Hello, everyone. Welcome to today's live broadcast, Elemental Analysis of Lithium-Ion Battery Cathode and Anode Precursor Chemicals, Lithium Iron Phosphate, and Silicon Graphite. I'm Bob Alberta, Special Projects Director for Spectroscopy, and I'll be your moderator for today's event. We are pleased to bring you this webcast, presented by Spectroscopy and sponsored by Agilent. I would now like to share a quick statement from our sponsor. Agilent is a leader in life sciences, diagnostics, and applied chemical markets. The company provides laboratories worldwide with instruments, services, consumables, applications, and expertise, enabling customers to gain the insights they seek. Agilent's expertise and trusted collaboration give customers the highest confidence in the company's solutions. We have a few important announcements before we begin. This webcast is designed to be interactive, and we encourage you to ask questions during the event. You can submit your questions by typing them in the Q&A box, which can be found at the bottom of the video player. You can then also enlarge the slide window by clicking on the small icon in the bottom right corner of the media player. The slides will advance automatically during the event, and if you have any technical problems viewing or hearing this presentation, please click on the question mark Help widget in the top right of your presentation window. I would now like to introduce today's speakers. We are so pleased to welcome Dr. Seema Singha and Dr. Emma Qi. Seema is an application scientist at Agilent Technologies, supporting the ICP-OES, MP-AES, and AA instruments. She has over 15 years of hands-on experience in the field of atomic spectroscopy. Before joining Agilent in 2017, she managed an agricultural chemistry laboratory for the analysis of water, soil, and plant tissue samples using an ICP-OES. She received a PhD degree in chemistry from the University of Illinois at Chicago and has numerous peer-reviewed publications. Emma is an application scientist working at Agilent Technologies in the Optical Atomic Spectroscopy Group. Emma has focused her last 12 months on streamlining the preparation and elemental analysis of lithium-ion battery materials using ICP-OES. Thank you both for joining us today. Please take it away. Thanks for inviting us today to talk about lithium-ion battery, cathode and anode materials. In the first part of the webinar, I will talk about determination of elemental impurities in lithium iron phosphate cathode materials via ICP-OES. I will discuss about lithium-ion battery industry chain and where elemental analysis fits into this industry. I will focus on to the cathode materials and workflow for lithium iron phosphate sample analysis on ICP-OES. I'll talk about the challenges and solutions for sample preparation and method development, and then show you some results from some real-life sample analysis on lithium iron phosphate samples. Lithium-ion battery industry chain has four segments: upstream, midstream, downstream, and recycling. Elemental analysis is done in step number one, two, and four. In the upstream, we do impurity testing on raw materials, and in the midstream, cathode materials, anode materials, and electrolytes are tested for quality control. A lot of research and development is also done at this part to create a better battery, and once the batteries are spent, the recycling is done to collect the metal and use those in the upstream again. There are various cathode materials used in the lithium-ion battery industry. Two most widely used cathode materials are lithium iron phosphate and ternary materials. Ternary materials are lithium plus three other elements. For example, lithium nickel cobalt manganate or lithium nickel cobalt aluminate. I compare the characteristics of the LFP and ternary materials in terms of energy density, safety, charging efficiency, life cycle, and cost. Lithium iron phosphate cathodes win in three categories. They're safer, life cycle is better, and cost is cheaper. That makes them very popular in the lithium battery industry. You have a better understanding of the lithium-ion battery industry and different cathode materials, let's talk about workflow in the lab. Lithium iron phosphate samples will be in solid form, and you have to turn them into liquid form before you analyze them on ICP. We used a microwave digester, and we took 0.1 gram of sample and added 6 milliliter of HCL and 2 milliliter of nitric acid, and digested the samples at 200 Celsius for 20 minutes. The final volume of the samples were 50 milliliter, that gives about 500 times dilution factor. We didn't see any precipitate or undigested material in the sample, but if you see any, don't forget to filter the sample before they're aspirated in the ICP-OES instrument. The next step in the workflow is method development. If you look at the lithium iron phosphate sample composition, you will see that lithium, iron, and phosphate concentrations are very high in the samples, and there are very trace levels of impurity. These samples are very challenging because they have high dissolved solids or salts, and there are high and low concentrations of elements coexisting in the same sample. There are easily ionizable elements at high concentrations and low concentrations. For example, you are looking for trace elements such as sodium and potassium in presence