The Binomial Distribution of Hydrogen and Deuterium in Arsanes, Diarsanes, and Triarsanes Generated from As(Iii)/[BH n D4-n ]− and the Effect of Trace Amounts of Rh(Iii) Ions

Journal of The American Society for Mass Spectrometry, Dec 2012

Recent studies of the formation of arsane in the borohydride/arsenate reaction demonstrate the occurrence of condensation cascades whereby small quantities of di- and triarsanes are formed. In this study, the isotopic composition of these di- and triarsanes was examined using deuterium labelled borohydrides. A statistical model was employed to construct the mass spectra of all diarsane and triarsane isotopologues (As2H n D4-n and As3H n D5-n ) from the mass spectra of isotopically pure compounds (As2H4, As2D4, As3H5, and As3D5). Subsequent deconvolution of the experimental mixed spectra shows that incorporation of hydrogen closely follows the binomial distribution, in accord with arsane formation. The H/D distribution in arsane, diarsane, and triarsane isotopologues is binomial in the absence of any interference. However, this is significantly altered by the presence of some transition metals; presented here, for the first time, are the effects of Rh(iii). The presence of Rh(iii) in the As(iii)/[BD4]− system entails the incorporation of hydrogen into the arsanes arising from the solvent, altering the expected binomial H/D distribution.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

http://link.springer.com/content/pdf/10.1007%2Fs13361-012-0489-5.pdf

The Binomial Distribution of Hydrogen and Deuterium in Arsanes, Diarsanes, and Triarsanes Generated from As(Iii)/[BH n D4-n ]− and the Effect of Trace Amounts of Rh(Iii) Ions

Enea Pagliano 0 2 Alessandro D'Ulivo 1 Zoltn Mester 0 Ralph E. Sturgeon 0 Juris Meija 0 0 National Research Council Canada , 1200 Montreal Road, Ottawa , ON K1A 0R6, Canada 1 CNR, Consiglio Nazionale delle Ricerche, Istituto di Chimica dei Composti Organometallici , Via G. Moruzzi, 1, 56124 Pisa, Italy 2 Scuola Normale Superiore, Piazza dei Cavalieri, 7, 56126 Pisa, Italy Recent studies of the formation of arsane in the borohydride/arsenate reaction demonstrate the occurrence of condensation cascades whereby small quantities of di- and triarsanes are formed. In this study, the isotopic composition of these di- and triarsanes was examined using deuterium labelled borohydrides. A statistical model was employed to construct the mass spectra of all diarsane and triarsane isotopologues (As2HnD4-n and As3HnD5-n) from the mass spectra of isotopically pure compounds (As2H4, As2D4, As3H5, and As3D5). Subsequent deconvolution of the experimental mixed spectra shows that incorporation of hydrogen closely follows the binomial distribution, in accord with arsane formation. The H/D distribution in arsane, diarsane, and triarsane isotopologues is binomial in the absence of any interference. However, this is significantly altered by the presence of some transition metals; presented here, for the first time, are the effects of Rh(III). The presence of Rh(III) in the As(III)/[BD4] system entails the incorporation of hydrogen into the arsanes arising from the solvent, altering the expected binomial H/D distribution. - G the most robust analytical tools for determination (and eneration of volatile compounds is regarded as one of speciation) of trace and ultra-trace levels of numerous elements [1, 2]. The main advantage of this technique lies with the possibility of separating the analyte from its matrix in the form of a vapor, thus allowing for better sensitivity, detection power, and less interference [2]. Chemical vapor generation using NaBH4 is a common practice in analytical chemistry and it finds routine application for the determination of many elements [3]. Furthermore, other boranes, such as NaBEt4, and more recently, triethyloxonium tetrafluoroborate, have been successfully applied for speciation of organometallic compounds [4] and vapor generation of anionic substrates such as Br, NO3, and NO2 [5, 6]. Despite widespread analytical application of hydride generation, attention to mechanistic aspects of this derivatization reaction has only recently arisen [710]. Reaction of inorganic As(III) with NaBH4 occurs via a concerted multistep process wherein all hydrogen atoms in AsH3 arise from different [BH4] molecules [11]. This mechanism, however, is strongly altered in the presence of noble metals such as Au(III), Pd(II), and Pt(II)their action on hydride generation cannot be completely explained as causing a decrease in the amount of hydride generated (yield interference) [12]; they play an active role in the transfer of hydrogen from [BH4] to the arsenic (mechanistic interference) [13]. Furthermore, another source of complexity in the formation of arsane arises due to condensation reactions that occur at high substrate concentration; the formation of polyarsanes, such as diarsane (As2H4) and triarsane (As3H5), is due to a reaction of the form [14]: Herein, we investigate the formation of these polyarsanes generated under non-analytical conditions in the absence and presence of Rh(III) ions. In particular, this work illustrates how complex mass spectra can be deconstructed and interpreted with the aid of simple mathematical models based on molecular symmetry considerations [15, 16] so as to infer valuable information about the mechanisms of arsane condensation. Reagents and Materials Sodium tetrahydroborate pellets (Alfa Aesar, Ward Hill, MA, USA) and sodium tetradeuteroborate (Cambridge Isotope Laboratories, Andover, MA, USA; x(D) 0 0.99) were employed to prepare stock solutions of 0.2 M NaBH4 and 0.2 M NaBD4 in 0.1 M NaOH. In order to vary the deuterium mole fraction in the reducing agent, different mixtures of the above solutions were also prepared. Arsane generation was performed in septumsealed vials using inorganic As(III) from Oak Ridge National Laboratory (Oak Ridge, TN, USA). Rhodium (III) (1000 mg/L in 10 % HCl) was purchased from SCP Science (Baie DUrf, Qubec, Canada) and used for the interference experiments. All other reagents were analytical grade. GC/MS Method A Hewlett-Packard gas chromatograph (Wilmington, DE) equipped with a Hewlett-Packard 5973 mass spectrometer was operated in splitless mode and fitted with a DB-5 capillary column (30 m0.25 mm i.d. 5 m). The GC was operated under the following conditions: injector temperature: 160 C; oven: isothermal at 35 C for 4 min, 15 C/min up to 200 C; transfer line program: initial temperature 150 C, 30 C/min up to 250 C; He gas flow of 1.2 mL/min was chosen. MS detector was working in positive EI at 70 eV (full-scan mode). The pressure of the source was 2 mPa. Generation of Arsane and Polyarsanes For generating arsane and polyarsanes, the following non-analytical procedure was adopted: 1 mL 800 mg/L As(III) in 1 M HCl was transferred to a screw capped vial (4 mL; Pierce Chemical Co., Rockford, IL, USA) fitted with a PTFE/silicone septum. The vial was closed and the headspace flushed with nitrogen. Hydride generation was subsequently performed by injecting 0.2 mL of 0.2 M [BH4]/[BD4] through the septum. GC/MS analysis was achieved by sampling a 0.5 mL volume of the vial headspace. Interference of Rhodium(III) on Arsane Generation To test the mechanistic interference of Rh(III) on arsane generation, the following procedure was adopted: 2 mL of 1 mg/L As(III) in 1 M HCl +20 L 1000 mg/L Rh(III) in 10 % HCl were placed into a 4 mL vial. After flushing the headspace with nitrogen, 0.2 mL of 0.2 M [BD4] in 0.1 M NaOH was injected into the solution using a plastic syringe fitted with a stainless steel needle to pierce the septum. A 0.4 mL volume of vial headspace was taken for analysis by GC/MS and the mass spectrum of the generated arsanes was monitored. The same procedure, omitting the spike of the interfering metal, was followed for reference purposes. Four independent replicates of this experiment were performed in order to assess the difference between the mass spectrum of the arsanes generated in the presence and absence of Rh(III). Health and Safety Considerations Due to the toxicity of the gaseous arsanes generated, all experiments should be conducted in a fume hood, and extreme care should be taken to avoid any exposure to these substances. Results and Discussion Mass spectra of arsenic hydride isotopologues show significant overlap (e.g., AsD+ and AsH2+, etc.). Moreover, it is not feasible to synthesize pure partially deuterated isotopologues or separate them by conventional gas chromatography. Nevertheless, the determination of the mole fraction of each isotopologue in a mixture can be achieved mathematically by applying a statistical mass balance model [15, 16]. Using this method, it was possible to delineate the mechanism of formation of diarsane and triarsane in the As(III)/[BH4] system and to discover a strong perturbation to the mechanism of arsane generation due to the presence of Rh(III). Statistical Model for the Mass Spectra of As2HnD4-n and As3HnD5-n Statistical reconstruction of partially deuterated hydride has been reported previously for deuterated homologues of diborane [17], methane [18], ethane [19], water, selane [15], arsane, stibane, bismuthane, stannane, and germane [16]. Here, a model proposed by Meija et al. [16] was used for this purpose. For the application of this model to diarsane (H2AsAsH2) and triarsane (H2AsAsHAsH2) it was assumed that the arsenic backbone (AsAs; AsAsAs) behaves as a virtual super-atom; in this way it is possible to consider diarsane and triarsane as tetra- and pentahydrides (EH4 and EH5). A statistical mass balance method bases the reconstruction of EHnDm mass spectra on experimental EHn+m and EDn+m mass spectra. All possible EHnDm fragmentation pathways can be written as combinations of events, such as loss of H (kH,i), loss of D (kD,i), loss of HD (kHD,i), loss of H2 (kH2,i), and loss of D2 (kD2,i). For each reaction, a probability coefficient k is assigned (see Figure 1). The abundance of a particular molecular ion/fragment ion with composition EHnDm can be written as follows: where f(EHnDm) is the sum of all events comprising the fragmentation of EHnDm: nnmCC11 kH;i nmmCC11 kD;i nmC2 nC1 mC1 kHD;i Therefore, f+(EHnDm) is obtained from all precursor fragments, EHn+Dm+. Using the above recursive equations, it is possible to express the observed abundance of each fragment ion in terms of the probability coefficients ki. Figure 1 shows the pictorial representation of all of these events for compounds EHnDm with n + m 5. The numerical values for coefficients ki are obtained from the mass spectra of the EHn+m and EDn+m compounds. As an example, from the mass spectrum of EHn+m, one can write the following: kH;2 fEHnm 1 kH2;1 fEHnm ; From Equation (5) it is possible to obtain coefficients kHi as a function of the observed ion intensities Ii. Unfortunately, considering also the equations arising from EDn+m (for the coefficients kDi), this is always an undetermined system of 2(n + m) equations in 2(2n + 2 m 1) unknown variables. However, one can assign all the possible combinations in ki [0; 1] to an arbitrary set of 2(n + m 1) values for ki, and then calculate the remaining reaction probabilities from Equation (5). To and f+(EHnDm) is the sum of all events comprising the formation of EHnDm from other ions. For the molecular ion f+ 0 1 by definition, whereas for the fragment ions with formula EHnDm the following can be written: fEHnDm f HEHn1Dm fEHn1Dm f DEHnDm1 fEHnDm1 f HDEHn1Dm1 fEHn1Dm1 f H2 EHn2Dm fEHn2Dm f D2 EHnDm2 fEHnDm2: estimate the probability of HD elimination, the following formula is used [16]: The choice of the arithmetic mean to estimate the loss of HD is conventional, and no difference in the results is observed if the geometric mean is utilized instead. In addition, all sets of ki with negative values were discarded in order to obtain physically meaningful results. All sets that satisfy the above condition are then used to reconstruct the mass spectra of the isotopologues. Mass spectra obtained from the various sets of k present no substantial differences and the average results of the constructed mass spectra of As2HnD4-n and As3HnD5-n isotopologues are given in Tables 1 and 2. As a consequence of the statistical model, the individual kset values do not reflect any physical properties of the fragmentation. Rather, they represent a mathematical fit consistent with the input data. However, it is possible to select only k-sets which satisfy certain physical properties of the system. For example, one can discard sets with kH and kD values differing by more than 50 %. This condition is in agreement with the isotopic effect observed in the mass spectra of AsH3 and AsD3 and has previously been used for arsanes [16]. The data reported in Tables 1 and 2 were calculated by omitting this condition, which does not alter the resulting mass spectra of di- and triarsanes. However, larger uncertainties are obtained (corresponding to the worst-case scenario). Although the statistical model suffers limitations when applied to small molecules (it fails to predict the mass spectrum of HOD with great accuracy [15]), it has proven to be effective for heavier molecules such as selane [15], arsane, stibane, bismuthane, stannane, and germane [16]. In practice it is difficult to obtain a mass spectrum of pure EDn+m from NaBD4 in D2O and NaOD or DCl. The omnipresence of H2O in the ambient air inevitably leads aExperimental mass spectrum. Each experimental set of data represents the mean of three independent determinations. Uncertainties represent the standard deviation of replicate measurements. bStatistical model. Each set of data represents the mean