ISSN:0003-2654    

DOI:10.1039/AN97398FX021

出版商:RSC    

年代:1973

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97398BX023

出版商:RSC    

年代:1973

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97398FP061

出版商:RSC    

年代:1973

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97398BP067

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

JUNE, 1973 THE ANALYST Vol. 98, No. I 167 Pyridylazonaphthols (PANS) and Pyridylazophenols (PAPS) as Analytical Reagents Part I.* Synthesis and Spectroscopic Examination of Reagents and Some Chelates BY D. BETTERIDGE AND D. JOHN? (Chemistry Department, University College of Swansea, Swansea, Glarnorgan, SA 2 8PP) 2-(2-Pyridylazo)- l-naphthol (o-a-PAN), 2- (2-pyridylazo)phenol (o-PAP) and 4-(2-pyridylazo)phenol ($-PAP) have been prepared. They and 1-(2- pyridylazo)-2-naphthol (O-p-PAN) and their chelates with various transition metals have been examined and characterised by infrared spectroscopy and mass and nuclear magnetic resonance spectrometry. The purity and structures of the reagents have been established and it was confirmed that they are terdentate ligands. A REVIEW by Anderson and Nicklessl has shown that analysts have not been slow to realise the importance of 1-(2-pyridylaz0)-2-naphthol (o-P-PAN).Since Cheng and Bray’s original work,2 there have been many developments with respect to both the use of o-P-PAN and the development of related reagents. These reagents react with many cations to form intensely coloured complexes that are eminently suitable for spectrophotometric determinations, chelatometric end-point detection and solvent extraction. The sensitivities are comparable with those obtained with dithizone (diphenylthiocarbazone) , and the versatility is comparable with that of 8-hydroxyquinoline. The great advantage of o-P-PAN is that its solutions and also solutions of its complexes are unusually stable for such a sensitive reagent.Reagent solutions can be kept for several months without change in the absorbance (Galik,3 and D. Betteridge, unpublished work), which should make o-P-PAN well suited for use in auto- mated determinations. The disadvantage of using PAN-type compounds at present is that their solution chemistry is very complex. Slight alterations of conditions, when modifying a published procedure, can result in irreproducible results or even in no results being obtained. For example, Galik3 has shown that copper is not extracted from a sulphate medium at pH 4 although it is extracted with other common anions from the bulk of the electrolyte. It has also been found that care has to be taken when extracting manganese, although a useful procedure for determining manganese in high-purity zirconium, etc., has been pub- l i ~ h e d .~ We shall show later that this care is necessary because of the formation of a hydroxy species, of the form MnR,OH, which is very dependent upon pH. These difficulties can be easily overcome if the basic chemistry of the system being used is understood. Accordingly, we have undertaken the study of four related compounds and their reactions with metal ions. Two prime considerations have been borne in mind: the chemistry of the system considered should be examined in considerable detail, and the information obtained and the methods used to obtain it should be of value to analysts. The reagents examined are insoluble in water and they form water-insoluble complexes that can often be extracted.Hence the work of Sommer and co-workerss-9 on water-soluble complexes of 4- (2-pyridylazo) resorcinol (PAR) and 4-(2-thiazolylazo)resorcinol (TAR) is to some extent paralleled and the possibilities of solvent extraction are considered more fully. The reagents are 1- (2-pyridylazo)-2-naphthol (o-p-PAN), 2-(2-pyridylazo)-l-naphthol (o-a-PAN), 2-(2-pyridylazo)phenol (o-PAP) and 4- (2-pyridy1azo)phenol ($-PAP). This paper deals with the nature of the reagents and some chelates. Subsequent papers in this series will deal with methods for investigating the equilibria of complex formation and analytical procedures. * For Part I1 of this series, see p. 390. t Present address: BP Chemicals (U.K.) Ltd., Llandarcy, Swansea. @ SAC and the authors. 377378 BETTERIDGE AND JOHN : PYRIDYLAZONAPHTHOLS AND [Analyst, Vol.98 EXPERIMENTAL REAGENTS- PAN-type compounds have been synthesised by the Chichibabinlo reaction of coupling 2-pyridyldiazotate with an appropriate naphthol or phenol under an inert atmosphere. Low yields and ill defined products usually result. Pollard, Nickless and Anderson,ll Anderson and Nickless12 and Andersonl3 have shown that the coupling of 2-hydrazinopyridine, syn- thesised by the procedure of Fargher and Furness,l4 with a suitable quinone results in relatively pure products of definite composition being produced in over-all yields of 40 to 60 per cent. We confirm these findings with the reservation that 2-hydrazinopyridine and o-quinones are usually unstable so that some experimental skill is required in order to obtain the desired results. The final purification stages were very difficult, and the products were subjected to repeated recrystallisation and, when appropriate, sublimation until the microanalyses and mass spectrum indicated that the compound was pure or that further attempts at purification would be valueless.2- (2-PyridyZaxo)$kenoZ (o-PAP)-Aqueous solutions of 2-hydrazinopyridine and 1,2- benzoquinone acidified with perchloric acid were allowed to react together to give o-PAP. The 1,2-benzoquinone was prepared by reaction of catechol in anhydrous diethyl ether with tetra- chloro-o-quinone, which had been synthesised by chlorination of catechol. The 1,2-benzo- quinone is unstable and was used immediately. 2-Hydrazinopyridine was prepared by allow- ing hydrazine hydrate and 2-chloropyridine to react together.'* The solid first isolated was o-PAP perchlorate, m.p.170 "C. Successive recrystallisations of this perchlorate salt from aqueous ethanolic mixtures gave deep red, granular crystals that melted at 128 "C. The lower melting-point indicated that these granular crystals were no longer the salt, but pure o-PAP, which was verified by microanalysis, thin-layer chromatography and infrared spectro- scopy. The microanalytical results were as follows- C H N Calculated for C,,H,N,O.HClO,, per cent.. . . . 44.1 3.4 14-0 Found, per cent. .. .. .. .. . . 66.7 6.2 20.6 Calculated for C,,H,N,O, per cent. .. . . 66.0 4-6 21.0 4- (2-PyridyZazo)$kenoZ (p-PAP)-Commercially available 1,4-benzoquinone was purified by recrystallisation from light petroleum of boiling range 40 to 60 "C and allowed to react with a solution of 2-hydrazinopyridine in perchloric acid.The product was washed with water, dissolved in methanol - formic acid (10 + 6) and ammonia was added to re-precipitate the $-PAP. The microanalytical results were as follows- C H N Calculated for C,,H,N,O, per cent. .. . . 66.0 4.6 2 1.0 Found, per cent. .. .. .. .. . . 66.0 4.6 21.6 2-(2-PyridyZazo)-l-napht~oZ (o-a-PAN)-This reagent was prepared by the reaction of 1,2-naphthoquinone with 2-hydrazinopyridine under acidic conditions- 0 bH In order to obtain a pure sample of 1,2-naphthoquinone, it was synthesised by oxidising 1,2-aminonaphthol hydrochloride with iron( 111) chloride.16 The crude, commercially available 1,2-aminonaphthol hydrochloride was purified as described by Conant and Corson.16 Crude o-a-PAN was precipitated by neutralising the ice-cold acidic reaction mixture with ammonia, but repeated recrystallisations failed to yield a pure product.Purification was effected by sublimation at 122 "C. Mass-spectrometric analysis indicated that there was a slight im- purity (about 1 per cent.) of a compound similar to o-a-PAN but with a higher relative molecular mass (28 a.m.u.). The microanalytical results were as follows- C H N Calculated for C,,H,,N,O, per cent. .. . . 72-3 4.4 16.9 Found, per cent. .. .. .. .. . . 72.1 4.4 16.6June, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I 379 1- (2-PyridyZaxo) -2-naphthol (o-P-PA N)-Impure o-P-PAN was obtained commercially and purified by repeated recrystallisations from aqueous ethanol (m.p.140 "C). The micro- analytical results were as follows- C H N Calculated for C,I;HllNIO, per cent. .. . . 72.3 4.4 16.9 Found, per cent. . . .. .. .. . . 72.6 4-3 16.6 4-(2-PyridyZazo)-l-~aphthoZ (o-a-PAN)-This compound was prepared by reaction of 2-hydrazinopyridine with 1,4-naphthoquinone. It could not be purified by the methods outlined above as it formed tars on recrystallisaton and exploded rather than sublimed. No further work was carried out on it. MASS SPECTRA- focusing mass spectrometer. The mass spectra of reagents and chelates were obtained on an A.E.I. MS9 double- INFRARED SPECTRA- The infrared spectra of the solid reagents and chelates were obtained from caesium iodide pressed discs with a Perkin-Elmer 225 grating infrared spectrometer.The caesium iodide was spectroscopically pure. NUCLEAR MAGNETIC RESONANCE SPECTRA- Nuclear magnetic resonance spectra were obtained with a Varian 100 H.A. spectrometer. Spectra of reagents-Saturated solutions of o-a-PAN, o-P-PAN and o-PAP in deuterated chloroform were concentrated enough for spectra to be measured, but it was necessary to use dimethyl sulphoxide as a solvent for $-PAP. Spectra of chelates-Equal volumes (5 ml) of a solution of the reagent (60 mg) in carbon tetrachloride and a buffered aqueous solution of a metal ion (30 mg) were equilibrated. The organic phase was removed and dried over molecular sieves. Other experiments showed that under these conditions of excess of the metal ion at a suitable pH, all of the organic reagent in the organic phase would be converted into the chelate.RESULTS AND DISCUSSION A detailed study of the mass, infrared and nuclear magnetic resonance spectra of the reagents and of some chelates of common transition metals was carried out with several objects in mind- to confirm that no errors of identification had been made; to see if the marked difference between o-a-PAN and o-P-PAN could be explained; to check the generally held views on bonding in the chelates; to provide a reference for the chelates of less common metals to be examined later, which chelates may be anomalous. MASS SPECTRA- A detailed discussion of the mass spectra of the reagents and of their chelates with manganese(II), cobalt (11), nickel(II), copper(I1) and zinc(I1) has been presented elsewhere.17 The two steps in the main fragmentation pattern of the reagents are loss of azo-nitrogen with the fusion of the pyridine and phenol or naphthol groups, and rupture of the compound formed with the release of pyridine and phenol or naphthol fragments.Alternatively, the pyridine group was lost first and the azo-nitrogen second. The 1 : 2 chelates, typically, lost first a whole ligand molecule and then the 1 : 1 complex that remained fragmented to give a metal - pyridine complex and a diazophenol or naphthol group. It was found that o- and +-hydroxyl groups could be detected with certainty so that the identification of the reagents was confirmed. It was also found that the spectra of the chelates provided information that was of value in the interpretation of the results of solution studies.It was possible to check the stoicheiometry of the chelates but not to ascertain their structure.380 BETTERIDGE AND JOHN : PYRIDYLAZONAPHTHOLS AND [Analyst, v0198 INFRARED SPECTROSCOPY- of investigation of chelate compounds.18~19 INFRARED SPECTRA OF REAGENTS- The infrared spectra of the complex molecules are extremely complicated and the extensive overlap of bands makes detailed interpretation very difficult. The assignments for the spectrum of pyridine are well established,20-22 as are those for a-naphthol, p-naphthol and phen01.~3-~6 Examination of the spectra of these individual compounds shows that there are several absorption bands in similar regions. This effect can be anticipated owing to the expected similarity of the ring vibrations and the C-H stretching and bending frequencies of the compounds.It is this type of similarity that leads to the extensive overlapping of bands in the reagent spectra and to the corresponding complexity. Spectrum of 2-(2-pyridyZazo)-l-naphthoZ (o-a-PAN)-The principal bands and their assignments are shown in Table I. In contrast to mass spectrometry, infrared spectroscopy is well established as a means TABLE I INFRARED SPECTRUM OF o-a-PAN Peaklcm-l Intensity* Assignment 3050 m Aromatic C-H stretch 1620 to 1400 s (multiplet) N=N, C=N, C=C stretch 1400 to 110 s (multiplet) C-0 stretch, O-H deformation, pyridine C-H deformation 900 to 400 s (multiplet) C-H in-plane deformations 400 to 200 m (multiplet) C-H out-of-plane deformations * m, bands of medium intensity; and s, bands of strong intensity.990 S Pyridine C-H deformation The spectrum consists principally of aromatic C-H and C=C absorptions, together with CO, OH and CN absorptions. Many of the bands that arise from these absorptions will inevitably overlap and give rise to the high degree of complexity observed. The well defined band at 3050 cm-l is characteristic of aromatic C-H stretching fre- quencies [Fig. l ( a ) ] . Coupled with this band are the expected aromatic C=C ring vibration frequencies a t about 1590, 1500 and 1450 cm-1. The shoulder at 1580 cm-l could be assigned to the presence of conjugated rings, and is often taken as an indication of such a system. The spectral range from 1620 to 1400 cm-l is further complicated by the presence of the C=C and C=N stretching frequencies of the pyridine ring.According to Bassignana and Cogrossi,26 Fig. 1. Infrared spectra: (a) aromatic C-H stretching for o-a-PAN; ((I) 1600 to 1400cm-l region for o-or-PAN; and (c) 1600 to 1400cm-1 region for p-PAPJune, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I 381 the N=N stretching frequency could also be observed in the region of 1410 & 30 cm-1. Because of the complexity of this region, positive identification of the band, which is usually of medium intensity, is extremely difficult. Another interesting feature of this region of the spectrum is the occurrence of a well defined shoulder peak at 1610 cm-1, which is also observe4 in the spectrum of o-@-PAN [Fig.l(b)]. In view of the known tautomerism of o-hydroxyazo compounds, it is possible to assign this band to the presence of the C=O stretching frequency of the hydrazo form, it?.