JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER I99 1 VOL. 6 457 Rapid Stopped-flow Microwave Digestion System Vassili Karanassios,* F. H. Li B. Liu and Eric D. Salint Department of Chemistry McGill University 80 I Sherbrooke Street West Montreal Quebec H3A 2K6 Canada A prototype system for stopped-flow microwave assisted wet digestions has been developed. A coiled Teflon PFA tube serves both as a sample container and as a digestion vessel. A sample plug consisting of a water slurry mixed with an acid mixture is pumped into the coil. Sample flow is stopped the coiled tube is sealed (by closing an input and an output valve) and microwave power is applied for 2 min for digestion of the sample. Methodology was developed using powdered botanical reference samples and was tested with powdered botanical and biological reference materials.The digests were analysed by inductively coupled plasma atomic emission spectrometry. In addition to comparisons with certified values the results were compared with those obtained by conventional open-vessel hot-plate digestions by open-vessel microwave digestions and by digestions taking 32 min by using the coiled tube system. Precise and in many instances quantitative digestions were obtained using a net digestion time of 2 min. Elemental recoveries were sample type and digestion time dependent and were found to be comparable with and sometimes superior to those obtained when using a 3 h long hot-plate digestion. In this preliminary study characteristics limitations and future directions are discussed. Keywords Microwave digestion; flow system; elemental analysis; powdered sample; inductively coupled plasma atomic emission spectrometry Microwave assisted wet digestions offer an alternative to traditional (i.e. open vessel hot-plate) time-consuming sample dissolution techniques. Since the first description’ of the use of microwave radiation as an energy source in acid the method has attracted considerable attention and has been successfully applied to a variety of sample types. Included in these are botanical biological and geological materials and foodstuffs. For more details the book by kngston and Jassie3 and recent reviews4y5 may be consulted. Although open- and closed-vessel acid digestions have been developed the high-pressure sealed-bomb approach is the most widely The success of bomb digestions is exemplified by the range of applications reported and the availability of commercial instrumentat ion.7 4 In spite of the advantages offered by closed-vessel digestions sample preparation still remains a multi-step and labour-intensive procedure. The labour involved per digestion is the same whether Parr bombs are heated in electrical furnaces for hours or sealed Teflon vessels are exposed to microwave energy for minutes. In an attempt to increase sample throughput further reduce labour and cost and facilitate automation wet digestions can be forced to occur as a sample stream slowly flows through a microwave oven. This is an approach that gives rise to the concept of flowing stream digestions which has been successful with blood samples.’ These were digested in 30 s while slowly flowing through a microwave oven.’ In order to accommodate powdered samples that require much longer digestion times (i.e.32 min) the flow may be interrupted for a period of time resulting in stopped-flow digestions. Ease of automation is perhaps the most important advantage of this approach. The goal of this work is the development of stopped-flow digestion instrumentation and simple methodology for the dissolution of powdered samples. Of particular interest are rapid (ie. less than 5 min) and reproducible extractions of ‘environmentally available’ elements* in botanical samples of environmental concern. In our system a Teflon PFA tube serves both as a sample container and a digestion vessel.Besides the advantages mentioned above this closed-vessel approach minimizes the risk of sample cross * Present address Department of Chemistry University of t To whom correspondence should be addressed. Waterloo Waterloo Ontario N2L 3G1 Canada. contamination when digesting multiple samples simultane- ously and of acid fumes attacking the oven components. Therefore it eliminates the need for special precautions such as oven coatings acid scrubbers or evacuated oven chamber^.