The different matrices/instrument conditions employed for each analysis and the elements (and their isotope used) measured by each method

are described in Table 1. For the Thermo XSERIES 2 ICP–MS the typical normal mode conditions were as follows: extraction voltage was typically –100 V, Gemcitabine order Rf Power 1400 W, focus voltage 12.0 V and nebuliser gas flow rate (using a Burgener Miramist nebuliser) 0.83 L/min. Dwell times were 50 ms for each element and 10 ms for internal standards, with 50 sweeps per replicate and three replicates per sample. The instrument was tuned on a daily basis to ensure optimisation. When using the Thermo XSERIES 2 ICP–MS in collision cell mode, typically using a collision cell gas flow of 3.5 mL/min of 7% hydrogen in helium. For the ICAP Q ICP–MS the typical normal conditions were as follows: extraction voltage was typically –120 V, Rf Power 1400 W, and nebuliser gas flow rate (using a PFA nebuliser) 1.05 L/min. Dwell times were 1 s for 9Be and 0.05 s for 72Ge, with 20 sweeps per replicate and three replicates per sample. The instrument was tuned on a daily basis to ensure optimisation. Creatinine was determined by an automated alkaline picrate method (Cocker et al., 2011), using an ABX Pentra

400 spectrophotometer (HORIBA ABX UK, Northampton, UK). An internal QC material made from a pooled urine sample and stored frozen in 1 mL aliquots was used. The QC sample was thawed Fossariinae at PLX-4720 in vivo room temperature before use and analysed after each calibration.

All QC results fell within the acceptable range. Where available, certified reference materials (CRMs) were analysed at the start and end of each analytical run, and again after every 20 samples. Certified reference materials used were ClinChek levels 1 and 2 (lot 923 Recipe, Germany) for all elements except for beryllium which used ClinChek levels 1 and 2 (lot 122 Recipe, Germany). In addition Lypocheck, urine metals Level 1 (lot 69141 Bio-Rad Laboratories, Hemel Hempstead, UK) was used for mercury and those elements analysed in CCT mode elements for which these CRMs were used are stated in Table 2. For elements where no CRM was available, a blank urine sample (from another unexposed source) was spiked with that element and kept frozen at −20 °C (as well as a portion of the blank sample) until ready for analysis to be used as internal quality control (these are referred to as ‘pool samples’ in Table 2). The samples diluted with hydrochloric acid as per Method 4 (Ag, Ir, Nb, Os, Pt, Rh, Ru, Ta and Te) had pool samples spiked at two different concentrations (50 ng/L and 200 ng/L). Rarer elements (Au, Ce, Dy, Er, Eu, Gd, Hf, Ho, In, La, Lu, Nd, Pr, Sm, Tb, Tm, Th, Y and Yb,) diluted in nitric acid as per Method 5 had pool samples at one concentration (between 0.1 and 100 μg/L depending on the likely abundance found in a urine sample).