Anal. Chem. 2000, 72, 1186-1191
Direct Determination of Dibutyl and Monobutyl Phosphate in a Tributyl Phosphate/Nitric Aqueous-Phase System by Electrospray Mass Spectrometry C. Lamouroux,*,† H. Virelizier,† C. Moulin,† J. C. Tabet,‡ and C. K. Jankowski†
DCC/DPE/SPCP/Laboratoire d’Analyse et Synthe` se Organique, CEA Saclay, BP2, 91191 Gif sur Yvette, France, and Laboratoire de Chimie structurale Organique et Biologique, Universite´ Pierre et Marie Curie, Paris VI, UMR 7613, 75252 Paris, France
Electrospray ionization mass spectrometry was tested for its potential use in the quantification of monobutyl phosphate (H2MBP) and dibutyl phosphate (HDBP), two degradation products of tributyl phosphate (TBP), the extractant used in the nuclear fuel reprocessing known as the PUREX process. Detection and quantification of these phosphate esters by electrospray in positive and negative ionization mode are reported in this study. This fast and reliable method, which does not require any preliminary sample extraction, appears to be very attractive for process control. Negative ionization mode gave abundant [M - H]- ions for both HDBP and H2MBP products. Thus, the concentration of H2MBP between 0.1 and 10 g/L in concentrated aqueous nitrate solutions can be precisely determined. Moreover, the concentration of HDBP up to 1 g/L in a TBP matrix was evaluated in this mode. For HDBP concentrations below 1 g/L, detection in the positive ionization mode appeared to be attractive. TBP and HDBP cluster detection allowed quantitative HDBP determination. Indeed, small amounts of HDBP in commercial TBP (60 mg/L) could be directly quantified using the specific [2TBP, HDBP + H]+ cluster at m/z 743. In nuclear technology, tributyl phosphate (TBP) is the most frequently used solvent in liquid-liquid extraction for fuel reprocessing. This extraction, known as the PUREX process (Plutonium Uranium Refining by EXtraction), is still considered as the most convenient method to retreat the spent fuel.1-3 The problem of solvent degradation in the nuclear industry is not only due to chemical reactions but also to intense radiations, which significantly increase the decomposition process. Therefore, when exposed to acidic conditions (nitric aqueous phase) and ionizing radiations (U, Pu, fission products), TBP undergoes degradation giving mainly dibutyl phosphate (HDBP) and to a lesser extent †
DCC/DPE/SPCP/Laboratoire d’Analyse et Synthe`se Organique. Universite´ Pierre et Marie Curie. (1) Burger, L. L.; McClanahan, E. D. Gamma Radiol. 1958, 50 (2), 153-154. (2) Burr, J. Radiat. Res. 1958, 8, 214-221. (3) Adamov, V. M.; Andreev, V. I.; Belayev, B. N.; Markow, G. S.; Ritari, M. S.; Shil′nikov, A. Y. Radiokhimiya 1992, 34 (1), 185-197. ‡
1186 Analytical Chemistry, Vol. 72, No. 6, March 15, 2000