---- A20 Although the free carbonyl band is to be H y drazo expected iii the 1700 cm-1 region, such a shift is often observed in hydrogen-bonded systems. - Examination of the spectra of o-PAP and @PAP shows that there is no absorption of any kind at this wavelength [Fig. l(c)]. This result is in keeping with Hadzi’s observation that o-hydroxyazophenols are true azo compounds and that naphthols exist as taut~mers.~’ The strong bands observed in the region of 1100 to 1300 crn-l are probably due to the C-0 stretching and O-H deforination frequencies. Complications in this region are the expected ring vibrations and C-H deformations of the pyridine ring at about 1200 cm-1.At lower frequencies, in the range 1000 to 500 cm-l, are a large number of strong absorptions that can be attributed to the C-H deformations of both the pyridine and naphthol rings. The presence of two strong bands at 990 and 705 cm-l is indicative of the presence of the pyridine ring. Most of the other bands overlap to such an extent that assignment would be extremely speculative. The remaining region of the spectrum below 500 cm-1 consists mainly of the C-H out-of-plane deformations. A notable absence from the spectrum is a band in the region of 3500 cm-1 corresponding to an O-H stretching frequency. It is known that in the presence of strong intramolecular hydrogen bonding this band would become broad and very weak.The absence of this band would therefore suggest that a-PAN has a strong intramolecular hydrogen bond and this is also suggested by the nuclear magnetic resonance spectrum (see below). Spectrum of 1-(2-pyridyZaxo)-2-naphthol! (o-P-PAN)-As expected, the spectrum of o-p-PAN is very similar in nature and complexity to that of o-a-PAN. The identification of individual bands is again difficult because of the high degree of overlap, but the aromatic stretching frequency at 3060 cm-l is well defined. The presence of extensive intramolecular hydrogen bonding is again evident, due to the absence of a band in the 3500 cm-l region, corresponding to the hydroxyl stretching mode. Because of the similarity of the spectra of o-a-PAN and o-P-PAN, a table of frequencies and their assignments for o-@-PAN can be considered to be identical with that for o-a-PAN (Table I).Spectrum of 2- (2-pyridyZazo)phenoZ (o-PA P)-The principal absorption bands and their assignments are listed in Table 11. TABLE I1 INFRARED SPECTRUM OF o-PAP Peak/cm-l Intensity* Assignment 3020 m Aromatic G H stretch 1600 to 1400 s (multiplet) C==C, G N , N=N stretch 1300 to 110 s (multiplet) G O stretch, O-H deformation, pyridine C-H deformation 900 to 500 s (multiplet) C-H in-plane deformations 460 to 200 s (multiplet) C-H out-of-plane deformations 980 S Pyridine C-H deformation * rn, bands of medium intensity; and s, bands of strong intensity.382 BETTERIDGE AND JOHN PYRIDYLAZONAPHTHOLS AND [Analyst, VOl. 98 Many of the absorption bands that arise from the phenol ring will be similar to those of the a- and ,&naphthol rings, so that this spectrum is similar to those of o-a-PAN and- o-p-PAN.The aromatic C-H stretching frequency band at 3020 cm-l is well defined. The aromatic C=C ring vibrations are observed in the region 1600 to 1400 cm-1, which region is again complicated owing to the presence of C=N and N=N bands. The remaining regions of the spectrum are also similar to those of o-a-PAN and o-p-PAN, and consist of bands that arise from C-0, 0-H and C-H stretching and deformation modes. The presence of strong intramolecular hydrogen bonding is indicated by the absence of absorption in the region of 3500 cm-l. In this respect, o-PAP is again similar to o-a-PAN and o-P-PAN, and interpretation is also supported by nuclear magnetic resonance studies.Spectrum of 4-(2-py&’yZaxo)phenoZ (0-PAP)-Below 2400 cm-l, this spectrum is virtually identical with that of o-PAP, but above this frequency there is a big difference. In the region of 3400 to 2400 cm-l there is an extremely broad, very strong absorption band (Fig. 2 ) . This band must be due, in part, to aromatic C-H stretching absorptions, but is so large that some other contributions must be involved. It seems likely that the broadening is caused by intermolecular hydrogen bonds formed by polymeric association of the reagent. This type of association is known to give rise to very broad, concentration-dependent bands in the 3-pm region. Such a change from intramolecular to intermolecular bonding can also be inferred from nuclear magnetic resonance spectra.cm-’ Fig. 2. Intermolecular hydrogen-bond absorption for p-PAP Attempts to investigate the presence of this intermolecular bonding more closely were made by obtaining reagent spectra for saturated solutions in chloroform. The only absorption observed was a well defined band a t 3020 cm-l that corresponded to aromatic C-H stretching. The broad band previously observed in the spectral region of 3400 to 2400 cm-1 was no longer observed. Because of the limited solubility of the reagent in this solvent, spectra were obtained for saturated solutions in acetone. These solutions were prepared by using spectroscopically pure acetone dried over molecular sieves. The spectra were very similar to those obtained for chloroform solutions, but an additional broad, very weak band in the region of 3600 cm-l was also observed.Although acetone was used as a blank solution, this band can be assigned to the hydroxyl stretching frequency only tentatively because of the ease with which acetone picks up moisture. The solutions in both chloroform and acetone were very dilute, and the disappearance of the broad absorption band is consistent with the presence of intermolecular bonds that would be broken on dilution. Although these results tend to confirm the presence of intermolecular bonding, conclusive proof would involve a complete concentration-dependent study. Unfortunately, the relative insolubility of the reagent in non-polar, non-hydroxylic solvents would seem to preclude this.June, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS.PART I 383 The spectra of the chelates of each reagent are similar and resemble each other more closely than the spectrum of the parent reagent. The general similarity of all the chelate spectra suggests that each metal has a similar effect on the vibrations of the ligand. For convenience, the spectra are divided into four regions, which are discussed separately. 4000 to 2800 cm-l region-A common feature in the spectra of all the chelates, and of each reagent, is the presence of a well defined band in the region of 3050 cm-l. That this band is stable in position in the spectra of both ligands and chelates indicates that it is due solely to aromatic C-H stretching. INFRARED SPECTRA OF THE CHELATE COMPOUNDS- 2800 to 1600 cm-1 region-This region is largely featureless in all the spectra studied.1600 to 900 cm-1 region-This skeletal region possesses a large number of broad, very strong bands in the spectra of both chelates and ligands. It is to be expected that several of the ligand absorption bands that occur in this region will undergo shifts on chelation. Identification of these shifts, normally to lower frequencies, is very difficult because of the large degree of overlapping of bands. A number of workers have f ~ ~ n d ~ ~ s ~ ~ that one of the more prominent shifts observed in this region has been that of the C-0 stretching frequency from about 1200 cm-l to about 1100 cm-l. Although in the present spectra the positive identification of this shift is difficult, a broadening effect near 1100 cm-l in each spectrum suggests that such a shift has occurred.One of the major features of this region is the appearance of a broad, very strong band at about 1340 cm-l in the spectrum of each chelate investigated (see Table 111). TABLE 111 POSITION OF CHELATE -N=N- FREQUENCY* (cm-1) Metal o-a-PAN O- p-PAN O-PAP Ni 1330 (b, s) 1335 (b, s) 1345 (fs, s) Zn 1355 (b, s) 1335 (b, s) 1360 (fs, s) Mn 1330 (b, s) 1330 (b, s) - * b, broad; s, strong; and fs, fairly sharp. I t has been suggested by U e n ~ , ~ ~ in his study of the chelate compounds of o-hydroxy- azobenzene, that this band is due to the shift to lower frequencies of the azo stretching band. 950 to 200 cm-1 regio.n-It is in this region, and in particular the medium to far infrared region, that the metal - oxygen and metal - nitrogen frequencies are likely to be found.These bands are usually well defined, but it is possible that they may be overlapped by other strong absorptions such as those resulting from the C-H deformations of the pyridine or enol ring. METAL - NITROGEN INFRARED FREQUENCIES- The general region expected for metal - nitrogen vibrations is the spectral region from 300 to 150 cm-1.31932 Bands found in this region of the spectra of chelates, and which are not present in the ligand spectra and cannot be assigned to any form of shift, are considered to be metal - nitrogen stretching vibrations. The frequencies are summarised in Table IV. TABLE IV METAL - NITROGEN FREQUENCIES* (cm-l) o-a-PAN O- 8-PAN O-PAP Ni Zn Mn 240 (sh, m) 244 (m) ; 219 (vs, sh) 243 (sh, m); 222 (m) 245 (sh, w) ; 225 (sh) 243 (m) ; 228 (sh, m) 244 (sh, s) ; 222 (sh, s) 247 (sh, w) ; 220 (vs, sh) 243 (s, sh) ; 228 (sh) 243 (sh, m) ; 223 (sh, w) * sh, sharp; w, weak; m, medium; s, strong; and vs, very strong.METAL - OXYGEN INFRARED FREQUENCIES- The highest frequencies are associated with double bonds, while those in the region of 400 to 600 cm-1 are associated with single bonds. The lowest frequencies, below 300 cm-l, arise either with heavy oxide ligands or with co-ordinated oxy-anions. In the present study, two distinct regions of metal - ligand absorptions were observed. Metal - oxygen stretching frequencies occur between 1000 and 250 cm-l.384 BETTERIDGE AND JOHN PYRIDYLAZONAPHTHOLS AND [Analyst, VOl.98 imposed with shiftedv pyridine C-H deformations. TABLE V METAL - OXYGEN FREQUENCIES* IN THE 600 O-E-PAN O- p-PAN Ni 622 (sh, s) 622 (w) Zn 615 (sh, s) 630 (sh, m) Mn 625 (sh, s) 626 (w) 600 to 650 cm-l regi~n-Lecomte~~ suggested that the absorptions may be due solely to metal - oxygen stretching vibrations. However, it has since been shown that this suggestion is an over-simplification, and that the metal - oxygen stretching is probably coupled with ring or C-H deformations (J. M. Williams and D. Betteridge, unpublished work). The bands observed in this range are shown in Table V, but for o-PAP chelates they may be super- TO 650 Cm-’ REGION O-PAP 638 (sh, s) 640 (s) 636 (sh, s) * sh, sharp; w, weak; m, medium; and s, strong. 400 to 500 cm-l wgion-The absorption bands observed in this region are generally accepted as being solely metal - oxygen stretching frequencies.They are listed in Table VI. TABLE VI METAL - OXYGEN FREQUENCIES* IN THE 400 TO 450 cm-1 REGION O-a-PAN O- p-PAN O-PAP Ni 430 (w) 442 (m) ; 429 (m) 485 (sh, m) Zn 423 (w) 437 (sh, s) 478 (sh, m) * sh, sharp; w, weak; in, medium; and s, strong. Mn 420 (w) 438 (sh, s) - Several attempts have been made to correlate metal - ligand frequencies with stability constants or electronegativities. Such a correlation has little value because too few chelates have been examined. NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY- Nuclear magnetic resonance spectroscopy has been applied to the study of chelate compounds on only a relatively few occasion^.^*-^^ The technique is handicapped by the difficulty of obtaining resolvable spectra for those chelates which contain a paramagnetic or ferromagnetic central metal atom.Mass spectrometry and infrared spectroscopy have been found to be more versatile, as these techniques permit the ready investigation of both reagents and chelates. The spectra are complex and those obtained on a 60-MHz instrument could not be resolved. NUCLEAR MAGNETIC RESONANCE SPECTRA OF LIGANDS- and their tau (7) values and assignments are shown in Table VII. 1 - (2-PyridyZaxo)-2-naphthoZ (o-p-PA N)-The proton signals obtained in the spectrum For convenience of discussion the protons are labelled as follows- HO H(3) The hydroxyl proton in o-/3-naphthol is observed at r 3.90. The corresponding peak with o-p-PAN is observed at T -5.70.(The positions of these hydroxyl protons were verified by deuterium oxide exchange.) This large shift can be attributed to strong hydrogen bonding in the reagent molecule. Generally, it is accepted that the greater the extent of hydrogen bonding, the greater is the “hydrogen-bond shift.”37 The hydrogen bonding was proved to be intramolecular, as no signal shift occurred on dilution of the solvent.June, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I TABLE VII NUCLEAR MAGNETIC RESONANCE SPECTRUM OF o-P-PAN 385 7 1.59 2-17 2.35, 2-45 2.68 2-93 3.30, 3-40 - 6-70 Peak Integration Complex multiplet 2 Complex multiplet 2 Multiplet 3 Multiplet 1 Doublet 1 Broad multiplet 1 Doublet 1 Assignment f# c (lower is f, upper is c) e, 8 4 5, 6, 7 d 3 OH Apart from the hydroxyl proton signal, all other proton signals are observed in the range 1.0 to 4.0 (Fig.3). The resulting complex spectrum is expected, as signals for aromatic and heteroaromatic protons are usually observed in this range. In order to assign the signals in this range successfully, a scale-expanded spectrum was provided. The interpretation is summarised as follows. 