^ It also offers additional safety protection to personnel by limiting exposure to hazardous acid fumes and by minimizing reagent handling. Experimental Instrumentation A schematic diagram of the instrumentation required for an ‘ideal’ stopped-flow microwave digestion system is shown in Fig.1 . The prototype system developed in this work consists of a peristaltic pump two high pressure valves a microwave oven a ‘tube assembly’ a pressure gauge and a temperature transducer. A list of equipment suppliers is provided in Table 1. The safety precautions required for routine operation of the ‘ideal’ system shown in Fig. 1 for example installation of waveguide attenuators in all access ports and of a pressure release valve are currently being addressed. The heart of this system is a conventional microwave oven. A slightly modified domestic-type commercially avail- able microwave oven was used. The oven has an internal volume of 0.9 ft3 (25.5 1) and operates at the standard 2.45 GHz frequency. The power is adjustable in 9 steps from 1 (‘low’ corresponding to a power of approximately 72 W) to 9 (‘high’ or ‘full’ which is equivalent to about 720 W) in I I Pump I tube I I Sample I oven I Perista I t ic I collector Waveguide Waveguide pump I attenuator attenuator Fume hood Slurry 1 Schematic diagram of an ‘ideal’ stopped-flow digestion L - - - - - - - - - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Fig.1 system458 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1991 VOL. 6 Table 1 List of instrument suppliers Digestion system- Microwave oven Microwave meter Pump tubing Digestion tube Pressure gauge ICP spectrometer- Spectrometer Spray chamber Nebulizer Peristaltic pump Toshiba Model EXR- 1690C Toshiba Minato-Ku Tokyo Japan Holaday Model HI-1 800 Holaday Industries Eadon Prairies MN USA Mandel Scientific Guelph Ontario Canada Teflon PFA Cole-Palmer Chicago IL USA Cole-Palmer Thermo Jarrell-Ash Model ICAP-6 I Franklin MA USA Technical Service Laboratories Ontario Canada Technical Service Laboratories Gilson Miniplus3 Model 3 12 Gilson Medical Electronics Middleton WI USA increments of 81 W. The power delivered from the magnetron to the oven was not calibrated for this work and was assumed from the manufacturer’s specifications.Partial power is delivered to the oven by automatically adjusting the amount of time full power is applied to the magnetron. For example for this oven and for power level 2 full power is applied to the magnetron for about 3 s power is then turned off for about 12 s and this 15 s cycle is repeated often as required to complete the desired heating time (i.e.2 min). The heating time is programmable via front panel push- buttons and can be set from 1 s to 99 min 99 s. The oven was modified by placing an electrical fan on its side as shown in Fig. 2. This was done in order to vent hot air during operation and to help cool the tube at the end of a digestion. In addition two 3/8 in ( e 9 . 5 mm) holes were drilled in the back of the oven. These serve as entrance and exit ports for the digestion tube. As a further safety precaution the oven was operated inside a fume hood. Two types of tubing constitute the ‘tube assembly’. These are identified as ‘pump tube’ and as ‘digestion tube’ in Fig. 1. The pump system consists of two conventional peristaltic pump tubes with an internal diameter (i,d.) of 0.090 in (-2.3 mm) and an outer diameter (0.d.) of 0.1 575 in (-4.0 mm).These are connected to the digestion tube with a T- joint. Slurry and acid(s) are pumped into the digestion tube (Fig. 1) utilizing two channels of a six channel peristaltic pump- A Teflon PFA tube (perfluoroalkoxy a chemically inert non-porous tetrafluoroethylene with a fully fluorinated alkoxy side chain7 with an 0.d. of 1/4 in (-6.35 mm) and an i d . of 5/32 in (-4.0 mm) serves as the digestion tube. According to manufacturer’s specifications it can withstand pressures of 42 1 psi (1 psi-6.894 x lo3 Pa) and tempera- tures of 260 “C. The tube 13.8 ft ( ~ 4 2 0 cm) in length between the input and output valves was coiled in six turns (the diameter of the inner turn of the coil was about 9 cm) and was placed facing the magnetron as shown in Fig.2. This arrangement was chosen in order to take full advan- tage of the power delivered from the magnetron to the sample. A sample plug is pumped into the centre of the coil leaving about 50 cm of air on both ends of the tubing. No Slurry and acid Waveguide in Solution out \ Grounded fan screen 3‘ ietron Fig. 2 Actual microwave oven and coiled-tube set-up portion of the sample acid slug extends outside the microwave cavity. During digestion the sample slowly rotates inside the coiled tube either clockwise or anti- clockwise due to pressure differences developed on either side of the sample plug. Fortuitously this rotation serves as a stirring mechanism and also helps to reduce the effects of non-uniform heating due to ‘hot spots’.These are due to inhomogeneous microwave fields and are typically ob- served in domestic microwave oven^.