Use of Microwave-Assisted Evaporation for the Complete Recovery of

Anal. Chem. 2000, 72, 2908-2913

Use of Microwave-Assisted Evaporation for the Complete Recovery of Volatile Species of Inorganic Trace Analytes Dirk D. Link and H. M. “Skip” Kingston*

Department of Chemistry and Biochemistry, 308 Mellon Hall, Duquesne University, Pittsburgh, Pennsylvania 15282-1503

Solutions must often be evaporated prior to analysis either to preconcentrate the analyte or to eliminate an incompatible matrix component. Elimination of the halogen-based acids HCl and HF using traditional evaporation methods poses recovery problems because of volatilization of the target analyte as the chloride or fluoride species. A new sample preparation chemistry for trace analysis, where losses of analyte due to volatilization during the evaporation process are minimized, is explored using the unique heating mechanisms of the microwave-assisted evaporation process. The heating mechanisms of hot plate evaporation and microwave-assisted evaporation are compared, and temperatures throughout the evaporation process using each method are predicted and experimentally verified. Because the solution actually cools during microwave-assisted evaporation, volatilization due to overheating at dryness is minimized. Elemental standard solutions and SRM soil and tissue digestates were evaporated using a hot plate method and a newly developed reduced-pressure microwave-assisted evaporation apparatus. Redissolution and analysis of the residue by ICPMS showed that complete recovery was achieved using microwave-assisted evaporation while losses of several classically volatile analytes occurred using hot plate evaporation. Many ultratrace detection techniques, such as inductively coupled plasma mass spectrometry (ICPMS), require homogeneous samples for analysis. To transform solid samples into homogeneous solutions, dissolution methods using concentrated mineral acids at elevated temperatures and pressures are the most common sample preparation methods in use. Mineral acid decompositions typically use nitric acid as the main oxidizing acid, often in combination with nonoxidizing acids such as hydrofluoric acid and hydrochloric acid. The widespread use of the halogenbased acids HCl and/or HF in digestion procedures increases the complexity of sample processing, especially for detection of analytes at ultratrace levels. The use of these acids for sample decomposition often results in solution matrixes that present compatibility problems with ICPMS or GFAAS analysis, which necessitates the removal or adjustment of the matrix prior to analysis. The traditional chemistries involved in recovering ana* Corresponding author: (phone) (412) 396-5564; (fax) (412) 396-5359; (email) [email protected].

2908 Analytical Chemistry, Vol. 72, No. 13, July 1, 2000

lytes during matrix removal procedures have been established using conventional heating methods. In light of a newly developing, fundamentally different mechanism of matrix removal, these traditional chemistries must be reevaluated. For the total decomposition of solid samples, HF is required for matrixes that contain silicates and certain refractory materials. 1-13 HCl, because of its dissolution and complexing ability, is used to aid in the decomposition of typical environmental samples as well as alloys and some refractory materials.2-4,7-10,13,14 Recent studies have also shown that when a total decomposition is not required, the use of HCl increases the leachable quantities of certain analytes from environmental samples.15-17 Moreover, several standard environmental sample preparation methods, such as EPA SW-846 leach methods 3015A and 3051A, and total decomposition method 3052 require HCl in the acid mixture for complexation and stabilization of certain elements in solution.18,19 (1) Headridge, J. B. In CRC Crit. Rev. Anal. Chem. 1972, 3, 461-490. (2) Bock, J. S. A Handbook of Decomposition Methods in Analytical Chemistry, 1st ed.; John Wiley and Sons: New York, 1979. (3) Sulcek, Z.; Povondra, P. Methods of Decomposition in Inorganic Analysis; CRC Press: Boca Raton, FL, 1989. (4) Willard, H. H.; Rulfs, C. L. In Treatise on Analytical Chemistry; Kolthoff, I. M., Elving, P. J., Eds.; Interscience: New York, 1961; Vol. 2, pp 10271050. (5) Bajo, S. In Preconcentration Techniques for Trace Elements; Alfassi, Z. B., Wai, C. M., Eds.; CRC Press: Boca Raton, FL, 1992; pp 3-31. (6) Vandecasteele, C.; Block, C. B. In Methods for Trace Element Determination; John Wiley and Sons: New York, 1993; pp 9-53. (7) Bogen, D. C. In Treatise on Analytical Chemistry; Kolthoff, I. M., Elving, P. J., Eds.; John Wiley and Sons: New York, 1982; Vol. 5, pp 1-22. (8) Scott, W. W. Scott’s Standard Methods of Analysis, 5th ed.; Van Nostrand: New York, 1939. (9) Kingston, H. M.; Jassie, L. B. In Introduction to Microwave Sample Preparation: Theory and Practice; Kingston, H. M., Jassie, L. B., Eds.; American Chemical Society: Washington, DC, 1988; pp 93-154. (10) Anderson, R. Sample Pretreatment and Separation; John Wiley and Sons: New York, 1987. (11) O’Haver, T. C. In Trace Analysis: Spectroscopic Methods for Elements; Elving, P. J., Winefordner, J. D., Kolthoff, I. M., Eds.; John Wiley and Sons: New York, 1976; Vol. 46; pp 63-78. (12) Van Loon, J. C. In Selected Methods of Trace Metal Analysis; Elving, P. J., Kolthoff, I. M., Winefordner, J. D., Eds.; John Wiley and Sons: New York, 1985; pp 77-111. (13) Zehr, B. D. Am. Lab. 1992, 24, 24-29. (14) Sulcek, Z.; Povondra, P.; Dolezal, J. CRC Crit. Rev. Anal. Chem. 1977, 8, 255-323. (15) Florian, D.; Barnes, R. M.; Knapp, G. Fresenius J. Anal. Chem. 1998, 362, 558-565. (16) Link, D. D.; Walter, P. J.; Kingston, H. M. Environ. Sci. Technol. 1998, 32, 3628-3632. (17) Link, D. D.; Kingston, H. M. Environ. Sci. Technol. 1999, 33, 2469-2473. (18) USEPA. Fed. Regist. 1998, 63, 25430-25438. 10.1021/ac991369d CCC: $19.00

