摘要:

OCTOBER, 1969 THE ANALYST Vol. 94, No. I 123 Controlled-potential Coulometers Based Electronic Units Part I. Development of Equipment* upon Modular BY G. PHILLIPS AND G. W. C. MILNER (Analytical Sciences Division, Atomic Energy Research Establishment, Harwell, Near Didcot, Berkshire) The construction of controlled-potential coulometric equipment is described in which modular electronic units are used. A digital voltmeter is used with the equipment for measuring and integrating the electrolysis current. It is readily adaptable and can be used manually or automatically, with a sample changer and print-out unit. THE construction of new electronic controlled-potential coulometers has been reported by several authors,lJJ~~ following the original paper by Boomans in 1957. In each case the instrument is designed to perform two basic functions during the electrolysis.Firstly, it maintains the potential of the working electrode at a selected value with reference to a standard electrode in contact with the solution, and, secondly, it integrates the electrolysis current passing during each determination. The most commonly used integrator takes an analogue proportion of the electrolysis current and integrates this in a resistance - capacity circuit in conjunction with an operational amplifier. This amplifier is a critical component of the coulometer, and needs careful maintenance to ensure low drift. It has the disadvantage of going out of balance when the coulometer is overloaded. In a more recent alternative integrator an a.c. frequency converter is used to convert a d.c.voltage into an a.c. signal, which is measured with an electronic counter.6 An instrument based on this principle has been built .and tested at A.E.R.E., Hanvell, and details are given in this paper. Successful applications of this instrument in analysis are reported in Part I1 (page 840). DESCRIPTION OF THE COULOMETER- The development of digital methods of data handling and the need for flexibility in the construction of complex experimental equipment has led to the availability of comprehensive unitised systems of electronic equipment. One such system is the Harwell-designed “2000 series” of electronic units.’ A selection of these units, together with a coulometer unit built to the same design principles and a Solartron digital voltmeter, have been used to construct a controlled-potential coulometer.A circuit diagram of the coulometer unit (2220) is shown in Fig. 1. The three-electrode electrolysis cell is shown schematically on the extreme left of the diagram. The potentiostat compares the potential difference between the working electrode and the reference electrode with that of an adjustable reference potential derived from a zener diode circuit (designated “set volts” in Fig. 1). In operation any difference between these two potentials is adjusted to zero by a current passing through the electrolysis cell via the working electrode and counter electrode. This current passes through a lOO-Ll precision resistor, which is connected in the input circuit of the digital voltmeter. The other circuits in the coulometer unit comprise a timing circuit to control the sampling rate of the digital voltmeter and an on - off system operated by a manual remote relay. A unit block diagram of the controlled-potential coulometer in the manually operated form, showing signal connections, is given in Fig.2. The five “2000 series” units used fit on a shelf, which can be accommodated within a standard 19-inch rack. The coulometer unit 2220 has three operational modes, viz. , “set volts,” “stand by” or “remote,” and “operate.” * Paper presented at the Second SAC Conference 1968, Nottingham. 0 SAC; Crown Copyright Reserved. 833834 PHILLIPS AND MILNER : CONTROLLED-POTENTIAL COULOMETERS BASED [Autalyst, Vol. 94 - Potentiostat -Timing and D.V.M. trigger circuit-Manual/Remote+ on - off system --IN Fig.1. Circuit diagram of the coulometer unit 2220: W, working electrode (platinum, mercury or silver) ; C, counter electrode (usually platinum) ; R, reference electrode; and D.V.M., digital voltmeter (for values of components see Appendix I) In the “set volts” position the required working potential is set by adjustment of a 10-kSZ helipot on the front of the unit. The value of this potential is displayed on the digital volt- meter by using the “auto” operational mode. In the “operate” position the coulometer Solartron digital voltmeter LM 1420.2 Register Scaler Coulometer Clock-pulse U Power I_ unit 2103 21 I2 2220 generator 2015 2 L43 Fig. 2. Block diagram of the controlled-potential coulometer unit 2220 receives a train of pulses from the clock-pulse generator 2143 at a rate of 5 s-1.The pulses are inverted and used to operate the “remote sample” of the digital voltmeter at a rate of 2.5 s-1. The electrolysis is initiated with the first pulse received after switching on and is continuous until the coulometer unit is switched to the “stand by” position. The electrolysis current passes through a 100.0-R precision resistor in the coulometer unit, the potential drop across which is sampled by the digital voltmeter. Sampling of the electrolysis current commences with the second pulse received after switching on and subsequently with every other pulse until the coulometer unit is returned to the “stand by” position. A countOctober, 19691 UPON MODULAR ELECTRONIC UNITS. PART I.DEVELOPMENT OF EQUIPMENT 835 corresponding to the integral of the area under the electrolysis current - time graph is accumulated serially in the scaler 2112 via a control and gate circuit connected to the rear of the digital voltmeter (Fig. 3). The scaler records four times the number of digits indicated on the digital voltmeter display. In operation, therefore, the digital voltmeter displays current readings, which can be used to follow the course of the electrolysis visually, and the scaler displays the current integral. With the digital voltmeter on the 2-V range, a reading of 1 V corresponds t o a current of 10 mA, and a reading of 0-001 V to a background current of 10 PA. The digital voltmeter can be calibrated internally against a standard cell and the coulometer is, therefore, inherently accurate and, with the setting described above, gives lo6 counts C-l.Any change in the range setting of the digital voltmeter, or of the value of the precision resistor, or of the sampling rate set by the clock-pulse generator, produces a pro rata change in the calibration value. With the above settings the scaler can accumulate 2 x 106 counts or about 20 micro-equivalents; the overflow register shown is necessary only above this level. It is essential for accurate results and optimum precision to match the electrolysis current to the dynamic range of the digital voltmeter. On the 2-V range this is from 22-99mA to 10pA. The range can be extended in the upward direction by using a 50-042 precision resistor in the electrolysis circuit, and in the downward direction by the range settings on the digital voltmeter.Sample D.V.M. board I LM 1420-2 Binary coded decimal information from digital voltmeter - Fig. 3. Control and gate circuit LM 1420.2 (for values of components see Appendix 11) AUTOMATIC OPERATION- Controlled-potential coulometric titrations are normally terminated when the current has fallen to a stable background level of at least three orders of magnitude lower than the initial current. For a given reaction, the time required to reach this level is dependent mainly upon the surface‘area of the working electrode, the volume of electrolyte and the rate of stirring and, to a much lesser extent, upon the temperature and the concentration of the material being electrolysed. If the parameters of the electrolysis cell are fixed, it is feasible to terminate a controlled-potential coulometric titration after a fixed time. The coulometer unit has been designed to operate either manually or by means of suitable binary levels when set in the “stand by” position.These levels can be obtained from a “2000 series” pre-set scaler 2166 connected to receive pulses from the clock-pulse generator and to give a 5-V negative-going level at the end of the set time. The working potential is set manually as described above and the coulometer unit switched to the “stand by” position. The electrolysis time required to reach the background current level is then set on the pre-set scaler 2166. Electrolysis is initiated by means of the re-set button on the scaler 2166, which continues to operate until the pre-set time has elapsed. The equipment then switches off leaving the current integral displayed on the scaler 2112.A controlled-potential coulometric determination is usually a two-step process, consisting of a pre-electrolysis at one potential, to condition the electrolyte, to remove an interfering reaction or to adjust the valency of the species of interest, followed by a change of potential to enable the electrolysis of the required constituent to be carried out. *4 fully automatic coulometer must, therefore, be able to apply two different pre-set potentials to the working836 PHILLIPS AND MILNER : CONTROLLED-POTENTIAL COULOMETERS BASED [Analyst, Vol. 94 electrode in a timed sequence. This can be achieved by arranging two single-stage automatic units to operate in sequence, as shown in Fig.4. The power unit and clock-pulse generator are common to both stages, and a small switching unit is required to switch the operation of the digital voltmeter from the first coulometer unit to the second. The sequence of operations is initiated by using the re-set button of register No. 1, which re-sets all integrating scalers and the pre-set timing scaler No. 1. Coulometer No. 1 commences to operate, and electrolysis at the first selected potential continues for N units of time set on pre-set scaler No. 1. After that time coulometer No. 1 is switched off, and pre-set scaler No. 2 is re-set automatically by the (N + 1)th pulse. Coulometer No. 2 then commences to operate at the second selected potential for N’ units of time set on pre-set scaler No.2. Coulometer No. 2 is switched off at N’ and electrolysis ceases, leaving the integral of each electrolysis step displayed on the corresponding scalers. A = B = C = D = E = F = G = N 0 o o o o o ( c M 1 Digital voltmeter LM 1420.2 Integrating scaler 21 17 la) Integrating scaler 21 17 [ I ) Coulometer (I ) Pre-set scaler 2 I66 (2) Pre-set scaler 2166 (I) Integrating scaler 21 17 (2a) H = K = L = M = N = P = J = Integrating scaler 21 17 (2) Coulometer (2) Gated clock-pulse generator 2166 Power unit 2015 Print-out control unit 2140 Stirrer control Sample changer Fig. 4. Automatic controlled-potential coulometer with sample changer The equipment can also be arranged to work in conjunction with a printer and sample changer (Fig.5) with the additional units shown in Fig. 4. The cycle of events is initiated manually by means of the start button on the print-out control unit 2140. This prints and re-sets the integrating scalers and triggers and sample-changing mechanism. The sample- changing mechanism raises the stirrer, gas line and electrode assembly clear of the electrolysis cell and rotates the next cell into position. The electrode assembly and attachments are then lowered into the new solution. The mechanism can be set to perform this operation once per cycle, or twice to permit rinsing of the electrode assembly. The time between the com- pletion of one electrolysis and the re-setting of pre-set scaler No. 1, which initiates the nextFig. 5. Sample changer [To face page 836October, 19691 UPON MODULAR ELECTRONIC UNITS.PART I. DEVELOPMENT OF EQUIPMENT 837 electrolysis, can be varied up to a maximum of 63 minutes. The time required for the mechanical movements is about 1 minute or less; the additional delay allows time for the removal of oxygen from the solution. The turn-table holds twenty electrolysis cells, and hence twenty automatic coulometric titrations can be carried out on irreversible systems not requiring a rinsing step, or ten on reversible systems needing rinsing of the electrodes. In this application the scaler units are changed to type 2117 to allow print-out access, and the gated clock-pulse generator 2116 is substituted for the 2143 to permit a delay to be introduced between determinations.ELECTROLYSIS CELLS- (a) Manua2 sample changing-Two types of electrolysis cells have been used with the above coulometer, one with a platinum-gauze working electrode, the other with a mercury-pool electrode. The cell for use with a platinum-gauze working electrode was of the type pre- viously described,2 in which cation or anion-exchange membranes (Permutit C20 or A20) were used to separate the three compartments. In practice it was found that leaks rapidly developed from the edges of the membranes because of attack by hydrochloric acid on the polystyrene cement used to seal the edges. This was replaced with a small circular gasket cut from &-inch thick neoprene sheet with cork borers. The cell for use with a mercury pool was of the type described by Jones, Schults and Dale.8 (b) Automatic sample changing-The cell used in the automatic sample changer for controlled-potential electrolysis with a platinum electrode or a mercury pool consists of a Perspex open-topped cylinder, 1& inches i.d.and 14 inches high. The outside of the cylinder is ledged to locate the cell in holes around the perimeter of the turn-table. The stirrer, gas inlet and electrode assembly are accommodated in a polythene cover which, in the lowered position, rests on top of the electrolysis cell. Possible damage to the electrolysis cells or electrode assembly by overdriving is avoided by means of a slipping clutch in the drive mechanism. When used with a mercury-pool electrode the cell requires 8 ml of mercury and the same volume of solution. For use with a platinum-gauze electrode a solution volume of 14 ml is required.INSTRUMENTAL OPERATIONAL PROCEDURES- (a) Electrolysis at a controlled potential-Switch the digital voltmeter to “auto” operation on the 2-V range, and the coulometer unit to “set volts.” Adjust the potentiometer control on the coulometer unit until the required working potential is displayed on the digital volt- meter. During this procedure random counts are recorded by the scaler and can be ignored. Switch the coulometer unit to “stand by” and the digital voltmeter to “manual” operation. Clear the scaler and register, and switch the coulometer to “operate.” The initial voltage readings are within the range of the digital voltmeter, i.e., less than 2.300 on the 2-V range. Allow the electrolysis to proceed until a stable background current is achieved, then switch the coulometer unit back to “stand by.” Note readings on the scaler and register, and calculate the weight of material electrolysed as follows- w= N x A x F ryt x 96,487’ where W is the weight of material electrolysed, g; N , the scaler and register readings; A , the atomic weight ; F, the coulometer factor, C counts-1; and n is the number of electrons involved in the reaction.(b) Determination of the Eo‘ of a reversible redox system-The necessary results for the construction of a coulogram and determination of the Eo‘ of a reversible redox system can be obtained by using the procedure described in (a) repetitively at intervals of working potential over the critical range. This is time consuming, as it is necessary to wait for equilibrium to be achieved at each applied potential value.The process can be speeded up by using the following procedure. Switch the digital voltmeter to manual operation, clear the scaler and register and switch the coulometer to operate. Adjust the potentiometer control on the coulometer unit until electrolysis commences at a slow rate, e.g., about 10 per cent. of the maximum rate. Allow electrolysis to proceed for about 1 minute, then carefully reverse the potential applied until the electrolysis rate is reduced to an insignificant level, or stops completely. Switch838 PHILLIPS AND MILNER : CONTROLLED-POTENTIAL COULOMETERS BASED [ A nalyst, Vol. 94 the scaler off and note the reading. Switch the coulometer unit to “set volts” and the digital voltmeter to “auto,” and note the applied potential.Return the coulometer unit to “stand by” and the digital voltmeter to “manual,” and switch on the scaler. Switch the coulometer to “operate,” and repeat the above process until no further electrolysis is obtained on increasing the applied potential. Plot the counts recorded against the applied potential or plot log Count = [oxidised] t Count = [reduced] against the applied potential and read off the Eo’ value. (c) Calibration check-With the digital voltmeter accurately calibrated on the 2-V range and the clock-pulse generator set to deliver pulses a t 200-millisecond intervals the coulometer should record 1 count per 10-6 C, or at 100-millisecond intervals 1 count per 5 x lo-’ C.This condition can be readily confirmed by substituting a 1 *35-V mercury cell and 100-0-Q precision resistor for the electrolysis cell, as shown in Fig. 6. On switching the coulometer unit to “operate” and allowing a few minutes for stabilisation, a constant current is obtained. The count accumulated in a fixed interval of time can then be noted and related to the number of coulombs passed. This procedure can be repeated at various current levels by adjustment of the potentiometer on the coulometer unit. This calibration check can be carried out manually or automatically. In the latter case, however, it is better to transfer time control to the scaler 2112, leaving the coulometer in continuous operation in order to avoid transient effects on switching on or off.This calibration procedure has been carried out repeatedly with a precision of e0.03 per cent. (coefficient of variation) and an accuracy within the same limits. Reference electrode (lead) Counter electrode (lead) Working electrode (lead) precision resistor 1-35 V mercury battery Fig. 6. Circuit for calibration check (d) Test of equipment with standard solutions-The equipment has been tested with standard solutions of iron, uranium, plutonium and copper. The results obtained are shown in Table I. TABLE I RECOVERIES OBTAINED WITH STANDARD SOLUTIONS Weight taken, mg Iron (Specpure), 1-592 Uranium, 9.680 Plutonium, 2.058 Uranium, 1.302 Copper, 1.009 *Uranium, 4.864 *Iron, 1-446 Precision mg per cent. freedom Weight found, (coefficient of variation), Degrees of .. 1.589 0.15 6 ..9.585 0.18 8 .. 2.059 0.10 5 .. 1.303 0.44 9 .. 1.009 0.20 8 .. 4-859 0.34 11 .. 1-442 0.18 11 * With the automatic sample changer. DISCUSSION The use of a modular system of construction for controlled-potential coulometric equipment has a number of advantages. The system can be designed by the user to be operated manually, semi-automatically or completely automatically to suit the nature of the work to be carriedOctober, 19691 UPON MODULAR ELECTRONIC UNITS. PART I. DEVELOPMENT OF EQUIPMENT 839 out. The method of integration used is free from the overloading difficulties associated with operational amplifier integrators used in most coulometric equipment. Skilled electronic knowledge is not required by the user, and faults can be rapidly identified and overcome by unit replacement.The equipment is more suitable for routine operation but can be used manually for the development of analytical controlled-potential coulometric procedures. The current limit of the equipment is 46mA and it is most suitable for determinations in the region of 20 micro-equivalents. There is no instrumental lower limit other than that set by background current effects. Appendix I LIST OF COMPONENTS FOR FIG. 1 = 100-R f 0.1 per cent. precision resistor = 1-MR resistors = 470-M resistors = 8.2-kR resistor R,, R,, R,,, R,,, G4, R,, R,,, R,, = 100-kR resistors = 100-R, l-W wire-wound resistor = 10-kQ resistors RlZ = 39-kn resistor RlsJ %a* %D* Rs7 = 2-7-kR resistors R14 = 10-kR helipot resistor RlS, R 2 7 , Rso, RS, = 4-7-kR resistors R 1 6 p R17 = 2.2-kR resistors RZOJ R,zJ Rsl, R88 = 18-kR resistors = 1-5-kR resistor = 22-kR resistors = 0-l-pF capacitor = l-pF capacitors = 0-Ol-pF capacitor = 150-pF capacitors c,, c,, c g , ClO, c,, = 56-pF capacitors Tl, Tl, = BCY31 transistors T,, Ts, T,, T, = C l l l transistors T4, T,, T,, T7 = OA2240 transistors Tll, T14 = AS213 transistors = AS221 transistors Tl, = 2N1309 transistor Tl, = AAYl1 transistor T20 = ACY19 transistor Rl %l Rs R, R7 R 4 2 R1lJ %6 R1oJ &?6) Rs8 %8 Cl CS c,, C8 RS& % J R41 ‘2, ‘4 Appendix I1 LIST OF COMPONENTS FOR FIG. 3 = 4.7-kQ resistor = 12-kSZ resistor = 27-kR resistor R4 = 8-2-kSZ resistor R,, R, = 18-kR resistors = 3.3-kR resistor = 47-kR resistor = 10-kR resistor = 470-R resistor = 390-pF capacitor = 47-pF capacitor Rl R, RS R7 R* R, R1o Cl CZ T,, T,, Ts = 2N1309 transistors T4 = AAYll transistor REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. Kelley, M. T., Jones, H. C., and Fisher, D. J., Analyt. Chern., 1959, 31, 488 and 956. Milner, G. W. C., and Edwards, J. W., U.K. Atomic Energy Autlzority Reflort A.E.R.E. R.3772, 1961. Rockett, J. J., U.K. Atomic Energy Authority Report A.E.R.E. R.3784, 1961. Harrar. J . E., and Shain, I., Analyt. Chem., 1966, 38, 1148. Booman, G. L., Ibid., 1957, 29, 213. Bard, A. J., and Solon, E., Ibid., 1962, 34, (9) 1181. Bisby, H., J. Instn Electronic and Radio Engrs, 1965, 29, 185. Jones, H. C., Shults, W. D., and Dale, J. M., Analyt. Chem., 1965, 37, 680. Received December 18th’ 1968 Accepted Afiril 4th, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699400833