of high concentration of lithium. There will be spectral interferences from iron and phosphorus, and there will be carryover or contamination. I'll discuss about how to overcome these challenges using various instrument features, internal standards, and accessories. The next few slides. The 5800 VDV ICP-OES is a dual view system with a vertical torch. The vertical torch can handle very high total dissolved solids with less maintenance. The torch is easy to install or deinstall using the torch loader. The analyst doesn't have to connect any gas tubing or perform torch optimization or alignment. Everything is automatically done once you close the torch loader. That saves time. The optics in the 5800 ICP-OES are optimized to provide you better detection limits and more stability. The default sample introduction system was used on the ICP-OES for the LFP sample types. We recommend using a fully demountable torch because lithium tends to degrade the outer tube of the torch over time. In case you want to replace the outer tube, you don't have to replace the whole torch. We added an accessory called argon humidifier to the sample introduction system to keep the nebulizer tip moist, because the high salt samples tend to form some deposit the tip of the nebulizer, and you might see some blockage when doing these type of sample analysis. To handle the effect of easily ionizable elements on certain elements, for example, sodium and potassium, we added rubidium as an internal standard in addition to helium. To remove carryover and maximize sample throughput, a smart feature of the instrument called Intelligent Rinse was used. Intelligent Rinse has a maximum rinse time, and it monitors the analytes in the rinse solution and decides how much rinse time it needs after each solution. For a very clean solution, Intelligent Rinse might move on to the next sample without using the maximum rinse time, thus saving the argon gas and the rinse solution and increasing sample throughput. During method development, the samples were screened using IntelliQuant. IntelliQuant is a semi-quantitative analysis. It can analyze up to 70 elements using pre-calibrated methods. IntelliQuant shows the data as a heat map, pie graph, or a bar graph. The heat map of a lithium iron phosphate sample shows the approximate concentrations of the primary salt elements and some traces. The numbers in the parenthesis are in ppm or milligram per liter. The color coding is: red means high concentration, orange is medium concentration, and light yellow to yellow is low concentration. IntelliQuant can also identify unknown spectral interferences using some kind of data analytics. In this lithium iron phosphate sample, IntelliQuant clearly identified that chromium 267 nm wavelength is interfered by phosphorus. It moved on to the second best wavelength, that is 205.560 nm, and took the data from that wavelength. With the knowledge of the approximate analyte concentrations and possible interferences from IntelliQuant, we are now ready to set up our quantitative analysis. This method was set up in dual view, where all the trace elements were analyzed in axial mode and the primary salt elements were analyzed in radial mode. The method parameters were mostly default, except for axial view, we used a little bit more RF power. It was 1.3 kW to get a little bit more sensitivity from the trace elements in axial view. All the trace elements were calibrated from 0 to 500 ppb in axial view, and the correlation coefficients were greater than 99.99% for all the elements. The calibration ranges for lithium, iron, and phosphorus in radial view were in the high ppm to match the sample concentrations and also to avoid any further sample dilution. In this dual view method, we were able to analyze for very low ppb concentration of the traces and very high ppm concentration of the salt element. Few slides ago, I talked about how IntelliQuant detected that the primary chromium line, 267 nanometer, had interference from phosphorus. If I still want to use this chromium line, I have to remove the spectral interferences. To remove the spectral interference, I used a method called Fast Automated Curve-fitting Technique, or FACT. FACT is a spectral deconvolution technique, and in order to do FACT, you have to analyze 3 solutions here: a blank and a pure analyte solution, that's chromium, and a pure interference solution, that's phosphorus. You load them into the FACT model for chromium 267 nanometer, and the FACT model will subtract the phosphorus interference from the combined peak. FACT can increase the data accuracy by removing the spectral interferences. The detection limits for the LFP method were studied in the blank and the salt matrix with a 5 ppb spike for the trace element. All the traces had some ppb to single-digit ppb detection limit, showing high sensitivity of the instrument. In the last column, the sample results are dilution adjusted and show the method is successful in analyzing low ppb to high ppm concentrations with a single sample preparation or dilution. To validate the method, a sample was spiked with 100 ppb of the trace element pre-digestion. The spike recoveries were within ±10%, even in the high salt concentration. The instrument was able to eliminate the matrix effect very efficiently. The instrument stability is an important factor to achieve reproducibility of the data. This graph