© 2000 American Chemical Society Published on Web 05/09/2000

Table 1. Elements That Have Been Documented To Be Volatile from Solutions of HCl or HF1-8,10,11,13,14,23,30,31a AlCl InCl SeCl,F a

AsCl,F MoCl,F SiCl,F

AuCl NbF SnCl,F

BCl,F OsF TaCl,F

BaCl PCl TeCl,F

CCl PaCl ThCl

CdCl PbCl,F TiCl,F

CrCl,F RaCl TlCl

GaCl ReCl,F VCl,F

GeCl,F RuCl,F WCl,F

HfCl SCl ZnCl,F

HgCl SbCl,F ZrCl,F

Cl indicates that the analyte has a documented volatile chloride species. F indicates that the analyte has a documented volatile fluoride species.

Large amounts of these acids in the solution matrix can present problems for the analysis of the sample. Many elements will combine with Cl- species in the ICPMS plasma to form polyatomic species which present isobaric interferences. For example, 40Ar35Cl+, and 40Ar19F + interfere with measurements of the monoisotopic elements 75As and 59Co, respectively, and 35Cl16O+ interferes with measurements of the primary isotope 51V (99.75% abundant). Also, even minor concentrations of HF may slowly etch the glass components of the ICPMS system, which not only limits their lifetime but also increases the matrix interferences due to additional Si entering the plasma. One possible solution to these problems is dilution of the sample solution. While dilution of the sample may provide a more compatible matrix, it is not appropriate when the concentration of analytes is reduced below the instrument detection limit and may not be suitable for ultratrace analytes. Therefore, postdissolution processing steps are routinely performed using some traditional form of evaporation to eliminate the matrix and concentrate the sample.20 The choice of heating method used to perform these evaporations is not a trivial one. There are ∼40 elements that have chemical forms that are potentially volatile (See refs 19 and 21 for a compendium of the chemistry of these analytes.). Problems associated with element volatility using traditional heating methods have been documented for the greater portion of the 20th century.2,4,5,7,10,22,23 Table 1 shows those analytes that have been reported to be volatile from solutions containing either HCl or HF, especially when evaporated. Problems in recovering the chloride species, which are notoriously volatile, have been welldocumented in the literature dating back to the 1930s (for example, see ref 22). However, the chloride forms of these analytes are not the only species that may be difficult to fully recover using traditional evaporation methods. The fluoride species of many analytes have also been shown to be volatile. Because of the large number of dissolution procedures that use either HCl or HF, incomplete recovery of analytes from these solutions due to volatilization is an important concern for analytical chemists. The volatilization chemistry of such analytes was established using conventional heating methods such as Bunsen burners, hot plates, or heating mantles, which involve the transfer of thermal energy through convection and conduction currents. This heating (19) Kingston, H. M.; Walter, P. J.; Chalk, S. J.; Lorentzen, E.; Link, D. In Microwave-Enhanced Chemistry: Fundamentals, Sample Preparation, and Applications; Kingston, H. M., Haswell, S. J., Eds.; American Chemical Society: Washington, DC, 1997; pp 223-349. (20) Treadwell, F. P.; Hall, S. B. In Analytical Chemistry; John Wiley and Sons: London, 1937; Vol. 1. (21) Walter, P. J.; Chalk, S. J.; Kingston, H. M.; Link, D. D., http://www.sampleprep.duq.edu/sampleprep, 1996-2000. (22) Hoffman, J. J.; Lundell, G. E. F. J. Res. Natl. Bur. Stand. 1939, 22, 465470. (23) Lundell, G. E. F.; Hoffman, J. I. Outlines of Methods of Chemical Analysis; John Wiley and Sons: New York, 1951.