^^^ Furthermore ow- ing to self-rotation the need for a sample container rotating device for example a rotating carousel typically used with conventional sealed-vessel digestion system~,~-~ is also eliminated. After repeated use for example about 100 times the external wall of the digestion tube showed visible signs of ageing such as yellowing particularly at the T- and valve- joints and was replaced for safety reasons. In addition a dark yellow-brown coating and black spots were visually observed inside the tube in particular in tube segments that are near the input and output ports and in the interior of the valves and the tube joints. Because these were primarily observed in parts of the tube assembly that are the furthest from the microwave energy flux or are external to the oven they were attributed to undigested sample.Some preliminary results indicate that the tube may have to be replaced more frequently owing to accumulation of undi- gested material (which may give rise to memory effects) rather than ageing or thermal/mechanical stress. The peristaltic pump tube was replaced depending on the work load about once a week. All measurements were performed using a Thermo Jarrell-Ash Model ICAP-6 1 inductively coupled argon plasma optical emission spectrometer. The ICAP-61 is a 0.75 m 34 channel polychromator system providing a spectral resolution of 0.048 nm (in the first order) and is capable of determining 32 elements simultaneously.The spectrometer was controlled by an IBM PC-compatible microcomputer using standard ThermoSpec software. A spray chamber a pneumatic V-type nebulizerlo and a peristaltic pump were used to introduce the digested samples into the inductively coupled plasma (ICP). Typical operating conditions were 1 .OO kW applied forward power 16 and 1.0 1 min-l for coolant and auxiliary argon flow rates respectively. The solution uptake rate was 2.0 ml min-l and the observation height was 15 mm above the load coil. Reagents and Procedures Standards reagents and standard reference materials All standards were prepared by serial dilution with distilled de-ionized water (1 8 MR cm-l specific resistivity prepared by feeding in-house distilled water through a Millipore purification system) of 1000 ppm (certified to k 1%) standard stock solutions (standards for atomic absorption spectrometry) of the respective element.Baker Instra-JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 1 VOL. 6 459 Analyzed inorganic acids were used throughout. A list of suppliers of chemicals and samples is given in Table 2. The methodology developed in this work relies heavily on the use of botanical reference materials provided by the Ontario Ministry of the Environment (MOE Table 2). Included in these are those designated as V85-1 Norway maple and White birch by the MOE. Particle sizes averaged 160 pm. Although there are no certified values for these samples the concentrations reported here are the average quoted by the MOE.These were obtained using a 3 h open- vessel hot-plate acid digestion procedure and were analysed by ICP atomic emission spectrometry (AES)." The MOE1l and other environment monitoring laboratories8 are inter- ested in trace elements that are environmentally available (i.e. not bound) so their digestions are typically not brought to completion. Their procedure 'quantitatively and repro- ducibly recovers most to all of the more toxic transition elements'. The methodology was tested using standard reference materials (SRMs). These include National Institute of Standards and Technology (NIST) (formerly National Bureau of Standards Table 2) Bovine Liver (SRM 1577) and Orchard Leaves (SRM 1 57 1). Certified values reported here were obtained either from certificates of analysis or from literature r e p o r t ~ .