 1surface covered with air-formed oxide; 2bare surface created by surface grinding of a rotating disk electrode of 1100 Al 0.5 cm diameter in 1.2 cm diameter Teflon holderwith 240 grit silicon carbide paper as the aluminum is rotated at 2000 rpm and cathodically polarized at * Electrochemical Society Active Member. Journal of The Electrochemical Society, 153 9B375-B383 2006 0013-4651/2006/1539/B375/9/$20.00 © The Electrochemical Society B375 ) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 54.90.228.213 Downloaded on 2016-10-27 to IP