Role of Hydroboron Intermediates in the Mechanism of Chemical

H3N−BH3) in the acidity range of 0.2−10 M HCl were several orders of magnitude lower than those predicted by kinetics laws and obtained in the...
0 downloads 0 Views 211KB Size
Download PDF
Anal. Chem. 2004, 76, 6342-6352

Role of Hydroboron Intermediates in the Mechanism of Chemical Vapor Generation in Strongly Acidic Media Alessandro D’Ulivo,* Massimo Onor, and Emanuela Pitzalis

Laboratorio di Chimica Analitica Strumentale, Istituto per i Processi Chimico-Fisici, C.N.R., Area della Ricerca di Pisa, Via G. Moruzzi, 1, 56124 Pisa, Italy

Unknown and controversial aspects related to the mechanisms of hydrolysis of borane complexes and to the mechanisms of chemical vapor generation for trace element determination in strongly acidic media (0.01-10 M HCl) have been investigated and clarified. The overall hydrolysis rates of borane complexes (BH4-, H3N-BH3) in the acidity range of 0.2-10 M HCl were several orders of magnitude lower than those predicted by kinetics laws and obtained in the pH range of 3.8-14. The decomposition of the borane complexes takes place stepwise and proceeds through the formation of hydroboron intermediates, LxBH4-xn (x ) 1, 2, 3), where L could be one or more species among the donor groups H2O, NH3, OH-, and Cl- and n is the charge of the hydroboron species (n ) 0, +1, -1, depending on L). Some intermediates present surprisingly long lifetimes at elevated acidities and play a key role in determining both the overall hydrolysis rates of borane complexes and the reactivity of Hg(II), As(III), Sb(III), Bi(III), Se(IV), Te(IV), and Sn(IV) in chemical vapor generation for trace element determination. Atomic absorption experiments demonstrated that almost all trihydroboron species (LBH3n), dihydroboron species (L2BH2n), and monohydroboron species (L3BHn) play an active role in the generation of elemental mercury and stibine. Some of these intermediates are inactive or play a marginal role in the generation of arsine, bismuthine, and hydrogen selenide. Hydrogen telluride is preferentially formed by those hydroboron species, which are stable in strongly acidic conditions, while the same species are unreactive in the generation of stannane. The collected experimental evidence is in agreement with the general reactivity of the elements in chemical vapor generation techniques and, together with other literature data, definitely rule out the hypothesis of “nascent hydrogen” as a possible mechanism of chemical vapor generation by borane complex derivatization. Chemical vapor generation (CVG) atomic spectrometry represents one of the most powerful analytical tools for determination of trace and ultratrace amounts of the elements such as Hg, As, Sb, Bi, Se, Te, Ge, Sn, and Pb.1-3 The most popular reagent is * To whom correspondence should be addressed. E-mail: [email protected].

6342 Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

undoubtedly tetrahydroborate(III) (THB) (mainly sodium or potassium salts), but recently other borane complexes such as ammonia-borane (AB), tert-butylamine-borane(TBAB), dimethylamine-borane, trimethylamine-borane, and cyanotrihydroborate(III) have been employed for both analytical and diagnostic purposes in CVG.4,5 The use of THB has been extended also to the generation of volatile species of Cd,6-8 In,9 Tl,10,11 Cu,12 Au,13 Ni,14 Ag,13,15 Zn,13,16 and other transition metals.2 Studies on fundamental aspects of CVG are relatively few in comparison with those devoted to analytical applications to trace and ultratrace element determination and speciation. At present, nearly 30 years since its introduction as an analytical derivatization method,17-19 the mechanism of the reaction bringing the formation of volatile hydrides is still matter of debate. Deˇdina reported an excellent synthesis of this debate.20 He classified the different hypotheses of mechanisms into two classes, termed “nascent hydrogen” and “non-nascent hydrogen” mechanisms. The first is based on the assumption that the effective species in the derivatization process is the atomic hydrogen or nascent hydrogen, which is thought to be formed during the acid (1) Deˇdina, J.; Tsalev, D. L. Hydride Generation Atomic Spectrometry; Wiley: Chichester, 1995; Chapter 2. (2) Sturgeon, R. E.; Mester, Z. Appl. Spectrosc. 2002, 56, 202A-213A. (3) Puk, R.; Weber, J. H. Appl. Organomet. Chem. 1994, 8, 293-302. (4) D’Ulivo, A.; Loreti, V.; Onor, M.; Pitzalis, E.; Zamboni, R. Anal. Chem. 2003, 75, 2591-2600. (5) D’Ulivo, A.; Baiocchi, C.; Pitzalis, E.; Onor, M.; Zamboni, R. Spectrochim. Acta, Part B 2004, 59, 471-486. (6) Feng, Y.-L.; Sturgeon, R. E.; Lam, J. W. Anal. Chem. 2003, 75, 635-640. (7) Sanz-Medel, A.; Valde´s-Hevia y Temprano, M. C.; Bordel Garcı´a, N.; Fernandez de la Campa, M. R. Anal. Chem. 1995, 67, 2216-2223. (8) Guo, X. W.; Guo, X. M. J. Anal. At. Spectrom. 1995, 10, 987-991. (9) Busheina, J. S.; Headridge, J. B. Talanta 1982, 29, 519-520. (10) Yan, D.; Yan, Z.; Cheng, G. S.; Li, A. M. Talanta 1984, 31, 133-134. (11) Ebdon, L.; Goodal, P.; Hill, S. J.; Stockwell, P.; Thompson, K. C. J. Anal. At. Spectrom. 1995, 10, 317-320. (12) Sturgeon, R. E.; Liu, J.; Boyko, V. J.; Luong, V. T. Anal. Chem. 1996, 68, 1883-1887. (13) Luna, A. S.; Sturgeon, R. E.; Campos, R. C. Anal. Chem. 2000, 72, 35233531. (14) Guo, X. M.; Huang, B.; Sun, Z.; Ke, R.; Wang, Q.; Gong, Z. B. Spectrochim. Acta, Part B 2000, 55, 943-950. (15) Matousˇek, T.; Deˇdina, J.; Vobecky´, M. J. Anal. At. Spectrom. 2002, 17, 52-56. (16) Sun, H.; Suo, R.; Lu, Y. Anal. Chim. Acta 2002, 457, 305-310. (17) Schmidt, F. J.; Royer, J. L. Anal. Lett. 1973, 17, 17-23. (18) Pollock, E. N.; West, S. J. At. Absorption Newslett. 1973, 12, 6-9. (19) Fernandez, F. J. At. Absorption Newslett. 1973, 12, 93-97. (20) Deˇdina, J. In Flow Analysis with Atomic Spectrometric Detection; Sanz-Medel, A., Ed.; Elsevier: Amsterdam, 1999; Chapter 8. 10.1021/ac040078o CCC: $27.50