1 .o L l 2.0 1 0 r Fig. 3. Nuclear magnetic resonance spectrum of 0- 8-PAN The maximum effect of the electron-donating properties of the naphthol hydroxyl group is observed for proton 3, which is in an ortho position. The position of proton 3 does not permit any long-range coupling, and as there is only one adjacent proton then the expected signal would be a distinct upfield doublet. Such a doublet is observed at T 3.30 to 3.40, which also integrates for only one proton.The ortho coupling constant for protons 3 and 4 was found to be 10 Hz. In order to locate proton 4, it was therefore required to find a doublet signal with a splitting of 10 Hz. Ideally, this doublet should be a mirror image of doublet 3, but because of the position of proton 4 some long-range coupling from proton 5 would be expected. Such a doublet is observed at r 2.35 to 2.45. The complex multiplet at an average r value of 2-93 corresponds to the proton d, which is the y-pyridine proton. The complex pattern occurs owing to the two ortho couplings of protons e and c and the meta coupling of proton f. The multiplet observed in the region of T 1.60 integrates for two protons. These protons are the cc-pyridine proton f, and the p-pyridine proton c.The low field position of these protons is typical of cc-pyridine protons,% and is due to the effect of the pyridyl nitrogen atom. The low field position of proton c is a result of the proximity of the azo group which, like the pyridyl nitrogen, causes de-shielding effects. The only other pyridine proton to be accounted for is e, which is hidden under the multiplet in the region of r 2.17.386 BETTERIDGE AND JOHN : PYRIDYLAZONAPHTHOLS AND [Analyst, vol. 98 The four remaining naphthol protons, 5, 6, 7 and 8, can be assigned as follows: three of them, 5, 6 and 7, are located in the multiplet at T 2.58, while the other is further downfield at T 2.17 owing to the effect of the azo ErouD.2 - ( 2 - P y r i ~ Z a z o ) - l - n a ~ ~ ~ t ~ o Z (o-a-PAR)LThe features of the spectrum are summarised in Table VIII. TABLE VIII NUCLEAR MAGNETIC RESONANCE SPECTRUM OF o-a-PAN 7 1.67 1.64 2.20 to 2-65 2.98 2.95 to 3.05 3.07 to 3.17 - 6.22 Peak Singlet (much fine structure) Singlet (much fine structure) Complex multiplet Multiplet Doublet Doublet Broad multiplet (weak) Integration 1 1 6 1 1 1 1 Assignment f e, 6. 6, 7, 8 d 3 4 OH c The protons are labelled as follows- OH H(8) For a-naphthol, the hydroxyl proton was observed at T 3-65, whereas for o-&-PAN the signal is observed at T -5.22. This is a clear indication that hydrogen bonding occurs, and that the hydrogen-bond shift is comparable with, but slightly less than, that observed for o-P-PAN.An interesting feature of this signal is that it is much less pronounced than the corresponding signal for o-P-PAN, which may be due to a combination of relaxation, exchange and structural effects.39 The pyridyl and naphthol protons of o-a-PAN are in an environment similar to that in which they occur in o-/I-PAN. As a consequence, the signals are observed at approximately the same positions, the variations being slight. Z-(Z-PyriayEazo)p~enoZ (o-PAP)-The signals of the spectrum obtained for this compound are listed in Table IX. TABLE IX NUCLEAR MAGNETIC RESONANCE SPECTRUM OF o-PAP 7 Peak Integration Assignment 1.26 to 1.30 Doublet with much splitting 1 f 1-94 to 2.00 Doublet (fine structure) 1 G 2.07 to 2.20 Multiplet 2 el 3 2-60 to 2.70 Multiplet 2 dl 6 2-82 to 3.00 Multiplet 2 4, 6 - 2-86 Broad singlet 1 OH The protons are labelled as follows- OH The intramolecular hydrogen bonding expected to be present in this reagent is confirmed by the large hydrogen-bond shift of the hydroxyl proton to r -2.86.June, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS.PART I 387 A significant feature of this spectrum is that the signal of the pyridine proton f is well removed from the others. It occurs as a doublet at T 1-26 to 1.30 with much fine structure due to ortko, meta and para coupling of the protons e, d and c, respectively. Another feature of the spectrum is the decreased de-shielding of proton c, which is located as a doublet at T 1-94 to 2.00. This decreased de-shielding is due to the increased aromaticity of the phenol ring and the consequent loss of bond fixation.The increased aromaticity is further reflected in the position of proton 3. This proton, which is in conjugation with the azo group, is in a similar environment to proton c and the corresponding signal is contained in the multiplet in the range T 2.07 to 2.20. It is likely that the signal for proton 3 contributes to the central largest peak at T 2-14. The remainder of this multiplet, which integrates for two protons, consists of the triplet of the pyridine proton e. The remaining y-pyridine proton d is observed as the usual multiplet in the range T 2-50 to 2-70. Also contained in this multiplet is the phenol proton 5 , in which the triplet peaks have coupling constants matched by protons 4 and 6. The electron-donating effect of the phenol hydroxyl group should be noticeable at both the o- and $-positions of the phenol ring.As a result, both protons 4 and 6 should be shielded, the effect being greatest for proton 6. This proton occurs as a doublet at T 2.91 to 2.99, the coupling constant being 8 Hz, which is matched for proton 5 . This doublet is part of a multiplet, the remainder of which is the expected triplet of proton 4. 4- (2-Pyridylaxo)$henol (p-PA P)-The signals observed in the spectrum are summarised in Table X. TABLE X NUCLEAR MAGNETIC RESONANCE SPECTRUM OF @-PAP 7 Peak Integration Assignment 2.95 to 3.04 Doublet 2 2, 6 2.07 to 2.16 Doublet 2 3, 6 1.91 to 2.66 Multiplet 3 c, d , e 1.30 to 1-36 Highly split doublet 1 f - 0.46 Singlet 1 OH The protons are labelled as follows- The spectrum of this reagent is less complicated than those of the reagents already discussed because of the equivalence of the phenol protons (Fig.4). The very strong upfield doublet at T 2.95 and 3.04 integrates for two protons, and can be assigned to protons 2 and 6. These protons are ortho to the hydroxyl group, and are equivalent, thus giving rise to identical signals. These two protons are split by the two equivalent protons 3 and 6, which are located by the powerful doublet at T 2.07 to 2.16. This doublet is the mirror image of the first doublet and has the same splitting of 9 Hz. The pyridyl protons have the usual type of signals in the expected regions. Proton f has the downfield multiplet expected. The other pyridine protons, e, d and c, are located in the highly complex series of signals in the range T 1.9 to 2.6.One of the noticeable features of the spectrum is the occurrence of a sharp singlet signal a t T -0.46, which can be assigned to the hydroxyl proton (Fig. 4). The position and shape of this signal suggests that the introduction of a para-hydroxyl group leads to a decrease in hydrogen bonding. This evidence, coupled with the results of the infrared studies, indicates that the decreased hydrogen-bond shift is due to the presence of intermolecular hydrogen bond- ing, in contrast with the intramolecular bonding in the other reagents. NUCLEAR MAGNETIC RESONANCE SPECTRA OF CHELATES- The spectra of the nickel(I1) and zinc(I1) chelates of o-N-PAN, o-P-PAN and o-PAP have been obtained. These spectra were found to be very different from those of the parent388 BETTERIDGE AND JOHN : PYRIDYLAZONAPHTHOLS AND [Analyst, VOl.98 ligands. The spectra of the nickel chelates were very poor, with small, ill defined signal peaks. The spectra of the zinc chelates were much clearer, with strong, well defined signals. Unlike the mass and infrared spectra, similar patterns were not observed with the different metals, but the three zinc spectra were very similar, as were the nickel spectra. Detailed interpretation of these spectra is exceedingly complicated, as the review of the nuclear magnetic resonance characteristics of paramagnetic molecules by Eaton and Phillips3s clearly indicates, and was not attempted. I 4 I 7 Fig. 4. Nuclear magnetic resonance spectrum of +-PAP GENERAL CONCLUSIONS The foregoing discussions show that these three spectroscopic techniques can contribute greatly to the understanding of the chemistry of chelates.In addition, the methods can be used to confirm many of the conclusions obtained from solution studies, and as such are useful complementary techniques. It seems that mass spectrometry and infrared spectroscopy are the most useful, as nuclear magnetic resonance spectrometry is somewhat restricted. It is clear that the compounds have now been correctly defined and that the compound previously reported4* as 4-(2-pyridylazo)-l-naphthol ($-%-PAN) is in fact 2-(2-pyridylazo)- 1-naphthol. However, these techniques do not provide a ready explanation for the large difference in pK values between o-a-PAN and o-/?-PAN, which led to the initial mis-identi- fication.The infrared and nuclear magnetic resonance spectra both indicate strong intra- molecular hydrogen bonding of comparable strength. I t is clear from the infrared spectra that the reagents act as terdentate ligands so that the basic stereochemistry of the chelates is fixed. There are, however, eight basic geometrical isomers of the reagent, depending on whether the ring systems are cis or trans to the diazo-nitrogen group and the relative positions of the pyridine-nitrogen and the hydroxyl group. With the aid of models (Prentice-Hall Framework Molecular Models), it is possible to envisage that two of these eight positions, by virtue of hydrogen bonding, will be much more favoured than the others, but that one might result in a stronger hydrogen bond than the other.It is therefore possible that the difference in the pK values is explained by the geometrical configuration of the reagent. The spectroscopic techniques applied do not give any information about this possibility, which will have to be checked by an X-ray determination or by a full equilibrium study in order to isolate the relative contributions of the enthalpy and entropy terms to the free energy change. We are grateful to the T. and E. Williams Scholarship Fund for a maintenance grant to one of us (D.J.).June, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I REFERENCES 389 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.32. 33. 34. 35. 36. 37. 38. 39. 40. Anderson, R. G., and Nickless, G., Analyst, 1967, 92, 207. Cheng, K. L., and Bray, R. H., Analyt. Chem., 1955, 27, 782. Galik, A., Talanta, 1969, 16, 201. Donaldson, E. M., and Inman, N. R., Talanta, 1966, 13, 489. Chromy, V., and Sommer, L., Ibid., 1967, 14, 393. Sommer, L., and Hnilickova, M., Ibid., 1969, 16, 83. Sommer, L., and Ivanov, V. M., Ibid., 1967, 14, 171. Sommer, L., Ivanov, V. M., and Novotna, H., Ibid., 1967, 14, 329. Sommer, L., and Novotna, H., Ibid., 1967, 14, 457. Chichibabin, A. E., Zh. Russk. Fiz.-khim. Obshch., 1920, 50, 512. Pollard, F. H., Nickless, G., and Anderson, R. G., Tulanta, 1966, 13, 725. Anderson, R. G., and Nickless, G., Proc. Soc. Analyt. Chem., 1966, 3, 149. Anderson, R. G., Ph.D. Thesis, University of Bristol, 1966. Fargher, R. G., and Furness, R., J . Chem. Soc., 1915, 107, 691. Fieser, L. F., Org. Synth., 1943, Collect. Vol. 11, 430. Conant, J . B., and Corson, B. B., Ibid., 1943, Collect. Vol. 11, 33. Betteridge, D., and John, D., Talanta, 1968, 15, 1227. Nakamoto, K., “Infrared Spectra of Inorganic and Co-ordination Compounds,” John Wiley, New Adams, D. M., “Metal - Ligand and Related Vibrations,” Edward Arnold (Publishers) Ltd., Kline, C. H., and Turkevich, J., J . Chem. Phys., 1944, 12, 300. Corssin, L., Fax, B. J., and Lord, R. C., Ibid., 1953, 21, 1170. Wimshurst, J . K., and Bernstein, H. J., Can. J . Chem., 1957, 35, 1185. Hunsberger, I. M., J . Amer. Chem. SOC., 1950, 72, 5626. Friedel, R. A., Ibid., 1951, 73, 2881. Barnes, R. B., Liddell, U., and Williams, V. Z., Ind. Engng Chem., Analyt. Edn, 1943, 15, 659. Bassignana, R., and Cogrossi, C., Tetrahedron, 1964, 20, 2361. Hadzi, D., J. Chem. SOC., 1956, 2143. Ueno, K., and Martell, A. E., J . Phys. Chem.. 1955, 59, 998. Charles, R. C., Freiser, H., Friedel, R., Hilliard, L. E., and Johnston, W. D., Spectrochim. Acta, Ueno, K., J . Amer. Chem. Sac., 1957, 79, 3066. Baldwin, D. A., Lever, A. B. P., and Parish, R. V., Inorg. Chem., 1969, 8, 107. Gill, N. S., Nuttall, R. H., Scaife, D. E., and Sharp, D. W. A, J . Inorg. Nucl. Chem., 1961, 18, 79. Lecomte, J., Discuss. Furaday SOC., 1950, 9, 125. Day, R. J., and Reilley, C. N., Analyt. Chem., 1964, 36, 1073. Eaton, D. R., and Phillips, W. D., in Waugh, J. S., Editor, “Advances in Magnetic Resonance,” Sawyer, D. T., and Smith, B. B., Inorg. Chem., 1960, 8, 379. --- , “High Resolution Nuclear Magnetic Resonance,” McGraw-Hill, New York, 1959, Bernstein, H. J., Pople, J. A., and Schneider, W. G., Can. J . Chem., 1957, 35, 1487. Emsley, J. W., Feeney, J., and Sutcliffe, L. H., “High Resolution Nuclear Magnetic Resonance,” Betteridge, D., Todd, P. K., Fernando, Q., and Freiser, H., Anulyf. Chem., 1963, 35, 729. York, 1963, p. 216. London, 1967. 1956, 8, 1. Academic Press, London, 1965, p. 103. pp. 400-421. Volume 1, Pcrgamon Press, Oxford, 1965, p. 534. Rcceived A$ril2nd, 1970 Accepted November 22nd, 1972

