Freseniu$' Joumal of
Fresenius J Anal Chem (1992) 343 : 852 -- 862
© Springer-Verlag 1992
Analytical chemistry in nuclear technology* H. J. Ache Institut fiir Radiochemie, Kernforschungszentrum Karlsruhe, Postfach 3640, W-7500 Karlsruhe 1, Federal Republic of Germany
Received September 12, 1991 Summary. The objectives of analytical chemistry in nuclear technology are discussed. The analytical techniques and methods commonly used in the various phases of the nuclear fuel cycle, mining and fuel fabrication, reprocessing and nuclear waste management are described and their advantages and disadvantages are demonstrated. The question of applying in-line analytical instruments in this area is addressed and some techniques which show a proven potential for such an employment are defined.
Introduction
In this paper the special features of "Analytical Chemistry in Nuclear Technology" will be discussed and it shall be demonstrated in which way the problems and working procedures of the analytical chemists in this field are unique and different from those employed in other technology areas. As in any other modern technology, analytical chemistry in nuclear technology has to provide the control of the proper chemical composition of the starting materials and the control of the chemical parameters of the ongoing process, and finally it is responsible for the quality control of the end products obtained. In contrast, however, to most other technologies a great portion of the materials which the analytical chemist has to handle in nuclear technology is radioactive. This in turn requires that he must observe certain safety procedures. He must be protected from the radiation emitted from the radioactive materials, whereby the amount of shielding will be determined by the level of radiation. Furthermore, if the radioactive material is an e-emitter the substance has to be contained in such a way that no small particles, aerosols etc. which contain e-emitting materials can find their way into the lab atmosphere, where they eventually can be inhaled by lab personnel. Therefore, the radioactive material to be analyzed will have to be handled in more or less heavily shielded gloveboxes or even remote controlled in c~-tight hot-cells. Thus in addition to the usual criteria for selecting a particular analytical technique for a given problem, such as high sensitivity, accuracy, sufficient detection limits, no interferences from other components in the sample, and being economical with regard to required labor, investment and operating costs, other considerations such as the compatibility of the technique with remote handling conditions * Presented at the 33rd IUPAC Congress, Budapest, August 1722, 1991.
in a glove-box or hot-cell environment, reduced radiation exposure of personnel and prevention of secondary radioactive waste and time-consuming waste handling will be determining factors. The consequences of these and other aspects will be demonstrated in the following by focussing on the analytical work required in various phases of the nuclear fuel cycle. The nuclear fuel cycle as shown in Fig. I can be subdivided in the four areas: mining and fuel fabrication, nuclear energy production in the reactor, reprocessing of nuclear fuel and nuclear waste management. Analytical chemistry involved in normal nuclear reactor operation, which is mainly the monitoring of low-level radioactivity in effluents, as well as the analytical techniques in safeguarding nuclear materials (although they represent important and interesting developments of analytical chemistry) cannot be considered within the scope of this paper. Thus the discussion will be centered on the three sections mining and fuel fabrication, reprocessing of nuclear fuel and nuclear waste management.
Mining and fuel fabrication
In this part of the nuclear fuel cycle, the first step is the mining of the uranium ores with a uranium oxide (U308) content of 0 . 0 4 - 0 . 5 % . From these ores a uranium oxide concentrate is obtained (yellow cake) which is passed to the conversion plant, in which the oxide is converted to U F 6. In a subsequent enrichment process the concentration of the fissionable uranium isotope of the mass 235 is increased
Supply
~ . ' ~ .~ .'*' ] J
Waste -'%~'~Management
m
Fig. 1. Nuclear fuel cycle. - - - U; . . . . . Pu;
fission products
853 Table 1. Concentration ranges and activity levels in nuclear fuel reprocessing Uranium minerals (0.04-0.5% U)
1
Uranium concentrate ("yellow cake", U3Os)
i
UF6
Typical standard specifications U 40% in % relative to U = 100 Mo V205 P205 C1, Br, I
<0.6 <1.0 <6.0 <0.5
As C03 B SiO2
Table 3. Quality control of UOz-powder for fuel fabrication U, Pu
XRF, Potentiometry, gravimetry
U, Pu isotopic composition C1, F, B
Mass spectrometry, after extraction of U, Pu from bulk with TBP-kerosene Pyrohydrolysis of nuclear fuel with 02 gas [saturated with H20 at elevated temperature
(tt00oc)]
<2.0 <2.0 <0.2 <4.0
Measurement of B as BF4-0xazine complex spectrophotometrically after treatment of H3BO3 with aqueous HF which gives BF2 Measurement of C1- and F - by ion chromatography with conductometric detection
UO2 (87.7% U) or UOE-PuO2 (U > 80%, Pu __<20% by weight) Total H
Heating (600° C) in argon flow (+ Oz) trapping of H20 formed, followed by electrolytic decomposition of H20 Heating in Oz-stream, determination of CO2 formed Gravimetry
U-analysis techniques employed in mining and fuel fabrication Analytical method
U-concentration range
Application
C
Spetrophotometry
1 -200 mg/1
Starting material
Gravimetry Potentiometry
1 - 5g 2 - 5 g/1
U/O Ratio Lanthanides and metals
XRF
100 gg
Routine
Fluorimetry
1 mg/1
Traces in effluents of production
( I Certification
ICP-emission spectroscopy
Table 2. Standard specifications of U O 2 for fuel fabrication Maximum concentration of impurities in gg/g U B
<
Mo
<250
4
Si
< 250
Cr
< 200
Th < 10 A1, Cu, Ta, Fe, Pb, Mg, ) Zn, Sn, Ti, W, V etc. ~ <250 Eu, Dy, Gd, Sm
<
F-, C1-
<350
C Moisture
1
Fig. 2. View of analytical laboratory at the Institute of Radiochemistry, Nuclear Research Center Karlsruhe, equipped for handling non-irradiated MOX fuels
100 0.4 weight %
U concentration 87.7%
from its natural abundance of approx. 0.7% to 2 - 4%. After converting U F 6 to UO2, the latter can be used for fuel element production (Table 1) [1]. The U-analyses accompanying these processes are mostly required to certify the standard specifications for uranium concentrate (Table 1), U F 6 and UO2. They are listed in Table 1 [2]. Table 2 gives standard specifications for UOzpowders for fuel-pellet production [3]. The analytical techniques involved to verify these specifications are fairly routine and summarized together with their main area of application in Table 3. Since uranium is a weak radioactive material, no particular shielding is necessary in these operations. More modern nuclear fuels, however, are made up of mixed oxides of uranium and plutonium (MOX fuels) [4].