~ J ~ All concentrations reported here are the average of five repeats and refer to dried samples. Digestion procedure Approximately 0.35 g of powdered sample is accurately weighed in a glass beaker and 20 ml of de-ionized water are added. The resultant slurry is continuously and rigorously stirred and is pumped with the input and output valves open (Fig. I) into the digestion coil simultaneously with about 20 ml of an acid (i.e. HC1 or HN03) or an acid mixture [ie. aqua regia HN03+HC1 (1 +3 v/v) or HN03+H202 (4+ 1 v/v)]. The valves are then closed and the sample is heated either at high power for 2 rnin or using a 'power programme' for a pre-determined period of time for example full power for 2 rnin immediately followed by power level 5 for 3 min.After digestion the sample is allowed to air cool for several minutes. The output is then opened to relieve the pressure and the effluent is quantita- tively transferred into the collection vessel. Samples with residues were filtered (using a Whatman filter paper No. 42) the filter was washed with distilled water and the filtrate was diluted to volume (i.e. 50 ml). The solutions were then analysed using an acid-containing blank. Acidified multi-element standard solutions contain- ing the elements of interest were used for calibration. Depending on sample type power programme and digestion time the digest may contain an unspecified amount of residue. The amount of residue remaining in the digest and the digestion tube increases with decreasing digestion time for a given power programme.The residue which was attributed to undigested organic material such as amino acid^,^>^ does not typically contain any trace ele- Table 2 List of suppliers of reagents. SRMs and samples Standards and reagents- Standards Acids NIST SRMs Fisher-Scientific Fair Lawn NJ USA J. T. Baker Phillipsburg NJ USA National Institute of Standards and Technology Gaithersburg MD USA Ontario Ministry of the Environment (MOE) Rexdale Ontario Canada SRMs and samples- MOE samples m e n t ~ . ~ > ~ J ~ However unless the material is allowed to settle before analysis it may clog the nebulizer thus necessitating an extra filtration step. In addition because some of the material adheres to the inner walls of the tube the input and output valves and the T-joints trace elements accumu- late and eventually give rise to memory effects.When sequentially digesting the same sample type undigested material was removed from the tube by a high- speed flush with a water plug. The digestion tube was then cleaned by pumping through an acid plug followed by multiple rinses with a water plug. The total time for the digestion cooling and cleaning cycles was about 5 min. When changing sample types the tube was cleaned by exposing a water-acid mixture to microwave energy for 2 rnin followed by multiple rinses with a water plug. The total time for this cleaning cycle was about 5 min. These procedures reduce analyte concentrations in blanks to levels below the detection limit of the ICP spectrometer.For the open-vessel work approximately 0.35 g of sample was accurately weighed in an Erlenmeyer flask and mixed with 20 ml of distilled de-ionized water and 20 ml of acid. The flask covered with a watch-glass was placed in a Pyrex vacuum desiccator (without the desiccant) which was subsequently placed in the microwave oven where it was exposed to full power for a net exposure time of 6 min. Power was applied in 2 min intervals followed by a 2 rnin cooling period for a total time of more than 10 rnin per digestion. In order to vent acid fumes and to provide additional protection of the oven and its associated elec- tronics the vacuum port in the lid of the desiccator was connected to Tygon tubing. The tube was routed through one of the holes drilled in the back of the oven to the fume hood.An open 100 ml beaker containing 50 ml of water was placed inside the oven in order to protect the magnetron from reflected power17*J2J4 and to ensure the use of experimental conditions that are the same as those reported by other w o r k e r ~ . ~ J ~ Safety Owing to the use of potentially hazardous microwave energy and strong acids at elevated pressures and tempera- tures safety was a key consideration in operating this system. The following safety criteria were set minimum radiation leakage (ie. 1 mW cm-2) measured at 5 cm from the oven15 or any tube or cable emerging from it,3 no acid fumes in the microwave cavity and maximum operating pressure and temperature of 125 psi and 230 "C. Although the tube assembly was safely tested and briefly operated at pressures as high as 200 psi the choice of a lower operating pressure for routine operation was dictated by the weakest links in the tube assembly more specifically the tube to valve and the tube to T-joint connections (Fig.1). Accord- ing to the manufacture's specifications this is only 125 psi. Microwave oven Owing to the modifications to the oven (ie. fan duct and access holes) radiation leakage was an important concern. In addition the tube emerging from the oven might act as