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

840 Analyst, October, 1969, Vol. 94, fifi. 840-843 Controlled-potential Coulometers Based upon Modular Electronic Units Part 11. The Determination of Ruthenium by Controlled-potential Coulometry* BY G. WELDRICK, G. PHILLIPS AND G. W. C. MILNER A method is described for the determination of ruthenium by controlled- potential coulometry. Quadrivalent ruthenium as the binuclear [Ru20]8+ chlorocomplex in 5 M hydrochloric acid is reduced to tervalent ruthenium at the platinum electrode at a potential of +0.05 V versus the S.C.E. Milli- gram amounts of ruthenium can be determined with a coefficient of variation of 1.0 per cent. The method has been applied to uranium - ruthenium alloys and uranium carbide - ruthenium cermet materials. (A.IzalyticaZ Sciences Division, Atomic Energy Research Establishment, Harwell, Near Didcot, Berkshire) THE possibility of determining ruthenium by controlled-potential coulometry resulted from an observation in this laboratory that ruthenium interfered in the coulometric determination of uranium.Hydrochloric acid solutions appeared to be the most suitable for further investi- gation because the chlorocomplexes of ruthenium have been characterised and information is available on the absorption spectra of these comp1exes.l At the commencement of this study no references were available in the literature on the determination of ruthenium by coulometry, but during the course of this work Stenina, Krylov and Agasyan published controlled-potential2 and constant-current coulometric3 methods for ruthenium. The work of these authors is based, however, upon a mononuclear chlorocomplex of ruthenium, different from that used in our work.EXPERIMENTAL COULOMETRIC INVESTIGATION OF RUTHENIUM IN CHLORIDE MEDIA- Preparation of solutions-Ruthenium solutions, suitable for coulometry, can be prepared by alkaline fusion of ruthenium metal followed by extraction of the cooled melt with hydro- chloric acid. For this purpose 100-mg amounts of Specpure ruthenium sponge were fused at 600" C with a mixture of sodium peroxide and sodium hydroxide in an alumina crucible. The cooled melt was extracted into 5 M hydrochloric acid. This solution, which was brown, gave the absorption spectrum shown in curve D of Fig. 1, with peaks at 385 and 470 nm. This spectrum is typical of the quadrivalent ruthenium polynuclear [Ru,O] 6+chlorocomplex, but the presence of minor amounts of other ruthenium chlorocomplexes could not be ruled out solely on spectrophotometric evidence. Coulometry at a mercury electrode-Aliquots of the above solution containing about 1 mg of ruthenium were transferred to a coulometric cell containing a stirred mercury electrode.It was found that the solution could be readily reduced and the brown colour discharged at potentials of about 0 V versus the S.C.E. This reduction corresponded to a one-electron change, but was not reversible within the range of application of the mercury-pool electrode. The reduced solution, which was colourless, had the absorption form shown in curve E of Fig. 1, and was typical of ruthenium(II1). Further reduction of the colourless solution was possible at potentials more negative than OV versus the S.C.E., producing an intense blue colour, but the reaction did not proceed at 100 per cent.current efficiency, presumably because of the discharge of hydrogen ions. The reduction was reversible at the mercury pool, but it was not suitable for quantitative work. This preliminary work with the mercury- pool electrode indicated that further studies on the coulometry of ruthenium in chloride solutions would be more conveniently carried out with a platinum electrode. Couulometry at a Platirtum ebctrode-Aliquots of the above solution containing about 1 mg of ruthenium were transferred to a coulometer cell containing a stationary platinum-gauze working electrode. Coulograms for ruthenium were plotted over the acidity range from 1 to 7 M hydrochloric acid.It was found that the reactions were more rapid in the more * Paper presented at the Second SAC Conference 1968, Nottingham. 0 SAC; Crown Copyright Reserved.WELDRICK, PHILLIPS AND MILNER 841 Wavelength, nm Fig. 1. The absorption spectra for chlorocomplexes of ruthenium(III), (IV) and (VI) : A, ruthenium(VI), R u O ~ ~ + ; B, ruthenium(IV), Ru4+; C, ruthenium(IV), Ru202*+; D, ruthenium(IV), Ru20s+; and E, ruthen- ium(III), Ru3+ concentrated acid solution, but with 7 M hydrochloric acid the cation-exchange membrane separators between the electrolysis cell compartments were attacked. With 5 M hydrochloric acid, however, less difficulty was experienced and the reactions were sufficiently rapid to be useful for the analysis. In 5 M hydrochloric acid the ruthenium(1V) [Ru2OI6+ chlorocomplex - - I u 1-0 Applied potential, V versus S.C.E.Fig. 2. Coulograms of ruthenium in 5 M hydrochloric acid was readily reduced, with an E,' of 0-365 V zleysus the S.C.E. (curve A, Fig. 2), giving a colour- less solution. This reaction was not reversible, but re-oxidation to a brown coloured solution could be achieved at a much higher potential with an E,' of 0.805 V zlemwus the S.C.E. (curve B, Fig. 2). This second reaction was reversible (curve C, Fig. 2). It was apparent from these reactions that the ruthenium species in the second brown solution, obtained by electrochemical oxidation of Ru3+, was not the same as the species in the original solution. An absorption spectrum on the second brown solution confirmed the presence of Ru4+ (curve B, Fig.1). The coulometric measurements indicated that all of these reactions involved one-electron changes, but the reversible Ru3+ - Ru4+ change was not faster than the irreversible reduction of the ruthenium(1V) [Ru20]6+ chlorocomplex, which could be achieved in 45 minutes in a 10-ml cell, or 30 minutes in a 5-ml cell. From these investigations a method for determining ruthenium, based upon fusion of the sample in alkaline peroxide, extraction into 5~ hydrochloric acid and a controlled- potential coulometric reduction at the platinum electrode, appeared to be feasible. The842 WELDRICK, PHILLIPS AND MILNER : CONTROLLED-POTENTIAL COULOMETERS [Analyst, Vol. 94 coulometric reduction was slow, taking about 45 minutes in a volume of lOml, compared with the more usual 20 to 25 minutes for a system such as Fe3+ - Fe2+.However, the fusion and dissolution were quite rapid and, as no valency adjustment or separation was required, the over-all time for analysis was reasonable (about 90 minutes). In an attempt to decrease the time required for the coulometric reduction, the possibility of using the Meitess end-point procedure was investigated. This procedure is dependent upon the electrolysis current obeying an equation of the form where i, is the initial current and it the current after time t. It was found that a current - time plot of the reduction of ruthenium showed two distinct rates of reaction, the first rapid and complete in less than 10 minutes and corresponding to about 50 per cent.of the titre, the second slow and complete only after about 45 minutes. With such a system the Meites end- point technique cannot be applied to the first slope, which is curved, and there appeared to be little advantage to be gained from applying the technique to the second slope. It is interesting, however, to compare this finding with that of Stenina and Aga~yan,~ who postulate an inter- mediate binuclear compound of ruthenium( IV) and ruthenium( 111) during the constant- current coulometric reduction of ruthenium( IV) with the electrogenerated Ti3+. Quantitative determinations of ruthenium in ruthenium metal, uuranium - ruthenium alloy and uranium - ruthenium carbide-The E,' value found for the reduction of quadrivalent ruthenium as the [Ru,0l6+ chlorocomplex indicates that it should be possible to achieve 99-9 per cent.reduction at an applied potential of +0.185 V versus the S.C.E. This was attempted with 3-mg amounts of ruthenium, prepared as described above, in 5 M hydrochloric acid. It was found that at this potential the electrolysis current remained in the region of 50 to 100 pA for a considerable time, and that the time required to reduce the final 1 to 2 per cent. of ruthenium(1V) was unacceptably long. The rate of electrolysis could be increased and background currents of less than 10pA could be achieved if the applied potential was decreased to +0.05 V versus the S.C.E. At this potential, electrolysis times of 30 minutes were achieved and spectrophotometric examination of the reduced solution showed no trace of quadrivalent ruthenium.This over-potential, which is typical of reactions of irreversible behaviour, was used in all subsequent work. Blank values determined on aliquots of solution prepared in exactly the same way as the ruthenium solution were found to be about 1 per cent. of the titre for a determination based upon 3 mg of ruthenium. No evidence was found for the presence of kinetic or induced blanks. DISSOLUTION OF RUTHENIUM-CONTAINING SAMPLES- Crush and grind the sample to pass through a 100-mesh sieve. Weigh a portion containing not more than 100 mg of ruthenium and add it to 0.5 g of sodium peroxide contained in a 15-ml recrystallised alumina crucible. Mix by rotating the crucible at an angle of 45".Add 0.5 g of sodium hydroxide, in pellet form, to the contents of the crucible. Place the crucible in a muffle furnace at 600" C and cover with an alumina lid. Remove the crucible from the furnace after 15 minutes and allow it to cool. Add 1 rnl of water to the melt and replace the lid for 1 minute. Add a further 0.5 ml of water and swirl the contents of the crucible. Warm gently and continue swirling the crucible until the melt is completely dispersed. Add the solution, dropwise, to 10 ml of 5 M hydrochloric acid in a 100-ml beaker. Rinse the crucible with two 1-ml portions of concentrated hydrochloric acid followed by two 1-ml portions of water and add the rinsings to the beaker. Warm the solution in 50 to 75 ml of 5 M hydro- chloric acid on the hot-plate for about 10 minutes.Cool, and dilute to give 100 ml of solution 5 M in hydrochloric acid. Controlled-fiotential coulometric determination of ruthenium-Take an aliquot of the solu- tion prepared as described above containing 2 to 4mg of ruthenium, and transfer it to the electrolysis cell. Clean the platinum electrode for every determination by boiling it in nitric acid, rinsing in distilled water and igniting in the flame of a Meker burner. Carry out pre- electrolysis at +0.6V versus the S.C.E. until a background current of less than 10pA is obtained. Adjust the applied potential to +0.05 V versus the S.C.E. and electrolyse until a background current of less than 10 pA is again produced. Carry out a blank determination with reagents processed in the same manner as for ruthenium-containing samples.Ensure that the volume of the blank aliquot is the same as for the ruthenium solution. it = i,c-=t METHODOctober, 19691 BASED UPON MODULAR ELECTRONIC UNITS. PART 11. 843 Calculate the weight of ruthenium as follows- N x 101.07 x F 96,487 x where W is the weight of ruthenium, mg; N , the scaler and register readings; and F , the coulometer factor, C counts-l. RESULTS W = Results for the determination of ruthenium in Specpure ruthenium metal, uranium - ruthenium alloy and uranium - ruthenium carbide are given in Table I. Uranium would not be expected to interfere in the determination of ruthenium at an applied potential of +0*05 V and this is confirmed in Table I. Uraniumcannot, however, be determined by controlled-potential coulometry in the presence of ruthenium because of the reduction of ruthenium(II1) at the mercury pool at the potential necessary for the reduction of UO,:+ to U4+.This interference was avoided by heating aliquots of the solution with perchloric and sulphuric acids to remove ruthenium by volatilisation prior to the determination of uranium by controlled-potential coulometry. TABLE I DETERMINATION OF RUTHENIUM IN RUTHENIUM METAL, URANIUM - RUTHENIUM ALLOY AND URANIUM - RUTHENIUM CARBIDE Recovery, Coefficient of variation, Sample per cent. No of determinations per cent. Specpure ruthenium . . . . RU 99-8 8 1.0 Uranium - ruthenium alloy . . Ru 50.7 6 1.1 u 49.4 100.1 Uranium - ruthenium carbide . . Ru 29.5 U 68.3 C 2-3 100.1 6 1.0 DISCUSSION The analytical chemistry of ruthenium is complicated not only by the various valency states but also by the variety of complexes possible for each valency state.ly4 Spectrophoto- metric methods for the determination of ruthenium in chloride solution usually overcome these difficulties by reducing the sample solution to ruthenium(II1) and then re-oxidising to a ruthenium(1V) chlorocomplex with an absorbance peak at 485 nm.6 It has been shown, by a combination of controlled-potential coulometry and spectrophotometry, that dissolution of ruthenium metal and various ruthenium compounds in alkaline peroxide medium followed by extraction into hydrochloric acid gives a reproducible ruthenium( IV) complex.Subse- quent adjustment of the valency of the ruthenium is unnecessary, and the ruthenium can be determined directly by a controlled-potential coulometric reduction at +0.05 V versus the S.C.E. Further electrolytic reduction of ruthenium(II1) is readily achieved in chloride medium to give blue chlorocomplexes of ruthenium(II), but this process is not suitable for the analysis, probably because of a catalytic reaction in which hydrogen ions are reduced. A reversible Ru4+ - Ru3+ couple at +Oaf305 V veysus the S.C.E. can also be used for the con- trolled-potential coulometric determination of ruthenium, but this Ru4+ complex is not obtained by the method of dissolution described. This latter redox couple has been previously reported in a method for the potentiostatic coulometry of ruthenium.3 REFERENCES 1. 2. 3. 4. 5. 6. Woodhead, J. L., and Fletcher, J. M., U.K. Atomic Energy Authority Refiort A.E.R.E. R.4123, 1962. Stenina, N. J., Krylov, Yu. A., and Agasyan, P. K., Zh. Analit. Khim., 1966, 21, 1319. Stenina, N. J., and Agasyan, P. K., Ibid., 1966, 21, 965. Goldberg, R. N., and Hepler, L. G., Chem. Rev., 1968, 68, 229. Waterbury, G. R., and Metz, C. F., Talanta, 1960, 6, 237. Meites, L., Analyt. Chem., 1959, 31, 1285. Received December 18th, 1968 Accepted April lst, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699400840

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

844 Analyst, October, 1969, Vol. 94, @. 844-846 Pre-concentration of Carbonyl Compounds from their Medium Followed by Polarographic Determination of their Azomethine Derivatives BY M. D. BOOTH AND B. FLEET (Chemistry Department, Imperial College, London, S . W.7) A method for the pre-concentration of carbonyl compounds on a chro- matographic column via the formation of an azomethine derivative is described. The subsequent elution of the derivative is monitored polarographically. THE pre-concentration of carbonyl compounds from their medium is important in many fields, especially in food flavourings. Polarographic techniques are suitable for the deter- minations of this class of compound, and the optimum conditions for determining various types of carbonyl compound have already been rep0rted.l 9 2 9 3 3 4 In the course of a study on the chromato-polarographic determination of carbonyl com- pounds a method has been developed for the pre-concentration of lower aliphatic carbonyl compounds and their separation from aromatic aldehydes and ketones.Parsons6 has studied the rate of formation of 2,4-dinitrophenylhydrazones on a Celite column impregnated with the reagent. The derivative is eluted from the column with benzene and the reaction followed spectrophotometrically. For the analysis of the carbonyl content of butter fat and butter-fat volatiles, a good separation into classes, ie., alkanals, alk-2-enals and alk-2,4-dienals, was achieved for the volatile constituents, but the carbonyl content of the involatiles was too high to allow a complete separation.The present study has been based on the formation of the semicarbazone derivatives as in this instance the reagent itself is electro-inactive, whereas with 2,4-dinitrophenyl- hydrazine a well defined wave is obtained by reduction of the nitro-group. The optimum conditions for the formation and polarographic determination of semicarbazones has been established.132 The reduction of the azomethine group occurs via a 4-electron process to form the primary amine and urea. 0 0 II II R-CH=N-NH-C-NH, + 4e- + 3H+ + R-CH,-NH, + H,N-GNH, Y H+ As the reduction occurs through the protonated complex the polarographic wave shows a conventional dependence on pH with the limiting value of the current extending over the pH range 2 to 4.5. Below pH 2 the hydrolysis of the semicarbazones of aliphatic carbonyls is rapid; above pH 4-5 the rate of protonation becomes the limiting step.It is possible to form the derivative irt sit^ by adding the carbonyl compound to a buffered supporting electrolyte containing a large excess of the reagent. The use of a buffered medium for this reaction has an important advantage; the rate of formation of semicarbazones shows a peak- shaped dependence on pH with a maximum at about pH 4 ~ 5 . ~ As this pH also corresponds to the limiting current region for the reduction, an acetate buffer (containing 50 per cent. of ethanol) can conveniently be used. Although a recent study4 has shown that the Girard T derivatives of aliphatic aldehydes give better defined reduction waves than the corresponding semicarbazones, their formation and subsequent chromatographic separation are far less favourable.0 SAC and the authors.BOOTH AND FLEET 845 EXPERIMENTAL REAGENTS- Carbonyl compounds were obtained commercially. Standard solutions ( M) in analytical-reagent grade ethanol were prepared. Unless otherwise stated all reagents were of anal ytical-reagent grade . Buffer solution, 0.2 M sodium acetate - 2.0 M acetic acid. APPARATUS- Polarograms were recorded on a Radelkis polarograph type OH102 (Metrimpex, Hungary). The polarographic vessel was a Kalousek cell with a separated reference electrode (saturated calomel electrode). The capillary used had the following characteristics: outflow velocity m = 2.04 mg s-l and drop time t = 4.1 s, at the potential of the S.C.E.mercury pressure h = 60 cm. COLUMN PREPARATION- Two columns were used, the first a short Celite column (B.D.H., 30 to 80 mesh) impreg- nated with the reagent for the formation and concentration of the derivative, followed by the main chromatographic column for the separation. Silica gel (B.D.H., 60 to 120 mesh) was used for this column without any pre-treatment. Reaction column-Fifty grams of Celite were impregnated with a solution of 0.45 g of semicarbazide hydrochloride in 2 ml of 85 per cent. orthophosphoric acid (spgr. 1.75) diluted with 8 ml of water. This was transferred to a column (2 cm in diameter) fitted with a sintered- glass plug, and washed with 50ml of absolute ethanol. This gave a reaction layer about 10 cm in length. A slight loss of reagent occurs during the washing process.This column is capable of retaining carbonyl compounds from samples of up to 100ml total volume in ethanol containing 10 per cent. of water. For smaller samples, e.g., 10 ml, a 2-cm reaction layer is sufficient. In the latter instance l o g of Celite are adequate. Separation column-The second column (1 cm in diameter) contained sufficient silica gel to form a layer 10 cm in length. The reaction column was fitted into the top of the separation column, after the formation of the semicarbazones. PROCEDURE- The sample, containing between 0.1 and 1 pmoles of the carbonyl compound in up to 100 ml of ethanol containing 10 per cent. of water, is slowly percolated through the reaction column. The semi- carbazones are eluted with either absolute ethanol or ethyl acetate - ethanol (1 + 1).Aliquots of the eluent selected from retention data were diluted with an equal volume of the aqueous acetate buffer and the polarogram measured between -0.5 and -1.7 V ueysus S.C.E. The semicarbazone derivatives are formed on the reaction column. RESULTS AND DISCUSSION Several solvent systems were examined with the aim of separating the semicarbazones of the lower aliphatic aldehydes and ketones. The range of solvents, however, was limited in that the polarographic reaction wave was only well defined in polar solvents. Hence, various mixtures of ethanol, dimethylformamide and ethyl acetate were chosen for study. The semicarbazones of acetone and acetaldehyde were chosen as representative of the lower aliphatic carbonyls and benzaldehyde and several substituted benzaldehyde semicarbazones for the aromatic series.All of the solvent systems examined gave incomplete separation of the semicarbazone derivatives of acetone and acetaldehyde. However, ethyl acetate - ethanol (1 + 1) was found to be a good solvent for the separation of the aromatic semicarbazones from the aliphatic fraction. Both acetone and acetaldehyde semicarbazones were sparingly sohxble in this solvent and thus remained at the top of the silica gel column. After elution of aromatic derivatives the column was washed with ethanol alone when the acetone and acetaldehyde semicarbazones were eluted quantitatively within the first 10 ml. It was possible to achieve a reasonable degree of separation of benzaldehyde semi- carbazone and the 4-methoxy and 4-chloro derivatives with the ethanol - ethyl acetate846 BOOTH AND FLEET solvent system.Measurement of successive 2-ml aliquots of the eluate showed distinctive peaks as the components emerged, but there was some overlapping. Although it would almost certainly be possible to improve this separation, e.g., by increasing the length of the column or modifying the eluting solvent, this was not studied because of the doubtful use- fulness of this particular separation. A summary of the optimum conditions for the separ- ation of certain mixtures of carbonyl compounds is given in Table I. TABLE I OPTIMUM CONDITIONS FOR THE SEPARATION OF MIXTURES OF CARBONYL COMPOUNDS Carbonyl compound Acetone 4-Chlorobenz- aldehyde Acetone Acetaldehyde Benzaldehyde 4-Chlorobenz- aldehyde 4-Methoxy- benzaldehyde AFTER FORMATION OF THEIR SEMICARBAZONE DERIVATIVES Separation column 10 cm, flow-rate 1 ml minute-l Carbonyl added, pmoles 0.25 0.5 0-25 0-25 0-50 0-50 0.50 Carbonyl recovered, pmoles 0.20 0.49 0-20 0.20 - - - Recovery, per cent.80 f 2 98 f 2 80 f 2 80 f 2 - - - Eluting solvent Ethanol Ethanol - ethyl acetate Ethanol or Ethanol - ethyl acetate - di- methylform- amide Ethyl acetate - ethanol Separation Complete* None t Partial$ Reten tion volume, ml 5 7 Both compounds approximately 18 10 8 6 * 4-Chlorobenzaldehyde derivative detected in 5 to 10-ml fraction. Acetone semicarbazone eluted with ethanol, after 20 ml of ethanol - ethyl acetate had been t The polarograph waves were ill defined in the ethanol - ethyl acetate - dimethylformamide 1 The semicarbazones would appear to exist as broad drawn-out bands on the column.run through the column. solvent mixture. Consequently some degree of overlap was observed. It is also possible to use a flow-through polarographic cell7 to monitor the eluate. Although the present work was unsuccessful in developing a chrornato-polarographic method for the separation of aliphatic carbonyl semicarbazone derivatives, it was felt that the development of a technique for pre-concentration of aldehyde and ketone semicarbazones from their medium and their separation into broad classes is of analytical importance. An important consideration here is the fact that a wide range of aldehyde and ketone semi- carbazones have similar diffusion current constantsJ2 hence it is possible to apply the above method for total carbonyl content. REFERENCES 1. 2. 3. 4. 6. 6. 7. Fleet, B., and Zuman, P., Colln Czech. Chem. Commun., 1967, 32, 2066. Fleet, B., Analytica Chim. Acta, 1966, 36, 304. Lund, H., Talanta, 1965, 12, 1066. Fleet, B., and Keliher, P. N., Analyst, 1969, 94, 659. Parsons, A. M., Ibid., 1966, 91, 1082. Jencks, W. P., “Progress in Physical Organic Chemistry,” Volume 2, John Wiley and Sons Inc., Fleet, B., Soe Win, and West, T. S., Proceedings of Technicon Symposium, “Automation in Analyti- Received February 20th, 1969 Accepted April 17th, 1969 New York, 1964. cal Chemistry,” 1967.