As plutonium is a strong c~-emitter, it is essential to contain the M O X material during the whole course of analysis in specially designed glove-box compartments. Each of the analytical techniques listed in Table 3 has to be individually performed in a separate glove-box, including sample preparation, aliquoting, weighing, drying, etc. For this purpose, the sample is moved in a channel from one box to the other without being exposed to the atmosphere. In Fig. 2 an analytical laboratory at the Institute of Radiochemistry, KfK, Karlsruhe, equipped for this purpose is shown. The final quality control of a M O X pellet includes the determination of 1.) fission products and neutron poisons; 2.) chemical purity and stoichiometry; 3.) density, structure and sinter; 4.) shape and surface structure. Only the first two items require chemical analysis. Basically, the same techniques as listed in Table 3 are being employed, while the U and Pu isotopic composition is determined by thermionic mass spectroscopy after separation of
854 Table 4. Concentration and activity ranges of samples originating from the PUREX process
Uranium Plutonium Fission products
500 to 5 - 10 -s g/1 250 to 10- 9 g/1 10- 3 to 10-11 Ci/1
Nuclear fuel reprocessing
Fig. 3. View of ICP-AES set-up in a glove-box for analyzing MOX fuels (Institute of Radiochemistry, Nuclear Research Center Karlsruhe)
these two elements from the bulk material via extraction with TBP-kerosene [5]. One aspect needs some more attention. In order to meet the specifications 28 trace elements, mostly metals and lanthanides, present in the ppm range have to be determined. A direct inductively coupled plasma atomic emission spectroscopy (ICP-AES) of the solution of the sample in HNO3 is not possible because of the large background produced by the presence of the heavy metals which affects particularly the evaluation of the lanthanide emission lines (U and Pu show approx. 2,400 lines in the relevant spectral range between 240 to 450 nm). Therefore, in a modified procedure the sample is dissolved in HNO3, U and Pu is quantitatively extracted by TBP-kerosene and the aqueous sample subjected to ICP emission spectroscopy [6]. In this way the detection limit for the lanthanides is improved by the factor 20, due to the reduced heavy metal background. The detection limits attained are sufficient to meet the specifications for the lanthanides and thorium. The recent development of ICP-mass spectrometry (ICPMS) for trace element determination may also prove to be useful for nuclear fuel analysis; however, preliminary results from some groups [7] seem to indicate that because of matrix interferences an extraction step removing uranium and plutonium may be necessary to achieve low detection limits for the analytes. In this context it seems appropriate to refer to the complications which occur when analytical instrumentation is incorporated into glove-boxes or hot-cells. This is exemplified by Fig. 3, where e.g. the massive filter arrangement with cooling tower on top of a glove-box is shown which is supposed to filter the hot Pu-containing plasma gases, when an ICP-AES is operated in a glove-box compartment [8]. In the case of the ICP-MS combination only the nebulizer is located in the hot-cell and the aerosol extracted to the instrument.