ISSN:0003-2654    

DOI:10.1039/AN9739800377

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

390 Analyst, June, 1973, Vol. 98, p p . 390-411 Pyridylazonaphthols (PANS) and Pyridylazophenols (PAPS) as Analytical Reagents Part 11." Spectrophotometric and Solvent-extraction Studies of Complex Formations BY D. BETTERIDGE AND D. JOHN? (Chemistry Department, University College of Swansea, Swansea, Glamorgan, SA2 8PP) The reactions between 2-(2-pyridylazo)-l-naphthol (o-oc-PAN), 1-(2- pyridylazo)-2-naphthol (o-P-PAN) and 2-( 2-pyridy1azo)phenol (0-PAP) with manganese(II), zinc(I1) and lanthanum(II1) and 4-(2-pyridylazo)phenol (#-PAP) with cobalt(II), nickel(I1) and copper(I1) have been studied. The spectrophotonietric procedure based on linear extrapolation, as used by Sommer, has been critically evaluated. The results from the spectrophoto- metric method have been used to predict the optimum conditions for solvent extraction.It is shown that this procedure is a valuable approach for systems in which hydroxy-complexes are common. SEVERAL studies have shown that o-P-PAN is very useful for the extraction of many cations.1-3 However, the extraction systems are often more complex than some of the early work suggests, and at present few detailed solvent-extraction studies have been made. Preliminary studies showed that 2-(2-pyridylazo)-l-naphthol (0-a-PAN) is often more advantageous than 1-(2-pyridylazo)-2-naphthol (o-P-PAN) and that 2-(%pyridylazo)phenol (0-PAP) and 4-(2- pyridy1azo)phenol (@-PAP) also form extractable complexes. The conventional method of determining equilibrium constants from solvent-extraction experiments is time consuming. It is possible, in principle, to determine the complex formation constants by other means, to determine the partition coefficient of the reagent and complex experimentally and to use the values obtained to predict the extraction graphs.One practical difficulty is that the con- stants that are obtained from solvent-extraction data relate to the aqueous phase saturated with the organic solvent and as, almost by definition, the complex is only sparingly soluble in water, the equilibrium constants must be determined in some other medium, e.g., 1 + 1 dioxan - water, so that they cannot be used directly for predicting extraction equilibria. Spectrophotometric methods can partially overcome this difficulty because solutions can be used that are so dilute that the complex can be maintained with such a small proportion of organic constituent that the values of the equilibrium constants are close enough to those obtained by solvent-extraction procedures to be interchanged.However, most spectrophoto- metric methods are based on Job's method, which can result in misleading or erroneous results, or both,4 and the study of a system over a wide range of conditions with these methods is extremely tedious. Recently, Sommer and co-workers5-11 have demonstrated that spectrophotometric procedures based upon linear extrapolation, when used with care, can be used most advan- tageously to study the complex equilibria of systems based on compounds analogous to the compounds discussed below. A practical advantage of their approach is that it is based upon the analysis of pH - absorbance curves so that no superfluous information is gathered and basic data are interpreted fully.In this paper, we assess this procedure by analysing several systems, for some of which results are already available for comparison, and show how it can be used for predicting extraction curves with an accuracy that is acceptable to the analyst. EXPERIMENTAL REAGENTS- Salts and solvents of analytical-reagent grade purity or better were used throughout. * For Part I of this series, see p. 377. t Present address: BP Chemicals (U.K.) Ltd., Llandarcy, Swansea. @ SAC and the authors. Parts I11 and IV will appear in the July issue.BETTERIDGE AND JOHN 391 1-(2-Pyridylaz0)-2-naphthol (0-p-PAN). 2- (2 - Py ridy 1 azo ) - 1 -naphthol ( o-a- P A N ) .2- (2-Pyridylazo)phenol (0-PA P) . 4-(2-Pyridylazo)phenol (p-PAP) . The above four reagent solutions were prepared and purified as described in Part I. Solutions in absolute or aqueous ethanol were prepared for spectrophotometric studies and in carbon tetrachloride or chloroform for solvent extraction. The concentrations are given in the procedures. The solutions were taken to be standard and the validity of this assumption was checked occasionally by spectrophotometric titration of a standardised metal-ion solution. Metal-ion solutions-Solutions of manganese(II), zinc(II), nickel(II), cobalt(I1) and lan- thanum( 11) were prepared and standardised titrimetrically with EDTA by standard procedures. Bufer solutions-Standard chloride, phthalate, phosphate, borate and hydroxide buffer solutions to cover the pH range from 0 to 12 were used.Sodium perchlorate-Solid sodium perchlorate was used in order to maintain an ionic strength of 0.10 & 0.01. Radioisotopes-Manganese44 and zinc-65 were obtained from the Radiochemical Centre, Amersham. They were diluted and mixed with carrier manganese(I1) and zinc(I1) so that solutions of known concentration and radioactivity were obtained. APPARATUS- pH meter-Radiometer, Model M4C. Spectro$hotometers-Unicam SP500, SP600 and SP800 and Cary, Model 16, instruments were used as appropriate. Absorbance values used for the calculation of equilibrium constants were obtained on fixed-wavelength instruments. Radioactivity-A 1-inch well sodium chloride (thallium-activated) crystal connected to a photomultiplier and a 1DL 1700 scaler was used to measure the radioactivity.Samples of constant volume were taken so as to ensure constant geometry and sufficient counts were taken so as to ensure a statistical error of not more than 1 per cent. on all except the very low count-rates, when an error of 10 per cent. was accepted. Computer-An IBM 1600 computer was used for the calculation of constants based on the spectrophotometric data and the statistical analyses of them. The standard program for linear regression in two variables 1620/6.0.27 from the 1620 General Program Library was used. PROCEDURES- Spectrophotometric determination of acid-dissociation constants-A 10-ml aliquot of a solution of o-a-PAN (1 x M) or $-PAP (1.25 x 10-4~) in 12.5 per cent.aqueous ethanol was placed in a 25-ml calibrated flask and sufficient solid sodium perchlorate to maintain an ionic strength of 0.10 & 0.01 was added. The solution was then made up to the mark with buffer solution. The solution was mixed well, the spectrum recorded and the absorbance and pH were measured. A pH - absorbance curve was plotted for each of the reagents at suitable wavelengths and the pK, values were calculated from the inflection point as determined from a graph of the differential AAIApH against pH. The pK, values were also calculated from the same pH - absorbance curve by the procedure described below. Spectrophotometric determination of formation constants-pH - absorbance curves were obtained with both the metal ion and the reagent in excess.The general procedure was to place 10 ml of ethanolic solution, 10 ml of metal-ion solution and 5 ml of buffer in a 25-ml calibrated flask that contained sufficient solid sodium perchlorate to maintain the ionic strength of the final solution at 0-10 -+ 0.01. These solutions were mixed thoroughly and the colour was allowed to develop to a maximum (a few minutes). The absorbance remained constant for the period of the measurement and showed little change after 24 hours. The spectra were recorded, the absorbances were determined at suitable wavelengths and the pH values were measured. The measurements were carried out a t room temperature. The method of continuous variations12 and the mole-ratio method13 were also applied to some systems.Solvent extraction and determination of acid-dissociation constants-The reagent solutions were made up in carbon tetrachloride (for o-a-PAN, o-p-PAN and o-PAP) or chloroform (for $-PAP) to concentrations of 1.0 x 1.0 x 2.0 x and 5.0 x 1 0 - 4 ~ ~ One-centimetre cells were used. M) in absolute ethanol or o-PAP (2.5 x392 BETTERIDGE AND JOHN : PYRIDYLAZONAPHTHOLS AND [Analyst, Vol. 98 respectively (@-PAP is insufficiently soluble in carbon tetrachloride for the distribution ratio to be measured accurately). A 5-ml aliquot of reagent solution was placed in a vial and 60 ml of buffered aqueous phase of ionic strength 0.10 & 0.01 were added. The solutions were shaken overnight in a box shaker that was maintained thermostatically at 24 &- 1 "C. The phases were allowed to separate and the absorbance of the organic phase was measured with the Unicam SP600 instrument and that of the aqueous phase, after adjustment of the pH to zero with concentrated hydrochloric acid, with the Cary 16 instrument.Calibration graphs were used to convert the absorbance values into concentration values. The distribution ratio of the reagent, D, was calculated from D = (CT - c W ) v W / c W v O D = CoVw/CwV, for D > 1, or for D < 1, where CT is the total concentration of the reagent, C, and C, are the concen- trations of the reagent in the organic and aqueous phase, respectively, and V , and Vw are the volumes of the organic and aqueous phase, respectively. Log D was plotted against pH and the partition coefficient of the reagent, KDR, and acid-dissociation constants, Kal and K,,, were calculated from1* D = KDR { [HI/KaI + 1 + KaJ [HI 1 Solvent extraction of metal ions-Reagent solutions were 5.0 x M in carbon tetra- chloride.M were diluted to a suitable concentration when the extraction was followed spectrophotometrically. When the extraction was followed radiochemically, a solution that was as dilute as the specific activity of the isotope would allow was used and a 0.2 M solution of sodium perchlorate was prepared. Equal volumes of organic and aqueous phase were used, the latter consisting of 2 nil of metal-ion solution, 2 ml of buffer solution and 4 ml of sodium perchlorate solution. The solutions were placed in a vial and shaken overnight, the layers were separated and the radioactivity of each phase was measured.The total activity in the two phases was computed and if it was less than 80 per cent. of the total activity in the vial, the results were discarded. Stock metal-ion solutions of 5 x RESULTS AND DISCUSSION The following symbols are used- HR, H,R+, R- CCXI a, P f the neutral, protonated and ionised forms of the reagent, respectively the total analytical concentration of species x the molar absorptivity of species x the total absorbance IMlo/lMlw, the distribution ratio [HR],/ [HR], or [MRn]o/ [MRnIw, the partition coefficient of the reagent [HR] [H+]/ [H2R+], the first acid-dissociation constant [R-] [H+]/[HR], the second acid-dissociation constant [MR,]/[M] [R-I2, the over-all formation constant of MR2 and extractable complex, respectively [MRI / CMI [RI WR2I [H+l2/[M1 WRI2 WR2OHl / WR2I PHI [H,,R]~%~H,-,R] the fractional extent of completion of the reaction as written.x=o [MI /CMT0J Charges are omitted unless possible ambiguity would arise from their absence, REAGENT EQUILIBRIA IN THE ABSENCE OF METAL IONS SPECTROPHOTOMETRIC STUDIES- Principles of graphical analysis of absorbnnce versus pH curves-This method was re- examined by Sommer5 in 1964 and has since been applied to the study of a number of reagents.6~7 The principles involved are summarised below.June, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I1 393 The method involves a detailed algebraic study of the equilibria of the reagent, and in the present study these are taken to be: Kaz Ka, HR + R- + H+ (Equilibrium 1) H,R+ + HR + H+ (Equilibrium 2) On the initial assumption that these equilibria represent the true reagent equilibria, a series of linear algebraic transformations can be derived. Experimental values of absorbance and pH are introduced into these transformations, which then provide values of the acid- dissociation constants, Kal and Ka,, and also molar absorptivity values of individual reagent species.An observed lack of linea.rity in these transformations indicates that the assumed equilibria are incorrect. In Equilibrium 1, if it is assumed that only the neutral (HR) and anionic (R-) forms of the reagent are present, it follows that the total reagent C R = [HRI + LR1 = [RI {e+ l} The total absorbance, A , is given by A = EHR [HRI + ER [R-I concentration, CR, is given by ..' (1) .. .. Substitute for [R] from equation (1) and re-arrange to give .. Multiply throughout by l/eR and re-arrange to give - (A - €HRCR) CR A =-+ €R Ka,ERA A graph of CR/A zleysus [H]A is usually a straight line of Values of the acid-dissociation constant, Ka,, can be obtained sponding values of absorbance and pH. Alternatively, it can be argued that r and A = [HR] [EHR + 6 3 1 [HI .. .. .. .. (1) .. .. intercept l/en, as A > EHRCR. by calculation by using corre- .. .. .. .. .. .. .. .. which give This equation, on multiplication by l/gHR and subsequent re-arrangement, gives .. .. CR - 1 Ka, ( A - CRER). 