monobutyl phosphate (H2MBP). The presence of these products affects the TBP performance as an extracting solvent. HDBP and H2MBP react with the extracted metal to form complexes or precipitates that remain dissolved in the nitric aqueous phase. As a result, this leads to a decrease in the extraction yield.4-6 Deterioration of TBP in the system TBP/HNO3 (3 M) led to 0.02 M HDBP for an irradiation dose of 0.19 MGy and 0.26 M HDBP for a dose of 1.52 MGy.7 Generally, the irradiation dose received by the solvent during three years of duty has been estimated at 1 MGy. Thus, the solvent is regularly regenerated with basic treatment with sodium hydroxide or carbonate washing that allows the elimination of both acidic degradation products.8-10 The growing request for quality control of commercial TBP and predominantly the necessity for monitoring of the presence of its main degradation products, HDBP and H2MBP, led to the development of several analytical methods. Techniques currently used to perform such determinations are presented in Table 1. Gas chromatography11,12 is mainly used to realize measurements in the organic phase. However, this requires a previous derivatization, by methylation, because of the nonvolatile character of HDBP. Infrared spectroscopy13 is used for determination of the HDBP concentration in the solvent based on the PdO absorption at 1230 cm-1 with a poor detection limit of 150 mg/L due to interferences coming from the TBP presence. Liquid ion pair chromatography14,15 is also used to determine (4) Guedon, V.; Thieblemont, J. C.; Revel, Y. J. Nucl. Sci. Technol. 1994, 31 (1), 48-53. (5) Nowack, Z.; Nowack, M. Radiochem. Radioanal. Lett. 1979, 38 (56), 343354. (6) Shuyao, Y.; Yu, S.; Tianzhen, T. Radiat. Phys. Chem. 1989, 33 (6), 599602. (7) Adamov, V. M.; Andreev, V. I.; Belyaev, B. N.; Markov, G. S.; Polyakov, M. S.; Ritari, A. E.; Shil′nikov, A. Y. Kerntechnik 1990, 55 (3), 133-137. (8) Frydrich, C.; Frydrich, R.; Max, G. Isoprentaxis 1993, 28, 343-354. (9) Ginisty, C.; Guillaume, B. Sep. Sci. Technol. 1990, 25, 1941-1952. (10) Aranjo, B. F.; Matsuda, H. T.; Kvada, T. A. J. Radioanal. Nucl. Chem. Lett. 1992, 166, 75-84. (11) Hardy, C. J. J. Chromatogr. 1964, 13, 372-374. (12) Brignocchi, A.; Gasparini, G. M. Anal. Lett. 1973, 6 (6), 523-530. (13) Stieglitz, L.; Ochsenfeld, W.; Schieder, H. Kernforschungszeutrum Karlsruhe Rep EURFNR-633 (KFK 691), Federal Republic of Germany (14) Muller, J. P.; Cojean, J.; Deloge, A. Analusis 1985, 13 (4), 160-165. (15) Grant, K. E.; Mong, G. M.; Clauss, S. A.; Wahl, K. L.; Campbell, J. A. J. Radioanal. Chem. 1997, 220 (1), 31-35. 10.1021/ac990613y CCC: $19.00
© 2000 American Chemical Society Published on Web 02/17/2000
Table 1. Summary of Different Techniques Applied to the Determination of HDBP and H2MBP in TBP/Nitric Aqueous-Phase System methoda GC IS LC
ESI-MS
system
sample treatment
working range and accuracy
HDBP and H2MBP in TBP alkane mixture TBP hydrocarbon diluent system HDBP in organic TBP/ nitric aqueous-phase NaNO3 1 M simulated waste sample TBP, HDBP, H2MBP, etc. simulated waste sample organic/nitric aqueousphase system: HDBP in organic phase
CH3 derivative treatment
100-800 mM HDBP ( 18 mM
50 mM
11, 12
extrctn by carbon sulfur
>150 mg/L
150 mg/L
13
organic phase: extrctn by NaOH preconcn
10-100 mg/L
5 mg/L HDBP
14
organic phase: extrctn by NaOH preconcnq
HDBP 0.1-10 g/L, r2 ) 0.999; H2MBP 0.1-10 g/L; r2 ) 0.976
HDBP 0.1 g/L; H2MBP 0.1 g/L
15
no treatment
1-10 g/L, 1% deviation; 10-100 mg/L, 1% deviation 0.1-10 g/L
5 mg/L
b
0.1 g/L
b
H2MBP in nitric aqueous phase
no treatment
detection limit
ref
a GC, gas chromatography; IS, infrared spectrometry; LC, liquid chromatography; ESI-MS, electrospray ionization mass spectrometry. b This work.