After "burn-up" of the fuel elements in commercial power reactors they are "reprocessed" to recover the remaining fissile material, U-235 and Pu-239, and also to condition the higly radioactive fission products and other actinides produced in this "burn-up" process for safe final disposal. This is accomplished in the so-called Purex process [9] in which after dissolving the burnt-up nuclear fuel in HNO3, the extraction of U and Pu from the resulting aqueous solutions is achieved with a TBP kerosene mixture. The objective of the process is to obtain three product streams, one containing pure uranium with as little as possible Pu and fission product impurities, a second one consisting of pure Pu with a minimum of uranium and fission product impurities present and finally a third one consisting of the fission products with a minimum of U and Pu contaminations. In order to achieve this goal the various streams have to be monitored for losses of product and the presence of contaminations. In addition to the fact that the process control requires a large number of analyses ( 3 0 0 - 400 analyses per day) the main difficulty can be seen in the fact that the concentration ranges cover 7 orders of magnitude for U, 11 orders for Pu and 14 orders of magnitude for the radioactivity levels (Table 4). To amplifiy the analytical problems, no correlation exists between these parameters, and samples of widely different composition and activity concentration have to be analyzed [10]. It is clear that the choice of the most appropriate analytical technique in each individual case will depend on several criteria, most likely, however, on the optimal concentration range and the time required for analysis. Figures 4 and 5 show schematically some of the most pertinent parameters of the methods usually employed for U and Pu analyses. ICP-MS has not been included in this summary since routine analysis with this technique has yet to be carried out. Its application range will, however, fall into the ppb or sub-ppb range. Fission product determinations are mostly done by analyzing the radiation emitted from these radionuclides, i.e. by 7-spectrometry. Needless to say that most of these analyses have to be carried out behind heavy shielding. For accurate nuclear material accountancy, economical considerations and last but not least for nuclear safe guard purposes it is of utmost importance to have highly reliable and accurate data on the amount of incoming (input analysis) and outgoing (output analysis) fissile materials in a reprocessing plant. The standard method for U and Pu element determinations in the dissolver solution has in the past been the isotopic dilution mass spectrometry (IDMS) [11], which although it is considered to be the most reliable technique has the disadvantage of being very time-consuming. The various steps involved are: Concentrated nitric acid solution containing U, Pu and fission products etc. is spiked with U233 and Pu-242 or 244, the Pu quantitatively reduced to
855 Time
Methods, Standard Deviation (%} a
Counting
required [hi
100-10 cr
1
50- 3
1-0,5
mass Spectrometry (
ICP-AES
1-0.1
1-0.5
5-1
4-2 10-3
Photometry
4 ~
Speotrophotom. XBF
10~°
lOa
106
104
10-4
5-1
'1. NAA
1g~2
15-1
I
Spectroscopy
=
102
1-05 4-0.5
5- I
5-0.5
1
G~m./~ectro,_ . Gravim~ronhem :_,05-005
4-1
4 ~ -
1-0.5
10D
51
10z
i I
104
Pu Concentrationg/I
Fig. 4. Concentration ranges of U-analysis techniques Fig. 6. View of analytical hot cell (Institute of Radiochemistry, Nuclear Research Center Karlsruhe) Time required [h]
Methods, Standard Deviation (%1 Time Resolv.Laser Fluoresc.Spectrom.
~
Optical Fluorimntry
15-10
MassSpectrometry
1-O.I
~- ICP-AES (
5-1
1 -0.5 1 - 0.5 10-4
5:1
NAA
1-0.5
5-1
4-2
Spectrophotometry 5-0.5 ~" XRF
5-0.5
1
~" Electrochemistry 4 ]0'.~
]04
10.2
0.5- 0.05
Gravinletry__ 0.5-0.05 ×-Ahsnrptiometry
10J8
4-0.5
10o
i02
10~4
4-1 4
1-0.2
1-0.5
U Concentrationg/I
a
Fig. 5. Concentration ranges of Pu-analysis techniques
miner
Counts I
Pu(IV) with NaNO2, followed by a separation of U f r o m Pu on an anion exchange column. U and Pu are retained on this column while the fission products pass the column. Subsequently U is eluted with 8 tool/1 HNO3 and Pu with 0.5tool/1 HNO3. Isotopic analysis is accomplished via thermo-ionic mass spectrometry. The results allow the determination of the amounts of U and Pu present and of their isotopic composition. This is probably the most accurate technique for the U and Pu isotope analysis with a standard deviation of less than 1%. Since the radioactivity level of the dissolver solutions is extremely high, the chemical procedures have to be carried out in a hot-cell and with a specially designed mass spectrometer. Figure 6 shows a typical analytical hot-cell arrangement used for this type of analysis at the Institute of Radiochemistry, KfK, Karlsruhe. More recently another technique has become available which claims to be equal or superior to the IDMS with a considerably simplified sample preparation. It is the K-edge densitometry [12, 13] which requires for a quantitative analysis two X-ray photon transmission measurements, one immediately below and one immediately above the absorption edge energy of the element of interest (Fig. 7). The difference observed can be correlated to the element concentration. Since this method is relatively insensitive to matrix effects and independent of the chemistry of the sample normally no prior sample treatment is required. The domain of this technique is a concentration range of > 20 g heavy metal/1 and it is therefore ideally suited for the analysis of uranium in the dissolver solution, which for a light-water reactor may be in the range of 200 g/1. However, Pu concen-