mechanism causes overheating of the sample residue at dryness, leading to analyte losses. Only recently has the use of microwave energy to perform postdigestion evaporations been explored. Differences between the fundamental mechanisms of microwave heating and conventional heating provide microwave heating with advantages for evaporating samples to dryness. This unique heating mechanism, and the fundamental relationships between sample volume and the absorption of microwave energy, results in microwave evaporations being able to retain those elements traditionally defined as volatile. In this work, the last several decades of analyte recovery chemistry, including the reported volatility of analytes from solutions containing these complexing acids, are reevaluated. Postdigestion evaporation procedures using both conventional and microwave heating methods are compared. The advantages of using microwave energy to perform evaporations versus using conventional methods, and the theory behind these advantages, are described. Recovery data are presented that demonstrate the complete recovery of volatile inorganic species using microwaveassisted evaporation, which is in contrast to the losses experienced using hot plate evaporation. The advantages of microwave-assisted evaporation may establish a new tool for trace inorganic analysis where losses due to volatilization of many inorganic species are no longer a concern. EXPERIMENTAL SECTION Custom Evaporation Vessel Closure. Commercially available evaporation equipment capable of performing the unique studies presented here was not available. For this reason, a custom evaporation apparatus was designed and machined using fluoropolymer materials. A diagram of the custom-designed vessel closure is shown in Figure 1. The closure was made from 2-in.diameter PFA Teflonrod (Read Plastics, Gaithersburg, MD). The rod was cut into small sections (∼2 in.), and four holes (1/8-in. diameter) were drilled through the section. The bottom of the cylinder was machined into a conical shape. On the top portion of the cap, the four holes were widened by re-drilling (11/32-in. diameter), and the holes were tapped with 1/8 NPT threads. Threaded fittings (Fluoroware, Johnson Equipment Co., Oakdale, PA) were secured into these holes, allowing four sealed ports in each vessel for monitoring and gas flow purposes. To ensure a proper seal of the cap onto the vessel, a Teflon-coated O-ring was fitted over the conical section of the cap. During the evaporations, three of the four available ports were sealed with plug ferrules. A section of 1/8-in. PFA tubing, ∼6 in. in length, was secured into the fourth fitting, and the tubing from each vessel connected to the central vacuum tubing using special T-fittings. Pressure inside each vessel was reduced using a special laboratory vacuum/fume scrubber pump (FAM-40, Milestone, Inc., Monroe, CT) connected to the central vacuum tubing. Analytical Chemistry, Vol. 72, No. 13, July 1, 2000

2909

Figure 1. Diagram of the custom-designed microwave-assisted evaporation vessel closure.34