an antenna transmitting microwave radiation. Microwave leakage was extensively tested using a sensitive microwave power meter. The measured values were found to be below the levels mentioned previously typical values were below 0.2 mW cm-2. Temperature Unlike other workers who measured temperature in real- time during the course of digestion^,^^^^^ the initial concern of this study was the final temperature of the samples rather than a temperature versus time profile.The final tempera-460 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1991 VOL. 6 ture was determined in order to avoid thermal degradation of the tube. Temperature was measured using a copper-constantan thermocouple encapsulated within a thin layer of glass and was held in place in the middle of the digestion tube with a T-joint (Fig. 1). The thermocouple leads were attached to a shielded grounded cable. Regardless of power programme sample type and digestion time the final temperature was below the safety limit. This is as expected considering the maximum allowable operating pressure (1 25 psi) and the boiling-points of the acids and acid mixtures used in this work.Pressure Pressure was monitored in real-time using a conventional pressure gauge in order to establish the operating conditions at which the tube assembly can be operated safely. Readings were taken manually at 1 min intervals. The results for various acids 0.35 g of V85-1 botanical sample and full power are shown in Fig. 3. The steep increase in pressure observed when a mixture of HNO and H202 is utilized indicates rapid formation of gaseous products and al- though it may have to be correlated with temperature and gaseous H202 de-composition products it also suggests a faster attack of botanical samples. However it also limits the digestion time to less than 2 min.From these results it can be concluded that the maximum time for which V85-1 botanical samples can be safely digested at full power is less than 3 min. In summary pressure rather than temperature is the limiting parameter in digestions involving botanical samples in nitric acid or its mixtures as has also been reported by other worker^.^^^^^ As mentioned earlier depending on power programme and digestion time an unspecified amount of undigested organic material remains in the digestion tube and settles in the collection vessel. This necessitates an extra time- consuming and labour-intensive filtration step. In order to eliminate this step and to obtain more complete digestions (i.e. those that result in clear and colourless solutions with no visible signs of residue) longer digestion times may have to be used.250 200 .- 150 \ 2? cn 100 L 50 0 1.00 2.00 3.00 4.00 5.00 6.00 7.00 Digestion ti melmi n Fig. 3 Digestion tube pressure versus heating time for MOE V8 5- 1 botanical samples high power and different acids (see text for discussion). A HN03+H202; B aqua regia; and C HNO,. Horizontal broken line is the safety limit Results and Discussion From the preceding discussion it can be concluded that microwave energy effectively couples to acidified slurries in the digestion tube. The key question then becomes what is the best set of operating conditions (i.e. power exposure time and acid mixture) that will provide safe rapid and complete digestions? Digestion Time and Safety Pressure Digestion times can be extended without exceeding the safety pressure.For example as shown in Fig. 4 a digestion time of 5 rnin can be achieved by applying full power for 2 rnin followed by the application of power level 5 for 3 min. In order to reduce the amount of undigested material remaining in the tube further digestion times can be extended to over 30 rnin by applying full power for 2 min followed by continuous application of power level 3 for 30 min. Almost clear and lightly coloured solutions signifying more complete digestions have been obtained using this programme V85-1 samples and aqua regia. Acid Mixtures As has been amply demon~trated,~*~>~J~ a variety of acid mixtures can be used for the destruction of organic matter.16 In this work HNO HC1 a mixture of HNO and H202 and aqua regia were tested with V85-1 botanical samples with various degrees of success.Because aqua regia resulted in the least amount of undigested material (on visual inspection) in the shortest time without exceeding the safety pressure it was the reagent of choice and was used throughout. Even with aqua regia and the use of a power programme complete digestions require over 30 min an excessively long time for a rapid digestion system. However