1 A -EG + EHR A[H] " Equation (IA) should also yield a straight line when CR/A is plotted against l/A[H]. The intercept of this line represents the reciprocal of the molar absorptivity, EHR.Values of Ka, can be calculated from corresponding values of absorbance and pH. Alternatively, a value of eR, which can be found from other experiments, can be substi- tuted and &/A plotted against ( A - CR~R)/A[H]. A similar procedure can be adopted to solve graphically all of the subsequent transformations, where the linearity is not obvious.394 BETTERIDGE AND JOHN : PYRIDYLAZONAPHTHOLS AND [Analyst, vol. 98 These equations are therefore derived simply from consideration of mass balance and equilibrium constants, with the one important algebraic manipulation being effected by the multiplet ~ / E H R or ~/ER. In equilibrium 2, if it is assumed that [R-] is negligible compared with [H,R] and [HR], then and By using shown that CR = LHR1 + LHzR1 A = EHR[HR] + EH,R[H&] these equations and proceeding in the same manner as before, it can be and For equation (11), the graph of CE/A veysus l/A[H] should be a straight line.Similarly, the graph of CR / A versm [H]/A for equation (IIA) should be linear. Values of the first acid- dissociation constant, &, and of the molar absorptivities, EHR and c H 2 ~ , can be calculated from these equations. Absorbance curves and results of calculations-The absorbance veysus pH curves are shown in Fig. 1. The wavelength of maximum absorption for the various species and isosbestic points are given in Table I. The results obtained for each transformation are summarised in Table 11. The transformations are represented as straight-line equations of the form y = mx + c, where y = &/A and x = 1/A [HI or [H]/A multiplied by a power factor to keep x and y mainly within the range 0.1 to 10.The values obtained are in good agreement with those obtained from the inflection point of the pH - absorbance curve. It is slightly advantageous that the molar absorptivity is calculated simultaneously and that a check on the nature of the equilibrium is provided. If two protons were to be released simultaneously, for example, the observed transformations would be curved or extremely ill defined. The least-squares analysis permits the precision of the result to be calculated. It was found that because the method is based on an extrapola- tion, it is very sensitive to the choice of points used for the calculation.Inevitably, a t one end of the part of the pH - absorbance curve that is being used for analysis, the basic assump- tion that one species is of negligible concentration compared with the others becomes less valid. A point from this part of the curve will therefore be “bad” and because it will be at the end of the transformation it can exert a disproportionate effect on the least-squares analysis. Several checks are possible: (i) the point may be so “bad” that it falls outside the 95 per cent. confidence limits and can be rejected; (ii) the point can be discarded and the analysis carried out again, and if the result is then the same the point can be accepted and if it is markedly different it can be rejected; and (iii) a different transformation can be used and the results compared.Rejection by the second of these procedures is not entirely satisfactory and care was taken to reject not more than one result and to confirm, by examination of the experi- mental pH - absorbance curves, that this result was a marginal value. Checks of (i) and (iii) were always carried out. A practical disadvantage of the method, therefore, is that a larger number of “good” experimental points must be obtained than is necessary for the simpler spectrophotometric procedures. SOLVENT-EXTRACTION STUDIES- The experimental curves of log D versus pH were very similar to that already published for o-a-PAN.14 All of the curves indicated that a protonated species of reagent was formed under acidic conditions and an anionic species under alkaline conditions.The pK values obtained are given in Table 111. The logarithm of the partition coefficients for o-a-PAN, o-p-PAN, o-PAP and +-PAP are 5.20, 5.00, 3.75 and 2.53, respectively. It was found that although the partition coefficients for 0-a-PAN and o-P-PAN were an order of magnitude greater than those reported earlier, the acid-dissociation constants were in reasonable agree- ment. This agreement is surprising, as the values of the partition coefficients are used in the calculation of the acid-dissociation constants. With such high partition coefficients, theJune, 19731 1 1 - 1 1 1 1 1 1 1 1 1 1 1 1 1 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I1 0.9 0.8 0.7 395 (C) Ka 1 - HzR' + HR + H" HR i R-+ H+ pKaq= 2.70 - - 1 -2 1.0 08 0.6 0-4 0.2 2.0 0.4 0 PH Fig.1. Absorbance versus pH graphs: (u) 5.0 x M $-PAP in 6 per cent. aqueous ethanol; M o-a-PAN in 40 per (b) 1.0 x lo-* M o-PAP in 5 per cent. aqueous ethanol; and (c) 4.0 x cent. aqueous ethanolTABLE I ABSORBANCE MAXIMA, MOLAR ABSORPTIVITIES AND ISOSBESTIC POINTS H,R+ HR R- Isosbestic point A r \ Reagent wnm E )czlnm E &bm E H,R 4 HR+ + H+ HR + R- + H+ P-PAP 402 2-38 x 104 358 1.89 x 104 442 2-10 x 1 0 4 373 382 O-PAP 352 1.62 x 1 0 4 328 1-56 x lo4 480 9-16 x 103 337 410 o-a-PAN 365 1.63 x 104 486 1.62 x 1 0 4 520 2.08 x lo4 476 - TABLE I1 MOLAR ABSORPTIVITIES AND ACID-DISSOCIATION CONSTANTS OF REAGENTS FROM LEAST-SQUARES FIT OF TRANSFORMATIONS (I), (IA), (11) AND (IIA) Correlation Standard error of Acid-dissociation Molar absorptivity constant ER PKa, Transformation Reagent y = w + c coefficient I p-PAP y = 0.101x + 4.756 0.992 0.0046 0.183 2.10 x 104 7-66 f 0.02 0-PAP y = 0.023~ + 1.003 0.990 0*0010 0.057 9-16 x 103 8.77 f 0-02 o-a-PAN J.J = 0 * 0 8 1 ~ + 4.808 0.988 0.0045 0.048 2.08 x 10' 10.17 & 0.02 IA $-PAP 0-PAP o-E-PAN I1 P-PAP O-PAP o-a-PAN IIA P-PAP 0-PAP o-a-PAN y = 0.264~ + 5.056 y = 0.018~ -+ 6.638 - y = 0.183~ + 4.203 y = 0.030~ $- 6.189 y = 0.055~ + 6.154 y = 0.044~ + 5.561 y = 0.041% + 6.621 y = 0.099~ + 6.182 0.979 0.975 - 0.980 0.992 0-993 0.978 0.989 0.990 0.0192 0.0014 - 0.0142 0-0015 0.0020 0.0033 0~0020 0.0049 0.198 0.091 I 0.057 0.054 0.018 0.150 0.035 0-023 EH R 1.98 x 104 1-51 x 104 2.38 x 1 0 4 - EHIR 1.62 x 10' 1.63 x lo4 €HR 1-80 x 104 1.62 x 1 0 4 1.51 x lo1 PKa, 7.78 f 0.02 8.92 f 0.04 - PKa, 3.00 f 0.03 2.72 f 0.06 2-73 f 0.03 2.94 f 0.07 2.64 f 0.05 2.60 f 0.05 P h'a, No calculations were carried out for o-a-PAN by using transformation (IIX) because of the proximity of the Amax.values of the reagent species HR and R-. * Z U c-c 0 E * Z U n b x $June, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I1 397 presence of a small amount (e.g., 0.1 per cent.) of a coloured impurity that is less extractable than the reagent can profoundly affect the experimentally determined value of the partition coefficient, but would scarcely affect the log D versus pH curve when the value of the dis- tribution ratio is less than 100. This effect would result in the experimentally determined pKa, value being greater than and the pKa, value being less than the true values.As there is an internal consistency in both sets of studies, which should prevent miscalculation, we can offer no explanation for the discrepancy unless there is some slight difference between the samples of reagent-grade carbon tetrachloride used in the two studies. The values obtained in this study were used throughout this work. TABLE I11 ACID-DISSOCIATION CONSTANTS AT AN IONIC STRENGTH OF 0.10 Reagent Method* O- p-PAN SP. Pot. S.E. S.E. o-E-PAN Pot. S.E. S.E. SP. SP- SP. SP. o-PAP SP. SP* SP. #-PAP SP. SP. SP. S.E. S.E. Medium 20% aqueous dioxan 50% aqueous dioxan Chloroform - water 50% aqueous dioxan Water Water - carbon tetrachloride Water - carbon tetrachloride 50% aqueous methanol 40% aqueous ethanol 40% aqueous ethanol 50% aqueous methanol 5 yo aqueous ethanol 5y0 aqueous ethanol Water - carbon tetrachloride 50% aqueous methanol 5% aqueous ethanol 5% aqueous ethanol Water - carbon tetrachloride - PK, 1.9 <2 2-9 2.9 2-54 3.0 3.1 2.90 2.29 2.67 2.70 1.85 2.68 2.68 2.68 2-47 2-97 2.86 3-58 PKa, 12.2 12.3 11.5 11.2 10.74 9.1 9.5 9.63 10.00 10.17 10.23 9.42 8.84 8.79 8.68 8-20 7-72 7.76 7-55 Reference 15 16 17 18 14 14 14 This work 19 This work, graphical This work, conventional 20 This work, graphical This work, conventional This work 20 This work, graphical This work, conventional This work * Sp.= spectrophotometric; Pot. = potentiomctric; S.E. = solvcnt extraction. COMPLEX FORMATION EQUILIBRIA SPECTROPHOTONETRIC STUDIES- The graphical method that is used for the determination of acid-dissociation constants can be applied in the study of the formation 01 complexes.The method consists in first postulating the chelation reaction and the species formed over a particular pH range, then algebraically deriving transformations that would necessarily be followed if the postulates were correct. Typically, these transformations are straight lines, and false assumptions are rcadily detected by the presence of curvature or random scatter when experimental values are substituted into the transformations. All parts of the simple absorbance - PI-I curve can be subjected to such analysis, so that the various conditions of chelation can be deduced over a wide pH range. Normally, absorbance-pH curves are obtained in the presence of excess of reagent and excess of metal ions, so that the absorbance can be expressed generally as A = f(PH) cma,cx A = .WM)~H,C~ A =~(CR> p ~ , cM Sommer and co-workers have summarised a large number of basic equations for a variety of chelation reactions in the presence of various concentrations of reagent and metal ions, and have since applied the method successfully to different systems.These systems include the reactions of 4-(2-thiazolylazo)resorcinol (TAR) and 4-(2-pyridylazo)resorcinol (PAR) with thallium and uranyl ions.8-f0 The reactions of copper(II), lead(II), cadmiuni(II), zinc(I1) and bismuth(II1) with TAR and of the lanthanides with PAR have also been reported.6911 and398 BETTERIDGE AND JOHN : PYRIDYLAZONAPHTHOLS AND [.4nd,St, VOl. 98 We have applied the method to the reactions of zinc(II), manganese(II), cobalt(II), nickel(II), lanthanum(III), copper(I1) and titanium(1V) with o-a-PAN, o-b-PAN, o-PAP and p-PAP and compared the results when possible with those obtained by different methods.Most of the findings are given below but some will be dealt with in subsequent papers. GRAPHICAL ANALYSIS OF ABSORBANCE CURVES- The general approach can be illustrated by the derivation of a few basic equations. One common reaction is chelate formation when the reagent is predominantly in the neutral form. The reaction can be represented as *KYR1 This reaction can be carried out with either metal ions or the reagent in excess and the un- reacted reagent may or may not contribute to the absorbance at the wavelength of maximum absorbance of the chelate.In the presence of excess of metal ions--In this instance, [MI = Cy, the total metal-ion concentration, and if it is assumed that E H R is negligible a t the wavelength being used, then Furthermore, if it is assumed that the concentration of the intermediate chelate species, MR, is negligible, then Re-arrange equation (5) and substitute for [MR,] : M+2HR + MR2+2H .. .. .. * * (3) The effects of these variations are considered below. .. * - (4) C R = [HR] + 2[MR,] . . .. .. - (5) A = EMR, [MRJ .. .. [HR] = CREMR, - ZA EMR, Divide equation (4) by equation (6) and substitute for [MR,]/[HR] by introducing the reaction constant, so that or Multiply through by ~ / A E ~ I R , and re-arrange : When the reagent makes a contribution to the absorbance, the expression for the total absorbance requires an additional term : A = EMR,[M&] EHR[HR] 6 - ... . . . (4a) The steps outlined above then give . . (IIIA) In the presence of excess of reagent-Under these conditions, [HR] = C,, the total reagent -- CR - - 1 . ( A - CREMK) . [HI2 A - EMR, + "KMR CMEMR, A [HR] ' ' concentration, and if the reagent absorption is negligible, then ' * (7) .. .. .. .. A [MR,] =- EMR, and if [MR] is negligible compared with [MI and [MR,], then so that C M = [RI] + [MR,] . . .. .. - - (8) Divide equation (7) by equation (9), substitute for [MR,]/[M] and re-arrange, to giveJune, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I1 399 Multiply through by ~/AEMR, and re-arrange, so that A- EMR, *KMR,c~REMR, [HI2 .. .... .. (IV) 1 1 CId - - + ~ - . When the reagent absorbance is not negligible, an extra term can be introduced. In practice, we found that either the reagent absorbance was negligible or it represented such a large contribution that internal compensations on the spectrophotometer had to be made. Another important type of reaction is the hydrolysis reaction KOH MR2 + OH- + MR,OH- .. .. .. . . (10) If it is assumed that the hydroxy-complex makes no contribution to the absorption, then in the presence of both excess of reagent and excess of metal ions the total absorbance, A , is given by A = EMR~ [MR2I For excess of metal ions, the total reagent concentration, CR, is given by For excess of reagent, the total metal-ion concentration, CM, is given by In the presence of excess of reagent: C R = 2[MR,] + 2[MR,OH] C M = [MI + PR2I + [MR,OHI .... - - (V) ions is academic because. or The derivation of equations in the presence of excess of metal in most systems, the metal hydroxide would be formed preferentially and be precipitated: The transformations for these and other systems are given in Table IV; some are in logarithmic