HDBP in aqueous nitric or organic phases. For the organic phase, an extraction of HDBP by NaOH is required to reduce the concentration of TBP. The detection limit in these conditions is ∼0.1 g/L. For H2MBP, liquid chromatography is the only method currently used, with a detection limit of 0.1 g/L but with poor reproducibility. The other techniques, e.g., capillary electrophoresis16,17 or ionic chromatography,18 are more restricted and therefore not suitable for reprocessing matrixes. Compared with all these different techniques, the use of electrospray mass spectrometry seems to provide a reliable, simple, and convenient method for a fast monitoring of HDBP and H2MBP in TBP/aqueous nitric phase without any previous separation or derivatization. Techniques of ionization at atmospheric pressure, i.e., electrospray (ESI-MS) or atmospheric pressure chemical ionization (APCI-MS), have already been used to study organophosphorus nonvolatile compounds.19 The purpose of this work is to establish potentialities offered by the use of direct infusion ESI-MS in terms of efficiency, sensitivity, and speed for the determination of HDBP and H2MBP in complex matrixes. EXPERIMENTAL SECTION Chemicals. Dibutyl and monobutyl phosphates were purchased from Panchim. Dibenzyl phosphate was provided by Lancaster and tributyl phosphate came from Marsan (Monaco). Nitric acid, sodium nitrate, and all spectrometric grade solvents used were obtained from Merck. Sample Preparation. (a) Organic Phase. Panoramic irradiation of pure TBP was performed with a 60Co γ radiation source in air at 22 ( 2 °C in order to simulate solvent degradation in contact with fission products in fuel reprocessing. The dose rate was 6.25 kGy/h, which corresponds to a total absorbed dose of ∼1 MGy simulating three years of solvent alteration in the reprocessing (16) Petru, A.; Rajer, P.; Ceeh, R.; Kuruc, J. J. Radioanal. Nucl. Chem. Articles 1989, 129 (2), 229-232. (17) Bocek, P.; Dolnik, V.; Dehl, M.; Janak, J. J. Chromatogr. 1980, 195, 303305. (18) Lash, R. P.; Hill, C. J. J. Liq. Chromatogr. 1979, 2 (3), 417-429. (19) Black, R. M.; Read, R. W. J. Chromatogr., A 1998, 794, 233-244.
process. The irradiated phase in contact with the nitric aqueous phase was washed three times with distilled water. An aliquot of these solutions is suitably diluted to be infused in ESI-MS. Solutions of HDBP 0.5 and 0.05 g/L in MeOH were prepared for standard addition method. A total of 10-50 µL volumes of these known solutions were respectively added to 1 mL of irradiated TBP diluted 1/10 000 and commercial TBP diluted 1/100 in order to determine HDBP concentration. Dibenzyl phosphate (HDBz) was introduced in the solvent phase as an internal standard according to the investigated concentration range of HDBP. The use of perdeuterated HDBP standard would have been more suitable but there is no simple way to synthesize this molecule. (b) Nitric Aqueous Phase. Solutions of H2MBP in the aqueous phase (of 5 M nitrates) were prepared. The influence of acidity on H2MBP detection was studied by varying the ratio of HNO3/NaNO3 while maintaining the total nitrate concentration at 5 M. The standardization range was realized from solutions of H2MBP of 0.1, 0.5, 1.0, 2.0, 5.0, and 10 g/L concentrations in (HNO3, 3 M)/(NaNO3, 2 M) suitably diluted 1/5000 in a MeOH/ H2O (70/30) solvent system. Mass Spectrometry. All mass spectrometric analyses were performed using a triple-quadrupole instrument model Quattro II (Micromass, Manchester, U.K.). Electrospray was used in positive and negative ionization modes without any modification, and typical operating conditions are as follows: The sample was introduced into the source at a rate of 10 µL/min with a syringe pump (Harvard Apparatus, Cambridge, MA). The instrument was operated by applying a voltage of (3.5 kV to the capillary. Source temperature was maintained at 80 °C. Source and most interface parameters were tuned at the beginning of the experiment in order to obtain the best signal sensitivity and were kept constant during each set of experiments. The electron multiplier was fixed at 650 V. Full-scan modes by scanning the quadrupole from minimum 50 to maximum 1000 with scan duration of 4 s and with an acquisition time of 2 min were used. The skimmer cone voltage value was studied and will be discussed in the next section. The identification of noncovalent clusters was established by using tandem mass spectrometry (MS/MS). Collision-induced Analytical Chemistry, Vol. 72, No. 6, March 15, 2000
1187
Figure 1. Electrospray mass spectra of TBP solutions: (a) irradiated TBP diluted 1/10 000 in 70/30 MeOH/H2O (ESI-); (b) 10-2 M TBP in 80/20 MeOH/H2O, 2% HCOOH spiked with 10-5 M HDBP (ESI+).