it is questionable whether complete digestions are necessary in order to obtain 100°/o recovery of trace elements in botanical samples. Elemental Recoveries and Digestion Time In order to address this question V85-1 botanical samples were digested with aqua regia for 2 4 8 16 and 32 min. The digested samples were analysed for Al Ba Cd Cu Fe Mg Mn and Zn. The average recovery was found to be about 100% irrespective of digestion time. Therefore trace elements are fully recovered in 2 rnin when digesting V85-1 samples at full power.However in order to substantiate this conclusion further recoveries of individual elements must be investigated. The effect of the matrix (i.e. sample type) on elemental recoveries will be examined subsequently . In order to obtain ‘total’ digestions and to establish a ‘reference’ value MOE V8 5- 1 botanical slurries were digested for 32 rnin by applying a power programme. These results their respective MOE values and an inter-compari- son (presented as Yo recovery with respect to the MOE values) are shown in Table 3. With the exception of A1 and Ba which show relatively high recoveries with respect to the MOE values the average concentrations and the standard 200 180 - 160 - .- 60 I I I - 2.0 3.0 4.0 Digestion time/min 5.0 Fig.4 Tube pressure at various power levels values in paren- theses are applied power in W A 9 (720); B 8 (639); C 7 (558); D 6 (477); E 5 (396); F 4 (315); G 3 (234); and H 2 (153)JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 I VOL. 6 46 1 Table 3 Analysis of 0 5 - 1 vegetation samples Concentratiodpg g- Recovery (O/O) Element A1 Ba Ca c u Fe Mg Mn Zn Coiled Coiled tube L* tube St 1 1 4 k l l 5 4 k 5 15k0.9 13k0.4 12 000 2 550 1 2 k 2 14-t 1 234-t 15 255-t 13 13 000 + 400 2 3 0 0 k 71 2300k88 61 + 2 67 f 2.5 1 5 0 f 8 140 f 6.4 *L Long 32 rnin digestion time. t S Short 2 rnin digestion time.$Average I all elements. §Average 2 excluding A1 and Ba. MOE value l o o k 11 10k0.7 12 000 k 580 13k0.8 230k 18 2 1 O O k 130 6 0 k 4.6 140k11 Average 1$ Average 24 Tube L MOE 114 150 108 92 102 109 102 107 110 103 Tube S MOE 54 130 100 108 111 109 111 103 103 107 Tube L tubes 21 1 115 108 86 92 100 91 114 115 99 deviations (SDs) reported here are in agreement with the values quoted by the MOE; elemental recoveries for most elements tested were about 100% (Table 3). In order to test for the recovery of trace metals when short digestion times are used V85-1 samples were digested for 2 rnin at high power. The results are also shown in Table 3. With the exceptions of Al which is under-recovered and of Ba which is over-recovered (both with respect to the MOE values) the concentrations (Table 3) and SDs are in agreement with those quoted by the MOE.A comparison of elemental concentrations obtained when using long ( i e . 32 min) and short (ie. 2 min) digestion times suggests that recovery of A1 depends on digestion time. This was verified by digesting V85-1 samples for 2 4 8 16 and 32 min. The results are shown in Fig. 5. The recovery of A1 increases with digestion time and plateaus at about 16 min. From these results it can be concluded that the hot-plate digestions used by the MOE under-recover A1 and that an increase in digestion time increases its recovery with this system. Similar conclusions can also be drawn for Ba (Table 3) an increase in digestion time is expected to result in a small increase in recovery.In order to substantiate further the results obtained when short digestion times are used and to provide a basis for compa~-ison,~J* V85-1 samples were digested in the micro- wave oven using open vessels. It is worth mentioning that when using open vessels about 100°/o recoveries are obtained from botanical and biological samples in less than 6 min.3J2 The recoveries obtained when using short digestions with the tube system and open vessels show striking similarities. In both instances [Fig. 6(a) and (b)] A1 is under-recovered Ba is over-recovered as are all other 120 I 1 40 ' I I I 1 I I I 0 5 10 15 20 25 30 35 Digestion time/min Fig. 5 Effect of digestion time on A1 recovery from MOE V8 5- 1 botanical sample using aqua regia and coiled tube digestion elements.From the results shown in Fig. 6(c) it can be concluded that recovery in the tube system in 2 rnin is almost as much as in open vessels in 6 min thus demonstrating the validity of short digestion times in a tube-based system. A comparison of short and long digestion times reveals that with the exception of A1 and Ba short digestion times provide quantitative recoveries for most