form because this form is more convenient for use in calculations. All of the transformations can be expressed as a linear function, although some contain two variables, [HI and [HR], in one term. In these instances, as for example transformation (111), EMR, was calculated from another transformation, (IV), or by simpler conventional means, and this was used to calculate [HR] by means of equation (6). The calculation procedure is very similar to that used for the acid-dissociation constants and the same checks were applied.However, even greater care is necessary because although a pH - absorbance curve reflects the course of several reactions, which allows these reactions to be detected and studied, it is easier to introduce “bad” points into the calculations. A further check was therefore always carried out. The values of the calculated constants were used to derive a pH - absorbance curve and this curve was then compared with the experimental curve. Zinc(l1) chelates of o-a-PAN, o-P-PAN and o-PAP-pH - absorbance curves obtained for the chelates of each reagent in the presence of excess of metal ions and of reagent are shown in Fig. 2. The absorption maxima of the zinc(I1) chelate with o-a-PAN, o-P-PAN and o-PAP occurred at 548 and 590, 514 and 550, and 530nm, respectively, the chelates of the naphthol derivatives each having two maxima, and the pH - absorbance curves were obtained a t 590, 550 and 530 nm, respectively.These curves indicate that each chelate has a stoicheiometry of metal to ligand of 1 : 2. At the upper limits of pH, the absorbance values became inconsistent and no smooth curve could be drawn. Hence, over the pH range 6 to 8 in which the chelate is being formed, the basic reaction would appear to be Zn + 2HR + ZnR, + 2H Transformations (111) and (IV) were therefore used for the reactions with o-a-PAN and o-PAP and transformation (IIIA) was used for o-P-PAN, as the reagent contributed to the absorbance. The results are given in Table V, which shows that there is good agreement between the molar absorptivities and stability constants determined in the presence of both excess of reagent and excess of metal ions.The agreement between these values, together with the linearity of transformation (IIIA) and hence of transformation (VIA), confirm that the chelation reactionTABLE IV BASIC COMPLEX FORMATION REACTIONS AND LINEA.R TRANSFORMATIONS Principal reaction Condition M + ZHR+MR,+ZH Excess of metal ions Excess of reagent Excess of metal ions Excess of reagent MR, + OH+MR,OH Excess of reagent M + HRsMR + H Excess of metal ions Excess of reagent Excess of metal ions Excess of reagent M + 2R+MR, M + R+MR No. 111 IIIA IV VI VIA VII V VIII IX X X I W 2 4 M w Y U 0 M n b W M,June, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS.PART I1 401 suggested is correct and that the assumptions made in the derivation of the equations are valid. Each chelate investigated therefore has a stoicheiometry of metal to ligand of 1 : 2, and is involved in a chelation reaction of the type PH Fig. 2. Absorbance zwsw pH graphs for zinc(I1) chelates: (a) in the presence of excess of metal ions- c M / M C R / M X/nm 0-a-PAN . . . . 2.0 x 1 0 - 4 4.0 x 10-6 590 0-P-PAN . . . . 2.0 x 1 0 - 4 5.0 x 10-5 550 O-PAP * . . . 2.0 x 10-4 6.0 x 10-6 530 c b f / M C R / M X/nm 0-M-PAN . . . . 2.0 x 10-6 2.0 x 10-4 590 o-P-PAN . . . . 2.0 x 10-6 2.0 x 1 0 - 4 550 0-PAP .. . . 2.0 x 10-6 2-0 x 10-4 630 (b) in the presence of excess of reagent- The values of the molar absorptivities and stability constants determined in this study were found to be consistent with the results of other Mangnnese(l1) chelates of o-a-PA N , o-P-PAN and o-PAY-The absorbance maxima of the nianganese(I1) chelates of o-a-PAN, o-P-PAN and o-PAP occurred at 646 and 686, 616 1 .o (a 1 5 6 7 8 9 1 0 1 1 1 2 PH Fig. 3.Absorbance versus pH graphs for manganese(I1) chelates: C m / h l C R / M h/nm (a) in the presence of excess of metal ions- 0-LX-PAN .. . . 2.0 x 10-4 4.0 x 10-6 686 0-PAP .. .. . . 2.0 x 10-4 4.0 x 10-6 630 o-P-PAN . . . . 2.0 x 10-4 4.0 x 10-6 546 (b) in the presence of excess of reagent- c a r / M Cn/M hlnm 0-LX-PAN .. . . 2.0 x 10-6 2.0 x 10-4 686 0-PAP .. . . .. 2.0 x 10-5 2.0 x 10-4 530 o - p - ~ ~ ~ .. .. 2.0 x 10-6 2.0 x 10-4 546 1 12TABLE V RESULTS FOR THE CHELATES OF ZINC(II) Standard error of -mate y = m t : + c coefficient slope, m Correlation y = 0.087~ + 4.515 1.000 0*0003 0-015 y = 0-804X + 8.349 1.000 0.0052 0.096 y = 0.066~ + 2.212 1.000 0.0002 0.016 y = 0.037% + 2.222 1.000 0-0002 0.0 15 y = 0.076~ + 4.165 1.000 0.0003 0.025 y = 0.453~ + 4.208 0.999 0.0055 0*108 pKa, = 10.20 (o-a-PAN); 12-20 (O-F-PAN); 8-80 (0-PAP).Condition Reagent Excess of metal ions o-a-PAN O- p-PAN 0-PAP Excess of reagent o-a-PAN O- 8-PAN O-PAP Condition Reagent Excess of metal ions o-a-PAN 0- 8-PAN 0-PAP Excess of reagent o-a-PAN 0- 8-PAN 0-PAP TABLE VI RESULTS FOR CHELATES OF MANGANESE(II) Standard error of Correlation r y = 0.619~ + 4.945 1.000 0.0006 0.01 1 y = 0.045~ + 2.455 1.000 0-000 1 0.009 y = m % + c coefficient slope, m estimate y = 0.874~ + 4.823 1.000 0.0006 0-029 y = 0.504~ + 9.201 1.000 0.0012 0.059 y = 0.086~ + 2.443 1.000 0-0002 0-015 y = 0.265% + 4.808 0.999 0-0032 0-132 pKa, = 10.20 (0-a-PAN) ; 12.20 (0-8-PAN) ; 8-80 (o-PAP).EMIR, x 4-43 4.75 2-40 4.52 4-50 2.40 EYR, x 4-15 4.04 2-17 4.09 4.07 2-08 log KMR, 19.13 f 0.07 21.63 f. 0.08 15-52 f 0.09 19.12 f 0.07 21.45 f 0.08 15-48 f 0.09 log KMR, 13-54 f 0.06 15-77 f 0.06 10.52 f 0.02 13.27 f 0.05 15-69 f 0.08 10-45 f 0.05 ?- z U 0 X Z n ? pJune, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I1 403 and 546, and 530 nm, respectively, the chelates of the naphthol derivative each having two maxima, and the pH - absorbance curves (Fig. 3) were obtained at 586, 546 and 530 nm, respectively.Over the pH range 6 to 9 the curves are similar in appearance to those found for the zinc(I1) chelates and the same analysis was applied. The results are shown in Table VI. At higher pH values, the absorbance of the neutral chelate was found to decrease with increase in pH. The decreased absorbance was usually accompanied by the formation of a precipitate, which was presumed to be a hydroxy-form of the chelate. With o-a-PAN and o-PAP, it was not found possible to correlate absorbance with pH, and it was observed that the absorbance tended to zero on standing the solution. The manganese(I1) chelate of o-P-PAN, however, gave reasonably smooth absorbance curves in the presence of both excess of reagent and excess of metal ions. The values tended to vary on standing the solution, but it proved possible to use them in transformation (V) (Fig.4). From these values, KMR,(OH) was found to be The linearity suggests that the hydrolysis reaction that was proposed is reasonable, but because of the variation noted above, we have some reservations about these conclusions. 0.2 - ( a ) 0 - -0.2 -0.4 - - 0.2 l,o.21 (b) ;l.l 0 -0.2 -0.4 -0.6 9.7 9.8 9-9 10.0 10.1 10.2 PH Fig. 4. Transformation (V) for the hydrolysis reaction of Mn(o-P-PAN),: (a) excess of reagent; and (b) excess of metal. Conditions as in Fig. 3 Lalzthanum(III) chelates of o-a-PAN, o-P-PAN and o-PAP-The chelation reactions of lanthanum( 111) with o-a-PAN, o-P-PAN and 0-PAP were investigated and absorbance curves in the presence of both excess of reagent and excess of metal ions were obtained.Absorbance measurements were made at 550 and 514 nm, these wavelengths correspond- ing to the absorbance maxima of the o-a-PAN and o-PAP chelates, respectively. The absorbance curves are given in Fig. 5 and show that the extent of chelation is greatly affected by hydrolysis of the lanthanum species, particularly when the pH is above 7.5. The occurrence of a sharp hydrolysis effect above this pH may be expected because of the equilibrium position of lanthanum hydroxide, indicated by21 Log [La3+] = 23-02 - 3pH When the reaction of o-p-PAN with lanthanum(II1) was investigated] only a slight colour change occurred. As the formation of the o-p-PAN chelate may be expected to occur at lower acidities than those for the a-PAN chelate, this lack of reaction may be due to prefer- ential hydrolysis of the lanthanum(II1).Unlike the absorbance curves for the manganese( 11) and zinc(I1) chelates, those shown in Fig. 5 give no direct indication of stoicheiometry. However, as the absorbance values in the presence of both excess of reagent and excess of metal ions are similar, the formation of 1 : 1 chelates can be anticipated. The neutral reagent species, HR, predominates over the pH range of chelation and so the suggested chelation reaction is KMR La3++HR + LaR2++H+ The validity of this postulate was tested by using transformation (IX) (Fig. 6).404 BETTERIDGE AND JOHN : PYRIDYLAZONAPHTHOLS AND [Analyst, Vol. 98 Fig. 6. Comparison of experimental and theoretical absorbance curves for LaIIl(o-a-PAN) [upper graphs: (a) cxccss of metal; and (b) excess of reagent] and LaIII(o-PAY) [lower graphs: (c) excess of metal; and (d) excess of reagent]. Solid lines, experirncntal; and brokcn lines, theoretical.Concentrations : (u) CR 5.0 x M, CX 5.0 x M ; (b) CR 4.0 x 1 0 - 4 M, CM 4.0 x 10-5 M : ( c ) CR 5.0 x 10-6 M, CM 5.0 x 10-4 M ; (d) CR 5.0 x 1 0 - 4 M, CM 5.0 x 10-5 M The formation of such an LaR2+ species is in agreement with the observation of Sommer and Novotna,ll who have reported the reactions of lanthanum with PAR. These workers also found that a 1 : 1 chelate was formed and that severe hydrolysis interfered in the accurate determination of absorbances at higher pH. I l l I l l I l l 0 10 20 30 40 50 60 70 80 90 100 [HI2 x lo--* Fig. 6. Graphs of transformation (IX) in the presence of excess of lanthanum(II1): (a) o-a-PAN; and (b) o-PAP.CE 6.0 x M and CM 5-0 x MJune, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I1 405 Agreement between the constants listed in Table VII is not quite as good as that observed for the manganese(I1) and zinc(I1) chelates, probably because of the ease of hydrolysis of lanthanum species. In order to determine precisely the hydrolysis effect, and to verify further the proposed chelation reactions, theoretical absorbance curves were determined by using the computed complex formation constants and known values for the hydrolysis of lanthanum(II1). TABLE VII MOLAR ABSORPTIVITIES AND STABILITY CONSTANTS OF LANTHANUM(III) CHELATE Reagent Transformation EMR Transformation log K M R o-E-PAN IIIA 2.50 x 104 VIA 7.16 f 0.02 IX 2-47 x 104 XI 7-08 f 0.04 0-PAP IIIA 1.43 x 104 VIA 6-76 f 0-04 IX 1.56 x 104 XI 6-58 f 0.04 pKa, = 8.80 (0-PAP) ; 10.20 (o-a-PAN).(i) I n the pyesence of excess of Zanthanum(III)-In this instance, the absorbance is limited by the concentration of the reagent. The absorbance, A , is given by A = E M R C H R ~ ~ ~ (1 - f) - - .. .. . . (11) where cVR = molar absorptivity of MR; CHRTot = total reagent concentration; andf = fraction of reaction completed. C R - C R f=- - ‘HKTot ‘R ‘MI? .. .. .. .. . . (12) 1 - 1 + CMR/CR where CR = unreacted reagent concentration. The proposed chelation reaction is * KMR M+HR + MR+H+ .. .. . . . . (13) The conditional reaction constant, *KhR, is given by In the presence of excess of lanthanum(III), CM = CYTot, the total metal-ion concen- tration, and where the side-reaction coefficients, a1 and p, are [HR]/C, and [M]/C,, respectively.In the present study, a1 = 1 and p, if the major hydroxy-complex is La(OH),, is given by TLaI [La] + CLa(OH),I or The equilibrium constant involving hydrolysis of La3+ is given by21 La3+ + 3H,O + La(OH), + 3H+ Log [La3+] = 23.02 - 3pH Values of obtained from equation (15) were used in equation (14) to calculate Cn values. Absorbances were then obtained by using equations (1 1) and (12).406 BETTERIDGE AND JOHN: PYRIDYLAZONAPHTHOLS AND [Analyst, VOl. 98 ion concentration and are given by (ii) I n the $resence of excess of reagent-Absorbance values are now limited by the metal- .. .. .. .. . . (17) 1 f = 1 + cMR/cM *K&R = - c ~ ~ [ H l - *KMR,~~ CMCR In the presence of excess of reagent, CR = CHRTot, the total reagent concentration, and As before, a = 1 and is given by equation (15).Values of CM obtained from equation (18) were then used to calculate absorbance from equations (17) and (16). The molar absorptivity, reaction, stability and dissociation constants used for the above calculations are summarised in Table VII. Theoretical absorbance curves were drawn and compared with the experimental curves (Fig. 5). Agreement between the theoretical and experimental curves is good, which confirms that the chelation reaction suggested is correct and that the decreased absorbance at high pH is due to hydrolysis of lanthanum species. Reaction of 4-(2-pyridylazo)phenol with cobalt(II) , nickeZ(II) and copper(II)-Spot tests show that $-PAP is selective in its reactions with metal ions, and will react principally with metals of Groups VIII and IB. Cobalt(II), copper(I1) and nickel(I1) react strongly, and the chelation reactions of these ions have been investigated.Absorbance curves were obtained in the presence of both excess of reagent and excess of metal ions. Absorbance measurements were made at 548, 520 and 517 nm, these wavelengths corre- sponding to the absorbance maxima of the copper, nickel and cobalt chelates, respectively. The