ISSN:0003-2654    

DOI:10.1039/AN9699400844

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

ArtaZyst, October, 1969, Vol. 94, p+. 847-4454 847 Some Observations on the Analytical Usefulness of Electrochemiluminescence for the Determination of Microgram Amounts of Aromatic Hydrocarbons* BY B. FLEET, P. N. KELIHER, G. F. KIRKBRIGHT AND C. J. PICKFORD (Chemistry Department, Imperial College, London, S . W. 7) P.M. .tube Measurement of the electrogenerated chemiluminescence of twenty-four aromatic compounds in dimethylformamide solution has been investigated as a potential analytical technique for their determination. The emission spectra, calibration results and limits of determination are presented. Experi- mental requirements and optimum conditions for the electrogeneration of the light-emitting species are described. EMISSION of radiation from solutions during electrolysis has been observed for solutions of metal salts,l Grignard reagents in ethers2 and alkaline solutions of fluorescent3 and chemi- luminescent4 compounds.The electrolysis of some aromatic hydrocarbons in solvents such as dimethylfonnamide and acetonitrile has recently been shown to result in emission of visible radiation.5 This phenomenon is referred to as electrochemiluminescence (ECL) , and the mechanism of the processes involved has been extensively studied by several workers.6 to l6 With the exception of the study by Cruser and Bard,17 and the application of electrochemi- luminescence measurement to the determination of polycyclic aromatic hydrocarbons in liver lipids by Bobr, Kozlov and Mikhailovskii,18 few results are available concerning the potential and range of application of ECL to the identification and determination of micro- gram amounts of organic compounds.In a previous short communi~ation~~ we have reported some initial observations on the analytical application of the technique of ECL measurement. This paper describes the apparatus developed for the study of the parameters governing ECL emission, and reports the ECL characteristics of a range of aromatic hydrocarbons and heterocyclic compounds. EXPERIMENTAL APPARATUS- A diagrammatic representation of the apparatus used is shown in Figs. 1 and 2. The wave form is applied to the working electrode of the cell by a function generator (Hewlett- Cel I Mono- chrom. Signal generator , PA meter Fig. 1. Diagrammatic representation of apparatus used * Paper presented at the Second SAC Conference 1968, Nottingham.0 SAC and the authors. Oscilloscope848 FLEET et al. : OBSERVATIONS ON THE ANALYTICAL [Analyst, Vol. 94 Packard Corporation, HP3300A) two-channel triangular, square or sine-wave, variable fre- quency (d.c. to 100-kHz with single sweep and phase-lock facilities) and operational ampli- fier potentiostat with Philbrick PF85AU amplifiers and Philbrick P66A booster. The potentio- stat circuit is as described by Schwarz and Shahao with a differential input to single-ended output follower and provides up to 100-mA signal. ,Teflon covered steel support Teflon cell cap Cell Nitrogen inlet tube Platinum window counter electrode Silver reference probe 0.004" platinum he working electrode dix Fig. 2. Electrochemiluminescence cell assembly The ECL sample cell is a fused silica spectrofluorimeter cell (10 x 10 x 40 cm) with a machined Teflon stopper into which the three electrodes and nitrogen inlet and outlet tubes are sealed.The working electrode is a platinum helix supported on a steel rod coated with PTFE. This rod acts as electrical connection to the potentiostat. The counter electrode is a platinum sheet spot-welded to a platinum-rod contact. The reference electrode is a silver wire in the same cell. Although the silver electrode under these conditions is only a pseudo- reference electrode, it exhibits a potential that remains constant during the course of an experiment and that can be related to the potential of the S.C.E. (-0.33 V). The radiation emitted from the working electrode is viewed through the counter electrode by an 11-stage photomultiplier tube (EMI9601B) via a metal dielectric interference mono- chromator (Barr and Stroud Ltd., London).The 3-mm slit of this monochromator results in a spectral half-band width of about 25 nm, over a spectral range from 400 to 700nm. The photomultiplier photocurrent is displayed on an ultra-sensitive micro ammeter (RCA Type WV84C), oscilloscope (Hewlett-Packard 175A), or pen recorder (Servoscribe Type AE511, Smiths Industries). REAGENTS- DimethyZformamide (British Drug Houses Ltd. laboratory-reagent grade) was used as solvent throughout. It was dried over anhydrous potassium carbonate and then over a molecular sieve (Lind 5A). The dried solvent was then distilled under vacuum in a fractional distillation assembly. The distillate was stored in brown-glass reagent bottles until required. Nitrogea-Oxygen-free nitrogen (British Oxygen Co., oxygen-free grade) was freed from traces of oxygen by passage through a 2 per cent.solution of the sodium salt of anthraquinone- P-sulphonic acid in 2 N sodium hydroxide containing granulated zinc. The nitrogen was then dried by passage through concentrated sulphuric acid and magnesium perchlorate. The gas was then finally passed through a Dreschel bottle containing dimethylformamide to ensure saturation of the gas with solvent and minimise evaporation of solvent in the sample cell during the de-gassing procedure. EZectroZyte-Tetrabutylammonium perchlorate (TBAP) . This was prepared by neutral- isation of tetrabutylammonium hydroxide solution (B.D.H., general-purpose reagent grade) with 70 per cent.analytical-reagent grade perchloric acid. The product was recrystallised twice from a water - ethanol mixture and dried under reduced pressure at 95" C for 24 hours.October, 19691 USEFULNESS OF ELECTROCHEMILUMINESCENCE 849 Some of the organic compounds, the ECL characteristics of which were studied, were obtained commercially, while others were supplied as samples by courtesy of the British American Tobacco Company. The compounds were purified before use by vacuum sub- limation or recrystallisation. EMISSION SPECTRA- The emission spectra were measured for 10-3 M solutions of each compound in dimethyl- formamide containing 10-2 M TBAP, and the dissolved gases present were removed by alternate freezing and thawing cycles or by nitrogen purging.The solution was placed in the cell, and a 30-H~ square-wave signal of increasing amplitude was applied to the working electrode while the solution was gently stirred by nitrogen. When light emission was detected, the optimum frequency and wavelength of maximum emission were established, and the current - voltage (cyclic volt ammogram) characteristics examined. The dependence of the light- emission intensity on the applied voltage was also investigated. When no light emission was observed with a 30-Hz square wave, a low frequency (1 Hz) signal was applied to the electrode and the effect of stopping the stirring was investi- gated. Spectra were obtained for the sample solutions in dimethylformamide and TBAP electrolyte (10-2 or 10-1 M, depending on which concentration produced the greater light intensity). The experimentally established optimum frequency and applied voltage were used. The light intensity of the emission was found to reach a stable value for most of the compounds investigated within 10 seconds (at 30 Hz) or 30 seconds (at 1 Hz).The spec- trum was recorded manually, and the mean light intensity was measured at intervals of 20nm between 400 and 700nm. The calibration data (ECL intensity at A,,,. vcysus concentration) were obtained by preparing a range of dimethylf ormamide - TBAP solutions containing various sample concen- trations, and recording the steady value of the light intensity obtained for each solution under optimum conditions.Alternatively, the light intensity obtained initially on application of the optimum voltage was recorded for each solution at the wavelength of maximum emission. In each instance the calibration graphs were plotted as log,, intensity versus log,, sample concentration. GENERAL EXPERIMENTAL PROCEDURE RESULTS AND DISCUSSION A wide range of compounds was examined for electrochemiluminescence in dimethyl- formamide solvent. In all instances both the emission spectra and the current - voltage graphs were recorded. The potentials at which the peak intensity of light emission occurred was compared with the current peaks on the current - voltage graph (cyclic voltammogram) corresponding to the formation of the radical anion and cations. Of the compounds studied, some twenty-four of those that showed intense emission were chosen for further study and their ECL spectra shown in Fig.3. These spectra are uncorrected for the response charac- teristics of the EMI96OlB photomultiplier and the monochromator used. In most instances broad-band spectra that resemble the fluorescence spectra of the compounds are obtained. For several compounds (e.g., phenanthrene, 2,2'-binaphthyl), long wavelength emission not found in the corresponding fluorescence spectra occurs. Several compounds were shown to exhibit only weak ECL emission and were not examined further. N-Phenylpyrrole, NN'-di- 2-naphthyl-$-phenylenediamine, $-methoxybenzoylnaphthalene and acetyl-1-naphthylamine exhibit weak emission in this way. Compounds that were found not to exhibit ECL emission in dimethylformamide under the conditions used in this study include diphenylbenzidine, 2,2'-dihydroxybiphenyl, 2,2'-dimethoxybiphenyl, 1,lO-phenanthroline, acridine, benzanthrone, 1-naphthoic acid, 4-phenylazophenol, quinaldine, naphthalene-1,5-&01, 2-aminothiazole, 3-methylisoquinoline, azulene and various hydroxynaphthalenesulphonic acids.For the twenty-four compounds examined in detail, the results obtained are shown in Table I. Four distinct types of electrochemiluminescent behaviour were observed experi- mentally for the compounds under the experimental conditions used. The type of behaviour assigned to each compound in Table I is responsible for the principal emission of analytical use. Most compounds also show one or more of the other types of behaviour at much lower light-emission intensities.850 P-LEET e5 ah.: OBSERVATIONS ON THE ANALYTICAL LA nalyst, Vol. Y4 9,IO-Diphenylanthracene \ FI uoranthene Coronene 2-Phenylnaphthalene L Benzylbiphenyl Phenanthrene I I f So0 600 Rubrene R octatet raene I ,&Diphenyl- 1,357- 2,2’- Binaphthyl Di benzofuran Naphthalene L ZFluoronaphthalene Pyrene Wavelength, nm Fig. 3. ECL emission spectra of compounds examinedOctober, 19691 USEFULNESS OF ELECTROCHEMILUMINESCENCE TABLE I RESULTS Normal Fluorescence Compound nm max., nmalssa ECL max. , emission Coronene . . . . 440 417, 440 Rubrene . . .. 575 575 Anthracene . . . . 450 375, 395, 419, 9-Phenylanthracene 430 430 9,lO-Diphenyl- anthracene . . 420 430 Pyrene . . . . 400,480 410,460, 397, Fluoranthene .. 460, 525 438,462,482 Chrysene .. .. - 387, 400, 408, l18-Diphenyl- 1,3,5,7- Phenanthrene . . 520,' 410a 398, 420,448 Perylene . . .. 455 442, 462, 494, 520 Binaphthyl . . . . 420, 670 385 Benzylbiphenyl . . 440 330, 350 p-Terphenyl . . .. 510 360 Naphthalene .. 470 320 Dibenzofuran . . 475 410,440, 470 Carbazole . . . . 670 408,429 Phenothiazine . . 530 - 444,474 385 423, 450 octatetraene . . 480 - Range for calibration graph, M 10-3+ 10-6 10-3+ 10-7 5 ~ 1 0 - ~ 10-3+ 10-7 10-3+ 10-7 5*10-5 10-3+ 2-10-5 10-3+ 2-10-6 10-3+ 2.10-5 5-10-3+ 2.10-5 2.10-3+ 2.104 - Dibenzothiophen . . 540 - - 2-Methoxynaphthalene 450 - only above Brucine . . . . 490 - only above Thebaine . . .. 485 - only above 2-Fluoronaphthalene 450 - only above 2-Phenylnaphthalene 445 - only above Type of emission (see text) a a a a a a a d a a8, dl a a d d a d b C b c + d C C d a Optimum frequency, Hz 30 30 10 30 50 5 30 20 10 50 10 40 1 30 30 30 30 0.1 1 d.c.0.1 10 30 50 851 Preferred waveform Square wave Square wave Square wave Square wave Square wave Square wave Square wave Square wave Square wave Square wave Square wave Square wave Square wave Square wave Square wave Square wave Square wave -2+ -3 Triangular wave Square wave - 2.95 (stirred) -2+ -4 Triangular wave Square wave Square wave Square wave - 0.5+ - 3.5 The four types of behaviour are as follows. (a) Radical cation - anion annihilation. This appears to be the most common type of mechanism involved in ECL emission for aromatic hydrocarbons where the energy available from the cation - anion radical reaction R+ + R- -+ R* + R is sufficient to form the excited singlet state from which ECL emission can occur.This type of process is recognisable from the necessity that emission only occurs after a complete voltage sweep from negative to positive potentials corresponding to the formation of the anion and cation-free radicals. This type of process results in ECL, for example, for rubrene, fluoranthene and diphenylanthracene, when the possibility of forming reasonably stable free radicals exists. The introduction of functional groups precludes the emission by this process as the radical ions, particularly the cation, are no longer stable. (b) When the radical cation or anion reacts with some impurity species, e.g., traces of molecular oxygen or water.This type of emission can be recognised experimentally when the electrode is held at the potential required to generate either R+ or R-, rather than when a square-wave form is applied. A steady emission of light may then occur as the radicals formed diffuse into the bulk of the solution and react with the impurities. We have observed behaviour of this type for the compounds phenothiazine and dibenzophenothiophene. (c) When the light emission is only observed on repeated cycling of the potential. Clearly in this instance the products of the primary electrode reaction are responsible for the light emission as no light is observed on the initial scan. This type of emission is usually observed as a band at longer wavelengths than ECL emission by other processes.The emission intensity increases with time, but usually remains weaker than the ECL stimulated by the primary852 [Artalyst, Vol. 94 process. This type of emission has been observed, for example, for chrysene and 2-fluoronaph- thalene. For some compounds, however, it appears to be the principal mode of emission, e.g., with the alkaloids brucine and thebaine. (d) When light emission occurs at potentials beyond those at which normal solution electrolyte reactions occur, e.g., with carbazole and @-terphenyl. We are currently under- taking a study of this type of behaviour to elucidate the mechanism of the process or processes involved, but it seems probable that higher reduced or oxidised species, for example the dianion or even products resulting from the reduction or oxidation of the solvent or supporting electrolyte, are involved.CALIBRATION DATA AND LIMITS OF DETERMINATION- Typical calibration graphs prepared by the procedure described are shown in Fig. 4 for twelve of the compounds examined. It will be observed that when log,, intensity is plotted zleisz.4.s log,, concentration, linear calibrations are obtained. Deviations from linearity are observed in several instances in the higher concentration ranges and can be ascribed to con- centration quenching. The dynamic range of the technique is good, as many of the calibration graphs are linear over a 1000-fold range of concentrations. ECL emission was observed for brucine, thebaine, 2-methoxynaphthalene, 2-fluoro- naphthalene and 2-phenylnaphthalene only at concentrations greater than M.The lower concentration limit of the calibration graph for each compound represents the practical limit of determination with our apparatus. The absolute determination limits obtainable reflect the efficiency of the particular experimental assembly used, and can be improved by reduction FLEET et al. : OBSERVATIONS ON THE ANALYTICAL x u .- 2 - c1 - M A I Anthracene 2,2'- B i napht hy I 4 - - 3 - 2 - - I - - O l l 1 I I 1 10 loo I o c Concentration .pg ml-' 0 L Diphenylanthracene - 9-Pheny lant hracene 5 - - 4- - 3 - - 2 - - I - - 6.4 I I 0 loo lo( Phenoth iazi ne - 3 - A I - Concentration p g ml" Fig. 4. Calibration graphs for ECL emission versus concentration of compounds in DMFOctober, 19691 USEFULNESS OF ELECTROCHEMILUMINESCENCE 853 Fluoranthene x C W U C M .- v) .- -1 Pyrene 3 1 x C W C .- wl U .- 0" -1 Phenanthrene 4 1 3 - x c - 0 r c .- wl .- 3 A I - Concentration pg ml-' I ,8-Diphenyl- I ,3,5,7*0ctatetraene t Concentration pg ml-' Fig.4 (contd.) Calibration graphs for ECL emission versus concentration of compounds in DMF in the sample volume required, careful monochromator design and selection of photomultiplier and electronic components. It was observed that deviations in linearity in concentration - light intensity relationships were often observed in those instances in which simple cation - anion annihilation was not operable, i.e., those observed to fit into categories (b), (c) and (d). This at once imposes a limitation on the technique as only the relatively simple polycyclic hydrocarbons, in which the possibility of forming a relatively stable free radical exists, show this type of process.The introduction of functional groupings such as hydxoxyl and amino precludes this, as noted previously. On the other hand, this limitation provides a degree of selectivity over the conventional fluorescence technique. It would seem, therefore, that the wide range of "energy deficient " electrochemiluminescent processes, where the energy of interaction of cation and anion radicals is insufficient to form the singlet state, are not amenable to quantitative ECL determination or at least can only be determined over a fairly limited concentration range. It is possible, however, that the removal of the randomness of the quenching process, i.e., by the addition of a specific anion or cation quencher to form a radical at a potential such that interaction will lead to the triplet state, may provide the solution.An advantage of ECL over conventional spectrofluorimetry is that as there is no source of excitation no problems associated with scattered radiation arise. It might also be possible, by controlling the electrolysis potential, to determine selectively individual components of simple mixtures whose fluorescence excitation and emission spectra are similar. Work on the analysis of mixtures of this type is currently in progress.854 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. FLEET, KELIHER, KIRKBRIGHT AND PICKFORD REFERENCES Bancroft, W. D., and Weiser, H. B., J .Phys. Chem., 1914, 18, 762. Dufford, R. T., Nightingale, D., and Gaddum, L. W., J . Amer. Chem. Soc., 1927, 49, 1858. Shlyapintokh, V. Ya., Postnikov, L. M., Karpukhin, 0. N., and Veretilnya, A. Ya., Zh. Fiz. Kuwana, T., Electroanalyt. Chem., 1963, 6, 164. Hercules, D. M., Science, 1964, 143, 808. Feldberg, S. W., J . Amer. Chem. Soc., 1966, 88, 390. -, J . Phys. Chem., 1966, 70, 3928. Marcus, R. A., J . Chem. Phys., 1965, 43, 2654. Parker, C. A., and Short, G. D., Trans. Faraday Soc.. 1967, 63, No. 639. Hoytink, G. J., Faraday Disc., April, 1968, Newcastle. Visco, R. E., and Chandross, E. A., Electrochim. Acta. 1968, 13, 1187. Bard, A. J., Santhanam, K. S. V., Cruser, S. A., and Faulkner, L. R., in Guilbault, G., Editor, Zweig, A., Hoffmann, A. K., Maricle, D. L., and Maurer, A. H., J . Amer. Chem. Soc., 1968,90, 261. Zweig, A., and Maricle, D. L., J . Phys. Chem., 1968, 72, 377. Chang, J., Hercules, D. M., and Roe, D. K., Electrochim. Acla, 1968, 13, 1197. Anges, Y., and Signore, R., C. R. Hebd. Skanc. Acad. Sci., Paris, Ser. A,B, 1968, 266B, 870. Cruser, S. A., and Bard, A. J., Analyt. Lett., 1967, 1, 11. Bobr, V. M., Kozlov, Yu. P., and Mikhailovskii, G. E., DokZ. Akad. Nauk. SSSR, 1967, 175, 1159. Fleet, B., Kirkbright, G. F., and Pickford, C. J., Talanta, 1968, 15, 566. Schwarz, W. M., and Shain, I., Analyt. Chem., 1963, 35, 1770. Berlman, I. B., “Handbook of Fluorescence Spectra of Aromatic Molecules,” Academic Press, Lijinsky, W., Chestnut, A., and Raha, C. R., Chicago Med. Sch. Q., 1960, 21, 49. Khim., 1963, 37, 2374. “Fluorescence, ” Marcel Dekker, New York, 1968. New York, 1965. Received February 7th, 1969 Accepted March 31st, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699400847