FBR Dissolver Solution 157 g U/I 52 g Pu/I 2.10 lz Bq/I
107 ]°gCd
106 ~
U-K Edge ~
K
Edge
108 104 10a 102 b
!uK~eu,K°' I
,~
,~co
t
PuKcq
loo
12o
1~o
Energy [keV]
led
Fig. 7. a Experimental arrangement for K-edge densitometry; b spectrum of a fast breeder reactor fuel dissolver solution (Ref. [12, 13])
trations which may be lower by a factor of roughly 100 will not be detected. Therefore the K-edge densitometer is combined with an energy dispersive XRF spectrometer to a hybrid instrument (Fig. 8) [12]. In this approach the K-edge densitometry technique provides the reference basis in terms of an accurately determined uranium concentration, while the less reliable XRF analysis is only used to determine the U/Pu ratio via the intensity ratio of the fluoresced K~I Xrays from uranium and plutonium. The output analysis, i.e. the analysis of the pure product streams U and Pu, presents a smaller problem to the analyti-
856 K-XRF Analys
K-Edge
x,~ Detector
~ample i ragspm
Fig. 8. K-edge densitometer combined with XRF spectrometer (Ref. [12, 13])
NeutronDetectors:
Cf-252-Source(moderated)
Mepi Mepl = (I)ep i
~
t
cb~ h
-
-
=
Nabs =
[
epithenn.neu'eon multiplication factor concentrationofflssile material macroscopictherma] cross-section of the solution
Mep' =
f(cb'lNa')
]
Fig. 9. Monitor for solutions containing fissile materials
cal chemist, since in the absence of fission products the radioactivity levels are highly reduced and IDMS, K-edge densitometry as well as XRF may be suitable for this purpose. In process analysis in general the trend is to replace the so-called off-line analysis, i.e. sampling and analysis by insitu, preferably by in-line techniques. It is obvious that the employment of in-line instrumentation resulting in improved and more continuous process plant control could benefit reprocessing plant operation in a number of ways [14]: improved nuclear material management in the plant, reduced costs of safeguarding the plant and its process, reduced nuclear material losses to waste, reduced waste management costs, - reduced rework costs for out-of-specification product, reduced radiation dose to plant and support personnel, - economies in investment in support laboratories. However, progress in introducing in-line instrumentation or sensor technology into these plants has been relatively slow. Several adverse factors for sensor technology can be identified, some of them can be summarized as follows: Process plant operators are understandably very conservative, their approach to process control is inevitably dominated by the process control options of which they have experience and which have proven themselves to be reliable over a number of years. -~i'he PUREX is an extremely "well behaved" process in which the downstream processes can usually be expected to -
-
-
-
-
remain in control if the correct feed input conditions are achieved and maintained. Refitting older plants with new process control equipment is in most cases prohibitive in terms of costs, space available and licensing requirements. Cost of sampling and laboratory analysis is only a few percent of the operational cost of reprocessing in large commercial plants. The incentive for reliance on sensor technologies in these plants begins to recede when compared with the analytical reliability and process investigative resource which analytical support teams have been able to offer their customers. Considering the large number of analyses to be performed in a reprocessing plant, a certain relief has been provided by the analytical chemists, who developed a number of ingenious automated systems which were capable of handling the demand. Another very important aspect that sensor technology has not made a great impact yet, can be seen in the fact that a larger number of components necessary for in-line analytical instrumentation has become available only in the very recent past. This includes e.g. reliable lasers, fiber optics which can resist radiation, photodiode arrays, chemometric methods for data analysis etc. needed for the employment of spectroscopic in-line methods; or compact XRF units, neutron monitoring units for concentration measurements of fissile materials and others. Secondly, there was very little opportunity to subject this type of in-line instrumentation to a sustained, long-term test under realistic conditions. Test facilities were and are not available. A "hot analytical teststand" specifically designed for this purpose was constructed at the Institute of Radiochemistry, KfK, Karlsruhe [15]. However, due to the cancellation of the German national R and D program in reprocessing, it was not put into operation. Despite these somewhat discouraging prospects one should keep in mind that future reprocessing plants will differ distinctly from today's generation of plants. In future process design, the trend to pulsed extraction columns and centrifugal contacters, with ever shorter solute residence times and highly optimized advanced flow sheets, with PUREX cycles, reduced perhaps to only one, will be more favourable to in-plant sensors than to sampling and analysis. Considering the long design lead-times it seems necessary and justified to start further development efforts soon. Several concentration-measuring sensor systems have already reached a reasonable level of process credibility. These include among others: Neutron monitors for concentration measurements of fissile materials [16 - 