absorbance curves are shown in Fig. 7. These curves indicate that maximum chelation occurs at a pH greater than the pK,, value of the reagent. Also, very little chelation occurs at pH values below the pK,, value, which suggests that the neutral form of the reagent has little tendency for chelation and that chelation occurs via the anionic form.This represents an unusual type of chelation because the protons that are lost do not originate in the chelating groups themselves. Preliminary investigations of the nature of these chelates by continuous variation,l2 mole-ratio13 and s l o p e - r a t i ~ ~ ~ ~ ~ ~ methods showed that each chelate has a stoicheiometry of metal to ligand of 1 : 2. As the anionic form of the reagent is involved in chelation, then the suggested chelation reaction is M + 2R- + MR, Therefore, the application of transformations (VI) and (VII) was tried. The consistency of the calculated constants and linearity shown by the correlation coefficient in Table VIII indicated that this is the predominant reaction over the pH range 7 to 8.The decrease in absorbance observed at higher pH indicates the formation of hydroxy- chelate species : KOH MR2+OH- + MR,OH- This hydrolysis reaction is identical with that of the manganese(I1) and zinc(I1) chelates with o-p-PAN. The hydrolysis constant, KMR(OH), can therefore be calculated from identical transformations. The linearity of these transformations is shown in Fig. 8, and the hydrolysis constants were found to be 104.58 * O a o 2 and * Oeo2 for the nickel(I1) and cobalt(I1) hydroxy-species, respectively. SOLVENT-EXTRACTION STUDIES- The theory of the solvent extraction of chelates is fairly well established.24-27 If the distribution ratio, D, is defined as the concentration of the metal in the organic phase divided1.2 - 1.0 8 0-8 0.6 8 0-4 s .ff o a 0.2 6 7 8 9 1 0 6 7 8 9 1 0 1 PH Fig.7. Absorbance zlersus pH graphs for chelates of p-PAP in 5 per cent. aqueous methanol: (a) with copper(I1); (b) with nickel(I1); and ( c ) with cobalt(I1). A, excess of reagent: CM 2.5 x M, CR 1.0 X W 4 M ; B, excess of metal: Cx 1.0 x M, CR 2.5 X M ; c, excess of reagent: CM 2.5 X M, CR 1.0 x M ; D, excess of metal: CM 1.0 x M, CB 2.5 x M, CR 1.0 X lo-* M ; F, excess of metal: CM 1.0 X M ; E, excess of reagent: CM 5.0 X M, CR 5-0 X M TABLE VIII MOLAR ABSORPTIVITIES AND STABILITY CONSTANTS FOR COBALT( 11) AND NICKEL( 11) CHELATES Standard error of Condition Metal ion y = m x + c coefficient slope, e s t i m a t e rn E M R ~ x lo4 log KMR, Correlation Excess of metal ions Ni(1I) y = 0 .0 3 1 ~ + 3.820 0.994 0.00 1 1 0.088 5-23 8.95 f 0-08 W I I ) y = 0 . 1 9 8 ~ + 6-843 0.998 0.0044 0.205 2-92 7.88 f 0.07 Excess of reagent Ni(I1) y = 0.015~ $- 2.031 0-981 0.0009 0.1 46 4.92 8-69 f 0.10 Co(I1) y = 0 . 2 0 8 ~ + 3-887 0.999 0-0031 0.266 2.57 7-82 f 0.06 pKa, = 7.77.408 BETTERIDGE AND JOHN : PYRIDYLAZONAPHTHOLS AND [Analyst, VOl. 98 by the concentration of the metal in the aqueous phase, it follows, for the extraction reaction that MC+ + 2HRo + MR2(,, + 2H$ D = [MR210/([MR2]W + cM) = KDX/(l + cM/[MR21W) = KDX/(l + [Ml/p[MR21) = KDX/{l + l/(pKMR,a&D)[HRlz)) .. . . (19) In equation (19) and subsequent equations K,, and KDR are the partition coefficients of the chelate and reagent, respectively.%(D) = [R-l/c€IR = KalKa2/ { [HI2 + Kal [HI (1 + K ~ ~ v o / v w ) + Ka,Ka, 1 where CHR is the total concentration of reagent present initially in the organic phase of volume Vo and the other terms have been defined previously. It is assumed that the metal is present in the organic phase only as MR,. In the absence of side-reactions that involve the metal ion (/3 = 1) and incomplete chelate formation (D < KDx) and on the assumption that the reagent is present in the organic phase mainly as HR, this expression reduces to = K1\IR~KDXK~~[HRl~/I(~RIHla .. .. . . (20) For practical purposes, the percentage extracted, E , is often used, and is related to the distribution ratio by The pH value at E = 50 per cent. is designated as pH&, and the difference between the pH, values of two extractable chelates is a measure of their separability.If it is assumed that the partition coefficients for chelates of the same reagent are identical and there are no competing side-reactions, extraction curves (E veisus pH) for the same reagent and a series of metal ions of the same charge will have identical shapes. The relative degree of extraction at a given pH will be governed by the relative values of the formation constant ISMn,. Hence, if the values of the constants in equation (20) are known and a value of [HRIo is defined, curves of E v e w m pH can readily be calculated. PO H Fig. 8. Graphs of hydrolysis equations for cobalt(I1) and nickel(I1) chelates of p-PAP: A, cobalt(I1); and B, nickel(I1) The values of all the terms in equation (20) except KDR and KDx can be determined by the spectrophotometric methods described above.K D R must be determined independently. Therefore, in order to calculate the extraction curve, one must either assume a value of KDx or determine it independently. Two approaches seem reasonable: (i) assume that KDx = K,,, which was applied to o-PAP systems; (ii) assume that KDx is the same for similar systems, which was applied to o-or-PAN and to o-p-PAN systems, the value of lo4 being taken from previous work.14-18 The calculated curves were compared with those obtainedJune, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I1 409 experimentally (Fig. 9). As would be expected from the above assumptions and the additional assumption that KMRI remains constant, not all of the experimental and theoretical curves coincide. Nevertheless, they are close enough for the analyst to use the method as a guide in the choice of his experimental conditions.Fig. 9. Comparison of experimental and predicted extraction curves for the chelates of zinc and manganese: A, zinc; and B, man- ganese. Solid lines, experimental ; and broken lines, calculated. Chelates: 0, o-a-PAN; 0 , o-S-PAN; and x , o-PAP A more valuable guide is provided when there are competing reactions, /3 < 1. The simplified equation (20) must be modified by including /3 in the numerator. The competing reaction might arise from the addition of a masking agent, L, which forms a series of com- plexes with the metal ion. If the formation constants are known, it is a straightforward matter to calculate 1/p, which is given by 1 + X k , k , .. . kn[LIn. The spectrophotometric studies should reveal the presence of hydroxy-species, which can be taken into account in a similar way: n 0 l/P = 1 + KMR,(OH) [OH] + KMR,(oH), [OH], The extraction curve for the manganese(I1) - o-P-PAN system, in which the presence of hydroxy-complexes had been established spectrophotometrically, was calculated with the aid of the formation constants determined spectrophotometrically. The curve is compared with the experimentally obtained curve in Fig. 10. The agreement is satisfactory and clearly PH Fig. 10. Comparison of experimental and predicted extraction curves for Mn(o- ,%PAN),. Solid line, experimental ; and broken line, predicted410 BETTERIDGE AND JOHN: PYRIDYLAZONAPHTHOLS AND [Analyst, vOl98 demonstrates the need to select the extraction conditions with care.For more exact com- parison, the calculated and experimental pH values are given in Table IX. TABLE IX COMPARISON OF EXPERIMENTAL SOLVENT-EXTRACTION CONSTANTS WITH CONSTANTS PREDICTED BY SPECTROPHOTOMETRY Experimental - Reagent Metalion LogKf pH,,, o-a-PAN Mn(I1) 13.30 7.30 Zn(I1) 19-60 4.10 O- B-PAN Mn(I1) 16-13 7.86 Zn(I1) 22-17 4.86 O-PAP Mn(I1) 10.42 7.60 Zn(I1) 16.68 6.00 Predicted - 13.40 7-22 19.13 4.36 16.80 7-96 21.60 6.10 10.60 7-66 16.60 6-07 Log Kf PHI/, pKa, = 9.63 (o-x-PAN); 11.62 (O-p-PAN) ; 8-88 (O-PAP). KDx = lo4, assumed for each chelate. It is more usual to deduce the values of the constants in equations (19) and (20) from the experimental curve, and this procedure was followed for the manganese(I1) - o-P-PAN system so as to determine the values of KMR,(OH) and KMR,(OH),.The value of KDXKMR, was calculated from the points on the rising part of the curve of log D veysus pH and a value of 18 was calculated as a function of pH from equation (19) by using points from the descending part of the curve. fl was then examined as a function of [OH] and it was deduced that MnR2(OH),2- could not be detected, that MnR2(OH) was present and that log KMnR,(OH) had a value of 7.64. The agreement between this value and that obtained spectrophotometrically is good, and is a little surprising because of the low number of acceptable experimental points in the solven t-extracti on sys tern.CONCLUSIONS The determination of reliable stability constants is a tiresome and time-consuming procedure, which can often produce results that are of limited applicability. The most accurate results are generally agreed to be those obtained by potentiometric titrations, and these results now have the advantage of the availability of well tested computer programs28 that can be used in order to transmute the experimental points into results. However, inevitably, relatively large concentrations of reactants are used, which results in the use of non-aqueous solvents and the increased likelihood of the formation of polynuclear species. Methods based on solvent extraction and spectrophotometry have direct appeal to the analytical chemist as he may be able to set up his analytical method while measuring the stability constants.This is true of solvent extraction, although it may be difficult to obtain reliable constants without a great deal of experimentation and it is also true of the spectro- photometric method described above. The pH - absorbance curves upon which the method is based are essential in the development of the analytical procedure. These curves can also yield much information about the effect of varying the conditions, e.g., reagent concentration, pH and masking agent. In this respect, as well as in terms of accuracy, the method involving these curves is far superior to Job’s method and the related procedures that are commonly used. It has also been shown that the results can be used directly to predict solvent-extraction curves.The weakness of the method is that it depends upon extrapolation, and it requires careful experimental work and alteration of conditions so as to ensure that reliable results are obtained. We have concluded that it is just as time consuming as the other reliable methods, but we feel that because of the amount of information of direct analytical interest that is obtained, it is a very desirable method. Of the four reagents studied, o-a-PAN is clearly the most suitable as it forms highly coloured stable complexes at lower pH values than o-fl-PAN, o-PAP or @-PAP. o-fl-PAN and o-PAP form complexes at about the same pH values but the molar absorptivities of PAN complexes are greater than those of PAP complexes. Specific analytical applications will be described in subsequent papers.June, 19731 PYRIDYLAZOPHENOLS AS ANALYTICAL REAGENTS. PART I1 41 1 to one 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 28. 26. 27. 28. We are grateful to the T. and E. Williams Scholarship Fund for a maintenance grant of us (D.J.). REFERENCES Anderson, R. G., and Nickless, G., Analyst, 1967, 92, 207. Shibata, S., Analytica Chim. Acta, 1960, 23, 367. Rosotti, F. J. C., and Rosotti, H., “The Determination of Stability Constants,” McGraw-Hill, Sommer, L., Folia Fac. Sci. Natn. Univ. Purkynianae Bmo, 1964, 5, Part 1, 1. Hnilickova, M., and Sommer, L., Talanta, 1966, 13, 667. Chromy, V., and Sommer, L., Ibid., 1967, 14, 393. Sommer, L., and Hnilickova, M., Ibid., 1969, 16, 83. Sommer, L., and Ivanov, V. M., Ibid., 1967, 14, 171. Sommer, L., Ivanov, V. M., and Novotna, H., Ibid., 1967, 14, 329. Sommer, L., and Novotna, H., Ibid,, 1967, 14, 457. Vosburgh, W. C., and Cooper, G. R., J . Amer. Chem. SOL, 1941, 63, 437. Yoe, J. H., and Jones, A. L., Ind. E n g g Chem., Analyt. Edn, 1964, 16, 11. Betteridge, D., Todd, P. K., Fernando, Q., and Freiser, H., Analyt. Chem., 1963, 35, 729. Pease, €5. F., and Williams, M. B., Ibid., 1959, 31, 1044. Corsini, A., Mai-Ling Yih, I., Fernando, Q., and Freiser, H., Ibid., 1962, 34, 1090. Nakagawa, G., and Wada, H., J. Chem. SOC. Jafian, 1963, 84, 639; Chem. Abstr., 1964, 61, 1242h. Betteridge, D., Fernando, Q., and Freiser, H., Analyt. Chem., 1963, 35, 294. Anderson, R. G., and Nickless, G., Analyst, 1968, 93, 13. Pourbaix, M., “Atlas of Electrochemical Equilibria in Aqueous Solutions,” Pergamon Press, Harvey, A. E., and Manning, D. L., J . Amer. Chem. SOC., 1950, 72, 4488. Morrison, G. H., and Freiser, H. F., “Solvent Extraction in Analytical Chemistry,” John Wiley, Starf, J., “The Solvent Extraction of Metal Chelates,” Pergamon Press, Oxford, 1964. Laitinen, H. A. , “Chemical Analysis,” McGraw-Hill, New York, 1960. Marcus, Y., and Kertes, A. S., “Ion-exchange and Solvent Extraction of Metal Complexes,” Childs, C. W., Hallman, P. S., and Perrin, D. D., Talanta, 1969, 16, 1119. - Ibid., 1961, 25, 348. New York, 1961, pp. 47-51. 9 , Analytica Chim. Acta, 1967, 39, 469. -- Oxford, 1966. t , Ibid., 1952, 74, 4744. -- New York, 1962. Wiley-Interscience, New York, 1969. Received April 2nd, 1970 Accepted November 22nd, 1972