Custom Microwave Evaporation Rotor. Rotors are used in microwave-enhanced processes to hold the vessels in proper positions and to rotate the vessels through the microwave cavity for more homogeneous heating. Because the custom closure (described above) was not compatible with commercially available rotors, a custom rotor was also designed and constructed. The rotor top and bottom plates, ∼12-in. diameter, were cut from a stress-relieved polypropylene sheet of 1/2-in. thickness (Cadillac Plastic, Monroeville, PA). The top plate was modified to allow the fittings from the vessel closures to extend through the plate. Threaded polypropylene support posts (Read Plastics, Gaithersburg, MD), 1.5 in. in diameter and 7.25 in. in length, were made and used to secure the vessels between the rotor plates. Custommade polypropylene nuts, which were hand-tightened onto the support post, secured the top plate. The vessel closures were sealed onto the liners by securing the top plate. Evaporation of Standard Solutions. The evaporation of standard analyte solutions using a hot plate and the custom microwave apparatus was compared using an aqueous multielement standard solution diluted to 1 ppm in 10% HCl. The microwave-assisted evaporations were performed using the Ethos 900 Laboratory Microwave System (Milestone USA, Monroe, CT) with TFM Teflon vessel liners. The microwave unit is housed in a class 1000 clean laboratory environment. The following heating program was used: 500 W for 20 min, 0 W for 1 min, and 500 W for 20 min. hot plate evaporations were performed inside a fume hood in the general laboratory area using Pyrex beakers. A dial setting of 4 on the hot plate was used, and the temperature was such that the solutions did not boil during the evaporation. A 10mL aliquot of the 1 ppm standard was delivered into each vessel to yield a starting amount of 10 µg of each analyte. After evaporation to dryness, the residue remaining at the bottom of each vessel was redissolved by adding 2 mL of concentrated HNO3 and a small amount of 18 MΩ‚cm water. The dissolved residue 2910

Analytical Chemistry, Vol. 72, No. 13, July 1, 2000

was then rinsed into clean sample vials and diluted to 50 mL with 18 MΩ‚cm water. Typical evaporation times were ∼40 min using both evaporation methods. The microwave-assisted method, however, did not require continuous monitoring whereas the hot plate method did. Temperature measurement was performed using a fiber-optic probe (Luxtron, Santa Clara, CA) in a replicate sample which was not analyzed due to concerns of sample contamination. Evaporation of Standard Reference Material Digestates. SRMs 2710 (Montana SoilsHighly Elevated Concentrations) and 1566a (Oyster Tissue) were decomposed using closed-vessel microwave dissolution methods. A mixture of 10 mL of HNO3 and 3 mL of HF was used to dissolve the soil sample, and 10 mL of HNO3 was used to dissolve the tissue sample. Following the complete microwave-assisted dissolution, 2 mL of HCl was added to each sample for complexation and stabilization of certain analytes. The liners were transferred to the custom microwaveassisted evaporation rotor. Sample digestates were evaporated to dryness, and the residue was redissolved and diluted as above. Elemental Analysis. Samples were analyzed by ICPMS using a PlasmaQuad II (VG Elemental) and HP 4500 benchtop ICPMS (Hewlitt-Packard) in clean laboratory environments of class 100 conditions or better. Data analysis calculations were performed using raw count rate files for each analyte in Microsoft Excel. No interference equations were used. RESULTS AND DISCUSSION Volatilization of inorganic species may occur when the boiling point of the species is reached, as the temperature becomes high enough for the species to escape the solid or liquid sample. Following the dissolution of a sample, the resulting aqueous solution consists of trace metal positive ions along with complexing negative ions, such as nitrate and chloride ions, all of which are surrounded by bulk solvent molecules. Volatility from the solution

Figure 2. Temperature of solution as volume decreased during microwave-assisted and hot plate evaporation. The final point in the hot plate temperature profile is that of the beaker bottom at dryness.