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Analyst, October, 1969, Vol. 94, $$. 855-859 855 Voltammetric Studies with Different Electrode Systems Part IV.* Determination of Silver by Using a Silver-Molybdenum System BY V. T. ATHAVALE, M. R. DHANESHWAR AND R. G. DHANESHWAR (A nalytical Division, Bhabha Atomic Research Centre Modular Labovatories, Trombay, Bombay-74, India) Silver is determined at trace level (1 to 100 p.p.m.) in organic solvents and alkalis by using a semi-micro cell incorporating a silver - molybdenum system. The determination is camed out on a differential cathode-ray polarograph. With this electrode system only silver and mercury cations were reduced and the peak potentials were well separated. No chemical separations are involved. Factors affecting the silver reduction in organic solvents are discussed.SILVER is one of the impurities that influence photo-conductive effects in semi-conductors.1 A method for its determination in some organic solvents of high purity, notably methanol and acetone, which may be used in the preparation of semi-conductor materials, is therefore required. Polarographic reduction of silver and mercury occur together in almost all media and the separation of the two waves becomes very difficult, especially when a dropping-mercury electrode is used, as the dissolution wave of mercury is always obtained. The silver indicator electrode has been thoroughly examined in potentiometry,2 13 in which it is responsive mainly to silver and mercury ions. Silver used as the cathode in polarography may overcome the difficulties encountered with the dropping-mercury electrode, but it failed to give rise to reproducible curves on a d.c.polarograph, and an investigation was carried out by using a differential cathode-ray polarograph. The choice of reference electrode also presents difficulty in the polarography of silver ions, as neither a saturated calomel nor a mercury-pool electrode can be used. A molybdenum- wire electrode, which has been used as a reference electrode under diverse experimental conditions4s6s6~7s8 is primarily a pH electr~de,~J~ but while the pH remains constant, it func- tions as a reference electrode in polarography and potentiometry.ll In a polarographic study of silver reduction, the dropping-mercury electrode - molybdenum system has been used with partial success.l2 The molybdenum reference electrode was, therefore, used in silver voltammetry.The silver - molybdenum bi-metallic system has been studied in great detail,l3,14 and it was shown that of all of the cations only mercury and silver reduction curves were obtained; an S-shaped mercury wave was obtained at -0.40 V and a silver peak at -0.60 V, thus eliminating the possibilities of interference. Moreover, with this system, no dissolution wave of mercury was obtained as with the dropping-mercury electrode. With the bi-metallic system the construction of a semi-micro cell is facilitated and it can be used in mixed and non-aqueous solvents, unlike the dropping mercury - S.C.E. system, which gives irreproducible results because of irregular drop time. EXPERIMENTAL APPARATUS- used, at 30" 0.1" C.A differential cathode-ray polarograph, Model A, 1660 (Southern Analytical Ltd.) was * For details of earlier parts of this series, see reference list, p. 859. 0 SAC and the authors.856 ATHAVALE et al. : VOLTAMMETRIC STUDIES WITH [Analyst, Vol. 94 Semi-micro cell-A cell of 0.5-ml capacity was constructed by simply arranging silver (28 s.w.g.) and molybdenum (22 s.w.g.) wire electrodes in a cork that was inserted into a Pyrex glass tube of 1 cm diameter. An improved version, with facilities for inlet and outlet of nitrogen, as well as for inserting the burette tip, is shown in Fig. 1. The silver-wire electrode was prepared by sealing the wire in soft glass tubing. This electrode was cleaned with nitric acid (1 + 4) and the molybdenum electrode by rubbing with emery paper. Burette - 1.5 cm Fig.1. Improved semi-micro cell: anode, molybdenum 4 mm; and cathode, silver 3 mm REAGENTS- Stock silver nitrate solution, 0.1 M. Stock sulphosalicylic acid solutiolz, 1 M. Analytical-reagent grade solvents were used and were further purified by re-distilling three times. RESULTS AND DISCUSSION SILVER REDUCTION IN AQUEOUS SOLUTIONS- In the semi-micro cell, the length of both electrodes is reduced to one fifth of that of normal electrodes. In view of the possibility of mutual polarisation in the bi-metallic electrode system, the silver reduction was re-examined for the small electrodes. The curves obtained for the 0 and 100 slope component factors are shown in Fig. 2; this factor does not greatly affect the shape of the curve.Current - concentration linearity is observed for the range 0.25 to 1-00 mM silver con- centration, as shown in graph (a), Fig. 3. The peak potentials obtained vary between -0.55 and -0.60 V, being almost identical with those obtained with normal electrodes. It was also confirmed that no ions other than mercury(1) were reduced. The mercury(1) reduction wave intermingles with the residual oxygen wave starting from -0.2 V and does not interfere in the silver determination. SILVER REDUCTION IN 50 PER CENT. METHANOL- Current - concentration linearity was studied in 50 per cent. methanol, in which the solubility of the supporting electrolyte and the resistance of the solution are not affected as they are with the aqueous solution. Also in a pure non-aqueous solvent, the electrode behaviour becomes erratic.The current - concentration graph is a straight line over the range 0.25 to 1 - 0 0 m ~ silver concentration, as shown in Fig. 3 ( b ) . Even in 50 per cent. organic solvents, the peak currents are much smaller than those with the aqueous solution for 1.00 mM silver concentration (2-0 pA compared with 6.30 pA in the aqueous solution).October, 19691 DIFFERENT ELECTRODE SYSTEMS. PART IV 857 Potential, V Fig. 2. Silver peak on silver (3 mm) cathode: anode, molybdenum 4 mm; silver con- centration, 0.5 m ~ ; shunt scale factor, 1.5 x 10; amplification factor, 1 x 100; and start potential, 0.4 V. Curves A and B were taken for 0 and 100 slope component, respectively; i, (PA) = 2 x 104 x number of divisions x shunt scale factor x amplification factor; and E , = - 043 V for both curves : current in arbitrary units Ag+ concentration, M 01 1 1 I I 0 0.25 0.50 0.75 1.0 xIO' Ag+ concentration, M Fig.3. Current - concentration graphs for Ag+ ions: (a), in aqueous solution; and (b), in 50 per cent. methanol: semi-micro cell Moreover, at lower silver concentrations, the peaks are not well defined, as shown in Fig. 4. The peak potentials do not vary much compared with those for the aqueous solution, EP being -0-61 to -0.54V. These results are in agreement with those obtained by Lietzke and Stoughton,lS who reported that in 50 per cent. methanol the currents resulting from in- organic cation reductions were reduced but that the potentials were not much affected. Similar reductions in peak currents were also obtained with aqueous mixtures of other organic solvents.858 ATHAVALE et al.: VOLTAMMETRIC STUDIES WITH [Arcalyst, VOl. 94 t I I I I I I I I I 0.1 02 0.3 0.4 Potential, V Fig. 4. Silver peaks for different concentrations in 60 per cent. meth- anol: cathode, silver 3 mm; anode, molybdenum 4 mm; for curves A and B, shunt scale factor and amplification factor 1 x 10 and 1 x 100, respec- tively, and for curve C, 6 x 1 and 1 x 100, respectively. Curves A, B and C were taken for 1-0, 1-6 and 0.26 m M silver concentration, res- pectively; start potential, 0.3 V for all of the curves; and E,, -044,-043 and -0.61 V for curves A, B and C: current in arbitrary units Another important drawback arising from the use of mixed solvents is the persistent oxygen wave, which is difficult to eliminate and especially so when dioxan is used.As the peak currents and potentials become progressively unsteady with increasing percentage of organic solvent, it is not possible to use larger amounts of the solvent sample. For these reasons, it is desirable not to carry out the silver determination in mixed solvents. SYNTHETIC SAMPLES- The difficulty mentioned above can be avoided by using a larger amount of the solvent and then removing it by evaporation. In the solvents used, viz., methanol, acetone and dioxan, silver nitrate is very soluble,16 and therefore the recoveries could readily be tested. TABLE I SILVER PEAK CURRENT AND POTENTIAL IN EVAPORATED SAMPLES Cathode, silver (3 mm) ; anode, molybdenum (4 mm) ; Ag+ added, 0-25 ml of M solution; and supporting electrolyte, 0.1 M sulphosalicylic acid Sample Amount, ml Water .. .. . . 10 Acetone . . .. .. 10 Acetone , . .. . . 25 Dioxan . . .. .. 8 Methanol . . .. . . 10 Methanol . . .. . . 60 Sodium hydroxide, ~ I U . . 10 E,, v -0.55 -0.41 - 0.41 -0.31 - 0.33 - 0.33 - 0.58 i,, PA 1-60 1-63 1-52 1-52 1.60 1.50 1-60 Remarks To x ml of the sample, as indicated in column 2, contained in a beaker, 0-25ml of 10-*~ Ag+ was added, the solution was evaporated on a water-bath and the residue taken up in 10 ml of 0.1 M sulphosalicylic acid As indicated in Table I, 0.25 ml of M silver nitrate was added to varying amounts of the solvents and the solutions were evaporated on a water-bath. The residues were taken up in 1 O m l of 0.1 M sulphosalicylic acid.The peak currents obtained for aqueous as well as other solvents were nearly identical, but the peak potentials varied considerably (EP for dioxan is -0-31 V, for methanol -0-33 V and for acetone -0-41 V; for water and sodium hydroxide solutions it has the same value). When determining silver, it is desirable to confirm the silver peaks by standard addition.October, 19691 DIFFERENT ELECTRODE SYSTEMS. PART Iv 859 Different volumes (8 to l O m l ) of acetone, dioxan, methanol and sodium hydroxide solution were taken, and the procedure given above was followed. If it is required to determine lower concentrations of silver, then a greater amount of sample (25 to 50 ml) can be taken for evaporation and the residue dissolved in 5 or 3 ml of 0-1 M sulphosalicylic acid instead of 10 ml.TABLE I1 RECOVERY OF SILVER FROM THE VARIOUS SAMPLES BY USING A SEMI-MICRO CELL Cathode, silver (3 mm) ; anode, molybdenum (4 mm) ; and supporting electrolyte, 0.1 M sulphosalicylic acid ON A DIFFERENTIAL CATHODE-RAY POLAROGRAPH Sample material Silver added, pg Silver recovered, pg Error, per cent. Methanol .. .. .. 34-0 34.0 Nil Methanol .. .. .. 13.6 14.4 + 6.67 Acetone . . .. .. .. 34-0 37.0 + 8.82 Dioxan . . .. . . .. 33.0 36.0 + 9.09 Sodium hydroxide, 1~ . . .. 31.6 31.6 Nil Addition is calculated on l-g basis for the sample. As shown in Table 11, up to 10 p.p.m. of silver can be determined with an error of 10 per cent. for the initial sample solutions of 8 to 10 ml. The amount of silver recovered is the mean of three readings. With 50 ml of sample solution taken for evaporation and a final volume of 3 to 5 ml, it was possible to determine silver down to the 1 p.p.m.level. It was confirmed that mercury, even if present in amount three times that of the silver, does not interfere. The silver content of the alkali and solvents was found to be less than 1 p.p.m. REFERENCES 1. Willardson, R. K., and Beer, A. C., “Semi Conductors and Semi Metals,” First Edition, Volume 11. 2. Bishop, E., and Dhaneshwar, R. G., Analyst, 1963, 88, 424, 442 and 433. 3. Dhaneshwar, R. G., Ph.D. thesis, “Differential Electrolytic Potentiometry of Argentimetric 4. Athavale, V. T., Burangey, S. V., and Dhaneshwar, R. G., in Shallis, P. W., Editor, “Proceedings 5. --- , J . Electroanalyt. Chem., 1966, 9, 169.6. Apt;, V. G., and Dhaneshwar, R. G., Indian J. Chem., 1969, 7 , 416. 7. Dhaneshwar, R. G., “Proceedings of the Symposium on Electrode Processes,” University of 8. Dhaneshwar, R. G., and Kulkami, A. V., Curr. Sci., 1968, 3!, 241. 9. Athavale, V. T., Apte, V. P., Dhaneshwar, M. R., and Dhaneshwar, R. G., Indian J. Chem., 10. Athavale, V. T., Dhaneshwar, M. R., and Dhaneshwar, R. G., Indian J. Chem., in the press. 11. Dhaneshwar, M. R., Ph.D. thesis, “Voltammetric Studies Using Different Electrode Systems,” 12. Dhaneshwar, R. G., and Kulkarni, A. V., Indian J. Chem., 1966, 4, 633. 13. Athavale, V. T., Dhaneshwar, M. R., and Dhaneshwar, R. G., J. EZeclroanaZyt. Chem., 1967,14, 31. 14. Burangey, S. V., Dhaneshwar, M. R., Dhaneshwar, R. G., Dharmarajan, V., and Kulkami, A. V., 15. Lietzke, M. H., and Stoughton, R. W., J. Amer. Chem. SOL, 1967, 79, 2067. 16. Mellor, J. W., “A Comprehensive Treatise on Inorganic and Theoretical Chemistry,” Longmans, Academic Press, London, 1966, p. 229. Reactions,” University of Exeter, 1962. of the SAC Conference 1966,” W. Heffer & Sons Ltd., Cambridge, 1966, p. 446. Jodhpur, Jodhpur, 1968, p. 55. 1968, 6. 666. University of Poona, 1967. Bhabha Atomic Research Centre, B.A.R.C. Report 387, 1969. Green and Co., London, Volume 111, 1948, p. 466. NOTE-References 4 and 13 are to Parts I and 111, of this series, respectively. Received Fe&wary 7th, 1969 Accepted March 26th, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699400855