of 5107 simulations. Uncertainties represent the standard deviation of replicate measurements. aExperimental mass spectrum. Each experimental set of data represents the mean of three independent determinations. Uncertainties represent the standard deviation of replicate measurements. bStatistical model. Each set of data represents the mean of 8107 simulations. Uncertainties represent the standard deviation of replicate measurements. to a mixture of EDn+m with traces of EHDn+m-1. For example, a typical mass spectrum of As2D4 is 150 151 152 153 154 155 156 157 158 0:425 0:002 0:119 0:008 0:200 0:000 0:004 0:020 0:223 ; where the presence of signals at odd nominal mass values is indicative of the presence of protium-containing isotopologues. In the first iteration of the calculations, the spectrum of As2D4 is taken by simply ignoring the odd-mass signals. Then, the spectrum of As2HD3 is obtained using the statistical model. At this point it is feasible to subtract the obtained spectrum of As2HD3 from the experimental spectrum of crude As2D4 (shown above). This calculation loop is repeated until all odd-mass signals vanish. After three iterations this procedure converged to the following mass spectrum of pure As2D4: continuous lines portray the theoretical trend based on a multistep direct hydrogen transfer from tetrahydroborate to the The mass spectra of all isotopologues of diarsane and triarsane are shown in Tables 1 and 2. Contributions of As2H2D2 and As2H3D are omitted in this iteration procedure due to the extremely low levels of these isotopologues in the crude As2D4. The statistical model allows the calculation of the mass spectra of the isotopologues, as presented in Tables 1 and 2. The model has been validated by comparing generated results to the weighted two-band target entropy minimization model [15, 16] which produced consistent mass spectra of isotopologues of arsane, stibane, and selane. From the reconstructed spectra in Tables 1 and 2, it is possible to determine the quantity of the various isotopologues from their mixtures. This is achieved using the relationship: where Y is the mass spectrum (vector) of the mixture of isotopologues, F is the matrix reporting the mass spectra of the pure isotopologues (Tables 1 and 2) and A is the vector reporting the mole fraction of each component in the mixture. In this work, Equation (7) was solved for A with a least squares method using the LINEST function in Microsoft Excel. Moreover, agreement between the gravimetric mole fraction of deuterium in the reducing agent (NaBH4+ NaBD4) and the net mole fraction of deuterium in the generated isotopologue mixtures as calculated from the statistical model (R2 90.98) is obtained. Scheme 1. Formation of di- and triarsanes in a concerted reduction and condensation cascade (R 0 H or OH). Mechanism of Formation of Polyarsanes Reduction of arsenates and methylarsenates with boranes is a concerted multistep reaction wherein all OH groups from the As-substrate are replaced by H groups originating from different borane molecules [11]: AsOH3 3BX3H ! AsH3 3BX3OH The incorporation of hydrogen into the final hydride can be monitored using tetrahydroborate enriched in deuterium via plots of the mole fraction of each isotopologue generated versus the mole fraction of deuterium in the tetrahydroborate reductant [8]. The stoichiometry of the above reaction is respected only when arsenic compounds are present at trace levels in an excess of boranes (i.e., under analytical conditions). When the concentration of arsenic rises above the millimolar level, however, a complex set of side reactions takes place [14] with formation of condensation products: AsX2 ! X2As AsX2; X2As AsX AsX2 BH4 As2H4; As3H5X H; OH: In this work, we have investigated for the first time the mechanism of incorporation of hydrogen into the di- and triarsanes arising from these condensation cascades. Experimental conditions are chosen to avoid any unwanted H/D exchange with the solvent (H2O) [20].The mole fraction of deuterium in the NaBH4 and NaBD4 mixtures was obtained Table 3. Composite mass spectra of arsanes generated in the reaction of As(III) with [BD4] in the presence and absence of Rh(III) ions (A) and the corresponding isotopologue distribution (B) aEach set of experimental data reflects the mean of four independent determinations. Uncertainties represent the standard deviation of replicate measurements. bThe experimental isotopologue distribution is compared with the calculated binomial distribution corresponding to the same level of deuterium