shows the 9-hour stability study of a 100 ppb CCV. The recoveries were within ±10%, and the RSDs were below 2.3% for all the traces. To show the ruggedness and stability of the instrument, a 100 ppb spiked sample was also read back after every 10 samples for nine hours. The RSDs were below 2.2%, and there was no instrument drift from the high salt matrix. In total, the method analyzed 236 solutions with excellent QC coverage and without any kind of recalibration. To summarize all the results, I would say that Agilent 5800 ICP-OES was able to handle the challenging lithium iron phosphate samples with ease. Various smart features in the hardware and software were used to create a robust method that gave us very low detection limits, wide linear dynamic range, high reproducibility, and accuracy. If you are trying to set up the workflow for elemental analysis on different types of lithium-ion battery samples, these are some publications that might be helpful. You can download them by scanning the barcodes or searching for the publication numbers in red. Thanks for listening in. Up next, Emma will present the silicon-carbon anode material. My talk is about the determination of elemental impurities in silicon-carbon anode materials for lithium-ion batteries by ICP-OES. In this presentation, I'm going to talk about why we analyze silicon-carbon anode materials and what the challenges are there. Then I'm going to outline the methods for sample preparation and also the instrumentation. Then in the results section, I will talk about how the Agilent 5800 ICP-OES with the ICP Expert software is able to meet the challenges and offers good sensitivity, accuracy, and stability for elemental impurity analysis of lithium-ion battery anode materials. At the end is a brief summary. The global lithium-ion battery market is fast growing in recent years, mostly driven by the soaring demand for electric cars, large-scale energy storage, and as well as small personal electronic devices. The anode is an essential component of a lithium-ion battery, making up 5%-15% of lithium-ion battery cost. Conventional commercial lithium-ion batteries primarily use graphite-based anodes. The advantages of graphite as anode materials are its availability, excellent electronic conductivity properties, and low cost. The drawbacks include decreased rate capacity, low specific capacity, as well as safety risks. To overcome the downsides of the current anode materials, researchers and industries have been attempting to find alternative anode materials with higher performance over the years. Among the potential alternative anode materials being researched, silicon has stood out as a potential very promising anode material, with its very high theoretical storage capacity, working potential, and natural abundance. Overcoming the biggest drawbacks of silicon, namely volume expansion and low conductivity, the most promising silicon anode material is the silicon-carbon nanocomposite, which benefits from the very high volumetric and specific capacity of the silicon, while the carbon matrix can handle volume fluctuation and maintains the structural integrity and electrical stability of the material. Production of silicon-carbon anode materials has advanced to commercialization recently. It is claimed the new materials have 5 times the capacity of graphite and affords up to 50% more energy density than the conventional graphite for lithium-ion batteries. Whether it's conventional or the new generation anode materials, it is important for relevant industries, which is developers, suppliers, and users of the materials, to use robust testing methods for quality control, because the quality of anode materials is critical to the overall performance of the final product. Elemental impurities can be detrimental to lithium-ion battery capacity, rate performance, cycle stability, et cetera. As a leading producer of anode materials, China has issued 2 national standards, one for graphite anode materials and the other is for silicon-carbon. Both recommended standards require the analysis of trace elements such as iron, sodium, chromium, copper, et cetera, using ICP-OES, following microwave digestion of the samples. Based on this methodology in the standards, we analyzed 25 elements in a graphite and a mixture of the graphite and silicon at 9 to 1 mass ratio, which are representing a graphite anode material and a silicon-carbon anode material, respectively, using the 5800 VDV ICP-OES. To analyze graphite and silicon-carbon anode materials using ICP-OES is not without its challenges. The biggest challenge is related to the complex sample matrix due to the nature of the samples and digestion methods used, which can lead to high background signal and physical and spectral interferences. Another challenge is related to the very low but varied impurity levels in the samples. In the GB standards, graphite fixed carbon is specified as high as 99.7% to 99.95% or higher, while the silicon mass fractions specified for silicon-carbon materials are between 2% and 40%. This means the instrument need to be able to have high sensitivity and wide working range, as well as being able to handle high complex sample matrices. On top of these challenges, the undissolved solids in the microwave digested samples and the process to ensure their fully removal, while not introducing any new contaminations or causing