is not a problem because the solvated ions in solution do not have a vapor pressure or boiling point. As the number of solvent molecules decreases, the positive and negative ions will recombine. At dryness, the resulting residue will consist of recombined salts, all of which have a certain vapor pressure and boiling point. It is at this moment where control of the temperature is so important because overheating at this stage leads to losses of volatile analytes. For minimization of losses due to volatility, the control of temperature at dryness is of utmost importance. Traditional heating methods heat the vessel containing the sample, which in turn heats the solution. This mechanism eventually leads to overheating of the evaporating solution and the residue of salts, because as the evaporation proceeds, the vessel bottom approaches the temperature of the heat source (e.g., the surface of the hot plate). Because the temperature is out of control at dryness, the likelihood of volatilizing certain species is increased. Temperature data for the evaporation of a solution using a hot plate is shown in Figure 2. At dryness, the temperature of the residue approaches the temperature of the vessel bottom, which in this case was over 170 °C. Evaporation using the mechanisms of microwave heating, which have been discussed in detail in the literature,9,19,24-27 produces a completely different temperature profile than evaporation using conventional heating methods. Briefly, the microwave(24) Kingston, H. M.; Jassie, L. B. Anal. Chem. 1986, 58, 2534-2541. (25) Mingos, D. M. P.; Baghurst, D. R. Chem. Soc. Rev. 1991, 20, 1-47. (26) Mingos, D. M. P.; Baghurst, D. R. In Microwave-Enhanced Chemistry: Fundamentals, Sample Preparation, and Applications; Kingston, H. M., Haswell, S. J., Eds.; American Chemical Society: Washington, DC, 1997; Vol. 1, pp 3-54. (27) Smith, R. D. Electric Power Research Institute, Palo Alto, 1984. (28) Kingston, H. M.; Jassie, L. B. The Symposium on Waste Testing and Quality Assurance, Washington, DC, July 11-15, 1988; pp D44-D56. (29) Neas, E. D.; Collins, M. J. In Introduction to Microwave Sample Preparation: Theory and Practice; Kingston, H. M., Jassie, L. B., Eds.; American Chemical Society: Washington, DC, 1988; pp 7-32. (30) Mizuike, A. Enrichment Techniques for Inorganic Trace Analysis; SpringerVerlag: New York, 1983. (31) Gorsuch, T. T. Accuracy in Trace Analysis: Sampling, Sample Handling, and Analysis; National Institute of Standards and Technology: Gaithersburg, MD, 1976; pp 491-507. (32) Weast, R. C., Ed. CRC Handbook of Chemistry and Physics, 54 ed.; CRC Press: Cleveland, OH, 1973. (33) Dean, J. A., Ed. Lange’s Handbook of Chemistry, 12 ed.; McGraw-Hill: New York, 1979. (34) Kingston, H. M. Method and Apparatus for Microwave Assisted Chemical Reactions; US Patent 5,830,417, 1999.

Figure 3. Model of the amount of microwave power absorbed by various sample masses of water, 1 M HCl, and 29 M HF.9

transparent fluoropolymer vessel allows the energy to interact directly with the sample, where mineral acids rapidly convert the incoming microwave energy to heat. When evaporation of the solution matrix begins and a portion of the gas phase is permitted to escape the system, the matrix volume (mass) will decrease. Comparisons of the amount of microwave power absorbed among various sample masses of the same absorber have shown that as the mass of sample decreases, the amount of power absorbed by the sample also decreases.19,24-26,28,29 This relationship is based on the following equation, which relates the power absorbed to the mass of sample inside the microwave cavity:

Pabst ) KCp∆Tm where Pabs is the apparent power absorbed by the sample in watts, t is the time of irradiation, K is a constant relating calories‚s-1 to watts, Cp is the heat capacity of the sample, ∆T is the change in temperature, and m is the total mass of sample in the microwave cavity. Figure 3 shows the decrease in power absorbed by various samples as the quantity of the sample decreases. The power absorbed falls sharply below about 50-25 g of total mass in the microwave cavity. When extrapolated for smaller volumes, the data predict that absorption will theoretically fall to zero. During the evaporation process, this unique relationship between sample mass and microwave absorption leads to a decrease in temperature, which is a fundamental parameter controlling species volatility. This concept is illustrated in Figure 2. Despite using a constant irradiated power, as the sample volume decreased from 10 to 3 mL (the smallest volume that was accurately measurable), the temperature decreased by nearly 30 °C. The extent of power absorption for these small masses had not previously been experimentally verified. The temperature difference at dryness between microwave evaporation and hot plate evaporation is projected to be more than 150 °C. By lowering the temperature at dryness, potential loss of volatile species is avoided and more complete recovery of volatile analytes is achieved. Table 2 shows the boiling points of some potentially volatile chloride species. Some species that have been documented to suffer from volatility problems during evaporation include AsCl3 and SbCl5, with boiling points of 130.2 and 79 °C, respectively. During the hot plate evaporation, where the temperature exceeded Analytical Chemistry, Vol. 72, No. 13, July 1, 2000

2911

Table 2. Boiling Points of Selected Chloride Species32,33 compound

bp (°C)

compound

bp (°C)

AsCl3 SeCl4

130.2 170-196a

SbCl5 SnCl4

79 115

a

Table 3. Percent Recovery of Analytes after Partial Evaporation of a Multielement Standard in 10% HCl to ∼3 mL Using Hot Plate and Microwave-Assisted Evaporation, and Percent Recovery of the Same Solution Evaporated to Dryness Using Each Methoda partial evaporation (to ∼3 mL)

SeCl4 sublimes in the temperature range given.