incorporation, x(D). gravimetrically and the mass spectra of the resultant di- and triarsanes were deconvoluted to obtain the mole fraction of each isotopologue. Results are shown in Figures 2 and 3. It is evident that the incorporation of protium and deuterium in di- and triarsanes closely follows the binomial distribution, i.e., as illustrated by the continuous line in Figures 2 and 3. The binomial distribution originates from the multistep hydrogen transfer, as outlined for diarsane in Figure 4. The trends observed in the composition of diarsane and triarsane follow the hypothesis of a concerted multistep formation of arsane in a fashion already demonstrated for AsH3 [11], and they further validate the proposed condensation cascade mechanism [14]. Therefore, the formation of di- and triarsanes can be described as a combination of two independent events: condensation of two hydroxoarsanes and their reduction by borohydrides via the concerted transfer of hydrogen to the arsenic substrate, as summarized in Scheme 1. Mechanistic Interference from Rhodium(III) The interaction between transition elements and tetrahydroborate is of great importance in chemistry, not just in analytical chemistry. Tetrahydroborate is of interest due to its capability to store hydrogen and its hydrolysis is catalyzed by noble and transition elements [21, 22]. Reaction of transition elements with tetrahydroborate is also employed in the synthesis of nanoparticles [23, 24] and to generate volatile derivatives suitable for analytical purposes [10]. Noteworthy is also the possibility of using transition elements, such as Co(II), Ni(II) [25], and Rh(III) [26], to modify the reducing power of the tetrahydroborate. For example, [BH4] cannot reduce nitriles, amides, and olefins, but it can do so in combination with transition element halides [27, 28]. Recent observations of the effects of Au(III), Pd(II), and Pt(II) on generation of arsane [13] confirm a peculiar chemistry in this system, and in this study we report a similar behaviour with Rh (III). The first macroscopic observation regarding the interference of Rh(III) within the As(III)/[BD4] system is the increased incorporation of hydrogen from the solvent into the arsanes. Indeed, when As(III) is derivatized to arsane by [BD4] in aqueous media, AsD3 is the major isotopologue formed, 86 %, while trace (G1 %) AsH3 is observed. The presence of rhodium alters this composition to 63 % AsD3 and 13 % AsH3. Although this phenomenon could be interpreted as an H/ D exchange of the arsane catalyzed by the presence of rhodium, it would be inconsistent with the observation that the injection of preformed AsD3 into a vial containing Rh (III)/[BD4] does not alter the integrity of AsD3. This evidence suggests that this H/D exchange occurs during the generation of arsane. Therefore, the presence of Rh(III) causes a strong perturbation to the mechanism of hydrogen transfer from tetrahydroborate to arsenic, creating a loss in the typical binomial character (Table 3), i.e., the amount of AsH3 (among the other isotopologues) increases significantly in the presence of Rh(III). A possible interpretation of the results shown in Table 3 can be achieved by considering the reductive process as the sum of two competing events: the reduction by [BD4] which is binomial, and the reduction by a hydride-borane-Rh intermediate, present in the early stages of the reaction, and able to exchange hydrogen with the solvent. The postulation of the formation of a new reducing agent, with differing reductive power with respect to [BD4], is also supported by observations reported in the organic chemistry literature, as noted earlier. It was not possible to study how the mechanistic interference of rhodium ions propagates from arsane to diand triarsane generated via the condensation cascades under non-analytical conditions. This was due to the fact that the large concentration of As(III) necessary to produce polyarsanes creates prohibitive conditions for Rh(III), leading to significant precipitation and significant reduction in the yield of the volatile arsanes.


This is a preview of a remote PDF: http://link.springer.com/content/pdf/10.1007%2Fs13361-012-0489-5.pdf

Enea Pagliano, Alessandro D’Ulivo, Zoltán Mester, Ralph E. Sturgeon, Juris Meija. The Binomial Distribution of Hydrogen and Deuterium in Arsanes, Diarsanes, and Triarsanes Generated from As(Iii)/[BH n D4-n ]− and the Effect of Trace Amounts of Rh(Iii) Ions, Journal of The American Society for Mass Spectrometry, 2012, 2178-2186, DOI: 10.1007/s13361-012-0489-5