dissociation (CID) was performed with argon at a pressure of 2 × 10-3 mTorr. For optimum sensitivity and successful detection of dibutyl and monobutyl phosphates, the search for the most convenient conditions such as ionization mode, cone voltage, and solvent conditions was undertaken. Between each injection, the capillary tube was cleaned with 1 mL of solvent. Infusions of a large quantity of TBP could lead to a loss of sensitivity due to source pollution. Infusions of concentrated nitrate solutions did not notably affect sensitivity. RESULTS AND DISCUSSION Optimization of Electrospray Conditions. The purpose of this work is to quantify first HDBP (concentrated about 1-5 g/L in distillated residues or in low-level mg/L) in a TBP mixture and second H2MBP in a nitric aqueous phase. Optimization of electrospray conditions (ionization mode, solvent choice, cone voltage value) was undertaken to allow these determinations. (a) Ionization Mode. TBP is a triester of phosphoric acid and it does not possess any mobile hydrogen while HDBP and H2MBP are respectively the di- and monoesters of phosphoric acid with a distinctive acidic character. Thus, TBP is not detected in negative ion mode while H2MBP and HDBP (Figure 1a) will be easily detected under these conditions.20,21 The proton affinity values22 of these esters have been evaluated at 918, 900, and 870 kJ/mol, respectively, for TBP, HDBP, and H2MBP. Moreover, TBP forms and adduct with phosphate diester with great reproducibility; these adducts are well detected in positive ion mode. (b) Solvent and TBP Concentration Influence on Detection. The solvent was chosen in order to obtain the best sensitivity in each case. It was observed that a change in the organic solvent from MeCN to MeOH had little effect on sensitivity although the intensities of the solvent adduct ions are reduced with methanol.
Most significant improvements in sensitivity were obtained by adding HCOOH23,24 (formic acid). Therefore, the use of a mixture 80/20 mixture of MeOH/H2O containing 2% HCOOH allowed the best sensitivity for [M + H]+ ions while showing few sodium adducts. In negative ionization mode, the solvent conditions were optimized to obtain the best sensitivity of [M - H]- ion while reducing the detection of dimers and trimers. The intensities of [2M - H]-, [3M - H]-, and [4M - H]- ions decreased using successively acetonitrile, methanol, and at least water as solvent because of the acidity difference of these solvents in the gas phase. A 70/30 mixture of MeOH/H2O was used to determine the HDBP concentration in distillated residues. The organic phase is mostly composed of TBP. A study was performed on the influence of TBP concentration on the nature of the mass spectra in positive ion mode. Dimeric [2TBP + H]+ ions were observed according to the sample concentration25 infused. Dimer [2TBP + H]+ ion (m/z 533) is dependent on TBP concentration and reached 100% for TBP concentrations up to 10-3 M. TBP easily forms specific adducts with phosphate diester such as HDBP or HDBz as can be seen in Figure 1b. Indeed, addition of small amount of HDBP (10-5-10-6 M) in the concentrated TBP mixture led to the detection of a specific adduct at m/z 743 attributed to [2TBP + HDBP + H]+. It is important to mention that at this HDBP concentration and for TBP ∼10-2 M in solution, no molecular peak is detected for HDBP in negative or positive ion mode. The identification of this adduct was done by CID. CID spectra are first characterized by a loss of 210 uma or Th corresponding to an HDBP molecule; this is leading to [2TBP + H]+ (m/z 533) followed by the loss of a neutral TBP molecule (-266 ppm). The same structure is observed for the dibenzyl phosphate (internal standard) which led to m/z 811 assigned to
(20) Metzger, K.; Rehberger, P. A.; Erbeni, G.; Lehmann, W. D. Anal. Chem. 1995, 67, 4178-4183. (21) Ding, J.; Bukhart, W.; Karrel, D. B. Rapid Commun. Mass Spectrom. 1994, 8, 94-98. (22) Lesage, D. Thesis, Universite´ Paris VI 1995, pp 105-107.