© 2004 American Chemical Society Published on Web 09/30/2004

hydrolysis of THB. The nascent hydrogen mechanism was first suggested by Robbins and Caruso,21 who postulated that THB works as a source of nascent hydrogen:

BH4- + H3O+ + 2H2O f B(OH)3 + 8H

(1)

in a way similar to metal/acid reduction system, for example, Zn/ H3O+

Zn + 2H3O+ f Zn2+ + 2H2O + 2H

(2)

The nascent hydrogen formed by the above reactions is considered to be the effective derivatizing agent forming the hydride, MHn:

M(m+) + (m+n)H f MHn + mH+

(3)

The hypothesis of hydride generation according to the nascent hydrogen mechanism has never been demonstrated. However, the mechanism of nascent hydrogen in CVG using THB derivatization is still accepted as a model by several authors.1,2 Recently, Carrero et al.22 supported the hypothesis of the nascent hydrogen mechanism with the aid of indirect experimental evidence on arsine and methylarsine generation in the presence of thiols. Their conclusion, however, did not properly take into account experimental evidence reported by other authors as, for example, the role of thiolohydroborates(III), RS-BH3-, which are formed by reaction of thiols with BH4- as demonstrated by Brindle and Le by 11B NMR,23 and the results of experiments performed using deuterated reagents discussed below. The second hypothesis comprises all the remaining mechanisms, and it received experimental support, limited to some elements and in particular reaction conditions. In a recent paper, it has been demonstrated that CVG of stannane, stibine, and bismuthine by THB, AB, and TBAB derivatization, can be achieved without the need to decompose the borane reagent in a wide range of solution acidities in the pH range 9.6-12.7 for THB, 4.5-12.7 for AB, and 4.7-11.5 for TBAB.5 In these conditions, the fraction of reagent that is decomposed in the reaction volume of the CVG apparatus is insufficient to produce the minimum stoichiometric concentration of nascent hydrogen that is necessary to generate the hydrides according to reaction 3. The only possible conclusion is that both BH4- and R3N-BH3 are the effective derivatizing species, acting probably through direct hydride transfer from boron to analyte atom. In acidic or strongly acidic solutions (-1 < pH < 2), which are the most commonly employed conditions in analytical determinations, no hypotheses about the type of mechanism could be made. This is due to the decomposition of both THB and amineboranes, which can be predicted to be very fast in acidic solutions. Therefore, the amount of nascent hydrogen necessary to form the hydride could be, at least in principle, generated under acidic reaction conditions. (21) Robbins, W. B.; Caruso, J. A. Anal. Chem. 1979, 51, 889A-899A. (22) Carrero, P.; Malave`, A.; Burguera, J. L.; Burguera, M.; Rondon, C. Anal. Chim. Acta 2001, 438, 195-204. (23) Brindle, I. D.; Le, X.-C. Anal. Chim. Acta 1990, 229, 239-247.

Concerning the nascent hydrogen hypothesis, the most important question, mostly disregarded in the analytical chemistry literature, is whether the generation of nascent hydrogen is compatible with the chemistry of THB and borane complexes. Laborda et al.24 strongly criticized the possibility that atomic hydrogen could be formed during THB decomposition, on the basis of thermodynamic considerations. Considering that the estimated standard potential for the H+/H couple is E0(H+/H) ) -2.016, neither THB [E0(H3BO3/BH4-) ) -0.482 V] nor zinc [E0(Zn2+/Zn) ) -0.763 V] would be able to perform the reduction of protons to atomic hydrogen. The studies described in the literature clearly indicated that reaction 1 is not compatible with the experimental evidence. THB decomposes stepwise, in both acid and alkaline media, and the four hydridic hydrogens are lost sequentially by interaction with species containing protons (H3O+, H2O).25 This brings the formation of molecular hydrogen as clearly demonstrated by the studies performed with deuterated reagents. The acid decomposition of KBD4 in 3 M H2SO426 leads to formation of HD as the main product (>90%), with only 3-5% of both D2 and H2. Similar results are reported by the hydrolysis of NaBH4 in D2O, where the main product was again HD, which represents a fraction >90 27 and 95% 28 of the total hydrogen evolved. The homologous compound of THB, LiAlH4, a rather powerful reducing agent, has been used for the preparation of high-purity HD (>99%) by controlled hydrolysis of LiAlH4 with D2O at low temperature,29 indicating a mechanism of decomposition similar to THB. This ruled out the formation of atomic hydrogen as intermediate in the generation of molecular hydrogen. The generation of atomic hydrogen following THB hydrolysis, in aqueous media and under a wide range of pH conditions, is not compatible with the current status of knowledge reported in the literature. In other terms, the formation of atomic hydrogen is not feasible by acid decomposition of THB. It can be concluded that, whatever the mechanism of CVG of hydrides using THB and borane complexes, it must follow routes different from those postulated by the nascent hydrogen mechanism. In particular, the few experiments performed with deuterated reagents (THB and deuterated THB) clearly indicated that the formation of arsine30 and stibine31 in acid media is the result of the direct transfer of hydrogen from boron to analyte atom. This paper consists of two parts. In the first part, qualitative experiments on the kinetics of hydrogen evolution during hydrolysis of borane complexes (THB and AB), complemented with fundamental data and experimental evidence reported in the literature, allowed us to clarify some aspects related to the mechanism of borane complex hydrolysis and to the existence of hydroboron intermediates in strongly acidic conditions. In the second, the reactivity of hydroboron intermediates toward Hg(II) and some hydride-forming elements has been investigated by atomic absorption spectrometry interfaced with CVG (CVG-AAS). (24) Laborda, F.; Bolea, E.; Baranguan, M. T.; Castillo, J. R. Spectrochim. Acta, Part B 2002, 57, 797-802. (25) Wang, F. T.; Jolly, W. L. Inorg. Chem. 1972, 11, 1933-1941. (26) Mesmer, R. E.; Jolly, W. L. J. Am. Chem. Soc. 1962, 84, 2039-2042. (27) T. Freund, J. Inorg. Nucl. Chem. 1959, 9, 246-251. (28) Jolly, W. L.; Mesmer, R. E. J. Am. Chem. Soc. 1961, 83, 4470-4471. (29) Wender, I.; Friedel, R. A.; Orchin, M. J. Am. Chem. Soc. 1949, 71, 1140. (30) Pergantis, S. A.; Winnik, W.; Heitmar, E. M.; Cullen, W. R. Talanta 1997, 44, 1941-1947. (31) Freund, T. J Am. Chem. Soc. 1961, 83, 2279.

Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

6343

Figure 1. Schematic representation of the experimental setup employed for monitoring the kinetic of hydrogen evolution during acid hydrolysis of THB and AB.

For the first time, a reaction model is proposed for CVG of Hg(0) and volatile hydrides by derivatization with borane reagents in strongly acidic conditions (-1 < pH < 2), which attempts to reconcile most of the experimental evidence reported in the analytical chemistry literature with the chemical properties of borane reagents involved in the derivatization reaction. EXPERIMENTAL SECTION Instrumentation. The kinetics of hydrogen gas evolution following acid hydrolysis of THB and AB were performed using the apparatus reported in Figure 1. It consists of a laboratorymade thermostatic cell realized in poly(methyl methacrylate) (24 mm i.d. × 46 mm) with a screw cap fitted with two ports. One of the ports is for the syringe injection of THB or AB, and the other is for connection of the headspace volume of the vessel with the pressure sensor (DS Europe, model LP625-1, 3.5 bar full scale). A check valve was interposed between the port and the syringe to avoid expansion of the evolved gas inside the syringe. An onoff valve was interposed between the port and the pressure sensor in order to allow removal of headspace gas at the end of the experiments. The signal output of the pressure sensor was continuously monitored with a frequency of 20 s-1 and stored on a PC. Atomic absorption experiments were performed with a PerkinElmer model 2380 atomic absorption spectrometer. Electrodeless discharge lamps (EDL System II, Perkin-Elmer) were employed for all the investigated elements and were operated at the current recommended by manufacturer. Absorbance measurements were performed at the following wavelengths and spectral bandwidths: As 193.7/0.7 nm; Sb 217.6/0.2 nm; Bi 223.1/0.2 nm; Se 196.0/2 nm; Te 214.3/0.2 nm; Sn 286.3 /0.7 nm; Hg 253.6/0.7 nm. A quartz T-tube (130 mm long, 8-mm inner diameter) operated at room temperature was employed as an atomic cell for mercury. For hydride-forming elements, the atomizer was a miniature argon-hydrogen diffusion flame32 supported on a quartz tube (6.5mm inner diameter). The continuous flow generator was realized with a peristaltic pump (Ismatec pump head MS/CA4-12 fitted on a Masterflex L (32) Deˇdina, J.; D’Ulivo, A.; Lampugnani, L.; Matousˇek, T.; Zamboni, R. Spectrochim. Acta, Part B 1998, 53, 1777-1990.

6344 Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

Figure 2. Schematic representation of the mixing sequences employed in CVG-AAS experiments.

drive (H-7519-25). Ismatec Tygon microtubings of appropriate diameters were used for propelling reductant and waste sample solutions. Ismatec Viton microtubings of appropriate diameters were used for propelling HCl and sample solutions. Sample, acid (HCl, where employed), and borane reagent (THB or AB) solution flow rates were 4, 4, and 2 mL min-1, respectively. Three different mixing sequences were used as schematically reported in Figure 2. All the mixing T-junctions and X-junctions were from Ismatec (Kel-F, 0.8-mm i.d.). The mixing sequence reported in Figure 2A is a typical analytical setup. The acid sample and reagent solutions merged in T1 and were allowed to react in the reaction loop L1 (100 or 500 µL, Teflon or Teflon PFA, 0.5-0.8-mm i.d.). The reaction mixture coming from the reaction loop merged with argon stripping gas in T2, and the stripping of gaseous products continued in a 500-µL stripping loop (Teflon PFA, 0.8-1-mm i.d.). The reaction mixture was then delivered to the gas liquid separator (borosilicate glass, 60 mm long, 10-mm i.d.). An auxiliary flow of gas was introduced into the gas-liquid separator. The composition and the flow rates of gases varied with the atomizer used. For the quartz T-tube atomizer, the following were used: 300 mL min-1 Ar (added through T2 or T3) and 180 mL min-1 Ar as auxiliary gas added through the gas-liquid separator. For the miniature flame atomizer, the following were used: 125 mL min-1 Ar (added through T2 or T3) and 100 mL min-1 Ar and 175 mL min-1 hydrogen, both added through the gas-liquid separator as auxiliary gas. Gas flow rates were controlled by calibrated ball rotameters. The mixing sequences reported in Figure 2B and C were purposely designed to investigate the effect of delaying the analyte addition to the HCl-THB (or HCl-AB) reaction mixture. The delay time was controlled with a volume L0 (Figure 2C), which is interposed between T1, where the acid hydrolysis of THB or AB started, and T2, where the analyte was added to reaction mixture. The volume L0 can be varied among 0, 4, 15, 50, and 500 µL. The 15-, 50-, and 500-µL volumes, were realized with Teflon or Teflon PFA tubings (0.5-0.8-mm i.d.) while the 4-µL volume was realized by modifying a mixing block with an additional 0.8 × 8 mm channel.5 The minimum delay time, corresponding formally to L0 ) 0 µL, was realized by the mixing sequence reported in Figure 2B, where all the three solutions merged in X1. For both mixing

Figure 3. Reaction paths for THB hydrolysis in the acidity range of 0.1-1.1 M HCl in 12% water-88% methanol solution, and at low temperature, reconstructed according to the kinetic data and the experimental evidences reported by Wang and Jolly.25