170 °C, losses of these elements are likely. On the other hand, during the microwave-assisted evaporation, where the temperature decreased as the evaporation approached dryness and the boiling points of these volatile species are not reached, losses of these species are avoided. The relationship between sample mass and percent power absorbed has been observed by many authors. However, this observation has often been made in the context of protecting the magnetron from damage by the increased amount of power being reflected from the microwave cavity back toward the magnetron.29 This study predicts that this relationship can be used as an advantage for the retention of volatile inorganic species during the evaporation process. This relationship indicates that the sample volume will decrease to a point where no effective coupling between the incoming microwave energy and the sample occurs, as predicted by extrapolation of the Figure 3 plot from 25 to 0 g. At this point, no heating will take place and the evaporation process will stop. In practice, we observe this volume to be a single drop of solution of ∼0.1 g or less. For this reason, the special vessel closure and fume scrubber/vacuum pump were used to reduce the pressure inside the vessels during evaporation. Evacuation of the vessels actively cools the system by removing the energetic gas molecules, which minimizes the surface heating of the vessel walls by the hot gas. In addition to allowing the gaseous solvent molecules to escape the system, the reduced pressure lowers the effective boiling point of the solvent, allowing the complete evaporation of the final microquantity of the solution from the vessel bottom. While reducing the pressure also lowers the boiling points of the analytes, losses have not been observed, as will be shown. Subsequent studies have shown that allowing a stream of gas to flow through the vessel promotes more rapid evaporation. This is done by allowing one of the plugged ports to remain slightly open or by actively pumping a gas into the vessel while evacuating from another port. The vessel closure designed for this study can accommodate this application, which is appropriate for evaporation of larger sample volumes, as well. Partial Evaporation of Standard Solution. To test the theory that overheating at dryness results in losses of volatile analytes, a partial evaporation of a standard solution in 10% HCl was performed. Table 3 shows the recovery for a partial evaporation, using both hot plate and microwave, where a small portion (∼3 mL) of the solution was allowed to remain. For all analytes tested, recoveries of 98% or better, within the 95% confidence interval, were achieved, using either the hot plate or microwave evaporation methods. Excellent recovery was demonstrated, even for As, Sb, and Sn, which are notoriously volatile chloride species, because the ions remain solvated and the temperature does not increase out of control. This supported our preliminary hypothesis that allowing a small amount of solution to remain would prevent the salts from overheating. 2912 Analytical Chemistry, Vol. 72, No. 13, July 1, 2000

total evaporation (to dryness)

element

hot plate

microwave

hot plate

microwave

As Cr Cu Mn Ni Sb Se Sn V

97.5 ( 2.5 98.0 ( 1.0 98.1 ( 1.5 97.8 ( 1.8 97.2 ( 2.8 99.6 ( 1.4 95.3 ( 2.5 100 ( 2.9 97.5 ( 1.7

102 ( 4.1 101 ( 4.7 104 ( 4.0 101 ( 3.5 103 ( 4.2 104 ( 4.7 104 ( 4.9 104 ( 3.3 102 ( 4.5

79.3 ( 5.5 103 ( 3.6 105 ( 1.1 101 ( 1.4 102 ( 0.9 83.0 ( 2.3 31.1 ( 9.4 64.5 ( 5.8 97.5 ( 2.9

94.6 ( 7.0 104 ( 0.5 99.9 ( 0.7 105 ( 1.0 106 ( 1.7 101 ( 3.0 89.5 ( 9.7 104 ( 1.9 91.9 ( 6.5

a

3.

Uncertainties are expressed as 95% confidence intervals, with n g

Table 4. Concentration of Analytes (µg/g) in SRM 2710 Following the Microwave-Assisted Evaporation of the Digestate Compared with the Certified Total Concentrationsa element

microwave-assisted evaporation

certified concn

Ag As Co Cr Mo Sb V

31.6 ( 4.1 685 ( 34 10.9 ( 2.0 31.7 ( 2.7 18.1 ( 1.7 35.6 ( 5.2 73.5 ( 3.7

35.3 ( 1.5 626 ( 38 10b 39b 19b 38.4 ( 3.0 76.6 ( 2.3

aUncertainties are expressed as 95% confidence intervals, with n g 3. b Certificate values for Cr, Co, and Mo are reference values and do not have an associated uncertainty.