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

860 Analyst, October, 1969, Vol. 94, Pp. 860-863 Limit of Determination in Photometric Titrations with Self-indicating Systems* BY E. R. GRoENEVELDt AND G. DEN BOEF (Laboratory for Alzalytical Chemistry, University of A msterdam, The Netherlands) An attempt is made to predict the lower limit of concentration (the limit of determination) of a photometric titration based on a self-indicating reaction. When only instrumental factors influence the precision of the result of such a titration it appears to be possible to express the limit of deter- mination in terms of the precision of the absorbance reading and other characteristics of the apparatus (path length and cell volume) and of the reaction ( &) . The theoretical prediction has been checked by using two commercially available photometers and two reactions, viz., the titration of vanadium(1V) with cerium(IV), and the titration of cerium(1V) with iron(I1).The agree- ment between the theoretically predicted value of the limit of determination and the experimental value appeared to be satisfactory. THE limit of determination is defined in this paper as the amount of substance, or its con- centration, that will give a standard deviation of 5 per cent. in the result. A photometric titration with a self-indicating system is based on the light absorption of one or more of the substances involved in the reaction and does not, therefore, require an indicator. In this study we have only investigated the influence of the photometer on the limit of determination. Two commercially available instruments were used, vix., a Zeiss spectrophotometer PMQ I1 (deflection type) and a Zeiss Elk0 I1 filter photometer (substitution principle, especially suitable for the accurate measurement of small absorbances) .Simple mathematical expres- sions are derived for the calculation of the standard deviation in the result of a manual Photometric titration in terms of the standard deviation of the absorbance readings of the instrument. Other factors influencing the precision of the result of a photometric titration, such as chemical kinetics, are not considered. Finally the predicted limit of determination is compared with experimental results for titrations based on “ideal reactions,” in which instrumental errors are greater than chemical errors. MATHEMATICAL EXPRESSIONS- Let the reaction be x+s+xs, where X is the substance to be determined, S the reagent and XS the reaction product.Assuming that Beer’s law is valid for all substances, the absorbance at any moment during the reaction can be expressed by where b is the length of the cell and €1 and ci are the molar absorptivity and the concentration of the absorbing substances, respectively. If the reaction proceeds quantitatively to the right and is extremely fast, which occurs in practice, the titration curve will consist of two straight lines, $J and q, intersecting at the equivalence point, provided that correction for dilution is carried out. A = bZ;EiCi, * Paper presented at the Joint Symposium on Limits of Detection in Analysis, April 17th and lSth, t Present address : Forensic Laboratory, Ministry of Justice, The Hague.0 SAC and the authors. 1968, Enschede.861 GROENEVELD AND DEN BOEF The straight line Cp, before the equivalence point, is given by v + c p 1 . .. .. . . (la); bt (Exs - ex) V A = - the line q, after the equivalence point, by .. b k V A = ' v + Cq where t is the normality of the standard solution of S; V the volume of the solution at the beginning of the titration; v the volume of the standard solution of S added to the solution; and (cXs - eS) -2 , respectively, where ve is the and Cp and C, are constants of values- volume added at the equivalence point. Exbvet bv t V V If the slope factors of Cp and q are mp and m,, then . . .. - * (2) C E Am = mp - m, = bt7 , . where & represents the sum of the E values, each with the appropriate sign.In practice 9 and q lines are found by means of absorbance readings, A , taken after additions, v , of standard solution. Let the co-ordinates of line p be (xi,yl)p and those of line q be (xi,yi),, where x represents volume and y the corrected absorbance. The point of intersection of p and g, corresponding with the end-point volume, V e , is given by the equation .. .. . . * - (3)) jjp - jjq + mpz, - mqzq Am The m-values are obtained by the classical least squares method, with the usual assump- By analogy, the variance in the y-direction can be calculated for each of the lines Cp and g. Zyf - jjCyi - m(Cxiyi - Zyi) Ve = - where jjp is the average of the y-values of line Cp, etc. tions. The result2 is .. .. * - (a), s2 =- N - 2 where N is the number of points for Cp or q.reading) and s,2 (the variance in additions of the standard solution)- When precision burettes are used s i > m2sf, and s Y = s,. the variance in ve is given by- The variance of y can also be expressed in terms of s i (the variance in the absorbance s; = s i + m2s,2 . . .. .. .. . . (5). Applying the law of propagation of errors to ve in equation (3) and substituting sA for s,, siq rz - 2v, . . (6). 9 +--- (Am)2 N , Cx; - This equation can be simplified. The expressions in brackets in equation (6) are dependent on the number of absorbance readings of the lines p and q, on Ve and on the values of xi, which can be expressed in terms of Ve as follows. Assuming a regular distribution of the volume additions, xi, over the whole titration curve, e.g., for five absorbance readings on each of the lines Cp and q, volumes of 0.1, 0.3, 0.5, 0.7 and 0.9 Ve on line Cp, and 1.1, 1.3, 1.5, 1-7 and 1.9 Ve on line q, the mean values of the expressions in brackets, for different values of N, are NpmdNq .... .. . . 2 3 4 5 6 9 Values of the terms in brackets . . 5.0 2.3 1.5 1-1 0.9 0-6 . These values are slightly dependent on the position of the various points on the lines p and q.862 GROENEVELD AND DEN BOEF: LIMIT OF DETERMINATION [AfldySt, VOl. 94 Five or six points on each line are the usual number of points in practice, and a regular distribution of the points is always taken. Therefore, for five or six regularly distributed points on each of the lines fi and q the predicted variance in the end-point is 1 This assumes a constant value of S, along the lines p and q.This is especially true at the limit of determination, where absorbance changes are small, and sAP is also equal to sAq, resulting in sue = 24- q2 .. .. .. .. .. Am According to our definition, at the limit of determination sue = 0.05 Ve . . .. .. .. .; (7). From equations (7), (6b) and (2) the limit of determination, expressed in terms of the amount of substance X, is given by- .. .. .. s A 1 / 2 v vet = ~ 0.05 b& ' . and in terms of the concentration of X by ' . . (8b). The limit of determination depends, therefore, on the instrumental arrangement (sA, V and b), as well as on the optical properties of the substances involved (&). When only one of the substances involved in the reaction absorbs, the value of the difference in absorbance at the limit of determination can be calculated from equation (8b) .... sA d2 AA = EbCx =- 0.05 . . (9). In practice these predicted limits of determination can be obtained when instrumental errors alone are involved. Any chemical or kinetic influence will disturb the system. Experimental results are compared with these predicted values with the two photometers for two reactions. RESULTS OBTAINED WITH THE PMQ 11 SPECTROPHOTOMETER, DEFLECTION TYPE- Reproducible values of the standard deviation sA can only be measured in vibration-free and draught-free rooms. Some scale drift also occurs, especially during the first 2 hours after the lamp has been switched on. Many absorbance values at A = 0 were measured, visually and by photographing the scale at time intervals of 15 seconds, at different values of the amplification of the apparatus and at different wavelengths.It can be proved that no important change in sA occurs in the absorbance region of A = 0 to 0 ~ 5 . ~ Details are given in Table I. TABLE I VALUES OF S i AT A = 0 WITH THE PMQ I1 SPECTROPHOTOMETER AT DIFFERENT WAVELENGTHS AND AMPLIFICATION Wavelength, nm Detector Amplification PA absorbance units 0.21 x 10-8 0.37 x 0-73 x 0.88 x 1.12 x 10-6 0.03 x lo-* 860 Photocell O N 1 600 Multiplier O P P 600 Multiplier O N 1 600 Multiplier 10/1O/I 360 Multiplier OP/I 360* Multiplier 0/1P A tungsten lamp was used as a light source, except at 360 nm,* when a hydrogen lamp was used. As seen in Table I, with the tungsten lamp sj increases as the wavelength decreases. This can be explained by Planck's law for black radiation, when temperature fluctuations of the wire in the lamp, caused by voltage fluctuations, are assumed to be the main contribution to the fluctuations in A .The stability of the hydrogen lamp was, in our experiments, supenor to that of the tungsten lamp.October, 19691 IN PHOTOMETRIC TITRATIONS WITH SELF-INDICATING SYSTEMS 863 To illustrate the concept of limit of determination we chose the reaction of vanadium(1V) with cerium(1V) in 0.5 M sulphuric acid at 750 nm. This is the wavelength of maximum absorption by vanadium(1V) with E = 15; the other substances involved in the reaction do not absorb at this wavelength. (see Table I), in equation (8b) , the predicted value of the limit of determination is When b = 2 cm and sA = 4.6 x cx = 4.3 x 104.In Table I1 results are given for three series of titrations in this concentration range. TABLE I1 RESULTS OF THE TITRATION M vOi+ with 0-045 M cerium(1V) Experiment Series I Series I1 Series I11 No. ve, ml ve, ml ve, Titration of 4.25 x 1 0.081 0.088 0.101 2 0.083 0.097 0.098 3 0.092 0.096 0.103 4 0.081 0.097 0.094 5 0.093 0.106 0-093 6 0.080 0-097 0-096 7 0.087 - 0.101 8 0.096 0.086 - - 9 10 0.087 - - v e 0.0868 0.0968 0.0979 S,e, per cent. 6.4 6.9 4.0 AA 0.014 0-014 0.014 - - As seen, there is good agreement between the standard deviation in the experiments The limit of determination is, therefore, in perfect agreement with the statistically with the value of 5 per cent.taken as the starting point. predicted value. RESULTS OBTAINED WITH THE ELKO 11, FILTER PHOTOMETER, SUBSTITUTION PRINCIPLE- The same reaction was carried out with the Elk0 I1 photometer. The standard deviation of the absorbance reading at A = 0 was 4 x for the filter When b = 2 cm, sA = 4 x lo6 and E = 15, substituting in equation (8b), the limit of S75,4 selected for the titration of VOi+ with cerium(1V). determination is given by Two series of titrations were carried out with solutions of about this concentration. In the first, 8 ml of 11-2 x M VOz+ were titrated with 4 x M cerium(1V). The results for seven titrations were Ve = 0.2299 ml, sVe = 1.6 per cent. and AA = 0.003. M, and for eleven titrations Ye = 0.1281 ml, s,,, = 3.7 per cent. and AA = 04015. Again these results agree with the predicted values. The titration of cerium(1V) with iron(I1) was also investigated; using the absorption of cerium(1V) at 490 nm, with filter S49E and b = 2 cm, sA = 4 x and & = 12 we predict that cx = 4-7 x 10-6. Titrations with 4 x M cerium(1V) were carried out, and for twelve titrations sve = 5.8 per cent., the absorbance range during the whole titration being about 0-001 absorbance units. This reaction also behaves ideally. REFERENCES cX = 3.8 x In the second series the concentration of VOi+ was 5.6 x 1. 2. 3. 4. Mandel. J.. “The Statistical Analysis of Experimental Data,” Interscience Publishers, New York Nalimov, V. V., “The Application of Mathematical Statistics to Chemical Analysis,” Pergamon Groeneveld, E. R., Ph.D. Thesis, University of Amsterdam, 1967. Groeneveld, E. R., and den Boef, G., 2. analyt. Chm., 1966, 219, 328. and London, 1964. Press, Oxford, London, Paris and Frankfurt, 1963. Received July 16th, 1968 Accepted March 14fh, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699400860