sample loss, is quite testing for the operators, but the process is vital for the quality of the results. Any undissolved solid particles may block the nebulizer or affect sample metrics during ICP-OES analysis. With these challenges in mind, our answer is the advanced Agilent 5800 ICP-OES, together with its smart features in the ICP Expert software. It is well designed and equipped to handle the complex sample matrices of lithium-ion battery anode materials. The Vista Chip detector of the 5800 ICP-OES offers fast, simultaneous measurement of both major and trace elements, offers automatic background correction, as well as effective interference measurement. In the ICP Expert software, the default Fitted Background Correction, FBC, can accurately correct spectral interferences and improve data accuracy. Also, in the ICP Expert software, the unique smart feature of IntelliQuant screening, it simplifies method development by performing a quick semi-quantitative scan of the sample, detecting spectral interferences, and helping with wavelength selection. In the following slides, I'm going to go into more details about how some of these features can help with the anode material analysis. This slide shows the process of sample preparation for ICP-OES analysis. The method was based on the microwave digestion method outlined in the two GB methods for anode materials. A 99.99% graphite and a mixture of this graphite with a 99% silicon at 9 to 1 mass ratio, were used to represent graphite and silicon-carbon anode material, respectively. One gram of the sample was accurately weighed on a balance before being added to a dry, clean microwave digestion tube. After thoroughly mixed in aqua regia, which is hydrochloric acid and nitric acid at 3 to 1 molar ratio, the samples were then digested using a MARS 6 microwave digestion system at 200 degrees, holding for 30 minutes. After digestion, the contents in the digestion vessel was transferred to a 50 mL tube and made up to volume. The solution was then filtered through a 0.45 micron disk filter to remove undissolved particles. An Agilent 5800 ICP-OES, equipped with an SPS 4 autosampler, was used for subsequent analysis. Blank control samples were also processed and analyzed following the same procedure. All solutions were prepared in triplicate. To analyze the prepared samples on ICP-OES, the Agilent 5800 ICP-OES instrument was controlled using the ICP Expert v7.6 software. The instrument was fitted with a SeaSpray glass concentric nebulizer, double pass cyclonic spray chamber, an Easy-fit, fully demountable torch with 1.8 millimeter quartz injector. Sample introduction was performed using the Agilent SPS 4 autosampler. Prior to quantitation on ICP-OES, to simplify the method development process for quantitation, a unique feature in the Agilent ICP Expert software in IntelliQuant is a great and very useful tool that makes the method development process quick and simple. It allows users to run a quick semi-quantitative analysis of a sample by scanning the entire spectral range of up to 70 elements. The results are displayed in a color-coded periodic table heat map, allowing users to easily see which elements are present and at what concentration. The heat map here shows the semi-quantitative results of a quick IntelliQuant scan of a silicon sample solution. Here, we can see aluminum and iron are highlighted in orange as the most abundant elements, followed by calcium, highlighted in dark yellow, and some other elements in light yellow. This information is just what we need to decide the concentration range of our calibration standards. The IntelliQuant function also identifies any potential interferences and determines the optimal analysis wavelengths using a star ranking based on potential interferences. This figure on the right shows the prime wavelength for vanadium 292.401 is ranked with just 1 star. Hovering the mouse on the red question mark, a message indicates the potential spectral interference of titanium 292.396. The vanadium 309.310 is ranked with 5 stars, with a green tick. This informs us the wavelengths we can use for vanadium quantitation. The star ranking is very useful with the selection of quantitation wavelengths. To run the samples for IntelliQuant, there's no need for prior knowledge of the samples or expertise in spectroscopy. With minimal instrument setup required, the operators can spend less time for quantitation method development. A total of 25 elements in our samples were analyzed. Following sample preparation, digested graphite samples were directly analyzed on ICP-OES. silicon-carbon samples were diluted 10 times before quantitation, so that the higher impurity concentrations were within the working range. Calibration standard solutions, blank control samples, Continuing Calibration Verification and blank, as well as internal standard solutions, were all prepared in the same matrix as the samples. Instrument operating conditions were evaluated and optimized based on calibration, linearity, and detection limits. Axial viewing mode was used for all 25 elements, and it turns out all the conditions were the instrument default conditions, except for the longer rate time to accommodate the low impurity concentrations. This slide show the working range and linear correlation coefficient for 25 elements at selected wavelengths. Note the bigger working range for aluminum