(23) Zhou, S.; Hamburger, M. J. Chromatogr., A 1996, 755 (2), 189-204. (24) Zhou, S.; Hamburger, M. Rapid Commun. Mass Spectrom. 1995, 9 (15), 1516-1521. (25) Bell, A. J.; Despeyroux, D.; Muller, J.; Watts, P. Int. J. Mass Spectrom. Ion Processes 1997, 165/166, 533-550.
1188 Analytical Chemistry, Vol. 72, No. 6, March 15, 2000
Figure 2. Fragmentation scheme of tri- and diphosphate esters: (a) TBP in ESI (+); (b) HDBP in ESI (-).
[2TBP, HDBz + H]+. Moreover, it was noted that the intensities of these specific adducts are proportional to the diester concentration under the two following conditions; (i) TBP concentration up to 10-3 M and (ii) [TBP]/[HDBP] ratio maintained under 500. It is worth noting that this specific heteroadduct could allow the detection of trace levels (in terms of the nuclear industry) of HDBP in TBP. (c) Influence of the Skimmer Cone Voltage. To measure HDBP in TBP, the influence of the skimmer cone voltage value on TBP fragmentations was studied. The rise of the skimmer cone voltage gave three successive fragmentations for TBP (see Figure 2a). The fragmentation of the molecule is similar to that observed under γ radiation. Fragments at m/z 211, 155, and 99 appeared successively by loss of a butene group by the well-known McLafferty rearrangement. This type of fragmentation is characteristic of phosphate ester and has been already reported by several ionization techniques: electron impact (EI),26 ehemical ionization (CI),27 CID,28 and atmospheric pressure ionization (API).29 Thus, to realize the quantification of HDBP in a TBP matrix, it is absolutely essential to avoid TBP fragmentations. The fragment at m/z 211 interferes with the detection of HDBP. Moreover, detection on the specific cluster [2TBP, HDBP + H]+ is strongly dependent on the stability of this adduct and in this way on the cone voltage value. Around 20 V, there is no TBP fragmentation and the specific adducts between TBP and diphosphate esters are stable. By the same token, the quantification of concentrated HDBP solution can be realized in negative ion mode. The cone voltage value must be chosen to avoid dissociation of HDBP. Upon the rise of the cone value, phosphate diester HDBP gave successively two ions at m/z 153 and 7920,21 (Figure 2b). At 20 V, no (26) Sass, S.; Fisher, T. L. Org. Mass. Spectrom. 1979, 14, 257-259. (27) Barcelo, D.; Maris, F. A.; Geerdink, R. B.; Frei, R. W.; De Jong, G. J.; Brinkman, U. A. Th. J. Chromatogr. 1987, 65, 394. (28) Zeller, L. C.; Farell, J. T., Jr.; Kenttamaa, H. I.; Kuiv-alainen, T. J. Am. Soc. Mass Spectrom. 1995, 4, 125. (29) Harden, C. S.; Snyder, A. P.; Eiceman, G. A. Org. Mass Spectrom. 1993, 28, 585-592.