Total Evaporation of Standard Solution. When the standard solutions were taken to dryness, the hot plate method overheated the dried salt residue at the beaker bottom. As stated previously, losses of As and Sb, and others, could be expected because of their relatively low boiling points. The data in Table 3 show that overheating of the salts during the hot plate evaporation does lead to the volatilization of several elements. Losses of 20% for As and Sb and over 70% for Se occurred during the hot plate evaporation. In contrast, the microwave-assisted evaporation gave excellent recoveries, greater than 98%, for these elements. This study provides evidence that complete recovery of volatile analytes, such as As, Sb, and Se, is possible using microwave-assisted evaporation. Evaporation of SRM Digestates. Microwave-assisted evaporation was tested to determine the recovery of analytes from actual digested samples. These matrixes represent some of the difficulties discussed previously, such as high concentrations of HCl and HF. SRM 2710 Montana Soil, a highly contaminated soil, was totally decomposed using HNO3 and HF. SRM 1566a Oyster Tissue, was totally decomposed using HNO3. Concentrated HCl was added following each digestion for complexation purposes and the solution subsequently evaporated to dryness. Tables 4 and 5 show the results of analysis following the microwave-assisted evaporation of SRM 2710 and SRM 1566a digestates, respectively.

Table 5. Concentration of Analytes (µg/g) in SRM 1566A (Oyster Tissue) Following the Microwave-Assisted Evaporation of the Digestate Compared with the Certified Total Concentrationsa element

microwave-assisted evaporation

certified concn

Ag As Cd Co Cr Hg Mn Ni Pb V

1.53 ( 0.04 13.8 ( 2.2 4.13 ( 0.24 0.49 ( 0.02 1.76 ( 0.03 0.063 ( 0.033 11.5 ( 0.5 2.29 ( 1.02 0.38 ( 0.07 4.59 ( 0.27

1.68 ( 0.15 14.0 ( 1.2 4.15 ( 0.38 0.57 ( 0.11 1.43 ( 0.46 0.0642 ( 0.0067 12.3 ( 1.5 2.25 ( 0.44 0.371 ( 0.014 4.68 ( 0.15

a

3.

Uncertainties are expressed as 95% confidence intervals, with n g

As shown in Table 4, the results of analysis for the soil digestate are statistically equivalent to the certified concentrations at the 95% confidence interval, even for As and Sb. Table 5 shows that the results for the oyster tissue sample, whose elemental composition consists of most analytes below a level of 10 µg/g, are statistically equivalent to the certified values at the 95% confidence interval, even for As, Hg, and V. Using hot plate evaporation for these matrixes, losses of several volatile elements were experienced as is well-documented in the literature. Microwave-assisted evaporation demonstrated effective recovery from two matrixes of vastly different composition, with analytes ranging from the near-percent level down to the submicrogram per gram level. Conclusions. The results of this study show that microwaveassisted evaporation is an effective method for the complete recovery of many analytes that have previously been documented

to be volatile. The basis for the differences in the chemistry reported here and the chemistry reported for the last several decades lies in the fundamentally different heating mechanisms involved in the two methods. The concept that the sample temperature actually decreases as the evaporation proceeds using microwave energy is not well-recognized. The benefits that result from this unique concept have minimized the volatility concerns of many analytes when evaporating solutions containing HCl and/ or HF to dryness, as demonstrated by this study. Applications of this technique may establish a new analytical tool for minimizing analyte volatility using the unique mechanisms of the microwaveenhanced evaporation process. Fundamental volatility chemistries, further developments, and new applications of this technique will be posted on Duquesne University’s sampleprep website.21 Many standard sample preparation procedures should be reevaluated in light of this new mechanism. Use of these new capabilities provides an opportunity to optimize trace elemental preparation procedures by employing nontraditional microwave instrumentation that previously was unavailable to the analytical laboratory. This capability is needed as ultratrace sample preparation procedures are required for an ever-increasing number of sensitive instrumental techniques. ACKNOWLEDGMENT The authors thank Los Alamos National Laboratory for funding a portion of this research. We also thank Milestone USA for equipment support. Portions of this research and apparatus described herein are covered by U.S. patent 5,830,417, and other patents are pending. Received for review November 29, 1999. Accepted February 11, 2000. AC991369D

Analytical Chemistry, Vol. 72, No. 13, July 1, 2000

2913