elements tested the average recovery of Ca Cu Fe Mg Mn and Zn was 99% (Table 3). A comparison of short digestion times with the MOE values also reveals that except for Al the tube digestion system recovers more in 2 rnin than the hot-plate method does in hours thus providing substantial time savings. From the low SDs it can be concluded that highly reproducible digestions are obtained with this system as has also been reported for other microwave digestion ~ y s t e m s ~ * ~ - ~ Therefore for an element that is difficult to digest such as Al a correction factor may be applied with a reasonable degree of confidence. This is of prime impor- tance in environmental monitoring where total digestions are not required but reproducible recoveries short sample preparation times and straightforward and rugged digestion procedures are crucial.8 These demands have been met by the system and the methodology described here. In summary from the results presented thus far it can be concluded that for V85-1 a 2 rnin digestion provides elemental recoveries of about 100% for most elements. This is further documented and the effect of sample matrix on individual elemental recoveries is investigated by consider- ing other MOE botanical samples and NIST botanical and biological SRMs.Analysis of MOE Botanical Samples and NIST SRMs In order to evaluate 2 rnin digestions further this study was expanded to include other MOE botanical samples and NIST botanical and biological materials. The results for the botanical samples are shown in Table 4 for MOE White birch and Norway maple and in Table 5 for NIST SRM 1571 (Orchard Leaves) and NIST SRM 1577 (Bovine Liver). It should be borne in mind that a key objective in this work was elemental extractions which are on average compatible with those obtained when using hot-plate digestions. Based on previously drawn conclusions for V85- 1 poor recoveries were expected for A1 [Fig.6(a)]. This is so for all462 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1991 VOL. 6 I5O;1 5 ; r ; y j 15;;l 100 100 100 50 50 50 0 Al Ba Ca Cu Fe Mg Mn Zn Al Ba Ca Cu Fe Mg Mn Zn Al Ba Ca Cu Fe Mg Mn Zn L- 100 100 > 50 50 50 CC 150 100 50 Al Ba Ca Cu Fe Mg M n Zn " Al Ba Ca Cu Fe M g Mn Zn Al Ba Ca Cu Fe Mg Mn Zn ( i ) 100 100 = m . . . . . di B'a C'a C l Fk h;g d n Element Fig. 6 Elemental recoveries for V85-1 Norway maple and White birch. (a) MOE V85-1 tube/MOE recovery; (b) MOE V85-I open- vessel/MOE recovery; (c) MOE V85-1 tubelopen-vessel recovery; (d) MOE Norway maple tube/MOE recovery; (e) MOE Norway maple open-vessel/MOE recovery; cf) MOE Norway maple tube/open-vessel recovery; (g) MOE White birch tube/MOE recovery; ( h ) MOE White birch open-vessel/MOE recovery; and (i) MOE White birch tube/open-vessel recovery ~~~ Table 4 Analysis of MOE Norway maple and White birch botanical samples Norway maple White birch Coiled Open MOE Coiled Open MOE tube/ vessel/ value/ tube/ vessel/ value/ Element Pg g-' Pg g-' P g g-' Pg g-' Pi? g-' Pg g-' A1 Ba Ca Cd c u Fe Mg Mn Pb Zn 130k 19 18 k 0.76 28 000 k 1 700 9.8 k 0.31 420 k 16 2 500 k 140 5 6 k 10 120 f 4.1 42 f 1.9 - 100 k 5.8 18 k 0.59 28000k 1 100 460k 13 2 6000 f 82 5 3 f 18 130 k 3.4 4 4 k 1.6 - 9.6 k 0.44 180 14 28 000 8 420 2 200 47 95 40 - 24 k 2.6 84 k 0.95 14 000 k 450 0.8 a 1 5.6 f 0.77 5.7 k 1.1 61 f 2 .5 68 k 2.4 2 500 k 67 24 k 2 100 k 3.4 16000 k 370 1 kO.07 2 100 k 75 611 .t 16 180 f 8.5 704 f 19 220 k 7.3 - - 42 86 14 000 0.9 5 59 2 100 600 200 - Table 5 Analysis of NIST SRM Orchard Leaves and Bovine Liver Orchard Leaves Bovine Liver Coiled tube/ Element Pg g-' A1 120k21 Ba 41 k0.96 Ca 21 OOOk650 Cd cu 11 k l .l Fe 240 & 13 Mg 6200k 150 Zn 25 20.63 - Pb 41 +15 * Reference 12. NIST value/ 410* 46.3* Pg g-' 20 900 k 300 1 2 k 1 300 k 20 6 200 k 200 25 k 3 - - Recovery (tube/NIST) (Yo) 29 89 100 92 80 100 100 - - Coiled tube/ 12.4 f 2.02 0.61 k O . l l 140k6.6 193k8.2 288 f 6.3 669 k 27.1 10k0.24 128 + 3.3 Pg g-' 1.01 k0.17 NIST value/ Pfz g-' 35.5* 12426 270 k 8 605 k 9 10.3 2 1 130k 13 1.24* - 193 f 10 Recovery (tube/NIST) (Yo) 35 50 113 100 107 111 97 98 -JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 1 VOL. 6 463 sample types analysed [Fig. 6(d) and (g) and Table 51.In addition a recovery higher than 100% was expected for Ba. However this is true only for Norway maple [Fig. 6(6)]. Barium was fully recovered from White birch [Fig. 6(g)] and it was under-recovered from both NIST samples (Table 5). In order to verify the results obtain using the tube system and short digestion times MOE samples were also digested