sequences B and C, the remaining part of the apparatus, from L1 to AAS detection, is the same as the one employed for (A). Chemicals. Aqueous solutions of AB were prepared by dissolving borane-ammonia complex (Aldrich, assay 90%) in NaOH, followed by microfiltration on a 0.45-µm membrane. The AB solutions are stable at least for two weeks at room temperature. A 5 M THB was prepared by dissolving sodium tetrahydroborate (BDH, pellets, reagent for AAS, assay 96%) in NaOH, followed by microfiltration on a 0.45-µm membrane. THB solution is stable for 1 week only if stored in a refrigerator at 0-4 °C. The assay of AB solution and THB solution was determined by iodimetric method as reported elsewhere.4,33 Stock solutions of 1000 µg mL-1 concentrations of the analytes Hg(II), As(III), Sb(III), Bi(III), Sn (IV), Se(IV), and Te(IV) were prepared by dilution of Merck concentrates. All the other chemicals were analytical grade reagents or higher. Water purified with a MilliQ system (Purelab Pro, USF) was used in all the operations. Procedure. For the kinetics of hydrogen evolution, 4 mL of acid solution was placed in the reaction vessel of Figure 1 and the syringe containing 400 µL of 1 M THB or AB in 0.1 M NaOH was positioned in the injection port. The solution was maintained under vigorous magnetic stirring, forming a deep vortex, until the temperature was equilibrated. Then, acquisition of the pressure signal started and the reductant solution was injected after ∼15 s. The acquisition of the pressure signal continued up to the maximum time of 60 min. Part of the reductant solution remained inside the injection port and the check valve. To make the measurements independent of the irreproducibility of the injection, all the data were normalized with respect to the maximum pressure value observed during the experiment performed with THB. The ionic strength was kept at the 10 M level with LiCl in all the measurements with HCl. In the case of HClO4, the less soluble NaClO4 kept the ionic strength at 5 M level. CVG-AAS measurements were performed by using the mixing sequences reported in Figure 2. The reaction loop volume was L1 ) 500 µL for all the elements, except for Te where L1 ) 100 µL was adopted. The reference signal, S0, was obtained using the mixing sequence A, with HCl concentration of 1 M for As, Sb, Bi, and Hg; 4 M for Se and Te; and 0.1 M for Sn. The concentration of HCl, for experiments performed with the experimental setup of Figure 2B and C, was the same for the HCl decomposition solution (merging with THB or AB) and for the analyte solution and was varied in the range of 0.01-10 M for Bi, Sn, and Hg; (33) Mesmer, R. E.; Jolly, W. L. Inorg. Chem. 1962, 3, 608-612.

0.02-10 M for As; 0.05-10 M for Sb; and 0.1-10 M for Se and Te. The concentration of THB and AB was always 0.2 M, except for Hg where 10-3 M THB has also been tested. The NaOH concentration in the reagent solution was 0.005 M, except in the case of measurement with configuration A, where 0.1 M NaOH was employed. Reference signals to check S0 values were typically run before and after the set measurements performed to scan the acid range for each volume of L0. The analyte concentration was 0.2 µg mL-1, except for Hg, where 0.1 µg mL-1 was employed. Safety Considerations. In many experiments, concentrated solutions (up to 10 M) of strong mineral acid (HClO4, HCl) have been used. During AAS experiments, amounts of toxic elements and acid vapors (HCl) are introduced in the atmosphere; an efficient fume extractor positioned above the atomizer is necessary. The hydrogen evolution in the kinetic experiments on THB and AB hydrolysis generate a pressure buildup in the reaction cell headspace (∼1 bar). Therefore, the headspace gases must flow out by using the valve positioned between the cell and pressure sensor (see Figure 1) before opening the reaction cell. RESULTS AND DISCUSSION Literature Data on Acid Hydrolysis of THB. Most of the confusion emerging from analytical literature about the decomposition rate of THB in acid solution arises from considering the THB decomposition as a single-step process (for example, according to reaction 1). The overall hydrolysis rate constant, kHydr, for THB (second-order reaction) was estimated to be 1.6 × 106 L mol-1 s-1 at 25 °C in the pH range of 3.8-14.33 Agterdenbos and co-workers34,35 suspected, on the basis of their experiments, that the overall hydrolysis of THB in strongly acidic media is much slower than predicted by the kHydr value reported above. The most relevant studies on the mechanism of THB hydrolysis in acid media was performed by Wang and Jolly,25 but it has been disregarded in the analytical chemistry literature on CVG. Wang and Jolly25 reported a kinetic study of the intermediates formed in the stepwise acid hydrolysis of THB, which was performed by using 88% methanol-12% water solution at -78 and -36 °C, in the acidity range of 0.1-1.1 M HCl. The stepwise decomposition of THB could be described according to the reaction pattern reported in Figure 3. Considering the rate constants reported for each step, it is possible to observe that the loss of the first hydrogen is too fast to be detected by (34) Agterdenbos, J.; Bax, D. Anal. Chim. Acta 1986, 188, 127-135. (35) Narsito; Agterdenbos, J.; Santosa, S. J. Anal. Chim. Acta 1990, 237, 189199.

Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

6345

measuring the pressure of evolved hydrogen, even at -78 °C. The loss of the second hydrogen, from BH3 (nascent borane) or its aquated complex with H2O-BH3, could generate two hydroboron species, (H2O)BH2OH and the borane cations (H2O)2BH2+, the latter species being resistant to acid hydrolysis at low temperatures. The dihydroboron species are involved in an equilibrium (H2O)BH2OH + H+ h (H2O)2BH2+ (Ke ) 6.4 at -36 °C), where only (H2O)BH2OH undergoes hydrolysis at low temperature to form the monohydroboron species (H2O)BH(OH)2. Another important observation is that the rate constants for diand monohydroboron species are independent of acid concentration, at least at the low temperature employed. It must be underlined that the authors were able to prepare acidic solutions of all the intermediates reported in Figure 3. The validity of the reaction scheme reported in Figure 3 for THB has never been verified under reaction conditions typically employed in analytical CVG. Literature Data on Acid Hydrolysis of AB. Kelly and Marriot,36 investigated the kinetics of acid hydrolysis of AB in the pH range of 4.3-6.1. At 25 °C, the overall second-order kinetic constant was estimated to be kHydr ) 6 L mol-1 s-1. The proposed mechanisms for AB hydrolysis were essentially two. In the first, which has been considered to be the more appropriate according to the experimental evidence, it can be assumed that the ratedetermining step is the displacement of BH3 from ammonia, which follows the protonic attack to nitrogen:

Figure 4. Relative pressure increase monitored during acid hydrolysis of THB and AB at +25 °C and in various acid conditions. (a) Full presentation of kinetic experiments; (b) closeup details of the same experiments reported in Figure 5a.

H3N-BH3 + H3O+ h H4N+-BH3 + H2O f NH4+ + H2O-BH3 (4) The H2O-BH3 will follow the same decomposition pattern reported in Figure 3. In the second, which had been considered the less probable, the displacement of the H3N-BH2+ group, which follows the protonic attack on borane group, is accompanied by the release of molecular hydrogen:

H3N-BH3 + H3O+ h H3N-BH4+ + H2O f (H2O)(NH3)BH2+ + H2 (5) The (H2O)(NH3)BH2+ was assumed to be rapidly hydrolyzed to B(OH)3. The validity of the above kinetic model has never been verified for AB under reaction conditions typically employed in analytical CVG. Qualitative Kinetics Experiments on THB and AB Hydrolysis. To verify the validity of the kinetics descibed above for THB and AB in strongly acidic aqueous solution, the decomposition of THB and AB was followed by monitoring hydrogen evolution with the apparatus reported in Figure 1. The results of experiments performed at +25 and -10 °C are reported in Figures 4 and 5, respectively. Figures 4a and 5a show complete experiments while Figures 4b and 5b present a closeup vision of the same experiments that is useful to evidentiate details of the kinetics within few seconds after reagent injection. The experimental results reported in Figures 4b and 5b indicated that the (36) Kelly, H. C.; Marriott, V. B. Inorg. Chem. 1979, 18, 2875-2878.

