USGS-NWQL: I-2477-92: Determination of metals in Water by ICP-MS

Summary
Analytes
Revision
Data and Sites

Official Method Name

Methods of Analysis by the U.S. Geological Survey National Water Quality Laboratory - Determination of Metals in Water by ICP-MS

Current Revision

1999

Media

WATER

Instrumentation

Inductively Coupled Plasma - Mass Spectrometry

Method Subcategory

Inorganic

Method Source

USGS-NWQL

Citation

Garbarino, J.R., 1999, Methods of analysis by the U.S. Geological Survey National Water Quality Laboratory--Determination of dissolved arsenic, boron, lithium, selenium, strontium, thallium, and vanadium using inductively coupled plasma-mass spectrometry: U.S. Geological Survey Open-File Report 99-093.

Brief Method Summary

Sample solution is pumped by a peristaltic pump into a pneumatic nebulizer, which generates a liquid aerosol. This aerosol is transported by argon gas flow into a water-cooled spray chamber where the large droplets are removed by gravity and condensation. The small droplets are further swept into a radiofrequency ICP, where evaporation, molecular dissociation, atomization, and ionization occur. The ions are physically extracted from the center of the plasma by a differentially pumped vacuum system, through a water-cooled sampler cone and skimmer cone assembly, which is in physical contact with the horizontally mounted ICP.
The extracted ions are focused by an electrostatic ion lens assembly and accelerated in to a unit-resolution quadrupole MS. For a given combination of radiofrequency and direct current voltages applied to the quadrupoles, only ions of a specific mass-to-charge ratio pass through the quadrupoles and reach the detector. Rapidly scanning the voltages on the quadrupoles has the effect of rapidly scanning the mass spectrum at the detector. Selected mass values, where high ion signals are expected, can be pre-programmed to be skipped in the mass scan to avoid overload damage to the detector. The mass spectrometer alternatively can be operated in a peak-jumping mode, which rapidly changes the applied voltages to effectively jump to preselected masses instead of operating in the continuous mass-scanning mode already described.
Mass-scanning is suggested in this method because the recorded spectrum can be recalled and qualitatively examined for the presence or absence of any analytes or potential interferents.
The ion signal is detected by a continuous dynode electron multiplier of the channeltron type. The physical impact of an ion on the detector surface produces a pulse of electrons. The resulting signal is processed in a digital, pulse-counting mode. New or upgraded instrumentation with an "extended dynamic range" option allows operation of the detector in an analog mode for high analyte concentrations. In either case, the ion signal is electronically amplified, and the resulting data are processed by a multichannel analyzer, or alternative data system, and computer.

Scope and Application

This method is suitable for the single-element or multi-element determination of dissolved concentrations of 22 selected trace metals in water. The method is applicable to surface-water, ground-water, drinking-water, and precipitation samples that have a measured specific conductance of less than 2,500 uS/cm at 25^oC. The method is applicable to metals in the concentration range from 1.0 to 1,000 ug/L. The use of new or upgraded instrumentation with an "extended dynamic range" option can extend the upper concentration limit to 200 mg/L. Also samples with a specific conductance greater than the specified limit may be analyzed after appropriate dilution to conform to the specified limit; however, method reporting limits for the original sample will be increased according to the dilution factor. The determination of dissolved arsenic, boron, lithium, selenium, strontium, thallium, and vanadium in filtered, acidified natural water was later added to the original method. While originally published as two stand-alone reports, the two have been combined into a single file for the benefit of NEMI users.

Applicable Concentration Range

0.08 - 1,000 (undiluted)

Interferences

Physical interferences are associated primarily with sample introduction and are minimized by using the internal standardization technique. Isotopes measured in this procedure have been selected specifically to minimize spectral interferences from isobaric, doubly charged, and molecular ions. Multiple isotopes can be measured for selected elements that have potential isobaric or molecular ion interference. The analyst must be aware of these interferences because they might be present with certain types of sample matrices. Memory effects related to sample transport are negligible for most elements that are normally present in natural water. Thallium and vanadium did not recover to reagent-blank intensity levels within the rinse period; thallium and vanadium intensities were 10 and 2 times greater than reagent-blank levels, respectively. Consequently, the analyst must review all analytical results to ensure that errors from carryover are minimized. Sample matrix composition could also affect the bias and variability of ICP-MS determinations. The use of internal standardization compensates for most matrix effects, however, some matrix interferences remain problematic. Matrix composition can suppress the ionization efficiency of the plasma and result in negatively biased elemental concentrations.