dissociations were observed and intensities are quite satisfactory in terms of signal-to-noise ratio. Therefore, in both modes, it is important to use a low cone voltage (20 V) to be able to perform quantitative measurements. Direct Determination of HDBP in TBP Organic Phase. Two very specific HDBP determinations within the reprocessing framework are needed. The first is the determination of relatively large HDBP concentrations (1 g/L) in a TBP mixure such as in distillation residues. The second is the determination of low HDBP concentrations (0.01 g/L) in TBP such as needed for TBP purity control. In the former case, the analyses were easily performed by using negative ionization mode. Quantification of HDBP by the standard addition method in irradiated TBP solutions was performed as shown in Figure 3. Hence, due to the complexity of the matrix containing a large number of TBP degradation products at different concentrations, the standard addition method is required. HDBP concentration was determined by plotting the ratio of the intensities I (m/z 209)/I (m/z 277) (internal standard) versus concentration and extrapolating this ratio back to the origin. The calibration graph was plotted according to the least-squares method, and a correlation coefficient of 0.998 was obtained. Relative standard deviation was under 2%. The limit of detection is 0.1 mg/L, which corresponds to 1 g/L HDBP in the matrix due to the dilution factor (1/10 000). It should be noted that it is also possible, for raw evaluation, to make direct measurements using standards made of TBP. Because of the fact that under radiolysis only 20% of the TBP is decomposed mostly into HDBP, all others degradation products could be neglected. Under this dilution (1/10 000), the injected concentrations of TBP are small (less than 3 pg) and no sensitivity loss is observed due to source pollution. Although it is difficult to estimate the exact number of analyses that can be made, the instrument can be used for one week without source cleaning. This technique is simple, fast, and efficient to determine concentrations of HDBP above 1 g/L in TBP. It should be noted that a lower dilution (1/100) to improve the limit of detection gave nonlinear results for HDBP addition, Analytical Chemistry, Vol. 72, No. 6, March 15, 2000
1189
Figure 3. Standard addition method for the determination of HDBP in irradiated TBP diluted 1/10 000 in 70/30 MeOH/H2O with 10 mg/L HDBz: (a) no addition; (b) first addition (+10 µL of 0.5 g/L HDBP); (c) third addition (+30 µL of 0.5 g/L HDBP); (d) fifth addition (+50 µL of 0.5 g/L HDBP). Ia ) I (m/z 209)/I (m/z 277); Y ) 0.167x + 1.205; R2 ) 0.998.
Figure 4. Standard addition method for the determination of HDBP in commercial TBP diluted 1/100 in 80/30 MeOH/H2O, 2% HCOOH with [HDBz] ) 1 × 10-4 Mg (28 mg/L): (a) no addition; (b) first addition, [HDBP] ) 5 × 10-7 M (0.105 mg/L); (c) second addition, [HDBP] ) 1 × 10-6 M (0.21 mg/L); (d) third addition, [HDBP] ) 5 × 10-6 M (1.05 mg/L); (e) fourth addition, [HDBP] ) 1 × 10-5 M (2.1 mg/L). Ib ) I (m/z 743)/I (m/z 811); Y ) 0.548x + 0.321; R2 ) 0.998.
certainly due to interferences caused by the large TBP concentration. The stable aggregates already mentionned between these two esters are probably at the origin of the loss of linearity. However, in this precise case, the specific adducts between TBP and phosphate diesters can be favorably used to realize the second determination in positive ionization mode. Hence, trace amounts of HDBP in TBP (test of TBP purity) can be determined with good reproducibility by analyzing the heterocluster between TBP and HDBP, specifically the adduct [2TBP, HDBP + H]+ at m/z 743 characteristic of the addition of HDBP in concentrated TBP solution (Figure 1b). Variation of the intensity of this adduct is directly proportional to the concentration of HDBP in solution. The same behavior is observed with the internal standard HDBz driving to the [2TBP + HDBz + H]+ ion adduct at m/z 811 (Figure 1b). Detection of these specific ions 1190 Analytical Chemistry, Vol. 72, No. 6, March 15, 2000
has been used to determine the presence of a low quantity of HDBP in a TBP solution. In these special conditions and by using the standard addition method, the determination of trace of HDBP in commercial TBP has been investigated. By plotting I (m/z 743)/I (m/z 811) versus concentration and by extrapolating it back to the origin, HDBP concentration has been estimated at 58 mg/L (Figure 4). This determination has been achieved three times, and it has shown good reproducibility evaluated at 2%. This determination is in agreement with the one obtained in our laboratory by gas chromatography (51 mg/L ( 10%) with preliminary diazomethane derivatization of the sample. Consequently, this mode of detection can be used for the determination of trace amounts of HDBP in a TBP mixture and is particularly convenient for quality control measurements of commercial TBP. The intrinsic limit of detection for HDBP was evaluated at 5 mg/L.