and iron when 2 ppm concentration standard was added for both to cover their higher concentrations in the samples, which we determined during IntelliQuant screening, when we showed in our previous slide. As shown in this table, good linearity was achieved for all 25 elements at selected wavelengths. The calibration curve of iron in this figure show a perfect linear calibration coefficient of 1 across 6 calibration points from 5 ppb and 2 ppm. Iron is an important efficiency indicator for anode as an impurity. The excellent linear dynamic range of the 5800 ICP-OES we show here allows for accurate detection of higher iron concentrations. After the samples are analyzed on ICP-OES, when it comes to correct interferences from complex sample matrix. There are a number of background correction methods in the ICP Expert software the users can choose. The default option is Fitted Background Correction, FBC. Eliminating the need for users' input, FBC provides accurate correction of both simple and complex background structures. As the example here shows, the FBC accurately models and corrects the spectral interference of OH 377.401 on the copper 377.395. To determine the sensitivity of the ICP-OES instrument, limits of detection and limits of quantitation were measured. 10 blank sample matrix solutions were analyzed. The values presented in this table are the average of three results measured on three non-consecutive days and calculated as one gram graphite in 50 mL solution. For silicon-carbon, the LODs and LOQs are 10 times of these values to account for the dilution factor. The high sensitivity of the Agilent ICP-OES instrument is shown by the well below one milligram per kg LODs. For some elements, the values are at ppb level. The excellent low LODs and LOQs are achieved because the Agilent 5800 ICP-OES uses the industry's most advanced free-form optics. The results of impurities in graphite and silicon-carbon are shown on the two columns on the right. Quantitation of the elements in the 99.99% graphite sample and the silicon-carbon sample with 10% silicon and 90% graphite using the Agilent ICP-OES, were able to detect element concentrations as low as 10-100 ppb levels, as the values in red shown. This result meets the specifications in the two GB methods for elemental impurities in liquid, lithium-ion battery anode materials. Spike recovery tests are an effective way to evaluate the accuracy and reliability of the sample preparation method and or the analytical method, especially when no certified reference materials are available. In our case, the graphite samples were spiked with all 25 elements at 25 ppb or 1.25 milligram per kilo before microwave digestion. The silicon-carbon sample solutions were diluted 10 times after microwave digestion and spiked. For calcium, potassium, and sodium, two higher levels spiked concentrations, 50 and 100 ppb, were also used. Spike recoveries within plus or minus 10% of the expected value were achieved for all spike experiments. The results for graphite at 25 ppb spike concentration are shown in this figure. The good spike recovery results for graphite samples before digestion validated the accuracy of the sample preparation procedure, while the spike recovery results for silicon-carbon samples after digestion confirmed the accuracy of the 5800 ICP-OES method for the analysis of low-level impurities, despite the complex sample matrix. To check the stability of the instrument and long-term validity of the calibration, 260 measurements were completed over a 7.5-hour period without recalibration. A CCB and a CCV sample were measured between every 10 measurements of spiked and unspiked graphite samples. The concentrations of the 19 CCV measurements were plotted against time and shown in this figure. The concentrations are within ±10 of the expected value, while the RSDs for all wavelengths are below 2%, except for potassium and sodium at 3.31% and 3.55%, respectively. The precision of 138 measurements of spiked graphite samples were also excellent over 7.5 hours, with RSDs below 5%. The excellent long-term stability can be attributed to the vertical torch configuration and the solid-state RF system used in the Agilent ICP-OES, which provides excellent plasma robustness and stability. In conclusion, the Agilent 5800 ICP-OES proved to be well-equipped to handle complex sample metrics for routine QC analysis of graphite and silicon-carbon anode materials using methods based on currently available standards for these materials. The unique IntelliQuant feature in the ICP Expert software identifies spectral interferences and simplifies the method development process for complex samples. The Fitted Background Correction method provides fast and easy, accurate background correction for samples with both simple and complex backgrounds and spectral interferences. The high sensitivity, excellent accuracy, and stability of the Agilent 5800 ICP-OES instrument were demonstrated by the very low detection limits, good spike recoveries, and long-term stability. This work has now been published as an application note on the agilent.com website. If you're interested, you can scan this QR code and access it on our website. Thank you for your attention. All right. Thank you both for that informative presentation. Before we get started on the question and answer session, I would like to remind our audience how to submit