18]; fiber optic spectrometry [19 - 22]; X-ray techniques, K-edge absorptiometry [12, 13, 23]; inline gamma spectrometry of process streams [24]. In high-radiation areas, such as the first cycle in the PUREX process, the number of analytical techniques for continuous monitoring of the concentration of fissile materials in pulsed extraction columns as planned for future reprocessing plants is very limited. One method to accomplish this goal would be to make use of the fact that neutrons from an outside source penetrating a solution containing fissile material induce fission in the solution resulting in an increased epithermal and thermal neutron flux (Fig. 9). The epithermal (or thermal) flux multiplication factor, Mep i o r Mth, defined as the ratio of the actual neutron count rate observed with an otherwise identical solution, however, without fissile material present, is indicative of the concentration of fissile material in the
857
bypass.~r~g. '
/
"
fiber optical cable trDm light source
,rlediagbregm /' /
i
fiber opgeal cable to photometer
/
lamp housing
/
/
/ OaFz window
optical in-line flow cell
\ light source (100 W quartz rungs/ten halogen lamp) '
control disc
rotating filter wheel
shielding
ff _~
bypass (aq. ph) /
~ /
conductivity measuring flow cell (four electrode type) //
~ val
/ interference tilters
/ ] light barrier I Peltier cells
' housing with proamplifier and light barrier electronics
shielding
amplifier
sampling device
Fig. 10. Fiber optical modification of the interference filter photometer SPECTRAN adapted to the bypass stream of a U-extraction column
K
.u
Fig. 11. Wavelength dispersive X-'ray spectrometer (Braggpolychromator) with position sensitive X-ray detector (diode array). Simultaneous measurement of up to 40 elements. T X-ray tube (rhodium, 55 kV, 40 mA); C flowthrough-cell with boron carbide window; S entrance slit (single slit); K crystal (fixed position); A Diode array (1024 pixels = diodes)
858 solution. The sensitivity of this technique under the given experimental conditions as defined by a 10% change in Mepi was found to be 1.1 g U-235/1 [16-18]. A typical experimental set-up for measuring U and Pu concentration in their different oxidation states in the second or third P U R E X cycle, again utilizing pulsed extraction columns, is shown in Fig. 10 [22]. In this case a simple process filter photometer (Spectran) was used. It could be replaced by a more refined instrument using a holographic grid and photodiode array detectors to allow a simultaneous analysis of the components of the solution [U(IV), U(VI), Pu(IV) and Pu(III)]. Essential for the data analysis are chemometric methods based on statistical methods such as principal component analysis, PCA, principal factor analysis, pattern recognition (PARC) and target factor analysis (TFA) [211. One system among the various types of in-line X-ray techniques developed so far should be discussed very briefly. It is a wavelength dispersive X-ray fluorescence spectrometer which allows the simultaneous measurement of up to 40 elements ranging from copper to ruthenium (K lines) and from tungsten to the actinides (L lines) with high spectral resolution using a Bragg polychromator and a position sensitive X-ray detector (Fig. 11) [23]. It consists of a rhodium X-ray tube as excitation source, a flow-throughcell with a boron carbide window, a fixed LiF 100 analyzer crystal, a single slit mounted between the cell and the crystal, and a silicon diode array detector which can detect X-rays in the 2 - 2 0 KeV (0.06-0.6 nm) range. The compact and easily mountable system is especially assigned for the inline monitoring of uranium and plutonium in the low and medium active processing streams in hot-cells. From first experiments using uranium and thorium solutions, typical detection limits (10rain integration time) for the heavy elements such as plutonium are estimated to be in the 0.1 mg/ml range even in the presence of a large excess of adjoining elements such as uranium due to the high spectrometer resolution.
Table 5. Waste classification High activity waste (HAW) > 104 Ci/m 3 Origin: aqueous concentrates from first U-Pu extraction cycle Treatment: calcination or vitrification (borosilicate glass) e.g. Pamela facility Medium activity waste (MAW) 0.1 -104 Ci/m 3 Low activity waste (LAW) < 0.1 Ci/m 3
Table 6. Concentrations of relevant constituents in MAWC
In Fig. 12 a survey of the different types of waste generated annually by a 1,000 MWe nuclear fission reactor is shown. Since MAW (medium activity waste) constitutes the largest portion (in volume) of all waste generated the analytical chemistry associated with its treatment shall be discussed in greater detail [25]. Liquid medium activity waste is classified as waste containing approx. 1 0 9 - 1 0 t4 Bq/m 3 solution (Table 5). It requires special handling and radiation shielding. In the Nuclear Research Center Karlsruhe (KIN) this waste originates mainly from the reprocessing plant in form of nitric acid solution. It is generated as a result of the reprocessing of nuclear fuels and cleaning of the plant. After concentration via evaporation an MAW concentrate (MAWC) is obtained. The subsequent treatment, storage and vitrification of the waste requires a thorough knowledge of the elemental composition, in some cases of the isotopic composition, especially of uranium and plutonium as well as of the chemical form in which some of these elements, actinides or fission products, are present in the waste. Table 6 shows typical concentrations of relevant constituents in MAWC [26].