Quality Control Requirements

Quality-control samples area analyzed at a minimum of one in every ten samples. These QC samples include at least one of each of the following: blanks, quality control samples, third party check solutions, replicates, and spikes. Correlation coefficients for calibration curves must be at least 0.99. QC samples must fall within 1.5 standard deviations of the mean value. If all of the data-acceptance criteria in the SOPs are met, then the analytical data are acceptable.

Sample Handling

Description: 250 mL Polyethylene bottle, acid-rinsed. Treatment and Preservation: Filter through 0.45-um filter, use filtered sample to rinse containers and acidify sample with nitric acid (HNO₃) to pH < 2.

Maximum Holding Time

180 days from sampling

Relative Cost

Less than $50

Sample Preparation Methods

This method has 7 analytes associated with it.

Analyte	Detection Level	Bias	Precision	Spiking Level
Arsenic(7440-38-2) Arsenic As	0.070 mg/L	102% Rec (SL)	N/A	0.50 mg/L
Boron(7440-42-8) B Boron Boron (Boron and Borates only)	0.500 mg/L	77% Rec (SL)	N/A	1.00 mg/L
Lithium(7439-93-2) Li Lithium	0.030 mg/L	94% Rec (SL)	N/A	0.20 mg/L
Selenium(7782-49-2) Se Selenium Selenium and Compounds	0.100 mg/L	110% Rec (SL)	N/A	0.90 mg/L
Strontium(7440-24-6) Sr Strontium	0.008 mg/L	93% Rec (SL)	N/A	0.20 mg/L
Thallium(7440-28-0) Thallium Tl	0.005 mg/L	97% Rec (SL)	N/A	0.05 mg/L
Vanadium(7440-62-2) V Vanadium Vanadium (fume or dust)	0.080 mg/L	98% Rec (SL)	N/A	0.60 mg/L

Precision Descriptor Notes:	Spike recovery percentages were determined for the new elements in matrices that are representative of reagent water, surface water, and ground water. Seven replicate recoveries at 5 to 10 times the MDL (the low-level spike) and 75 mg/L (the high-level spike) were determined in each matrix over a period of about 1 week. Average recoveries in the reagent-water matrix ranged from 93 to 105 percent for all elements except boron, whose low-level spike recovery was 77 percent. Recovery variability for the low-level spike ranged from 4 to 11 percent, depending on the element and matrix.
Detection Level Note:	Method detection limits were obtained by analyzing seven aliquots of a laboratory-prepared multi-element solution with concentrations of the metals near the expected detection limit of the method. The seven aliquots were analyzed consecutively in a single analysis procedure. The results were blank-subtracted. The standard deviation of each analyte for the seven replicate analyses was multiplied by the appropriate t-value for 99- percent confidence level to obtain the corresponding method detection limit. For the analytes added to the original method: short-term MDLs were calculated based on guidance in USEPA CFR40 and represent pooled averages on the basis of four MDLs determined on different days over several weeks.

Precision Descriptor Notes:

Spike recovery percentages were determined for the new elements in matrices that are representative of reagent water, surface water, and ground water. Seven replicate recoveries at 5 to 10 times the MDL (the low-level spike) and 75 mg/L (the high-level spike) were determined in each matrix over a period of about 1 week. Average recoveries in the reagent-water matrix ranged from 93 to 105 percent for all elements except boron, whose low-level spike recovery was 77 percent. Recovery variability for the low-level spike ranged from 4 to 11 percent, depending on the element and matrix.

Detection Level Note:

Method detection limits were obtained by analyzing seven aliquots of a laboratory-prepared multi-element solution with concentrations of the metals near the expected detection limit of the method. The seven aliquots were analyzed consecutively in a single analysis procedure. The results were blank-subtracted. The standard deviation of each analyte for the seven replicate analyses was multiplied by the appropriate t-value for 99- percent confidence level to obtain the corresponding method detection limit. For the analytes added to the original method: short-term MDLs were calculated based on guidance in USEPA CFR40 and represent pooled averages on the basis of four MDLs determined on different days over several weeks.

Revision	PDF/Link
1999

The Water Quality Portal contains sample results from sites that use this method.

Download WQP data

Select data:

Sites only
Sample results only

File format:

Comma-separated
Tab-separated
MS Excel 2007+

Sites where this method was used

Loading Sites ...

USGS-NWQL: I-2477-92: Determination of metals in Water by ICP-MS

WHO WE ARE

LEGAL