questions. You can submit your questions by typing them in the Q&A box, which can be found directly below the video player. Our first question of the day is: What are the common issues you may encounter during sample preparation, and how can you manage them? Common issues you may encounter during sample preparations and how to manage them. There are indeed some things an operator should be mindful while preparing the anode materials for ICP-OES analysis. The first thing can be the constant drifting you find on the balance if you try to weigh the microwave digestion vessel directly on the balance. This is because the static charges on the digestion vessel. The solution is to not to weigh the digestion vessel on the balance, but to weigh the sample on the balance before transferring to the digestion vessel. Another thing to be mindful is the purity of the acids for digestion. You may find high impurity levels in the acids. Higher purity reagents may be needed. A final thing to be mindful is the filtration of the samples. It's suggested that a minimum of 0.45 psi micron filters should be used before ICP-OES analysis, because if not fully removed, even very fine particles can affect your sample matrix and complicate your results, even if they do not block the nebulizer. For particulate samples, how can you monitor and be alerted of any nebulizer problems like blockage? This question is about how to monitor and be alerted of any nebulizer problems, like blockage for particulate samples. If any undissolved particles in the samples cause a full blockage of the nebulizer, there's a Neb Alert function in the ICP Expert software you can use. To be alerted of the blockage, you can go to File and Option, and then in the Instrument tab, you can tick Enable Neb Alert and set lower and upper pressure limits. Since the system monitors the nebulizer back pressure, if the pressure moves out of the set range, an alert will pop up, indicating a potentially blocked nebulizer as soon as it happens, rather than finding out later when the QC fails. If the problem with unresolved, undissolved particles in the sample is not severe to cause full blockage of the nebulizer, you can still detect any issues by monitoring the internal standard over time to see if there's a constant drift. To diagnose the problem, go to the trend chart of the internal standard on the analysis page. The GB standards for anode materials specify other groups of elements, such as magnetic matter and restricted matter, to be determined using ICP-OES. Is your method covering these groups of elements? This is a very good question. In the 2 GB standards for graphite and silicon carbon materials, a number of groups of elements are determined using ICP-OES, including trace elements, magnetic matter, which is a sum of 5 elements such as iron, cobalt, chromium, nickel, and zinc, and also restricted matter, including cadmium, lead, and mercury. Each group of the elements is referred to a different sample preparation method. The method we used in this work is based on the method for trace elements analysis. If you need to determine the concentrations of the other groups of elements, please refer to the methods specified in the standards. The ICP-OES method developed in our work can still be used for the analysis of these elements after different sample preparation methods. All right. Thank you so much, Emma. Just a few questions now for Seema. First question: Did you try hot block digestion for the lithium iron phosphate samples? No, we didn't use hot block because it's an open vessel system, and at high temperature, some of the analytes, for example, mercury and selenium, might evaporate from the vial. Mercury is an important impurity monitored in the lithium battery industry, so we didn't want to risk it. Next question from the audience: Can you analyze different types of cathode materials in the same method? Yes, absolutely, keep in mind that some of the analyte concentrations can vary from sample to sample. In lithium iron phosphate samples, iron is at about 1,000 ppm, in the ternary materials, it's a trace element. You have to choose the appropriate wavelengths, viewing modes, and make sure there is no carryover or memory effect. If you really want to use 1 method for all, you can use a software feature called MultiCal. MultiCal will allow you to set up different concentration ranges for a different analyte wavelength. Great, our final question of the day before we wrap up: Do you have to calibrate for all the interfering elements for FACT modeling? You don't have to calibrate for any of the interference for FACT modeling. It's just the analyte that needs to be calibrated for, and the FACT models can be saved in the library, so you don't have to build the models every day either. All right. With that, we will wrap up. I want to thank the audience for attending and for participating in today's event. I would also like to thank our sponsor, Agilent, for making today's webcast possible. We would like to ask everyone in the audience to please participate in a brief survey. This survey will appear on your screen after today's presentation has ended. You'll receive an email alerting you when this webcast will be available for replay, and we invite you to forward that announcement to your colleagues who may have missed today's live event. We hope to see you all next time. Take care, everybody.