Concentration range, g/1
FClSO2NO~PO~. Oxalate EDTA Na Fe Tc U Pu
0.3 --0.5 0.8--2.6 2.0-10.9 200 - 350 < 0.5 < 0.5 <0.1 30-50 10 2 - 2 < 5 . 1 0 - 3 - 7 . 10 -a 2.6--7.6 8 • 10-4-9 • 10 -a
Table 7. Previously used method of determination FC1-
sol NO~ PO~-
Ion-selective electrode Ion-selective electrode Precipitation as BaSO4 (gravimetry) Reduction to NH4 (spectroph0tometry) Complex formation with molybdate ions (spectrophotometry)
Oxalate EDTA
Nuclear waste management
Element
Indirect method: complex formation with Cu, precipitation of non-complexed Cu 2÷ as hydroxide, Cu by AAS in filtrate
Properties such as pH, dry residue, free acid, total H +, and density were determined by standard procedures, while NH4~ was measured by distillation and titration, anionactive tensides by extraction and spectrophotometry, and TOC (total organic carbon) by oxidation to CO2 and titration. Organic components such as TBP (tributylphosphate), H D B P (dibutylphosphate) and the alkanes were determined after extraction by gas chromatography. Tritium is detected after distillation by liquid scintillation spectroscopy, fission and activation products by 7-spectroscopy, and the sum of c~- and fi-emitters by proportional counting. As can be seen from Table 5 and 6, the remaining components of the MAWC fall in two categories, anions and cations, whereby in the case of actinides also a determination of the most important isotopes is required. In the case of the anion analyses, an important development was the advance of ion chromatography. In Table 7 the older techniques [27], which have been in use for many
859 3Ot ~ aWc~or~ ~ission;rtoducts
z=Tff
I
L~ - O.3tPu
])
Pulse 6enerntor
~=7~ LxP~
k__A/ .-, 28.7t U
~;'~ I I
[~%. I )
Spent Fuel Elements
1 \
< 1% I > 99%olActivity ~,
MAW
8S
FC
i
Boxcar
I.tegr.ter I- A
--
I/_~___~ I~AI
FuradQy Eoge
ActivityInventory (ca. 2 Years alter FuelOischaroe)
I FissionProducts
~---~1°/o
j
Opticnl Cell
-lt
\ 13x 107CiJ J~ FissionProducls I Fission I 1 \ in Liquid 1ProductsL/"
I
Pulsed Oye Laser J
12.5t Claddingand StructuralMaterial
Claddingand
~=~1=ructur~lMatarial
| /
WasteVolumina
Fig. 14. Scheme of time-resolved laser fluorimetry. B S beam splitter; SID silicon diode; FC fluorescence cell; P M photomultiplier; PA preamplifier
HAW
Fig. 12. Annual amount of waste generated from spent fuel elements
from a 1000 MWe nuclear power plant
Conductivity
F-
10 z.zl
Oxalate -----
= 1 : 100 diluted =1: lOdiluted
Fig. 13. Typical ion chromatogram resulting from MAWC
years and which now have been replaced by ion chromatography (IC) or high-performance liquid chromatography (HPLC), are listed. From this table and Fig. 13, where a typical ion chromatogram of MAWC is shown, it becomes immediately obvious that the introduction ofIC, a technique where ions such as F - , C1- SO4z-, NO 3-, PO4a- and others can be determined in one batch without any prior sample preparation, is a tremendous time-saving step as compared to the older methods, which comprised e.g. for the SO 2determination, precipation, and drying of the sample followed by gravimetric determination. Typical detection limits and standard deviations are 10- s g/1 and 0.01 - 0.1%, respectively. Besides the obvious savings of time it is equally important that the new procedure also significantly reduces the exposure of personnel to radiation. An interesting case is the determination of EDTA (ethylenediaminetetraacetic acid). In nuclear technology it is used as a decontamination agent, and therefore it becomes
incorporated into the MAWC. Since it forms water soluble complexes with most heavy metals, one can expect that the heavy metals vitrified in MAWC can become mobilized if water is getting access to the vitrified MAWC. EDTA is analyzed by adding an excess of Cu(II) ions to the solution at pH 10 and heating the mixture, whereupon the soluble Cu(II) EDTA complex is formed. The noncomplexed Cu precipitates as Cu(OH)2. After filtration the filtrate in which the Cu EDTA is present as anion is analyzed by HPLC. The detection is done photometrically at 250 nm. Other anions present in the MAWC do not interfere under these experimental conditions. Detection limits and standard deviations are 1 mg/1 and 5 - 1%, respectively [28]. Among the cations the most interesting ones for the further treatment of MAWC are the actinides U and Pu, Fe, Na (and also the various Tc species since the latter are assumed to show a strong tendency for migration if water gets access to the vitrified MAWC). The following discussion will be limited to the U and Pu analysis. An overview of the methods available for U determination, their range of application, and the time required has been shown in Fig. 4. In view of the fact that the reduction of sample material allows a more efficient handling of sample, e.g. without the use of elaborate shielding, an attempt was made to develop extremely sensitive analysis techniques for U determinations. As can be seen from Fig. 4, fluorimetry is one of the most sensitive methods for U detection, a method which has, however, the disadvantage that the presence of impurities or temperature and pH changes cause fluorescence quenching. Thus the time-resolved laser fluorimetry by which these effects can be eliminated was developed for routine operation (Fig. 14) [29-31]. In this way it was possible to detect accurately less than 0.05 ppb of U. Another technique which requires more material which, however, on the other side, has the added advantages of being a true multi-element analysis, is atomic emission spectroscopy with excitation via ICP (inductively coupled plasma). Because of the larger sized samples required, more extensive radiation shielding was considered necessary. The plasma torch and related equipment was therefore set up in a hot-cell of 15 cm lead-equivalent and lead glass windows.