6346

Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

Figure 5. Relative pressure increase monitored during acid hydrolysis of THB and AB at -10 °C and in various acid conditions. (a) Full presentation of kinetic experiments; (b) closeup details of the same experiments reported in (a).

system response is probably not faster than ∼0.2 s, thus limiting the possibility to follow faster kinetics. This limitation arises from the finite time necessary for reagent injection and solutions mixing and for the equilibration of hydrogen pressure. To follow faster kinetics, the only way is to decrease the temperature, down to the lowest limit allowed by aqueous samples at 5-10 M saline content. The injection of the solution volume produces a signal as a consequence of the contraction of the headspace volume of the apparatus. This little sharp increase of the pressure in coincidence with the injection can be observed for example in Figure 5b for AB at 0.2 M HClO4. It can be taken as an indication of the background pressure level for all the measurements. Assuming the following reactions for the formation of the final products37,38

BH4- + H+ + 3H2O f 4H2 + B(OH)3 H3N-BH3 + H+ + 3H2O f 3H2 + B(OH)3 + NH4+

(6) (7)

The maximum pressure values recorded for THB and AB compare quite well with theoretical ratio of 4 to 3, indicating the

evolution of 4 mol of hydrogen for THB and 3 mol of hydrogen for AB, after complete decomposition of 1 mol of reagent (Figures 4a and 5a). The first comment on the obtained results is that the decomposition rates of both THB and AB are much slower than expected on the basis of the second-order rate laws d[CB]/dt ) kHydrCB[H+] (CB is the concentration of either THB or AB), reported in the literature. Assuming, for THB, at 25 °C kHydr ) 1.6 × 106 L mol-1 s-1, 33 the time necessary to decompose 99% of the reagent should be 14 and 0.29 µs at 0.2 and 10 M H+, respectively. For AB, assuming kHydr ) 6 L mol-1 s-1 at 25 °C,36 the time necessary to decompose 99% of the reagent should be 4 and 0.8 s at 0.2 and 10 M H+, respectively. The data reported in Figure 4b clearly indicate that the overall decompostion rates of both THB and AB in strongly acidic solution are several orders of magnitude slower than those observed at pH >3.8.33,36 From Figures 4 and 5 is evident that some drastic changes in the decomposition rates occur during the hydrolysis of both AB and THB. For AB, the loss of the first hydrogen is fast in comparison with the loss of the remaining two hydrogens. In the case of THB, hydrogen evolution suddenly slows down after the loss of two or three hydrogens, depending on acidity and temperature. The formation of dihydrobon and monohydroboron species seems therefore to be the key step slowing down the overall decomposition processes of both THB and AB. Details of the hydrogen evolution reported in Figures 4b and 5b indicate that, in general, HCl and HClO4 do not behave in the same mode. This could be due to the circumstance that chloride ion, in contrast to perchlorate ion, may compete with H2O and OH- in forming complex hydroboron intermediates. The most interesting evidence, valid for both acids, is that the decomposition rate decreases by increasing the acidity from 0.2 to 10 M H+, which is in agreement with the findings of Wang and Jolly25 on the formation of hydroboron intermediates possessing lifetimes much longer than BH4-at elevated acidities. However, the dihydroboron species, L2BH2n, formed by the loss of the first hydrogen of AB, appears to be more stable than the dihydroboron species arising from the loss of the first and second hydrogens of THB. The different stability between these dihydroboron species could be explained by taking into consideration the possibility that, in the hydrolysis of AB at pH k3 > k4. The confusion that is present in the analytical literature about the hydrolysis rate of THB arises from considering k1 as representative of the overall hydrolysis process of THB, while it only refers to rate of loss of the first hydrogen, which is first order in both [BH4-] and [H+]. Therefore, k1[H+] identifies with kHydr for low [H+] values, as those realized at pH >3.8. In strongly acidic conditions, k1[H+] becomes much higher than the relatively pH independent k2, k3, and k4, and the overall hydrolysis rate of THB is controlled by decomposition of one of the intermediates. Considering that for [H+] ) 1 M, 99% of the first hydrogen of BH4- is lost in less than 3 µs, it casts doubt on the probability that BH4- species can play an active role in most of derivatization reactions carried out in strongly acidic conditions. A similar conclusion applies also to AB hydrolysis, where in strongly acidic conditions (-1 < pH < 2) the formation of L2BH2n intermediates slows down the hydrolysis rate process by several orders of magnitude. The mechanism of hydrolysis of AB should be revised by assuming the existence of two parallel reaction paths taking place at low acidities according to reaction 4 and at high acidities according to reaction 5, respectively. Taking also into account the studies reported on the kinetics and mechanism of hydrolysis of some amine-haloboranes40 and amine-cyanoboranes,41 the sequence of intermediates that are formed in the AB hydrolysis in strongly acidic media could be k 2′′

k 3′′

k 4′′

(9) (L could be one or more groups among H2O, OH-, and Cl-). Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