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

864 Amdyst, October, 1969, Vol. 94, fq5. 864-870 The Atornic=ernission Spectroscopy of the Rare Earth Elements in a Separated Nitrous Oxide = Acetylene Flame* D. N. HINGLE,? G. F. KIRKBRIGHT AND T. S. WEST (Chemistry Department, Imperial College, Londorr. S . W. 7) Flame emission in the interconal zone of a separated nitrous oxide- acetylene flame is shown to provide a sensitive technique for the detection of the rare earth elements, scandium and yttrium. The relative intensities of the principal atomic lines of these elements emitted in this flame have been measured, and the detection limits at the lines most suitable for flame- emission spectroscopy are reported for aqueous and ethanolic sample solutions. The effect on the atomic-emission intensities obtained for each of the sixteen elements investigated in the presence of the other fifteen elements has been investigated.At the concentrations used in this general survey no serious spectral line interferences between the rare earths were observed because of the simple spectra excited by the hot, reducing fuel-rich flame. ATOMIC flame-emission spectroscopy has several potential advantages over other flame spectroscopic techniques for the determination of traces of the rare earth elements. This technique has not been adopted for the rare earths as fully as its potential sensitivity and selectivity warrant. Molecular-band emission spectroscopy in flames has long been applied to the rare earth elements, however, and many of the rare earth monoxides emit well defined band spectra in flames of moderate temperature such as the air - acetylene,lD2 oxygen - hydro- gen3s4 and town gas - oxygen6 flames.When individual rare earths are to be determined in complex rare earth matrices, however, serious radiational and physical inter-element effects are frequently encountered in analytical moleculax-band spectroscopy in such flames. These effects are caused by the complexity of the band spectra observed from rare earth mixtures and the refractory nature of many of the rare earth oxides. The hot, reducing fuel-rich oxygen - acetylene flame has been shown to promote dissociation of rare earth oxide species to provide excited rare earth atoms for analytical atomic ~pectroscopy.~ 9 7 Later application of a total-consumption nebuliser - burner fitted with a pre-mixing attachment has further enabled these workers* s9 to obtain superior detection limits for the rare earths in this flame because of the lower background and noise levels from the oxygen - acetylene flame at this burner.In general, the samples were introduced into the flame as ethanolic solutions of the rare earth perchlorate salts. Skogerboe, Heybey and Morrisonlo have reported atomic emission from alcoholic solutions of several rare earths introduced into a turbulent oxygen - hydrogen flame by a force-feed mechanical delivery system. Pickett and Koirtyohannll have investigated the emission of several of the rare earth elements in the pre-mixed nitrous oxide - acetylene flame by using a long-path atomic-absorption burner. In several recent papers from this laboratory separated flames have been demonstrated to possess several advantages over conventional flames in atomic-emission, atomic-absorption and atomic-fluorescence spectro~copy.~~ ,13914915,16 In particular, the interconal zone of the fuel-rich separated nitrous oxide - acetylene flame exhibits a high temperature and has strongly reducing properties.Its reducing nature and relatively low background and noise levels make the interconal region of this stable pre-mixed flame particularly suitable for the atomic-emission spectroscopy of elements that form refractory oxides in cooler conventional flames. This paper describes the application of this flame to the sensitive and selective detection of the rare earth elements. APPARATUS- The burner arrangement used has been described in detail elsewhere15; it consists of a circular, stainless-steel, water-cooled burner head with a circular slot of 0.50 mm wide and 11 mm i.d.The silica separator tube is 50 mm long and 21.5 mm i.d. A short silica side-arm * Paper presented at the Second SAC Conference 1968, Nottingham. t Present address : Murex Limited, Rainham, Essex. 0 SAC and the authors.HINGLE, KIRKBRIGHT AND WEST 865 is fused to the separator tube and this is fitted with a B19 ground-glass cone and socket, to which an optical-quality silica end-window is cemented. In operation, the primary reaction zone of the flame burns steadily at the slot of the burner head. The diffusion zone, where the incompletely combusted products of the primary reaction bum with atmospheric oxygen, is maintained at the top of the silica separator tube.The interconal zone of the flame is viewed by the monochromator and detector system via the window in the side-arm of the separator tube. The detector assembly and nebuliser unit of a Techtron AA4 flame spectrophotometer (Varian - Techtron Pty. Limited, Victoria, Australia) were used. The monochromator of this apparatus has a reciprocal linear dispersion at the exit slit, in the first order, of 3.3 nm. The photomultiplier detector was a Hamamatsu, type R213, spectral response 185 to 800 nm, and wavelength of maximum response 430 nm. Nitrous oxide pressure at the nebuliser was 7.5 p.s.i.g., corresponding to a flow-rate of 5-2 litres minute-l. The 1 nm minute-l spectrum scanning motor used in the earlier work16 was replaced by a synchronous motor to drive the monochromator at 10 nm minute-l.REAGENTS- The rare earths were obtained as their oxides (Johnson, Matthey and Co., London). All solutions were prepared by dilution of lo00 p.p.m. ethanolic stock solutions of the rare earth perchlorates. With the exception of cerium these stock solutions were prepared as follows. The required amount of the rare earth oxide was transferred to a 10-ml beaker and 8 ml of 60 per cent. analytical-reagent grade perchloric acid added. The solution was evaporated at a distance of about 5 cm below an infrared lamp (250 W, Philips) until first a clear solution formed and then crystals of the rare earth perchlorate were deposited. At this point the solution volume was 0-2 to 0.3 ml.The beaker was then placed in a desiccator containing silica gel to cool. When the solutions were cool crystallisation was complete and no free liquid remained. The crystalline perchlorate was dissolved and diluted with absolute ethanol to 100 ml to yield a 1000 p.p.m. solution, and stored in a polythene bottle. The cerium stock solution was prepared from AnalaR ammonium cerium nitrate, (NH,),Ce(NO,),. The required amount of this salt was weighed into a 25-ml beaker, and perchloric acid was added and the solution evaporated as described above. The evaporation process was repeated twice and the crystals obtained after the third evaporation were dissolved in absolute ethanol in the manner previously described. A 1O,o00 p.p.m. sodium solution was prepared by dissolution of sodium perchlorate in absolute ethanol. All rare earth solutions nebulised into the flame contained 200 p.p.m.of sodium. RESULTS AND DISCUSSION To assist interpretation of the flame-emission spectroscopy results, the effect of nebuljsa- tion of ethanolic rather than aqueous solutions on the nebuliser efficiency and solution nebu- lisation rate was first investigated. Samarium solutions in distilled water, 25, 50 and 75 per cent. ethanol - water solutions and absolute ethanol were nebulised into the separated nitrous oxide - acetylene flame at a flow-rate of 5-2 litres minute-l of nitrous oxide. In each instance 1Oml of solution were nebulised several times in the flame and the results obtained are shown in Table I. TABLE I EFFECT OF ETHANOL ON SOLUTION NEBULISATION RATE AND EMISSION INTENSITY FOR SAMARIUM AT 488.3 nm Emission intensity Volume entering Nebulisation rate for samarium at 488.3 nm Solution flame, ml minute-l relative to water relative to water Aqueous .. .. .. 0.27 1 : l 1 : l 50 per cent. ethanol . . 0.43 1 : 1.5 1 : 1.8 25 per cent. ethanol . . 0.3 1 1 : 1.2 1 : 1.4 76 per cent. ethanol . . 0.54 1 : 2.0 1 :3*1 100 per cent. ethanol . . 0.62 1 : 2.2 1 : 6.5866 HINGLE et d. : ATOMIC-EMISSION SPECTROSCOPY OF THE RARE EARTH [Analyst, VO~. 94 It is evident that the nebulisation efficiency is doubled when absolute ethanol is used. The final column in Table I shows the manner in which the use of organic solvent affects the emission intensity for samarium at 488.3 nm. Here slightly more than &fold enhancement in observed emission intensity is obtained for samarium introduced into the flame in absolute ethanol compared with that observed when an aqueous solution of the same concentration is nebulised.Similar enhancements can be observed for many of the other rare earth elements. This effect cannot be attributed to more efficient nebulisation for ethanol, and it is difficult unequivocally to ascribe the additional enhancement to one particular cause. It may possibly be attributed to a slightly higher flame temperature or increased reducing ability in the interconal zone obtained when an organic solvent is nebulised, or by a chemi- luminescent effect whereby a fraction of the rare earth atoms present in the flame plasma is formed directly in the excited state in the presence of organic solvent.EFFECT OF ADDITION OF SODIUM- Several authors17Js have described the effect of ionisation of atoms of various elements in the nitrous oxide - acetylene flame on their determination by atomic-absorption spectro- scopy. Manning,lS for example, has reported that 75 per cent. ionisation of europium atoms occurs when 100 p.p.m. europium solutions are introduced into a nitrous oxide - acetylene flame. It is common practice in flame spectroscopy to suppress thermal ionisation of this type by the addition of a high concentration of an element such as sodium or potassium, the ionisation potential of which is lower than that of the atoms of the element being deter- mined. The ionisation potentials of the rare earths lie between about 5.5 and 6-5 eV, so that appreciable ionisation occurs for many of the elements in the nitrous oxide - acetylene flame.Lower sensitivities, therefore, result from the smaller number of ground-state neutral atoms available for excitation. The effect of the addition of a more easily ionised element on the atomic emission of a typical rare earth element has been investigated. A 25 p.p.m. ethanolic solution of samarium (ionisation potential 5.6 eV) was used for this investigation. The results obtained are shown in Table 11. It was decided from these results that throughout the investigation all rare earth solutions nebulised should contain 200 p.p.m. of sodium to prevent any loss of sensitivity through partial ionisation. TABLE I1 EFFECT OF ADDITION OF SODIUM TO SUPPRESS IONISATION OF SAMARIUM IN THE SEPARATED NITROUS OXIDE - ACETYLENE FLAME Emission intensity at 488.3 nm, Solution arbitrary units 26 p.p.m.of samarium . . .. .. .. .. 37.6 25 p.p.m. of samarium + 10 p.p.m. of sodium . . .. 65 25 p.p.m. of samarium + 50 p.p.rn. of sodium . . .. 62 26 p.p.m. of samarium + 100 p.p.m. of sodium . . .. 66-6 26 p.p.m. of samarium + 200 p.p.m. of sodium . . .. 67.5 26 p.p.m. of samarium + 600 p.p.m. of sodium . . .. 67 26 p.p.m. of samarium + 1000 p.p.m. of sodium . . 67 .. SELECTION OF SUITABLE ATOMIC LINES FOR ANALYTICAL SPECTROSCOPY- With the 'exception of cerium and gadolinium, solutions were prepared that contained 100 p.p.m. of the rare earth element under study and 200 p.p.m. of sodium in ethanol. The cerium and gadolinium solutions contained 500 p.p.m.of cerium and 200 p.p.m. of gadolinium. These solutions were nebulised into the separated nitrous oxide - acetylene flame and the atomic-emission spectrum for each element was recorded. The fuel flow-rate was adjusted in each instance to obtain the maximum signal-to-background ratio for each element. For example, the optimum ratio for europium was obtained even from a fuel-lean flame, whereas for cerium and gadolinium a higher background, reducing fuel-rich flame was required to obtain useful atomic emission for these elements. The instrumental conditions used through- out were as follows. Slit width 100 nm (spectral band width 0-33 nm); scan rate 10 nm minute-1; recorder speed 1 inch minute-l; and nitrous oxide flow-rate 5-2 litres minute-l.The amplifier gain was adjusted in each instance to ensure that the emission recorded from the most intense line for the element concerned produced full-scale deflection at the recorder. The most intense useful lines were selected from the recorded spectra, and the lines most suitable for the quantitative determination of the elements were chosen taking into accountOctober, 19691 867 the flame background in the same spectral region. The relative intensities of the analytically useful lines were accurately measured at a slit width of 50 nm (spectral band width 0.16 nm). The results obtained are shown in Table 111. These readings were obtained by setting the monochromator at the wavelength of each line and nebulising the sample and blank solutions.TABLE I11 RELATIVE INTENSITIES OF THE MOST USEFUL RARE EARTH EMISSION LINE (In each instance the most intense line of an element is given a relative ELEMENTS IN A SEPARATED NITROUS OXIDE - ACETYLENE FLAME OBTAINED IN THE SEPARATED NITROUS OXIDE - ACETYLENE FLAME Element Lanthanum. . Cerium . . Praseodymium Neodymium Samarium .. Europium . . Gadolinium Terbium . . .. .. .. .. .. .. .. .. - intensity of one hundred) Slit 50 pm Wavelength, Relative nm intensity" 550.134 494.977 545.515 428-026 520.0 12t : 520-042$ 520.039: 495.136 5 13.342 493.974 492.459 492-453 463.424 468.345 494.483 495.478 488-377 476.027 2:;;: } 520-059 478-310 471.610 471.707 428.221 428.283 428.350 459.403 462.722 466.188 676.520 434-662 461.966 432-712 422-585 442-241 440.186 405-822 443-063 407.870 441.473 430.634 431.384 432-647 431.885 433.845 406- 159 449.308 100 72 60 31 100 100 56 52 46 100 41 39 35 50 130 75 75 68 62 60 35 100 77 61 4 100 80 72 72 59 54 50 47 43 36 33 17 100 67 62 25 25 Element Dysprosium Holmium ..Erbium . . Thulium . . Ytterbium . . Lutetium . . Scandium . . Yttrium .. Wavelength, Relative nm intensity* .. 421.172 100 418,678 46 404.599 46 419,485 31 458.937 11 . . 410-384 100 405.393 82 416.303 50 404.081 5 .. 400-797 100 415-110 58 397.304 397.360 } 27 408-765 16 460.662 9 .. 410.584 100 418.762 94 409.419 93 371.792 78 374-407 47 435.993 26 530.712 19 567.585 12 .. 398.798 346-436 555.648 .. 451.857 361.211 337.650 328.174 327-897 .. 391-181 402.369 390-749 402.040 399.661 326.991 .. 407.738 410.238 412-831 412.485 464.370 467-484 362.094 404-764 * Uncorrected for response characteristics of Hamamatsu R213 photo- multiplier and monochromator.t These are the only clearly defined assignable lines observed for cerium. $ Denotes unresolved lines. 100 9 8 100 34 26 22 14 100 91 83 71 11 5 100 92 78 71 43 34 25 11868 HINGLE et al. : ATOMIC-EMISSION SPECTROSCOPY OF THE RARE EARTH [Autalyst, Vol. 94 LIMITS OF DETECTION- Table IV shows the detection limit obtained in the separated nitrous oxide - acetylene flame with the instrumentation described here in 10, 25, 50, and 75 per cent. ethanol - water mixtures and in absolute ethanol for each of the rare earth elements. The detection limit was defined as that concentration of the element in solution producing a signal-to-noise ratio of unity.A slit width of 50 nm was used to obtain these results. The use of this relatively narrow band-pass (0-16nm) yields poorer detection limits in many instances than when a wider slit was used, but ensures minimal spectral interference from atomic lines of other elements present in the matrix to be analysed. The optimum fuel flow-rate was used for each element when its detection limit was determined. Atomic emission from the elements whose oxides are refractory in nature (e.g., cerium, praseodymium and gadolinium) was at a maximum in a fuel-rich reducing flame. For the less refractory rare earth elements, whose atoms are more readily formed by a purely thermal dissociation of the monoxide species, it was found that lower detection limits were obtained by using a leaner and hotter, less reducing flame. TABLE IV RARE EARTH DETECTION LIMITS IN SEPARATED NITROUS OXIDE - ACETYLENE FLAME Detection limits, p.p.m.Element La ce Pr Nd Sm Eu G d Tb Ho Er Tm Yb Lu sc Y DY Wavelength, nm 660.134 620.012 520.042 620.039 496-136 492.453 488.377 469.403 434.662 432.647 421.172 410.384 400.7 97 410-584 398.798 451,857 391.1 8 1 407.738 Ethanol, 100 per cent. 4 60 1.8 0.6 0.18 0.007 3 1.3 0.18 0.07 0.21 0.06 0.012 0.7 0.06 0.6 Ethanol, 75 per cent. 6 Ethanol, 50 per cent. 8 Ethanol, 10 per cent. 10 4 0.8 0.3 0.015 6 2-8 0.25 0.18 0.32 0.13 0.019 1.0 0.10 0.65 10 1.2 0.4 0.017 8 6 0.46 0.23 0.46 0-17 0.025 1.4 0-13 0-77 12.6 2.0 0.5 0.02 7.5 0.55 0.30 0.5 0.25 0.033 1.7 0.16 0.9 10 INTERFERENCES- The effect on the atomic-emission intensity obtained at the selected analytical lines for each of the sixteen elements investigated in the presence of the other fifteen elements was examined.A slit width of 500 nm (band width 0.16 nm) was used throughout to minimise spectral interference. Initially it was hoped that improved selectivity might result from adjusting the acetylene flow-rate to suit the element under study. For example, spectral interference from any nearby atomic lines of the more refractory rare earth elements on the chosen analytical line of a less refractory element can be minimised by using a stoicheiometric flame. Under these conditions, however, more serious interference may be encountered from the oxide band spectra of refractory rare earth elements than in a fuel-rich flame, and this replaces the interference caused by their atomic emission.This effect was observed in several instances. The acetylene flow-rate corresponding to that required for a slightly fuel-rich flame was, therefore, used throughout. Solutions containing twenty times the minimum detectable concentration of the element studied, or 100 p.p.m., whichever was less, were prepared. These solutions were prepared to contain 200 p.p.m. of sodium and 100 p.p.m. of the rare earth element, the interference of which was under examination. Each of these solutions was nebulised in turn in the separated flame, followed by a standard solution of the element at the same concentration (containing no interfering element) and a blank solution. The degree of any interference was assessed from the signals obtained for the standard solution and solution of the ion containing the foreign ion.In each instance the monochromator was set at the wavelength of the most useful analytical line. When serious interference was encountered, an attempt was made to find an alternative line for the element at which the foreign ion did not interfere. All The following procedure was used.October, 19691 869 the investigations of interferences were carried out with a 60 per cent. ethanol - water mixture. The results obtained are shown in Table V. TABLE V SPECTRAL INTERFERENCES FOUND FOR DETERMINATION OF EACH ELEMENT INVESTIGATED ELEMENTS IN A SEPARATED NITROUS OXIDE - ACETYLENE FLAME Element, p.p.m. Lanthanum, 100 . . Cerium, 300 . . .. Praseodymium, 100 ..Neodymium, 25 Samarium, 8 .. Europium, 0.34 Gadolinium, 100 Terbium, 100 . . Dysprosium, 7.5 Holmium, 4.5 . . Erbium, 9 .. Thulium, 3.4 . . Ytterbium, 0-5 Lutetium, 28 . . Scandium, 2.5 . . Yttrium, 14 .. .. .. .. .. a . .. .. .. .. .. .. - . . . Wavelength, nm 550.134 520.042 495-136 513.342 492,453 463.424 488.397 471.610 459.403 434.646 432.647 431-885 421.172 4 10.384 405.392 400.797 410.584 499.4 18 398.798 346.436 461.857 331.211 39 1- 18 1 407.738 410.238 Interfering species and wavelength of line, nm - 1 - Nd 495.029 Nd 496.067 Nd 495.246 Nd (100 p.p.m.) does not interfere Pr 492.459 Sm 492.404 Pr and Sm do not interfere Nd 488.381 Y oxide band Sc oxide band ,Nd and Sc do not interfere Y Gadolinium oxide band interferes The 462.722 and 466.188 europium lines are also in a gadolinium oxide band system and cannot be used as an alternative - Gd 432.712 Gd does not interfere Y 410-238 Dy 410.388 Y and Dy do not interfere HO 410.384 Ho does not interfere Er 398.766 Er does not interfere Pr 451-663 Gd 451.966 Pr and Gd do not interfere Gd 407-870 Er 407.788 Er and Gd do not interfere - - - Magnitude of interference for 100 p.p.m.of interfering ion, per cent. 4 - 35 80 200 10 25 3 20 - 10 - 36 12 - - 10 40 10 10 - 3 5 It was observed that several of the rare earth oxide samples used during this work contained detectable amounts of other rare earth elements. In these instances an apparent positive interference effect results when the element occurring as impurity in the sample whose interference is being investigated is the same as that being determined.This was corrected for by nebulising a solution of the interfering ion and subtracting the observed line- emission intensity at the wavelength of the analytical line of the element being determined. At the concentrations investigated it is evident that in only one case (samarium) do as many as three other rare earth elements interfere when the most useful analytical line is used. For praseodymium, neodymium, samarium, terbium, holmium, thulium, ytterbium, lutetium and yttrium even these elements which interfere at the most useful line do not interfere when an alternative (but usually less intense) line is selected. Lanthanum, gado- linium, dysprosium, erbium and scandium can be determined at the concentrations investi- gated without interference in the presence of any other rare earth elements at 100 p.p.m.CONCLUSION The principal difficulty encountered in the development of both flame-emission and absorption methods for the determination of the rare earth elements has until recently been the lack of suitable flames capable of dissociating the extremely stable rare earth monoxide species formed in most flames. Fassel, Curry and Kniselep have demonstrated the usefulness870 HINGLE, KIRKBRIGHT AND WEST of the hot, fuel-rich oxygen - acetylene flame in this respect for production of the atomic spectra of the rare earths. These authors concluded that in this flame thermal dissociation of the monoxide molecules does not appear to be the primary process in populating the flame with neutral (or excited) atoms. The dissociation of rare earth monoxide molecules in the pre-mixed nitrous oxide - acetylene flame reported here also requires a fuel-rich flame.As is also the case for the oxygen - acetylene flame, the fuel-rich flame is slightly cooler than the stoicheiometric flame, and if the dissociation mechanism were purely thermal in nature, greater line intensities would be expected in the stoicheiometric flame. The necessity for the use of a fuel-rich nitrous oxide-acetylene flame has also been demonstrated in the determination of the rare earth elements by atomic-absorption spectrosc0py.m Several workers21 s2, ,2s have suggested mechanisms to explain the need for fuel-rich reducing flames to dissociate molecular oxide molecules.The investigation reported here reveals that good sensitivity is available for the deter- mination of the rare earth elements by flame-emission spectroscopy in the separated nitrous oxide - acetylene flame. The use of a separated flame ensures suppression of flame background radiation from OH and the chemiluminescence of the reaction CO + 0 -+ CO,. The reducing atmosphere in the interconal zone of the fuel-rich flame is also protected from atmospheric oxidation in this flame, and this results in the presence of a greater total number of emitting sample atoms whose radiation is available for detection. Because of the high intensity of the emitted radiation from samples introduced into the flame, it is possible to retain good sensitivity while using a monochromator of good resolution at narrow slit widths to obtain good spectral selectivity.The sensitivities obtained compare well with those obtainable by flame-emission spectroscopy in the oxygen - acetylene flame at a total-consumption atomiser - burner, and even in only 10 per cent. ethanol - water solutions the detection limits obtained for neodymium, europium, holmium, lutetium, scandium, samarium, ytterbium and lanthanum are equal or superior to those obtainable by atomic-absorption spectroscopy.m Good selec- tivity is easier to obtain by measurement of the atomic-line spectra of the rare earths by flame-emission spectroscopy than by arc or spark-emission spectroscopy. The simpler spectra obtainable with the flame source as compared with the often complex spectra from arc and spark sources enable analyses of rare earth mixtures to be made with a simple low dispersion spectrophotometer.Even with a high resolution spectrograph the positive location of lines of the element required to be determined is frequently difficult, and the incidence of inter- ference from lines of other rare earth elements is high. We are grateful to Murex Limited for provision of study leave to one of us, D.N.H., and for the loan of the flame spectrophotometer used in this work. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. 16. 17. 18. 19. 20. 21. 22. 23. REFERENCES Lundegardh, H., Svensk. Kem. Tidskr., 1930, 42, 61; Chem. Abstr., 24, 2962. Pinta. M., J . Rech. Cent. Natn Rech Scient., 1962, 4, 260; Chem. Abstr., 47, 8611b. Piccardi, G., and Sberna, A., Atti. Accad. Naa. Linck. Rc., 1932. 15, 83; Chem. Abstr., 26, 5031. Rains, T. C., House, H. P., and Menis, O., Analytica Chim. Acta. 1960, 22, 316. Piccardi, G., Spectrochim. Acta, 1939, 1, 249. Fassel, V. A., Curry, R. H., and Kniseley, R. N., Ibid., 1962, 18, 1127. D’Silva, A. P., Kniseley, R. N., Fassel, V. A., Curry, R. H., and Myers, R. B., A d y t . Chem., Fassel, V. A., Kniseley, R. N., and D’Silva, A. P., Ibid., 1964, 36, 1287. Fassel, V. A., and Golightly, D. W., Ibid., 1967, 39, 466. Skogerboe, R. K., Heybey, A. T., and Momson, G. H., Ibid., 1966, 38, 1821. Pickett, E. E., and Koirtyohann, S. R., Spectrochim. Acta, 1968, 23B, 236. Kirkbright. G. F., Semb, A., and West, T. S., Talanta, 1967, 14, 1011. --- , Ibid., 1968, 15, 441. Hinile, D.’N., Kirkbright, G. F., and West, T. S., Ibid., 1968, 15, 199. --- , Analyst, 1968, 93, 622. Hobbs, R.,S., Kirkbright, G. F.. Sargent, M., and West, T. S., Talanta, 1968, 15, 997. Amos, M. D., and Willis, J. B., Spectrochim. Acta, 1966, 22, 1326. Manning, D. C., and Capacho-Delgado. L., Analytica Chim. Ada, 1966, 36. 312. Manning, D. C., Atomic Absorption Newsletter, 1966, 5, 63. Slavin, W., Appl. Spectrosc., 1966, 20, 281. Kirkbright, G. F., Peters, M. K., Sargent, M., and West, T. S., Talanta, 1968. 15, 663. Gibson, J. H., Crossman, W. E. L., and Cooke, W. D., AnaZyt. Chem., 1963, 35, 266. Amos, M. D., and Thomas, P. E., Analytica Chim. Acta. 1966, 32, 139. 1964, 36, 632. Received December loth, 1968 Accepted March 28th, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699400864