860 C/S
104 -- MAW 1:10 diluted .... 0.1 ppmTc --,- 1.0ppmT¢ ----lO.OppnlTc
A
~
I|
I I | | j |
/I ~ | ~ ~
! 10 3
102
260.S
26().9
26i.0
26i.1
26i.2
Wavelength[nm]
Fig. 15. ICP-AES experimental arrangement in a hot-cell (Institute of Radiochemistry, Nuclear Research Center Karlsruhe) used for MAWC analysis
Fig. 16. ICP-AES spectra of technetium in standard solutions and genuine MAWC
U238
Adapting the ICP technique for operation in a hot-cell, where reduced pressure and other restrictions on ventilation exist, is by no means a trivial matter, and elaborate modifications had to be made before satisfactory working conditions could be achieved. A photograph of the arrangement is given in Fig. 15. ICP-emission spectroscopy was also used for the determination of Tc. Typical spectra are shown in Fig. 16. The limits of determinations are usually: 1 rag/1 for Tc, 1 mg/1 for U and 1 rag/1 for Pu. In Fig. 5 the most common techniques for Pu determinations have been summarized as a function of the optimal concentration ranges at which they can be carried out, their standard deviation and the time required for the analysis. Mass spectrometric isotope dilution analysis is certainly the most accurate, but on the other hand also the most timeconsuming technique. It is therefore rarely used in waste analysis and if so only when accurate information on the isotopic composition is required. Neutron activation [32] starts with the evaporation of the solution. Subsequently the dry sample which has been sealed in a quartz tube is irradiated with thermal neutrons together with a standard. This is followed by V-spectroscopic measurement of the relevant activation products. This is again a rather time-consuming technique which requires the use of a nuclear reactor as neutron source. Therefore most of the time Pu element analysis is done by ICP-AES. The simultaneous determination of small amounts of U and Pu and the isotopic composition of U and Pu has so far required a rather elaborate mass spectrometric isotope dilution analysis as described above [5, 11]. It is a most accurate method, however, in MAWC analysis usually an uncertainty of 10% in the isotope composition is tolerable. Therefore it was of considerable interest to develop a less accurate but also less time-consuming technique for the analysis of the U and Pu isotopes. It was found that the isotopic line shift in the emission spectra (ICPAES) by using e.g. a 3 m monochromator (Jobin-Yvon THR 1500), produced a clear separation of the emission lines of the various U and Pu isotopes [33]. Typical examples are shown in Figs. 17 and 18. Presently the experimental error
424.437
(90°/°)
FWHM O.O037nm
424.375
4241305
4241415 4241435 WavelengthInto]
4241455
424.475
Fig. 17. ICP-AES spectrum of uranium (10% U-235)
56,6% 239 I
402,12
402.14
402.16
402.18
Wavelength[nm]
Fig. 18. ICP-AES spectrum of a Pu-isotope mixture
is still in the range of 10%, which would be considered satisfactory for MAWC analysis. It can however, be expected that further improvements can be made to increase the accuracy and thus make this technique available for the solution of additional analytical problems. A considerable demand for analytical techniques for the determination of trace amounts of actinides and their speciation exists within the framework of studies of the safe dis-
861 beam
beam
s EMG 102
quartz- -window
laser ~
. 308nm energy, 100mJ XeCI
energy
pulse
8%
/
teflon ._^ ~ ~O-ring flank=
~] / / ~ - - ~ - - s o l u t i o n
lead ~[71
,[[ ["--piezoelectric
spring~
~ disc P'i wtre
~ stainless
' \
steel
[ photodiode
]
[ photodiede
tefloninsulator
][preamplifier]
trigger
Io
Fig. 19. Scheme of experimental set-up of pulsed photoacoustic spectrometry
boxcar A
posal of vitrified high activity waste in geological depositories. Besides the time resolved laser fluorimetry which provides extreme detection limits for U, one technique which was found to be useful for detecting such small amotmts of elements is the pulsed laser induced photoacoustic spectroscopy (LIPAS), with an improvement of the limits of detection of three or more orders of magnitude as compared with conventional spectralphotometry [34-37]. LIPAS uses pulsed laser light to excite selectively analyte species. Radiationless deexcitation transitions of the excited species dissipating the energy in the solution lead because of the pulsed excitation to heat and pressure waves, which can rbe detected by pressure sensors and correlated to the analyte concentration as schematically shown in Fig. 19. In contrast to ICP-MS, which also displays an equal or even better limit of detection, the LIPAS technique can directly recognize the chemical state of the analyte without prior separation procedures.
Conclusion
In summary, it can be stated that analytical chemistry in nuclear technology faces indeed some unique problems, mainly caused by the large amount of radioactivity which the analytical chemist has to handle in this field. However, he can cope with this issue by making use of ingeniously designed sample taking systems, automatics and robotics, by using shielded glove-boxes and hot-cells to protect himself and others against radiation, or by in-line techniques and newly developed methods which require only miniscule amounts of material and in this way reduce radiation exposure. It may come as a surprise to some, however, that otherwise the methods are exactly the same as in any other comparable industrial laboratory, even radioanalytical techniques such as neutron activation analysis as one might expect to find more frequently applied in this area, are hardly used anymore perhaps with the exception of ~- and ~-spectrometry. But even there, modern techniques such as ICP-
MS have already surpassed the detection limits of a- and 7spectrometry for a number of radionuclides with half lives larger than a few hundred years. Thus in most aspects analytical chemistry in nuclear technology will in the future experience the same developments and profit from the progress in instrumental analysis as analytical chemistry in any other industrial area.