6347

Alternatively, at lower acidities, the displacement of NH3 group according to reaction 4 results in the reaction path already described for BH3 in reaction 8. The possibility for chloride to be involved in the formation of complex hydroboron intermediates is the most reasonable explanation of the different reactivities that can be observed between HCl and HClO4 reaction media, for the hydrolysis of both THB and AB. Generation of Volatile Species by Reaction with Hydroboron Intermediates. In addition to BH4- and H3N-BH3, the intermediate hydroboron species, reported in reaction 8 or reaction 9, possess from one to three B-H bonds and can be considered as potential derivatization species in hydride generation or reduction processes. To verify this hypothesis, the CVG-AAS from aqueous ionic species of Hg(II), As(III), Sb(III), Bi(III), Se(IV), Te(IV), and Sn(IV) has been tested under experimental conditions that can allow the reaction of the analytes with the different hydroboron species formed during acid decomposition of THB and AB. Hydrochloric acid has been used in the experiments, considering its widespread use in CVG. The obtained results should be considered valid only for this acid considering its influence on the hydrolysis rate of both THB and AB and on the formation of different hydroboron intermediates (see previous section). The experimental setup used for this purpose is represented in Figure 2C. The reagent is hydrolyzed with 0.01-10 M HCl in a first reaction volume, L0, for a controlled time: the contact time between the HCl and borane solutions for the 4-, 15-, 50-, and 500µL volumes were 0.04, 0.15, 0.5, and 5 s, respectively. These contact times include the time necessary for mixing the solutions, which were estimated to be ∼0.02 s.5 When the analyte solution is added to the flowing reaction mixture in T2, the analyte can interact only with those hydroboron species still present in the reaction mixture, the concentration of which should be dependent only on contact time and acid concentration. The minimum possible delay, corresponding to L0 ) 0 µL, was realized with the addition sequence reported in Figure 2B. In this case, the hydrolysis of reagents takes place in the presence of the analyte, which has therefore the possibility to react with all the hydroboron species that are forming along the reaction path, downstream to the mixing X1 junction. The only doubt is concerned with the possibility that the species BH4- may play a role in the derivatization process at L0 ) 0 mL. Considering that more than 99% of BH4- should be completely decomposed to LBH3n in less than 0.5 ms in 0.01 M acid media, the only probability to partially interact with the analyte is during the ∼20-ms mixing process step. This implies that the formation of volatile derivatives should also be completed within the same time. The lack of reliable kinetic data on the formation of volatile derivatives in CVG hinders the clarification of this point. According to the above consideration for L0 g 4 mL and 0.01 < HCl < 10 M, there are few chances for the BH4- species to play a significant role in the derivatization processes which are most probably carried out by other tri-, di-, and monohydroboron intermediates, among those reported in reaction 8. For AB, in the lower acidity range, the data collected at 25 °C allowed only a rough evaluation of the rate loss of the first hydrogen, k′2 >50 L mol-1 s-1, in 0.2 M HCl. For AB derivatization, it cannot be excluded that H3N-BH3 species may 6348 Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

Figure 6. CVG-AAS experiments for mercury using the mixing sequences reported in Figure 2B (L0 ) 0 µL) and C (L0 ) 4-500 µL) and using THB (0.2 or 10-3 M) and 0.2 M AB. Reference signal S0 obtained with mixing sequence A, with L1 ) 500 µL and 1 M HCl.

play a role in the derivatization reaction in the low-acid range (0.01-0.2 M HCl) when using L0 e 50 µL (t e 0.5 s). The reference signal, S0, is obtained using a typical analytical mixing sequence (Figure 2A) under reaction conditions allowing the achievement of maximum generation efficiency of the volatile products. The normalization of the signal with respect to S0 is useful to compare the sensitivity obtained in the specific experiment with those of a standard continuous flow apparatus under optimized conditions. Considering the high concentration of HCl employed for some experiments, the formation of volatile chlorides could be a process contributing to the formation of volatile species of the investigated elements. Some measurements were performed with configuration C, but replacing THB or AB solutions with either water or a dilute solution of NaHCO3. At 5 and 10 M HCl, no signals were detectable for As(III), Sb(III), Bi(III), Se(IV), Te(IV), and Hg(II), indicating that generation of volatile chloride did not contribute significantly to the measured absorbance in CVG-AAS experiments reported in next sections. Mercury. The relative sensitivity obtained from CVG-AAS experiments for mercury is reported in Figure 6. The sensitivity is hardly dependent on the experimental conditions. Using 0.2 M THB or 0.2 M AB, the relative sensitivity is not affected by the contact time between acid and reagent solution in L0, whatever the acidity. Decreasing the THB concentration by 2 orders of magnitude indicated a loss of sensitivity only for HCl > 2 M, which is more pronounced for the highest decomposition volumes (L0 ) 500 µL). The above results clearly indicated that some species containing B-H bonds, which are able to reduce Hg(II) to Hg(0), are still present at the end of L0 even after 5-s hydrolysis in acid conditions (-1 < pH < 2). The experimental evidence on the analytical behavior of mercury at trace level indicated that this element is the most reactive in the family of CVG elements. It can be reduced using a stoichiometric amount of THB in alkaline conditions at the nanomolar level42 or using micromolar amounts (42) Bramanti, E.; D’Ulivo, A.; Lampugnani, L.; Raspi, G.; Zamboni, R. J. Anal. At. Spectrom. 1999, 14, 179-185.

Figure 7. CVG-AAS experiments for Sb using the mixing sequences reported in Figure 2B (L0 ) 0 µL) and C (L0 ) 4-500 µL) and using 0.2 M THB and 0.2 M AB. Reference signal S0 obtained with mixing sequence A, with L1 ) 500 µL and 1 M HCl.

of THB42,43 and AB4 in acid conditions. Taking into account the discussion reported in the previous sections, the results of Figure 6 indicate that mercury species, Hg2+, its chloride complexes, or both are very reactive toward all the possible hydroboron intermediates formed by the decomposition of both THB and AB (reactions 8 and 9). This could explain the high reactivity of mercury and its relatively insensitivity to CVG conditions, in terms of acidity and reagent concentration. Antimony. The relative sensitivities obtained from CVG-AAS experiments for antimony are reported in Figure 7. For this element, the same conclusion reported for mercury is valid. The various antimony species that can be formed at different acidities easily react with all the possible hydroboron intermediates formed by THB and AB decomposition. Furthermore, as has been demonstrated in a recent work,5 BH4- and H3N-BH3 species are able to generate stibine in a neutral-alkaline pH range. In this case, the role of acidity and acid media is important, contributing to form Sb(III) species that are active toward the hydride transfer from the hydroboron species that are formed in those specific reaction conditions. According to the aqueous chemistry of antimony, by increasing acidity from strongly alkaline to strongly acidic conditions, the speciation of the element should change according to the sequence SbO2- f Sb(O)OH f SbO+, Sb(OH)2+ f Sb3+. Antimonite is unreactive to THB, Sb(O)OH presents limited solubility, while the compounds related to cationic species are very reactive toward THB and also toward the milder amineboranes.5 (43) Chen, Y.-W.; Tong, J.; D’Ulivo, A..; Belzile, N. Analyst 2002, 127, 15411546.

Figure 8. CVG-AAS experiments for Bi using the mixing sequences reported in Figure 2B (L0 ) 0 µL) and C (L0 ) 4-500 µL) and using 0.2 M THB and 0.2 M AB. Reference signal S0 obtained with mixing sequence A, with L1 ) 500 µL and 1 M HCl.