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Analyst, October, 1969, Vol. 94, $$. 871-4378 871 Emission Spectra Obtained from the Combustion of Organic Compounds in Hydrogen Flames BY R. M. DAGNALL, D. J. SMITH, K. C. THOMPSON AND T. S. WEST (Department of Chemistry, Imperial College, London, S . W.7) A study is presented of the emission spectra produced by nebulisation of organic liquids into a nitrogen - hydrogen diffusion flame burning in air, and into a laminar-flow pre-mixed air - hydrogen flame. Both flames exhibit low background emissions and that of the diffusion flame is particularly low even over the OH-band region. Carbon, hydrogen and oxygen-containing compounds exhibit intense bands for CH, CHO, C, and OH species. Nitrogen- containing organic compounds additionally display NH, NO and CN bands ; chlorine-containing compounds give CC1 bands in the diffusion flame ; sulphur compounds give CS and S , bands; and phosphorus compounds give HPO and PO bands.The spectral distribution of these bands and their intensities indicate that many types of organic compound can be characterised and identifred by direct observation of their emission spectra during combustion in these cool flames. THE measurement of the flame-emission spectra of organic compounds is a relatively new technique in analytical chemistry, but Gaydonl has already identified many emitting species (Cz, CH, CHO and CH,O) from flames supported by various organic compounds. Flame-emission detectors for use in gas chromatography have been de~cribed.~ s3 94 These detectors involve the use of the C, and CH emission from air or oxy-hydrogen flames.McCor- mack6 found that greater sensitivities were obtained by using a microwave-generated plasma rather than the flame as a method of identifying peaks from gas chromatograms. More recently McCrea and Lights have described the measurement of C , and CH-band emissions for the determination of hydrocarbons in methanol, and Robinson and Smith7 have given a more general description of the emission spectra of organic liquids in oxy- hydrogen flames. The purpose of this study was to investigate the emission spectra produced by combusting organic compounds containing nitrogen, chlorine, phosphorus and sulphur, in addition to carbon, hydrogen and oxygen, in low-temperature laminar-flow nitrogen - hydrogen diffusion flames and pre-mixed air - hydrogen flames.APPARATUS- For these studies a Unicam SP9OOA flame-emission - atomic-absorption spectrophoto- meter, fitted with an E.M.I. 9601B photomultiplier, a standard air - acetylene (rectangular) burner head and various quartz tube burner heads (see text), was used. The spectra were recorded with a Servoscribe recorder used over the 0 to 10-mV range. Fuel gas, hydrogen from a cylinder; diluent gas, nitrogen from a cylinder; and air from a compressor. EMISSION FROM THE NITROGEN - HYDROGEN DIFFUSION FLAME- This flame, which has previously been describeda r9 #lo S~~ in papers from this laboratory, was maintained on the standard Unicam 1.8 x 7.5-cm air - acetylene emission head. The only background emission from this flame corresponds to OH-band emission about 310 nm, and even this is only about one fortieth of that of a conventional pre-mixed air - hydrogen flame.RESULTS 0 SAC and the authors.872 DAGNALL et id.: EMISSION SPECTRA OBTAINED FROM COMBUSTION [Aflaiyst, vol. 94 Nitrogen was used as the nebulising gas (at 15 p.s.i.g., 4 1 minute-l) with the conven- tional SPSOOA nebulising system. Hydrogen was introduced, as usual, at the bottom of the burner. The CH and C, emissions, when nebulising methanol, increased as the hydrogen flow-rate increased, but at fairly high flow-rates the emission began to decrease. The nebuli- sation of certain organic compounds (e.g., benzene) caused the flame to become unstable at even moderate hydrogen flow-rates. The most stable, reproducible fiarne was obtained when the hydrogen flow-rate was set just above that necessary to prevent the flame from lifting off when nebulising distilled water.Emission measurements were taken with the top of the burner head level with the bottom of the monochromator slit. Thus measurements were taken from the coolest region of the flame8 (centre temperature about 280" C, outer temperature about 800" C) where least breakdown would be expected to occur and fairly large molecular fragments should exist. The thermal energy of the flame is insufficient to account for all of the observed emissions, most of which are of chemiluminescent origin. Solid organic compounds were not examined at this stage because it was necessary to dissolve them in various organic solvents and con- comitant complication of the observed spectra.Solutions were, however, used with the air - hydrogen flame. CARBON - HYDROGEN AND CARBON - HYDROGEN - OXYGEN COMPOUNDS- Table I summarises the band emissions seen during the aspiration of methanol, acetone, acetylacetone, isopropyl alcohol, benzene and formaldehyde into the diffusion flame. Obviously CH and CHO compounds predominantly yield C, and CH emission spectra and increase the OH-band emission. TABLE I BANDS OBSERVED FROM SOME REPRESENTATIVE C-H-o COMPOUNDS AND BENZENE Bands Main Most Compound observed * wavelengths, nm intense band Methanol .. .. .. .. .. c2 616 CH CH 431 CH 43 1 Acetone . . .. .. .. .. c* 516 CH Isopropyl alcohol . . .. .. c, CH 616 431 CH Acetylacetone . . .. .. .. c, 616 CH Benzene . . .. .. Formaldehyde (40 per cent.aqueous) . . Complex banded 390 to 570 43 1 CH CH 431 .. .. c, 616 c, - continuum * OH emission was observed at about 310 nm for all compounds and the intensity was about CH 43 1 - 2 to 4 times that caused by the flame background. MethartoZ-It will be seen from Fig. 1 that in the region 400 to 500 nm the only emissions observed from the combustion of methanol were CH (431 nm) and C, (main band head at 516 nm). The fact that the latter is almost as strong as the former, although the methanol molecule contains only one carbon atom, suggests complex exothermic mechanisms involving liberation of excited C, species. The OH emission about 310 nm is four times greater than for water. The continuum, which has an intensity of about one half that of the C, band, may be caused by the CO - 0, reaction or other unidentified breakdown products.A cetofie-The emission from acetone (also isopropyl alcohol and acetylacetone) was similar to that for methanol except that, relative to the C, band and the continuum, the CH band was more pronounced. The OH emission was about twice that observed for methanol. Bemefie-The C, emission was, as expected, more intense and was greater than the CH emission. Two weak bands at 274 and 267 nm caused by undecomposed benzene were discernible above the OH background, and emission of the latter at 310nm was similar in intensity to that for methanol.October, 19691 OF ORGANIC COMPOUNDS IN HYDROGEN FLAMES 873 Formaldehyde solution (40 per cent., aqueous)-The C, emission was negligible and the CH emission at 431 nm was surprisingly weak.A broad complex band was observed from 390 to 570nm. It can be concluded that the CH:C, ratio is not significant because the emissions are chemiluminescent and critically dependent on fuel flow-rate and nebulisation rate. m Y .- C 3 2 2 .- D L m >; .- m C 0 C U - I I I 400 450 so0 550 Wavelength, nm Fig. 1. The spectrum of methanol in the nitrogen - hydrogen diffusion flame : slit, 0.02 mm; gain, 3,lO; band width, 2 Wavelength, nm Fig. 2. The spectrum of 50 per cent. methanol - 50 per cent. 0.88 ammonia solu- tion v/v in the nitrogen - hydrogen diffusion flame: A, slit, 0.03 mm; 3,5; band width, 2. B, slit, 0-10mm; gain, 3,8; band width, 2 NITROGEN-CONTAINING COMPOUNDS- All organic nitrogen-containing compounds tested gave CN, NH and NO emissions.The CH : CN : NH :NO intensity ratio (see Table 11) would appear to be dependent to some degree on the structure of the compound. It is, however, important to note that, as with CH and 0 compounds, these ratios are dependent on the hydrogen and nitrogen flow-rates and the rate of nebulisation, as well as the optical system used. It was essential to maintain the gas flow-rates constant during these measurements. TABLE I1 CN : CH : NH : NO EMISSION INTENSITY RATIOS OF NITROGEN-CONTAINING COMPOUNDS IN THE NITROGEN - HYDROGEN DIFFUSION FLAME Emission intensity* CH, NH, NO, ckt 259 nm Compound 360 nm 431 nm 337 nm 0-88 Ammonia solution - methanol (1 + 1 v/v) 100 93 260 300 Acetonitrile . . .. .. .. . . 100 77 40 94 Butylamine . . .. ... . .. 100 390 78 160 Diaminoethane . . .. .. .. .. 100 92 77 170 Pyridine .. .. .. .. . . 100 210 50 140 * The CN, CH and NH emission intensities were all measured with a slit width of 0.03 mm and converted into gain 2,5. The NO emission was measured a t slit width 0-1 mm and converted into gain 3,5. The CN : CH : NH : NO ratio was then calculated after arbitrarily setting the CN emission intensity to 100 units. t The main 389 nm CN band was not used because it was four times more intense than the 360 nm CN band and made the measurement of the CN : CH : NH ratios difficult. Ammonia (50 per cent. 0.88 ammonia solution - 50 per cent. methanol v/v)-The main features of this spectrum (see Fig. 2) were a broad band extending from 600 to 330 nm with a maximum at 520 nm, weak CN and NH emissions peaking at 389 and 337 nm, respectively, and weak NO emission12 with main band heads at 215,227,237,248,259 and 272 nm.Weak C, and CH emissions were observed at 516 and 431 nm, respectively. The broad band from 600 to 330nm was attributed to NH, emission.12874 DAGNALL et al.: EMISSION SPECTRA OBTAINED FROM COMBUSTION [Analyst, Vol. 94 The CN emission must result from a chemiluminescent reaction between the methanol and ammonia (or breakdown products of methanol and ammonia). The NO-band emission which extended into the far ultraviolet must be caused by an energetic reaction because bands about 220 nm require an excitation energy of 5.5 eV, which is far beyond the energy available in the diffusion flame. As the NO bands were still present on nebulising aqueous ammonia solution, we must assume that some energetic reaction between nitrogen and oxygen species must occur at the edge of the flame to produce excited NO molecules, and that carbon species are not necessary. Acetonitrile-Weak C, and CH emissions were observed at 516 and 431 nm.The CH emission at 431 nm was 2.5 times more intense than the C, emission at 516 nm. The most prominent features of the spectrum were the intense CN bands at 360, 389 and 418nm (see Fig. 3). Weak NO bands at 215, 227, 237, 248, 259 and 272nm were also observed. Pyridine, btdylamine and diaminoethane-Spectra similar to that of acetonitrile were obtained. The only major difference was that C, and CH emissions were much stronger with respect to the CN emission and that the NH : CN intensity ratio had increased (see Table 11).CHLORINE COMPOUNDS- Carbon tetrachloride-The main features of this spectrum were intense C, emission and weak CH and CC1 emission at 277 and 279 nm (Fig. 4). The intense C, emission is difficult to explain as the carbon tetrachloride molecule contains only one carbon atom, but can be accounted for by re-combination processes. Chloroform-The chloroform spectrum was similar to that of carbon tetrachloride, except that the C-H emission was slightly more intense with respect to the C, emission. CN I ~ CN 337 360 380 400 430 516 Wavelength, nm Fig. 3. The spectrum of acetonitrile in the nitrogen - hydrogen diffusion flame : slit, 0-013mm; gain, 2,O; band width, 2 Wavelength, nm Fig. 4. The spectrum of carbon tetra- chloride in the nitrogen - hydrogen diffusion flame: A, slit, 0.01 mm; gain, 3,O; band width, 2.B, slit, 0-038 mm; gain, 3,lO; band width, 2 EFFECT OF ADDING AIR TO THE NITROGEN - HYDROGEN FLAME- C,, CH, NH, CN, NO and OH emissions, but to decrease the emission from GC1. The effect of adding air through a third jet in the burner base13 was to increase the THE LAMINAR-FLOW PRE-MIXED AIR - HYDROGEN FLAME- Previous workerss s 7 have used total-consumption burners which gave turbulent air and oxy-hydrogen flames. The organic emission extended throughout the flame. In the present study, when methanol was nebulised into a laminar-flow pre-mixed air - hydrogen flame, well defined blue primary reaction cones were observed just above the holes in the burner head.It is from these primary reaction cones that almost all the observed organic emission occurs. In general, the maximal emission intensity for most species occurred in the lean flameOctober, 19691 OF ORGANIC COMPOUNDS I N HYDROGEN FLAMES 875 with the hydrogen flow-rate just above the flash-back point. Under these conditions with the standard burner head used, the primary reaction cones were about 3 mm high. The length of these cones could be increased to 18 mm to fill the monochromator slit by increasing the hydrogen flow-rate, but this resulted in a decrease in the emission intensity. 0.5 cm * Burner stem Fig. 5. The lami- nar-flow pre-mixed air - hydrogen burner head Instead of using the standard burner head, the flame could be maintained more suitably on a 5 mm i.d.quartz tube fitted into the top of the burner stem with a suitable adaptor (Fig. 5). When methanol was nebulised, a single, well defined, primary reaction cone, 18 mm high, was obtained with the hydrogen flow-rate just above the flash-back point. This “tube” burner showed no carbon build-up, unlike conventional burner heads, and also remained quite cool in operation. Larger diameter “tube” burners (ie., 10mm) gave weaker emission and a tendency to flash-back, while smaller diameter burners ( i e . , 3 mm) caused the flame to lift off rather easily. In general, the emission intensities with the 5-mm tube burner were about 2 to 3 times greater than those obtained with the conventional burner head. The organic compounds under test were nebulised as 10 per cent.v/v solutions in methanol. This procedure was adopted for two reasons. Firstly, the flame was rather un- stable when nebulising certain compounds (e.g., benzene and pyridine) and, secondly, the nebulising properties of solutions containing 90 per cent. v/v of methanol should be similar. In fact the rate of solution uptake was found to be similar for a range of compounds (see Table 111). The spectra from the organic compounds are always accompanied by those from methanol in these experiments. TABLE I11 RATE OF UPTAKE OF METHANOLIC SOLUTIONS Compound .. Rate of solution uptake,* ml minute-1 Methanol . . .. .. .. .. 3-06 Acetonitrile, 10 per cent. .. . I .. 3-19 Butylamine, 10 per cent. .. .. .. 2.94 Triethylamine, 10 per cent. .. .. a . 2-94 Nitrobenzene, 10 per cent... .. .. 2.88 Pyridine, 10 per cent. .. .. .. 2-97 * About 0-7 ml minute-’ actually reaches the flame. The emission from the air - hydrogen flame was quite dependent on the hydrogen flow- rate and this was always set just above the flash-back point when nebulising methanol. (Under these flame conditions the CH and CN emission intensities were optimised, while optimal C, emission occurred at a slightly higher hydrogen flow-rate.) The air pressure was set at 15 p.s.i.g. (4 1 minute-l) by using the conventional Unicam nebulising system.876 DAGNALL et d.: EMISSION SPECTRA OBTAINED FROM COMBUSTION [Afidyst, Vol. 94 The C, and CH emissions from methanol constituted about 5 to 60 per cent. of the total C, and CH emission observed from the 10 per cent.sample - 90 per cent. methanol solutions: Thus, not surprisingly, the C, : CH ratio of a given substance was not meaningful. The background emission of the methanol solution was low at the main CN, NH and NO peaks. The OH emission at 310 nm was fairly constant (+20 per cent.) for all solutions examined. Methanol gave a spectrum similar to that from the diffusion flame (Fig. 1); the only major difference was that there was less of a continuum beneath the C, and CH peaks. The CH emission at 431 nm was only about eight times more intense than in the diffusion flame. This low ratio was surprising because of the considerable temperature difference between the two flames, and underlines the chemiluminescent nature of the emission. Acetone in methanol gave a spectrum similar to methanol, but the C, emission was more pronounced, while with benzene in methanol only C, and CH emissions were observed. Indeed, all compounds examined which contained only carbon, hydrogen and oxygen gave similar spectra, exhibiting C,, CH and OH emissions.NITROGEN COMPOUNDS- The ammonia in methanol gave a spectrum similar to that observed in the diffusion flame (Fig. 2), but far more intense, with NH,, NH and CN emission and weak C, and CH emission. The spectrum from acetonitrile in methanol was also similar to that in the diffusion flame (Fig. 3). The CN emission intensity at 389 nm when nebulising undiluted acetonitrile was fifteen times more intense than that from the diffusion flame, while the CH emission at 431 nm was only about six times more intense.With butylamine in methanol, the major difference was the increase in the NH : CH intensity ratio. All nitrogen compounds gave CN, NH and NO emission, the only significant differences between the spectra were the actual CN:NH:NO emission intensity ratios, and these are shown in Table IV. TABLE IV CN : NH : NO EMISSION INTENSITY RATIOS OF NITROGEN-CONTAINING COMPOUNDS IN THE PRE-MIXED AIR - HYDROGEN FLAME Compound 0-88 Ammonia solution -methanol (1 + 1 v/v) Acetonitrile . . .. .. .. .. Butylamine . . .. .. .. .. Diaminoethane . . .. .. . . .. Pyidine .. .. .. . . .. Diethylamine . . .. .. .. .. Triethylamine . . .. .. .. .. Nitrobenzene . . .. .. .. .. Emission intensity* CN, NH, NO, 100 280 800 100 38 40 100 89 126 100 64 90 100 24 31 100 62 67 100 51 67 100 107 270 f A \ 337 nm 259 nm 360 nm * The CN and NH emission intensities were all measured with a slit width of 0.02 mm and converted into gain 2,lO.The NO emission was measured at slit width 0.08 mm and converted into gain 3,lO. The CN : NH : NO ratio was then calculated after arbitrarily setting the CN emission intensity to 100 units. It would appear that the CN: NH ratio increases in the order ammonia < primary < secondary < tertiary amines, while the CN : NO ratio increases in the order ammonia < RNO, < RNH, < R,NH - R,N < RCN < 0 N; i.e., the CN: NO intensity ratio is high for compounds containing multiple carbon -nitrogen bonds and low for compounds containing G<z- and NH, groupings. its concentration over the range 0.005 to 10 per cent.CHLORINE COMPOUNDS- The flame assumed an intense green colour on aspiration of CHCI, and CC1, and intense C, and weak CH emissions were observed. No CCI emission was observed as in the diffusion flame, Zoc. cit. The CN emission at 389 nm from acetonitrile in methanol was directly proportional toOctober, 19691 OF ORGANIC COMPOUNDS IN HYDROGEN FLAMES 877 SULPHUR COMPOUNDS- With carbon disulphide in methanol, the flame assumed an intense purple colour and a broad continuum from about 570 to 250nm, and a peak about 385nm, were observed (Fig. 6). Above the continuum, some weak S, emission was present. Weak CS emission at 257nm and extremely weak C, and CH emissions were also observed. The thiophen spectrum showed only C, and CH emission, the C, emission was intense compared with the CH emission.This could be caused by incomplete breakdown of the thiophen in the primary reaction cone of the air - hydrogen flame. No S, or CS emission was observed. Wavelength, nrn Fig. 6. The spectrum of 10 per cent. carbon disulphide - 90 per cent. methanol v/v in the air - hydrogen flame : A, slit, 0-012 mm; gain, 3,4; band width, 2. B, slit, 0.02 mm; gain, 3,lO; band width, 2 PHOSPHORUS COMPOUNDS- A strong continuum with a peak about 530 nm was observed between 600 to 400 nm (Fig, 7) on aspirating tributyl phosphate solution. This was attributed to the HPO species.l~J4~~ Weak C, and OH emissions were observed above this continuum. In addition, strong PO emissions were observed in the ultraviolet, with main bands at 240, 246, 247, 248, 254 and 256 nm.The emission intensity at the most intense PO band at 246 nm was Wavelength, nm Fig. 7. The spectrum of 10 per cent. tributyl phosphate - 90 per cent. methanol v/v in the air - hydrogen flame: slit, 0.04 mm; gain, 3,lO; band width, 2 proportional to the tributyl phosphate concentration in methanol over the range 0.05 to 10 per cent. No CP emissionf2 was observed. When phosphoric acid (0.01 per cent. in methanol) was aspirated, strong HPO emission was observed in the visible and PO in the ultraviolet. CONCLUSIONS It is important to stress that the emissions from both flames are chemiluminescent in origin and are, therefore, markedly dependent on experimental variables in their intensities.878 DAGNALL, SMITH, THOMPSON AND WEST The intensities from the diffusion flame were, in general, an order of magnitude lower than those obtained with the pre-mixed air - hydrogen flame.It is possible, however, that too much sample was reaching the flame and quenching the emission, because on nebulising methanol the inner temperature of the flame decreased.8 Although the pre-mixed air - hydrogen flame generally gives higher sensitivities, it possesses a higher background and there are certain instances when spectra, e.g., S,, can be obtained in the diffusion flame, which can scarcely be observed in the pre-mixed air - hydrogen flame. The emission intensities from this flame should be greatly increased if undiluted compound were nebulised instead of 10 per cent. methanolic solutions. It should be possible to monitor the output of a gas-chromatographic column by using this type of flame without, however, encountering the nebulisation problem experienced in this study. These experiments show that many organic compounds can be typified by spraying them into one of these two flames and observing the simple band spectra thus obtained. When compounds of similar composition have been separated by gas chromatography, identification and quantitative measurement should be possible. Quantitative relationships were in fact observed for acetonitrile in methanol and tributyl phosphate in methanol wih the air - hydrogen flame. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Gaydon, A. G., “The Spectroscopy of Flames,” Chapman ,yd Hall, London, 1967. Grant, D. W., in Desty, H., Editor, “Gas Chromatography, Butterworths. London, 1968, p. 163. Braman, R. S., Analyt. Chem., 1966, 38, 734. Juvet, R. S., and Durbin, R. P., Ibid., 1966, 38, 666. McCormack, A. J., M.S. Thesis, Cornell University, 1963. McCrea, P. F., and Light, T. S., Analyt. Chem., 1967, 39, 1731. Robinson, J. W., and Smith, V., Analytica Chim. Acta, 1966, 36. 489. Dagnall, R. M., Thqmpson, K. C., and West, T. S., Analyst, 1967, 92, 606. , Ibid., 1968, 93, 618. --- , Ibid., 1969, 94, 643. Gaydon, A. G., and Pearse, R. W. B., “The Identification of Molecular Spectra,’’ Chapman and Mackison, R., Analyst, 1964, 89, 746. Brite, D. W., Analyt. Chem., 1966, 27, 1816. Davis, A., Dinan, F. J., Lobbett, E. J., Chazin, J. D., and Tufts, L. E., Ibid., 1964, 36, 1066. , I , Ibzd., 1968, 93, 72. --- I --- Hall, London, 1965. Received December 20th, 1968 Accepted May 2nd, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699400871