Figure 5. Calibration curve of H2MBP in 3 M HNO3/2 M NaNO3, 1/5000 70/30 MeOH/H2O. Ic ) 100I (m/z 153)/I (m/z 147); Y ) 7.593x + 0.081; R2 ) 0.999.
Direct Determination of H2MBP in Nitric Aqueous Phase. Similarly to HDBP, H2MBP gives intense [M - H]- ion at m/z 153. Reprocessing operations are usually performed with concentrated nitric aqueous phase (3-5 M); those of solvent regeneration use basic washing in the presence of sodium carbonate or sodium hydroxide. Instead the effect of nitrate salt concentration on the detection of H2MBP has been studied. It seems that when the concentration of nitrate exceeds 10-3 M, sensitivity drops dramatically. Thus, it was necessary to dilute strong nitric or nitrate H2MBP solutions to avoid the usually observed loss of sensitivity on H2MBP detection. The influence of acidity on H2MBP detection has also been studied. Variation of the HNO3/NaNO3 ratio from 5.0 to 0.25 with a total nitrate salt concentration of 5 M has shown that, in this range of pH, the variation of acidity on monobutyl phosphate detection is not a limiting factor. The general aspect of the spectrum is characterized by a base peak at m/z 62 [NO3]-, some adducts [HNO3, NO3]-, [NaNO3, NO3]- at m/z 125 and 147, respectively, and the molecular ion peak for H2MBP at m/z 153. One can note that a variation in acidity resulted in the inversion of clusters peak intensities at m/z 125 and 147 but did not affect the signal for H2MBP.
A standardization curve was made for concentrations of H2MBP between 0.1 and 10 g/L (six standard solutions) HNO3/ NaNO3 (3 M/2 M) as shown in Figure 5. Three measurements were realized each time with good reproducibility evaluated at 2%. This method was successfully used to quantify H2MBP in simulated solutions containing fission products (Zr 1 g/L). The obtained performances with ESI-MS with a limit of detection (LOD) of 0.1 g/L (3σ blank/slope) were compared to ion pair chromatography.15 In terms of sample preparation, ESIMS did not require any previous treatment except dilution. Moreover, the use of ESI-MS with direct infusion is a very quick method compared to liquid chromatography and it generates little waste. In conditions described here, the ruggedness of ESI-MS is perfectly suited to carry out a great number of analyses without cleaning. CONCLUSION By combining mass analysis with the ability to generate gasphase ions from labile polar and ionic species in solution, ESIMS has made possible many new and important advances in analytical chemistry. This study shows that ESI is a promising analytical method for specific detection of HDBP in a TBP matrix. Hence, compared to the other techniques, the ESI method is faster, is more reproducible, and allows adequate LOD. Furthermore, total quantification of H2MBP and HDBP could also be carried out from MS/MS experiments using the specific fragment [PO3]- at m/z 79 because of the lack of this ion in the TBP spectrum. Studies in positive ionization mode have allowed the detection of specific TBP-HDBP clusters in concentrated TBP solutions, which are very useful for the detection of small amounts of HDBP. Further work on the use of ESI-MS as a promising technique to determine dissolved HDBP transition metal complexes of nuclear interest is in progress.
Received for review June 9, 1999. Accepted December 20, 1999. AC990613Y
Analytical Chemistry, Vol. 72, No. 6, March 15, 2000
1191
