Laboratory analysis for beryllium samples has always presented a challenge to the analytical community.

The most frequently used beryllium analytical techniques in the U.S. are ICP-AES, inductively coupled plasma mass spectroscopy (ICP-MS), ICP-AAS, graphite furnace atomic absorption spectroscopy, and fluorescence. Some of the commonly used beryllium-specific analytical methods are OSHA IDG125G/ID206, NIOSH 7704, NIOSH 9110, and NIOSH 9102. Typical laboratory sample detection is based on the mass on the collection media such as the MCE filter in a sample cassette, wipe, and so on. For air samples, mass on the collection media is based on the amount of air drawn through the filter. The higher the air volume, the higher the probability of “capturing” more mass of interest on the filter; therefore, the easier it is for the laboratory analysis to “see” it. Lowering the OEL for beryllium means that the analytical detection limits for shorter-duration tasks like the OELs would be challenging. The lower the air volume, the higher the analytical sensitivity needed to achieve the desired limits of detection. Tables 1 and 2 show the relationships between beryllium air sampling rates, amount of volume collected, and the resulting lab detection limit. ISO 20581 and best practices set the detection limit for the analytical laboratory at 10 percent of the OEL. This means that if the target OEL is the ACGIH TLV of 0.05 µg, the lab should have an LOD of 0.005 µg (5 ng). According to the NIOSH Manual of Analytical Methods, fluorometry can achieve an LOD of 0.1 ng. Going by these tables, if the target OEL is 0.05 µg/m³, the sample duration is 30 minutes or less with a flow rate of around 2.0 liters per minute, and the analytical laboratory follows the ISO 20581 guidelines for the detection limit at 10 percent of the OEL, then the target OEL can currently be met only using molecular fluorescence. If the target OEL is from the new OSHA beryllium rule, with a PEL of 0.2 µg/m³, then at 2.0 L/min with sampling duration of 30 minutes the target OEL can be met using ICP-MS. The beryllium analytical methods mentioned here are not the only methods used for beryllium analysis. Microwave-induced plasma spectroscopy is another, and some DOE national laboratories are exploring the use of direct-reading instrumentation such as laser-induced breakdown spectroscopy. Beryllium direct-reading instrumentation is currently in the research and development phase. Typically, industrial hygienists are trained to manage carcinogenic risks to As Low As Reasonably Achievable (ALARA). However, given the laboratory analytical challenges, the implementation of ALARA for beryllium exposure management strategies could be modified to As Low As Accurately Detectible as described in Tables 1 and 2. This approach assumes that the personnel who manage beryllium risks have already considered elimination, substitution, and engineering controls and still must use beryllium for their projects. Laboratories currently performing industrial hygiene sample analysis follow the accreditation guidelines from ISO/IEC 17025. The AIHA Laboratory Accreditation Programs, LLC is the accrediting body utilized by most IH laboratories. Currently DOE 10 CFR 850 requires laboratory accreditation for beryllium sampling analysis or demonstration of equivalent quality assurance, but OSHA and the EU do not. (However, the new OSHA rule 1910.1024 does state that sample analysis be evaluated by a laboratory that can measure beryllium to an accuracy of plus or minus 25 percent within a statistical confidence level of 95 percent for airborne concentrations at or above the action level.) Participation in AIHA’s Beryllium Proficiency Analytical Testing program (BePAT) is a positive factor for clients who are looking for a lab to analyze beryllium samples. And accreditation, even when not required, allows clients to have greater confidence in laboratory data. However, accreditation by itself does not guarantee that the laboratory is experienced in handling beryllium samples. It is prudent for the client to inquire about a lab’s experience in dealing with the intricacies of beryllium analysis, data expectations, and the associated sample media. For example, some laboratories are familiar with air sampling filter cassettes but not with beryllium surface wipes or bulk samples. Another reason for communication is that most samples, in my experience, are lower than the RL. The client must then deal with censored data (results that are below the laboratory’s RL). Laboratory accreditation guidelines do not discourage the laboratory from reporting results that are below the RL. In these instances, the results are reported as “less than” the RL or “ND” (not detected), and cannot be reported as zero. (As mentioned earlier, no laboratory can currently detect beryllium at the zero level.) Before commencing a sampling event, it is important to ensure the following:

• the data quality objectives are established and understood by both the laboratory and the client
  
• the analytical laboratory’s capabilities and limitations are based on the client’s target OEL
  
• information is made available to the laboratory regarding the type of samples that are to be analyzed (air, swipe, vacuum, bulk) and the anticipated number of samples
  
• possible interferences, especially when using ICP-AES, are discussed
  
• lab QA/QC procedures and acceptance criteria are established
  

Agree on a turnaround time—how quickly can the results be provided based on when the laboratory receives the samples? This is important because “rush” samples typically cost more. However, most sampling jobs are limited by time and budget. In such cases, communication with the laboratory ensures that they can meet the project’s schedule. In some cases, the laboratory might have other projects that may not match up with the client’s schedule.