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Aaalyst, October, 1969, Vol. 94, $9. 879-883 879 A Method for the Determination of Lead in Blood by Atomic-absorption Spectrophotometry BY P. P. DONOVAN AND D. T. FEELEY (Public Analyst's Laboratory, Galway Regional Hospital, Galway, Ireland) A method for the determination of trace amounts of lead in blood is described. The organic material in the blood is oxidised by dry ashing at 500" C and a solution of the ash in dilute hydrochloric acid prepared. The lead in the ash solution is determined by atomic absorption at 217 nm in an air - propane flame after isolation by a double-extraction procedure with dithizone and ammonium tetramethylenedithiocarbamate as complexing agents. Recovery tests are carried out and the lead content of the blood of workers from a lead mine, determined by this method, is compared with results obtained by using the mixed colour method of the Analytical Methods Committee.Comparative tests with wet-oxidation and dry-ashing techniques are made on samples of blood to which known amounts of lead have been added. The interference caused by bismuth is also investigated. THE determination of lead in blood is an important factor in the assessment of the toxicological hazard to workers engaged in lead mining.1 Workers exposed to lead dusts or fumes usually absorb some lead and generally have raised blood-lead levels.2 KehoeS has stated that an individual with a blood-lead content in excess of 80 pg per 100 g of blood is in danger of developing lead intoxication. Accurate methods of analysis are, therefore, required for the indication of abnormal lead absorption.The lead content of the blood of workers in lead mines in Western Ireland is constantly monitored1 and more than three thousand determina- tions have been carried out in this laboratory over the past 2 years. As a result of these tests, workers with excessively high lead contents in their blood are placed under clinical observation and hospital treatment provided if symptoms of lead poisoning develop. Published methods for the determination of lead in blood generally involve destruction of the organic material followed by detennination of lead as dithiz~nate.~s~s~ Polarographic,' spectrographi~8~~ and atomic-absorption spectrophotometriclOpll methods have also been used for determination of lead in blood and other biological materials.In this connection the reports and recommendations of H~ffmanl~v~~ with regard to the need for further study on methods for the determination of lead in food by atomic-absorption spectrophotometry are noted. For the determination of lead in blood, digestion with concentrated acids is usually preferred to dry ashing4~5~6 but, as dry ashing is more economical in time and reagents, several tests were carried out to compare the results obtained with both methods. The investigation showed that although wet oxidation is more accurate, the dry-ashing method is quite accept- able for routine work involving many samples. In this respect it is noted that in a collabora- tive study14 on the determination of trace metals in animal feeds by atomic-absorption spectrophotometry comparable results were obtained with both wet-digestion and dry-ashing methods for the destruction of the organic material.Several methods for the determination of lead in the blood ash solutions by atomic absorption were tried, including direct aspiration of a solution of the ash in dilute hydrochloric acid, single extraction of lead dithizonate into isobutyl methyl ketone, and also single extrac- tion of the complex formed between lead and ammonium tetramethylenedithiocarbamate into isobutyl methyl ketone. However, more consistent recovery results were obtained with both dithizone and ammonium tetramethylenedithiocarbamate in a double-extraction procedure. In this method, the lead is extracted into dithizone in chloroform from an ammoniacal solution containing citrate and cyanide, returned to nitric acid and finally extracted into isobutyl methyl ketone as its complex with ammonium tetramethylenedithiocarbamate. This method gives satisfactory agreement with the mixed colour method of the Analytical Methods Committee.ls 0 SAC and the authors.880 DONOVAN AND FEELEY: A METHOD FOR THE DETERMINATION [Analyst, Vol.94 Disposable 5-ml plastic syringes are used to take the samples of blood from the mine workers, and the method described in this paper is designed for the determination of lead in about 5 ml of blood. The detection limit with this method is 2 pg per 100 g of blood (0.02 p.p.m.). APPARATUS- Atomic-absorption spectrophotometer-A Perkin-Elmer, model 303, atomic-absorption spectrophotometer equipped with a hollow-cathode lead lamp and a three-slot Boling burner is used.The operating parameters are: compressed air flow-rate 6 ml minute-l; wavelength 217 nm; slit position 4; lamp current 30 mA; solution uptake 4.2 ml minute-l; scale expan- sion x 10; and meter response 2. The propane flow-rate is adjusted during aspiration of water-saturated isobutyl methyl ketone to give a blue flame in which “three tongues of flame” are visible. M u , furnace-The furnace should be capable of operating at 500” C. Silica basins-These should be lead-free with a capacity of about 50 ml, and have a lip. Separating funnels-Capacity, 50 ml. Graduated flasks-Capacity, 50 ml. All reagents must be lead-free, and water must be distilled or de-ionised and lead-free. Hydrochloric acid, sp.gr.1.18. Dilute hydrochloric acid-Dilute 10 volumes of hydrochloric acid (sp.g-r. 1-18) to Nitric acid, sp.gr. 1-42. Dilute nitric acid-Dilute 1 volume of nitric acid (sp.gr. 1-42) to 100 volumes with water. Ammonia solution-Dilute 1 volume of ammonia solution (sp.gr. 04380) with 1 volume Chloro form-Anal yt ical-reagent grade. Potassium cyanide solution-A 10 per cent. w/v solution in water. Dithizone (dipheny1thiocarbazone)-Analytical-reagent grade. Concentrated dithizone solution-A 0.3 per cent. w/v solution in chloroform. Filter and Dilute dithizone solution-Dilute 1 volume of concentrated dithizone solution to Concentrated lead standard solution-Dissolve 1.60 g of lead nitrate in water, add 10 ml Dilute lead standard solution-Dilute 1 volume of concentrated lead standard solution EXPERIMENTAL REAGENTS- 100 volumes with water.of water. store in a refrigerator. 100 volumes with chloroform. of nitric acid (sp.gr. 1-42) and dilute to 1 litre with water. to 1000 volumes with dilute nitric acid solution. 1 ml of solution = 1 pg of lead. Prepare freshly as required. Thymol blue indicator solution-Triturate 0.4 g of thymol blue with 8-6 ml of 0-1 N sodium Acetic acid - sodium acetate solution-Dissolve 15 g of anhydrous sodium acetate in water, Ammonium tetramethyhnedithiocarbamate solution-A 1 per cent. w/v solution in water. Isobutyl methyl ketone-Water-saturated isobutyl methyl ketone. The following additional apparatus and reagents are required for the wet-oxidation Digestion flasks-Kjeldahl transparent Vitreosil flasks.Sul~huric acid, sp.gr. 1-84. Perchloric acid, s9.gr. 1.54. hydroxide and dilute to 1 litre with 95 per cent. ethanol. add 21 ml of glacial acetic acid and dilute to 1 litre with water. method. SAMPLE PREPARATION REAGENT BLANK- Carry out a blank test on all the reagents omitting only the sample. DESTRUCTION OF ORGANIC MATTER- Weigh the sample of blood into a silica basin, evaporate to dryness on a boiling water bath, heat over a low flame to volatilise as much of the organic material as possible and thenOctober, 19691 OF LEAD I N BLOOD BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 881 in a muffle furnace at 500” C for 4 hours. (This is normally sufficient for complete ashing, but if any charred organic material remains it can be removed by the addition of a slight excess of dilute hydrochloric acid solution, then evaporating to dryness on a boiling water bath and again ashing at 500” C for 30 minutes.) Remove the dish from the furnace, allow it to cool, place on a boiling water bath, add 2ml of hydrochloric acid (sp.gr.1-18>, heat for 3 minutes, add 20 ml of water, heat for a further 5 minutes, transfer it to a 50-ml graduated flask, wash the silica dish with distilled water, add the washings to the flask and dilute to 50ml with distilled water. SEPARATION OF LEAD- Transfer 25 ml of the sample obtained by the method described above under Sample preparation to a separating funnel, add 3 drops of t h p o l blue indicator solution, 3 ml of ammonium citrate solution and sufficient ammonia solution to give a green - blue colour, indicating pH 9.0 to 9.5.Add 1 ml of potassium cyanide solution and 5 ml of dilute dithizone solution; shake the funnel for 1 minute and allow to separate. Run the chloroform layer into a second separating funnel and reject the aqueous layer. Add 10ml of dilute nitric acid solution to the chloroform extract, shake for 1 minute, allow it to separate and reject the chloroform layer. To the nitric acid layer, add 9 ml of acetic acid - sodium acetate solution. Adjust the pH to the range 2.2 to 2.8 by the addition of dilute nitric acid solution or acetic acid - sodium acetate solution, by using pH paper (see Note 1). Add 1 ml of am- monium tetramethylenedithiocarbamate solution and 5 ml of isobutyl methyl ketone solution ; shake the mixture for 2 minutes, allow it to separate and reject the aqueous layer.Filter the organic layer through a 7-cm Whatman No. 541 filter-paper into a test-tube. Set the zero on the atomic-absorption spectrophotometer with the blank solution and measure the percentage absorption of the sample solution. Convert the percentage absorption into absorb- ance and read the number of micrograms of lead from a calibration graph. PREPARATION OF CALIBRATION GRAPH- Measure 0, 0.5, 1.0, 1.5, 2.0 and 3.0 ml of dilute lead standard solution into separating funnels, dilute to 10ml with dilute nitric acid solution and proceed as described under Separation of lead, beginning at the words “add 9 ml of acetic acid - sodium acetate solu- tion. . . .” Plot a graph relating the number of micrograms of lead to absorbance.The absorption graph is linear. A separate calibration graph is prepared for each batch of samples and standard solutions are introduced at regular intervals during the aspiration of the test samples to ensure that the instrument is operating with a constant sensitivity. NOTE 1- dilute nitric acid solution gives a pH of about 2.6. after preparation. pH adjustment is not usually necessary because the addition of 9 ml of acetic acid to 10 ml of The pH of this mixed solution should be checked RESULTS Several samples of blood of low lead content, each weighing 5 g , were ashed in the manner described. The ash was dissolved in hydrochloric acid and known amounts of lead added to each ash solution. The lead content of each solution was determined separately on four occasions.The results obtained are shown in Table I. TABLE I RECOVERY OF LEAD ADDED TO BLOOD ASH SOLUTIONS Lead added, pg per 100 g of blood 10 20 30 40 60 80 100 Lead found after subtraction of blank, pg per 100 g of blood r 1 11 8 10 9 20 17 19 21 29 28 30 29 37 40 41 38 58 60 59 59 81 80 79 81 102 99 100 98 The lead content of the blank sample of blood was 7 pg per 100 g.882 [Analyst, Vol. 94 INTERFERENCE CAUSED BY BISMUTH- As bismuth may cause some interference in the determination of lead by the mixed colour m e t h ~ d l ~ y ~ ~ a series of tests was carried out to investigate the effect of bismuth on the determination of lead by the atomic-absorption method. Known amounts of lead and bismuth were, therefore, added to blood ash solutions and the lead content determined by the method described.Each sample was analysed in duplicate and the results obtained are shown in Table 11. DONOVAN AND FEELEY: A METHOD FOR THE DETERMINATION TABLE I1 RECOVERY OF LEAD ADDED TO BLOOD ASH SOLUTIONS CONTAINING BISMUTH Lead found after Lead added, Bismuth added, subtraction of blank, pg per 100 g of blood pg per 100 g of blood pg per 100 g of blood (1) (2) 10 0 8 10 10 10 9 10 10 25 11 9 10 50 10 12 10 100 9 8 20 0 21 18 20 10 19 20 20 25 20 17 20 50 18 21 20 100 19 21 40 0 39 41 40 10 40 41 40 25 37 39 40 50 38 41 40 100 39 42 80 0 78 81 80 10 79 80 80 26 77 78 80 50 78 82 80 100 76 79 The results in Table I1 indicate that the presence of bismuth does not interfere with the determination of lead. No difference in absorption could be detected between the samples containing lead and bismuth and those containing lead alone.COMPARISON OF ATOMIC-ABSORPTION AND THE ANALYTICAL METHODS COMMITTEE’S MIXED The lead content of the blood of 150 workers in lead mines in western Ireland was deter- mined by using both the atomic-absorption spectrophotometric method and the mixed colour method of the Analytical Methods Cornmitteels on aliquot samples. Some of the results obtained are shown in Table 111. The lead content of the blank sample of blood was 9 pg per 100 g. COLOUR METHODS- TABLE I11 COMPARISON OF RESULTS OBTAINED BY ATOMIC-ABSORPTION AND MIXED COLOUR METHODS Worker No. 1 2 3 4 5 6 7 8 9 10 11 12 13 Atomic-absorption method, pg per 100 g of blood 9 9 14 17 23 24 32 36 40 62 64 67 64 Mixed colour method, pg per 100 g of blood 7 10 16 17 22 25 30 36 42 49 51 66 66October, 19691 OF LEAD IN BLOOD BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 883 The figures in Table I11 are representative of the results obtained and show that com- parable agreement is obtained between both methods.COMPARISON OF RESULTS OBTAINED BY WET OXIDATION AND DRY ASHING- Known amounts of lead were added to several duplicate blood samples, aliquot portions of which were dry ashed and wet oxidised. Sulphuric, nitric and perchloric acids were used in the wet oxidation and any precipitate that formed after the oxidation was dissolved by boiling with 10 ml of dilute hydrochloric acid. The lead content of the duplicate samples was determined and the results obtained are shown in Table IV.TABLE IV bMPARISON OF RESULTS OBTAINED BY WET-OXIDATION AND DRY-ASHING METHODS Lead added, pg per 100 g of blood 10 20 40 60 80 100 Lead recovered after subtraction of blank, pg per 100 g of blood Wet-oxidation method Dry-ashing method A r 1 9 9 19 18 38 38 58 57 78 76 96 95 The lead content of the blank sample of blood was 10 pg per 100 g. The results in Table IV show that the dry-ashing procedure gives recoveries that, on average, are slightly lower than those obtained with the wet-oxidation process. The recovery results compare favourably with those obtained by Gorsuch17 and the slight loss of lead is probably caused by co-precipitation and solubility factors in the wet-oxidation method, and volatilisation and retention on the silica crucibles in the dry-ashing procedure.The repro- ducibility factor with such minute amounts of lead must also be taken into consideration. CONCLUSIONS A method for the determination of trace amounts of lead in blood has been described. This method is shown to give consistent results and to be suitable for routine control work when the possibility of occupational hazards arise with regard to lead contamination. The presence of bismuth in blood does not cause interference in this method of analysis, We wish to express our appreciation for the co-operation received from the mining authorities concerned, who were most anxious to facilitate the work with technical and financial assistance. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. Donovan, P. P., Feeley, D. T., Canavan, P. P., Chem. & Ind., 1968, 51, 1802. Gibson, S. M., MacKenzie, J. C., Goldberg, A., Brit. J. Ind. Med., 1068, 25, 40. Kehoe, R. A,, J. R. Inst. Publ. Hlth Hyg., 1961, 24, 188. Bermen, E., Amer. J. Clin. Path., 1961, 36, 549. Vinter, P., J. Med. Lab. Technol., 1964, 21, 281. Monerieff, A. A., A r c h . Dis. Childh., 1964, 39, 1. Mosheva, N., Pracovni Lkk., 1966, 18, 69. Nifontova, M. V., and Ternovskaya, L. N., Lab. Delo., 1961, 7, 13. Netelson, S., J. Microchem., 1963, 7 , 448. Willis, J. B., Analyt. Chem., 1962, 34, 614. Pierce, J. O., and Cholak, J., Arch. Envir. Hlth, 1966, 13, 208. Hoffman, I., J. Ass. Ofl. Agric. Chem., 1967, 50, 917. Heckman, M., Ibid., 1968, 51, 776. Analytical Methods Committee, Analyst, 1959, 84, 127. Browett, E. V., and Moss, R., Ibid., 1965, 90, 715. Gorsuch, T. T., Ibid., 1959, 84, 135. -, Ibid., 1968, 51, 317. Received March 7th, 1969 Accepted April 26th, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699400879

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

884 Artalyst, October, 1969, Vol. 94, flfl. 884-885 Atomic-absorption Determination of Strontium in a Standard Plant Material : Comment on Results of Inter-laboratory Comparison BY D. J. DAVID (Division of Plant Industry, CSIRO, Canberra, A .C. T., Australia) It is claimed that the concentration of strontium in Bowen's standard kale (Analyst, 1967, 92, 124) is 101-13 f 2.38 p.p.m., on evidence that X-ray fluorescence values obtained subsequently to Bowen's inter-laboratory com- parison were in good agreement at 101 p.p.m. with atomic-absorption values obtained in this laboratory by using anion exchange and the addition method, but not reported separately from other atomic-absorption results by Bowen. Methods used in obtaining the neutron-activation results and these other atomic-absorption results implied in the inter-laboratory comparison should be examined for serious interferences.BECAUSE of the value of 101 p.p.m. for strontium obtained by Champion and Whitteml in their X-ray fluorescence analysis of Bowen's standard kale, it has become necessary to reveal the original atomic-absorption values contributed by this laboratory to Bowen for inclusion in his inter-laboratory comparison of results.2 These are shown in Table I. TABLE I ATOMIC-ABSORPTION VALUES FOR STRONTIUM IN STANDARD KALE DERIVED FROM SEPARATE PORTIONS OF THE POWDERED MATERIAL AND CORRECTED TO OVEN-DRY BASIS (80" TO 90" C FOR 18 HOURS) By addition method, p.p.m.- Mean 103-4 104.6 100.4 101.4 102.4 103.6 97.2 97.7 101.0 99.9 Against separate standards, p.p.m.- 106-6 105.6 106.2 107.0 105.8 METHOD For the first eight values in Table I, strontium was determined on 2-g portions of the powdered material received from Dr.Bowen, by ashing in a muffle furnace, dissolving the ash in hydrochloric acid, De-Acidite FF (acetate form) anion-exchange treatment of this solution and atomic-absorption analysis of the column effluent as described by David.3 The anion-exchange step was effected by using the same column and procedure as that used by David.4 The last four values in Table I were obtained from a comparison of atomic-absorption readings carried out on the effluents from the anion-exchange columns with those carried out on separate strontium standards in 0.1 N acetic acid. Absorption was measured at 460.7 nm and an air - acetylene flame was used (10-cm single slot, pre-mix burner).A moisture value of 5-61 per cent. (mean of twelve determinations) was used to convert the strontium values into the oven-dry basis shown in Table I. DISCUSSION Bowen's inter-laboratory study2 gives a grand mean for strontium of 84.1 10.7 p.p.m. (twenty determinations), a neutron-activation mean of 74-7 4-2 p.p.m. (two laboratories, six determinations) and an atomic-absorption mean of 88.1 & 10.2 p.p.m. (two laboratories, fourteen determinations). The atomic-absorption mean of 88.1 p.p.m. included all eight of the addition method results shown in Table I, but not the results arising from the use of 0 SAC and the author.DAVID 885 separate standards (personal communication from H. J. M. Bowen). A high result from the use of separate standards containing no addition of an easily ionised element is to be expected on grounds of repression of ionisation of strontium by large excesses of potassium and calcium in the sample solutions.These, together with the eight addition method replicates in Table I, give a value for strontium in the standard kale of 101.13 In addition to the results reported here, six other apparently reliable X-ray and flame-emission values in the range 100 to 106 p.p.m. of strontium in the standard kale have been reported to the author in a personal communi- cation from R. N. Whittem. The atomic-absorption values included in this mean of 101.13 p.p.m. are unlikely to be in error by more than a few parts per million, because severely depressing anions were replaced with acetate and any slight remaining enhancements or depressions compensated for by the use of the addition method. It is suggested, then, that the methods used to produce the other atomic-absorption values and the neutron-activation values for strontium reported in Bowen’s inter-laboratory study2 should be examined for depressive interferences amounting to 20 to 4-0 per cent. In the case of the atomic-absorption method it is likely that either no interference-suppressing agent, or an inadequate one, was added to the solutions. Champion and Whittem’s replicates for strontium were 101 and 101 p.p.m. 2.38 p.p.m. REFERENCES 1. 2. 3. 4. Champion, K. P., and Whittem, R. N., Analyst, 1968, 93, 550. Bowen, H. J. M., Ibid., 1967, 92, 124. David, D. J., Ibid., 1962, 87, 576. - , Ibid., 1964, 89, 747. Received March 12th, 1969 Accepted April 21st, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699400884

出版商:RSC    

年代:1969

数据来源: RSC