References
1. Lieser KH (1969) Einfiihrung in die Kernchemie. Verlag Chemie, Weinheim, pp 375ff. 2. Mainka E (1988) GIT Supplement 1/88:46ff. 3. Davies W, Gray W (1964) Talanta 11 : 1203 - 1211 4. Schlosser G, Manzel R (1979) Siemens-Forschungs- und Entwicklungsbericht, vol. 8, p 108 5. De Bi6vre P (Nov. 1989) JNMM XVIII, No 1:47-55 6. a) Geyer F, Mainka E, Mfiller HG (1988) J Nucl Mat 153:102-- 107 b) Clain AF, Mainka E, Ache H J, Sonderdruck Wissenschaftlicher AbschluBbericht 21. Internationales Seminar Juli 1986, Universitfit Karlsruhe 7. Vijayalakshmi S, Krishna Prabhu R, Mahalingam TR, Mathews CK (1989) 2nd Karlsruhe Int Conf on Analytical Chemistry in Nuclear Technology, Karlsruhe, June 5 - 9 , 1989, Book of Abstracts, Abstr No 43 8. Mainka E, Mfiller HG, Neuber J (1988) ASTM, Spec Techn Publ 951 : 146-155 9. Baumg~irtner F (ed) (1978) Chemic der Nuklearen Entsorgung, vol I--III. Verlag Thiemig, Mfinchen 10. Berg R, Ertel D, in ref9, pp 162--181 11. De Bi6vre P, Gallet M, Hendrickx F, Lycke W, Wolters WH, Eberhardt KR, Fassett JD, Gramlich JW, Machian LA, Mainka E, Wertenbach H (1984) Report KfK-3761, Eur 1991e 12. Otmer H, Eberle H, Matussek P, Michel-Piper I (1986) Report KfK-4012 13. Otmer H (1983) ESARDA Bulletin, No 4, p 19 14. Allan CG, Invited lecture presented at 2nd Karlsruhe Int Conf on Analytical Chemistry in Nuclear Technology, Karlsruhe, June 5-9, 1989
862 t 5. Ache HJ (I 988) Report KfK-4476, pp 85--104 16. Gantner E, Kuhnes U, Trundt D (1988) KfK-Nachrichten 20:117-121 17. Gantner E, Mainka E, Ruf H, Ache HJ (1984) In: Lain WS (ed) Analytical spectroscopy, Proc of 26th Conf on Analytical Chemistry in Energy Technology, Knoxville, Tenn, Oct 11 -- 13, 1983. Elsevier Scientific Publ Co, Amsterdam, pp 263-272 18. Gantner E, Kuhnes U, Ache HJ (1991) J Radioanal Nucl Chem, to be published 19. Bostick DT, McCue DD, Strain JE, Bauer ML, Dixon DM (1981) Proc Conf on Analytical Chemistry in Energy Technology, Gatlinburgh, pp 225--253 20. Groll P, R6mer J, R6der L, Persohn M, Schlosser B (1986) Anal Chim Acta 190:265 21. O'Rourke PE (1989) JNMM XVIII, No 1 : 8 5 - 9 5 22. Bfirck J, Kr~imer K, K6nig W (1990) Report KfK-4672 23. Neuber J, Braun R, 2nd Karlsruhe Int Conf on Analytical Chemistry in Nuclear Technology, Karlsruhe, June 5 - 9 , 1989, Book of Abstracts, Abstr No 94 24. Webster RK, Smales AA, Dance DF, Slee LJ (1961) Anal Chim Acta 24:371 25. Ache HJ (1988) J Radioanal Nucl Chem 124:415-430 26. Wertenbaeh H, Drobnik S, Mainka E, Unger M, Proc Conf on Nuclear Fuel Reprocessing and Waste, Paris, Aug 24--28, 1987 (in press)
27. Coerdt W, Mainka E (1985) Fresenius Z Anal Chem 320:503 28. K6nig W, Mainka E, Weis S, Geyer F, Mfiller HG, Unger M, Int Conf on Analytical Chemistry in Nuclear Technology, Karlsruhe, June 3--6, 1985, Book of Abstracts, p 165 29. School S, Mainka E, Hellmund E (1986) In: Lain WS (ed) Analytical chemistry instrumentation, Proc of 28th ORNLDOE Conf on Analytical Chemistry in Energy Technology, Knoxville, Tenn, Sept 3 0 - O c t 3, 1986. Levis Pub Inc, Chelsea, p 251 30. Hoeppener-Kramar U, Mainka E (1990) Report KfK-4782 31. Hoeppener-Kramar U, Mainka E (1989a) Fresenius Z Anal Chem 333:760-761 and (1990) Report KfK-4782 32. Hawa AH, v. Baeckmann A (1974) Report KfK-1888 33. Mainka E, Mfiller HG, Neuber J (1986) GDCh-Fachgruppe ,,Nuklearchemie", Vortragstagung, Regensburg, FRG, Sept 2 9 - O c t 2, 1986 34. Klenze R, Kim JI, Wimmer H (1991) Radiochim Acta 52/ 53 : 97-103 35. Bohnert R, Faubel W, Ache HJ (1990) Fresenius J Anal Chem 338:695-- 698 36. Adelhelm K, Faubel W, Ache HJ (1990) Fresenius J Anal Chem 338:259- 264 37. Stahr B, Faubel W, Menzler PM, Ache H J, 2nd Int Conf on Analytical Chemistry in Nuclear Technology, Karlsruhe, June 5 - 9 , 1989, Book of Abstracts, Abstr No 32