in open vessels. Elemental recoveries are shown in Fig 6(e) for Norway maple and in Fig. 6(h) for White birch. Much like the tube digestion [Fig. 6(d)] similar recoveries are obtained when using open-vessel digestions for Norway maple [Fig. 6(e)]. For White birch however the observed recoveries are considerably different [Fig. 6(g) and (h)]. No explanation can be offered at this time for this discrepancy. By comparing elemental recoveries obtained by the tube system with those of open vessels some subtle points begin to emerge.For example with the exception of Al both digestion procedures provide about equal recoveries for all elements for Norway maple [Fig. 6 0 1 . For White birch [Fig. 6(i)] open-vessel digestions recover more than the tube-based system suggesting that longer digestion times may be required for this sample. The data shown in Fig. 6(c) (f) and (i) indicate that the digestion time required for White birch [Fig. 6(i)] is longer than that required for V85- 1 [Fig. 6(c)] and Norway maple [Fig. 6 0 1 . This is much like other microwave digestion systems3y4 in which elemental recoveries are sample type dependent. For example the low Fe recovery obtained when digesting Orchard Leaves (Table 5 ) could be attributed to the siliceous material present in the leaves and the low Ba recoveries (Table 5) could be attributed to sulphur-containing species.The latter may produce sulphate which causes Ba to precipitate. Clearly more work remains to be done to substantiate these conclusions further and to establish the extent of the dependence of elemental recoveries on sample type and on operating system parameters. Conclusions Precise and with some exceptions quantitative extractions of trace elements in botanical and biological samples of environmental concern are obtained rapidly using straight- forward methodology simple and inexpensive instrumenta- tion and net digestion times of 2 min. Recoveries were found to be comparable to and sometimes superior to those obtained when using a 3 h hot-plate digestion.Elemental recoveries were found to be sample type and digestion dependent. The present shortcomings of the system stem from the use of fairly large volumes of acids and from delays arising during cooling of the digests and cleaning of the tube. These are under study and appear to be easy to solve. It is worth pointing out that if the tube is not thoroughly cleaned memory effects become an important consideration. The additional safety precautions required for routine oper- ation for example installation of waveguide attenuators in all access p0rts~9~ and of a pressure release valve,17J8 are being addressed as shown in Fig. 1. In the future the dependence of pressure temperature accuracy and elemental recoveries on sample/particle size acid mixture digestion time and power programme will be documented for a variety of botanical biological and geological SRMs.Furthermore computer control of the prototype system described here an implementation in- volving multiple digestion tubes and even a direct interface to the ICP are envisaged. Financial assistance from the Ontario Ministry of the Environment Project 45 3G is gratefully acknowledged. References 1 Abu-Samra A. Morris J. S. and Koirtyohann S. R. Anal. Chem. 1975 47 1475. 2 Barrett P. Davidowski L. J. Jr. Penaro K. W. and Copeland T. R. Anal. Chem. 1978 50 1021. 3 Introduction to Microwave Sample Preparation. Theory and Practice eds. Kingston H. M. and Jassie L. B. American Chemical Society Washington 1988. 4 Matusiewicz H. and Sturgeon R. E. Prog. Anal. Spectrosc. 1989 12 21. 5 Matusiewicz H. Spectroscopy 1991 6 38. 6 Sulcek Z. and Povondra P. Methods of Decomposition in Inorganic Analysis CRC Press Boca Raton FL 1989 ch. 6. 7 Burguera M. Burguera J. L. and Alarcon 0. M. Anal. Chim. Acta 1986 179 351. 8 Millward C. G. and Kluckner P. D. J. Anal. At. Spectrom. 1989 4 709. 9 Kingston H. M. and Jassie L. B. Anal. Chem. 1986 58 2534. 10 Legere G. and Burgener P. ICP Znj Newsl. 1985 11 447. 1 1 Boomer D. personal communication July 1990. 12 Nadkarni R. A. Anal. Chem. 1984 56 2233. 13 Schelkoph G. M. and Milne D. B. Anal. Chem. 1988 60 2060. 14 Aysola P. Anderson P. and Langford C . H. Anal. Chem. 1987 59 1582. 15 Toshiba Microwave Oven Service Data File No. 330-353 Toshiba Minato-Ku Tokyo Japan. 16 Gorsuch T. T. The Destruction of Organic Matter Pergamon NY 1970. 17 Kratochvil B. and Mamba S. Can. J. Chem. 1990 68 360. 18 Gilman L. and Grooms W. Anal. Chem. 1988 60 1624. Paper 1 /0053 7E Received February 5th 1991 Accepted May lst 1991