Bismuth. The relative sensitivities obtained from CVG-AAS experiments for bismuth are reported in Figure 8. In this case, it is evident that not all the hydroboron intermediates present the same degree of reactivity toward the possible substrates containing the analyte, the nature of which is controlled by the complex cationic chemistry of Bi3+ and BiO+. The similarity between the behavior of THB and AB at lower acidities (HCl < 0.1 M) indicates that H2O-BH3 and H3N-BH3, respectively, are likely to be the effective derivatizing species in these conditions. The rising of sensitivity in the acidity range of 0.01-0.1 M HCl is due to the formation of analyte substrates that are more reactive toward LBH3n species.5 On the contrary, it appears evident that analyte substrates are not reactive toward species obtained after prolonged decomposition times of THB at 5 and 10 M acidity, for example, L2BH2n and L3BHn. The decrease of sensitivity at higher acidities is more pronounced for AB, indicating even less reactivity of analyte toward the L2BH2n and L3BHn species that are formed by AB hydrolysis at elevated acidities. This experimental evidence confirms the different nature of di- and monohydroboron intermediates formed by acid hydrolysis of THB and AB, as already hypothesized in the sections dedicated to hydrolysis mechanism of THB (see reaction 8) and AB (see reaction 9). Arsenic. The relative sensitivities obtained from CVG-AAS experiments for arsenic are reported in Figure 9. The efficiency of arsine generation is strongly dependent on reaction conditions, including the type of borane reagent. For THB, it is evident that there is a strong dependence on decomposition volume L0, in particular in the acidity range of 0.1-2 M HCl. The effect of acidity on the activation of substrate can be excluded considering the strong influence of L0 at constant Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

6349

Figure 9. CVG-AAS experiments for As using the mixing sequences reported in Figure 2B (L0 ) 0 µL) and C (L0 ) 4-500 µL) and using 0.2 M THB and 0.2 M AB. Reference signal S0 obtained with mixing sequence A, with L1 ) 500 µL and 1 M HCl.

HCl concentration. The effect is therefore due to the presence of suitable hydroboron species at the end of the decomposition volume. The sharp decrease of sensitivity in this intermediate acidity range could be due to the rapid disappearance of LBH3n to form some particular L2BH2n species that is not suitable as a derivatizing species for As(III). Increasing the HCl concentration could produce the formation of different hydroboron intermediates L2BH2n and also the more acid-resistant monohydroboron species, L3BHn. The decomposition rate of this acid-resistant hydroboron species is fairly independent of both acid concentration and time at elevated acidities (HCl > 5 M; see Figures 4 and 5), and this is in agreement with the relative independence of sensitivity from L0 volume at HCl > 5 M. It is worth discussing that the results obtained by THB derivatization with L0 ) 500 µL decomposition volume resemble the behavior of arsenic species in CVG of arsines, reported by other author in the presence of L-cysteine.23,44 In particular, the rising of sensitivity at high acidities was interpreted by Carrero et al. 22 as evidence indicating arsine generation according to the nascent hydrogen mechanism. Apart from the discussion reported in the introduction about the impossibility of generating atomic hydrogen from THB acid hydrolysis, the evidence reported in Figure 9 cannot be explained by the reaction of As(III) with a single derivatizing species, including atomic hydrogen. In the reaction of both THB and AB with As(III), arsenic species appears to be converted more efficiently to arsine by hydroboron species that are more stable in concentrated HCl solution, L2BH2n and L3BHn (L ) H2O, NH3, OH-, Cl-). The results (44) Shraim, A.; Chiswell, B.; Olszowy, H. Talanta 1999, 50, 1109-1127.

6350 Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

Figure 10. CVG-AAS experiments for Se using the mixing sequences reported in Figure 2B (L0 ) 0 µL) and C (L0 ) 4-500 µL) and using 0.2 M THB and 0.2 M AB. Reference signal S0 obtained with mixing sequence A, with L1 ) 500 µL and 4 M HCl.

obtained with AB are quite different from those obtained using THB. In particular, the sensitivity loss obtained between 0.1 and 1 M HCl using THB is not observed using AB. This represents further, convincing evidence confirming that different di- and monohydroboron intermediates are formed by the hydrolysis of THB and AB in strongly acidic conditions, according to the reaction paths reported in reaction 8 and reaction 9. Selenium. The relative sensitivities obtained from CVG-AAS experiments for selenium are reported in Figure 10. The generation of hydrogen selenide is strongly dependent on all the experimental parameters including the type of reagent. With THB, the behavior of Se(IV) is not far from that observed for arsenic. The same conclusion also applies, with the only exception that the higher sensitivities obtained for Se(IV) with respect to As(III) at acidities lower than 0.1 M HCl indicate a more pronounced role of the species LBH3n in the formation of hydrogen selenide. For AB, it appears evident that the formation of hydrogen selenide is mainly originated by its decomposition products, reported in reaction 9, while the role of H3N-BH3 is negligible as indicated by the low sensitivities obtained at HCl < 0.2 M, whatever the size of L0. The dramatic difference between the sensitivities obtained by THB and AB at L0 ) 500 µL further confirms the different nature of di- and monohydroboron intermediates generated by the hydrolysis of THB and AB in strongly acidic conditions. Tellurium. The relative sensitivities obtained from CVG-AAS experiments for tellurium are reported in Figure 11. The low sensitivity obtained for hydrogen telluride generation at pH >0 cannot be addressed to the relatively high acidity of this compound (Ka ) 3 × 10-3). Indeed, at pH 1 and pH 0, the undissociated

Figure 11. CVG-AAS experiments for Te using the mixing sequences reported in Figure 2B (L0 ) 0 µL) and C (L0 ) 4-500 µL) and using 0.2 M THB and 0.2 M AB. Reference signal S0 obtained with mixing sequence A, with L1 ) 100 µL and 4 M HCl.

fraction of H2Te can be calculated to be 97 and 99.7%, respectively. The low sensitivities obtained for Te(IV) at the lower acidities seem to be due to the less pronounced reactivity of LBH3n and some L2BH2n species with respect to the dihydroboron and monohydroboron species that could be formed at higher acidities. This is even more evident from the behavior observed with AB, where the sensitivities approach zero below 1 M HCl and the maximum sensitivity is observed for HCl > 5 M after prolonged decomposition times. The role of H3N-BH3 is therefore negligible, while that of di- and monohydroboron intermediates reported in reaction 9 appears to be here of great relevance. Tin. The relative sensitivity obtained for stannane generation with THB is reported in Figure 12. It can be observed that the maximum sensitivity was obtained in the range of 0.05-0.1 M HCl. In a previous work, it has been demonstrated that stannane could be generated by the direct interaction of BH4- with Sn(IV) in the pH range of 10.6-12.7 and, by the direct interaction of H3N-BH3 with Sn(IV), in the pH range of 6.8-12.5.5 The reactivity of tin with THB reported in Figure 12 indicates that the hydroboron species, which are formed at acidity higher than 1 M HCl, dihydroboron and monohydroboron species are unreactive toward Sn(IV) species, which are present under the same reaction conditions. The most relevant role in obtaining the sensitivities reported can be addressed to LBH3n. For AB, the relative sensitivities obtained by using mixing sequences B and C were much lower than expected. Using the mixing sequence A with AB ) 0.2 M in either 0.1 or 5 × 10-3 M NaOH, S/S0 values in the range of 0.82-0.91 were obtained. Apart from clarification of this anomalous behavior, the trend observed for AB is the same as the one observed for THB. In this case, the

Figure 12. CVG-AAS experiments for Sn using the mixing sequences reported in Figure 2B (L0 ) 0 µL) and C (L0 ) 4-500 µL) and using 0.2 M THB. Reference signal S0 obtained with mixing sequence A, with L1 ) 500 µL and 0.1 M HCl.

predominant role in generating stannane should be ascribed to the species H3N-BH3. Considerations on the Effects of Ligand/Donor Species in CVG. In the present work, the effect of only few donor species, H2O, OH-, NH3, and Cl-, has been considered in the experiments performed on both the kinetics of THB and AB hydrolysis and on the reactivity of the analyte in CVG. In principle, any ligand/ donor species could produce chemical modification in a CVG reaction system, through the formation of complexes with both the analyte and the hydroboron species. It is likely that the formation of hydroboron-ligand complexes could play a significant role in many CVG systems using inorganic and organic additives. As recently reviewed, many organic and inorganic complexes of borane or substituted boranes are commercialized, synthesized, or identified,45 but the specific role they play in a given CVG reaction system cannot easily be predicted as discussed in the few examples reported below. The reactivity of Ge,23 As,22,46 and Sn47 in CVG techniques is strongly influenced by the presence of thiols. Among the effects produced by thiols, the change of pH interval at which the maximum sensitivity is obtained, signal enhancement, and a much better control of interferences, have been well recognized. The hydrolysis of THB in the presence of thiols, to form the thiolborane complex RS-BH3-,23 and the formation of thiol-analyte complexes48 were identified as possible mechanisms contributing to the modification of the reactivity observed in CVG. (45) D’Ulivo, A. Spectrochim. Acta, Part B 2004, 59, 793-825. (46) Chen, H.; Brindle, I. D.; Le, X. Anal. Chem. 1992, 64, 667-672. (47) Chen, H. W.; Yao, W.; Wu, D. X.; Brindle, I. D. Spectrochim. Acta, Part B 1996, 51, 1829-1836.

Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

6351

In the presence of Cl-, Br-, I-, and SCN-, the generation of hydrogen selenide by THB reduction of Se(IV) is improved by a catalytic effect; a better control of interferences has also been observed.49 Also in this case, the role played by the additives in modifying the reactivity of Se(IV) in CVG could be addressed to additive-analyte interaction because selenic acid, H2SeO3, can be converted into complexes of the type SeOX2, SeX4, SeX62- (X ) Cl-, Br-). These halogenated species could be more reactive than H2SeO3 toward those hydroboron species that are formed in the same conditions. However, the experimental evidence collected in this work indicates that chloride also affects the hydrolysis rates of THB and AB, suggesting that the additive-hydroboron species interaction could also play a role in hydrogen selenide generation. In any case, the presently available information does not allow discriminating among different possible hypothesis such as the activation of analyte substrate (analyte-additive complex formation), the formation of suitable hydroboron intermediates (borane-additive complexes), or both the effects taking place simultaneously. The identification of the specific role played by a given additive in CVG could be obtained only through specific experiments. Among them, the study of the kinetics of hydrogen evolution, described in the present paper, and CVG experiments using properly designed mixing sequences, described here and in a previous papers,5,49 seem to be useful complementary tools for the investigation of the mechanism of action of additives in CVG. CONCLUSION There is a strict correlation between the mechanism of THB and AB hydrolysis and the mechanism of formation of volatile derivatives. The stepwise decomposition of borane complexes generates intermediate hydroboron species whose chemical structure is strongly dependent on pH and solution composition. The rate laws reported in the literature for THB and AB hydrolysis in the pH range of 3.8-14 are not valid in the strongly acidic conditions commonly employed in analytical CVG (-1 < pH < 2), where both THB and AB decompose much slower than at pH >3.8. The formation of hydroboron intermediates, which are more (48) Le, X. C.; Cullen, W. R.; Reimer, K. J. Anal. Chim. Acta 1994, 285, 277285. (49) D’Ulivo, A.; Gianfranceschi, L.; Lampugnani, L.; Zamboni, R. Spectrochim. Acta, Part B 2002, 57, 2081-2094.

6352

Analytical Chemistry, Vol. 76, No. 21, November 1, 2004

resistant than BH4- and H3N-BH3 species to acid hydrolysis, is at the origin of the derivatization mechanisms in analytical CVG in acid media. The hydroboron intermediates, still maintaining one or more B-H bonds, are able to perform derivatization processes. The effectiveness of the derivatization process depends on both the nature of the substrate containing the analyte and the nature of hydroboron species. In HCl aqueous media, the species of Hg(II) and Sb(III) are very reactive and unselective toward all the hydroboron species. Other analyte species react preferentially with some particular hydroboron species. Analyte species of Te(IV) and Sn(IV) represent two extreme examples, the former reacting preferentially with those hydroboron species formed in strongly acidic media, while the latter is completely unreactive toward the same hydroboron species. The mechanism of derivatization reaction, in any case, can be addressed to the direct action of one or more hydrogen atoms bound to boron. The nascent hydrogen mechanism can be definitely ruled out by combining all the evidences presented in this paper with those already reported in the literature.4,5,26-28 Therefore, the decomposition of the borane complexes, under appropriate conditions, could be necessary to form not the nascent hydrogen, which cannot be formed by acid hydrolysis of borane complexes, but a wide number of different hydroboron species, BH3, LBH3n, L2BH2n, L2BH2n, and L3BHn (L could be one or more among H2O, NH3, OH-, and Cl-, and the charge n ) -1, 0, +1). One or more of these intermediates present a suitable degree of reactivity toward those analyte substrates that are formed under the same reaction conditions. In strongly acidic reaction media, THB, AB, and other borane complexes could be therefore considered as suitable precursors for the in situ generation of hydroboron species, which could be sometimes more reactive than the precursors themselves in the specific reaction considered. Several other concepts commonly adopted in the interpretation of the reactivity observed in analytical CVG along the past years, as for example the mechanism of action of chemical additives, masking agents and interfering species, should be revised at the light of these new evidences.

Received for review April 23, 2004. Accepted July 27, 2004. AC040078O