ISSN:0003-2654    

DOI:10.1039/AN9739800390

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

412 Analyst, June, 1973, Vol. 98, $9. 412415 Determination of Chloride in Aqueous Soil Extracts and Water Samples by Means of a Chloride-selective Electrode BY A. R. SELMER-OLSEN (Chemical Research Laboratory, Agricultural University of Norway, 1432 AS-NLH, Norway) AND A. 0IEN (Institute of Soil Science, Agricultural University of Norway, 1432 As-NLH, Norway) The chloride contents of water samples and soil extracts have been determined with an Orion chloride-selective electrode. The chloride content, at levels up to 100 mg l-l, is determined in a solution that is 0.5 M with respect to ammonium nitrate and 0.03 M with respect to nitric acid. Comparison of this method with a colorimetric AutoAnalyzer method showed no significant difference between the results obtained, but for water samples low in chloride (less than 3 mgl-l) the colorimetric method was more accurate.CURRENTLY, the most widely used methods for the determination of chlorides are the gravimetric method (precipitation as silver chloride) and the argentimetric and mercurimetric methods (in which different types of coloured or potentiometric indicators are used). Bremnerl recommends the well known Mohr titrimetric method for the determination of chloride in aqueous soil extracts. This method, however, is not very sensitive, and is also subject to errors caused by adsorption effects and over-titration. Davey and Bembrickz used a silver - silver chloride electrode for the determination of chloride in aqueous extracts of soil. A potentiometric determination is much more sensitive than Mohr’s method, and is therefore more suitable for determining small concentrations of chloride in soil extracts and water samples.Technicon3 recommend a colorimetric method based on the release of thiocyanate ion from mercury(I1) thiocyanate by an equivalent amount of chloride. A red colour is formed by the reaction of thiocyanate with iron(II1). This method is very sensitive and is well suited for large numbers of samples. The Orion chloride-selective electrode appears to be convenient for the determination of chloride in soil extracts and water samples as it is more sensitive than Mohr’s method and is portable, thus enabling measurements to be made in the field. In this investigation, different types of soil and water samples have been analysed by this electrode method, and the results are compared with those obtained by the Technicon AutoAnalyzer method.EXPERIMENTAL APPARATUS- An ion-specific meter, Orion Research, Model 401, with a chloride-selective electrode, Model 94-17, was used for the potentiometric determination of chloride in water and soil extracts. The measurements were made against the Orion double-junction reference electrode, Model 90-02. A Technicon AutoAnalyzer, with a manifold system according to O’Brien,* was used for the colorimetric determination of chloride in water and soil extracts. Its use is becoming more widespread. REAGENTS- All reagents used were of analytical-reagent grade. Standard solutions of chloride-A stock solution of chloride was prepared by dissolving 2.103 g of potassium chloride in water and diluting the solution to 1 litre so as to give a 0 SAC and the authors.SELMER-OLSEN AND 0 I E N 413 concentration of 1 g 1-1 of chloride.Standard solutions in the range 0 to 100 mg 1-1 of chloride were made by diluting appropriate volumes of the stock solution with ammonium nitrate and nitric acid solutions. The final concentrations of ammonium nitrate and nitric acid were 0.5 and 0.03 M, respectively. Ammonium nitrate solution, 0.5 M-Ammonium nitrate (40 g) was dissolved in distilled water and the solution was made up to 1 litre. Nitric acid, 0.3 M-concentrated nitric acid was diluted appropriately with distilled water. Ammonium nitrate - nitric acid reagent-A solution was prepared that was 5 M in am- monium nitrate and 0.3 M in nitric acid.PROCEDURE- SoiZ samples-A 10-g amount of air-dried soil (dried a t 30 "C and passed through a 2-mm sieve) was shaken for 15 minutes with 50 ml of 0-5 M ammonium nitrate solution. The extracts were either centrifuged, or filtered through Schleicher and Schull No. 589 white ribbon filters; 2 ml of 0-3 M nitric acid were added to 20 ml of the extract before measurement with the electrode. The measured values were then compared with those given by the standard graph. Water samples-Ammonium nitrate - nitric acid reagent (2 ml) was added to 20 ml of water and the chloride content was measured with the electrode. EXTRACTION TIME- When four soils of various textures were shaken with water or 0.5 M ammonium nitrate solution for periods of 5, 30 and 60 minutes, the chloride values obtained after shaking the mixtures for 5 minutes were identical with those obtained after the longer periods of shaking.Results were the same with both extractants, provided that 2 ml of 0-3 M nitric acid per 20 ml of extract were added to the 0.5 M ammonium nitrate extracts and 2 ml of ammonium nitrate - nitric acid reagent to the aqueous extracts. These results seem reasonable as the chloride ions are only lightly bound to the negative soil particles. EFFECT OF AMMONIUM NITRATE AND NITRIC ACID- According to Orion5 the presence of ammonia can cause some interference. To overcome this effect the soil was extracted with 0.5 M ammonium nitrate solution. At this high con- centration of ammonium ions, the relatively small amounts of ammonia present in the soil will cause no significant errors.Hydroxyl ions may also interfere, and concentrations greater than eighty times the chloride concentration cause errors5 Varying the pH of the 0.5 M ammonium nitrate solution (containing 10 mg 1-1 of chloride) by addition of nitric acid showed that the same results for chloride were obtained in solutions with pH values ranging from 1.6 to 5.0. The addition of 2 ml of 0.3 M nitric acid, which should allow for any carbonate extracted from calcareous soils, is advised. Davey and Bembrick2 also found that a mixture of ammonium nitrate and nitric acid increased the accuracy of the method. We used a higher level of acidity, however. RECOVERY TESTS- The recovery of known amounts of chloride, added before extraction to four soils of various textures, ranged from 99-5 to 102 per cent.The chloride content found, expressed as milligrams of chloride per 100 g of soil, was constant, at least in the range 5 to 40 g of soil per 50 ml of extracting solution. COMPARISON OF METHODS- A series of soil extracts was prepared, and the chloride contents were determined by both the electrode and the colorimetric AutoAnalyzer methods. If the extracts were highly coloured, blank values were obtained by running the samples with water instead of mercury(I1) thiocyanate reagent in the AutoAnalyzer method. Table I shows the results for thirteen soil samples treated as follows. A 10-g sample of air-dried soil was shaken with 50 ml of 0-5 M ammonium nitrate solution for 30 minutes and the mixture then centrifuged.The chloride content in the soil extract was determined by both the electrode and the Auto- Analyzer methods after nitric acid had been added to the extracts. A t-test gave no indication RESULTS AND DISCUSSION The ionic strength will also be nearly constant.414 SELMER-OLSEN AND 01EN : DETERMINATION OF CHLORIDE I N SOIL [A?ZU&St, VOl. 98 TABLE I DETERMINATION OF CHLORIDE IN SOIL EXTRACTS Chloride content in soil extract/ mg 1-1,* found by- Soil sample ignition, electrode Auto Analyzer No. Soil type per cent. PHmo method method 2 Sand 5.9 7.5 0.7 1.8 3 Sand 7.7 7-3 0.6 0.6 4 Sandy peat soil 75.3 4.6 11.0 13.0 5 Sand 5-7 5.5 1.9 0-6 6 Sand 5.9 7.5 1.4 1.2 7 Clayey sand 10.4 7.3 0.9 0.9 8 Sandy peat soil 38.0 4.6 10.0 9.4 10 Sandy clay 7.7 6-0 1.4 1.0 11 Sandy clay 8.0 5.8 1.0 0.9 Loss on - 1 Sandy peat soil 48-8 5.9 8.5 9.4 9 Sandy peat soil 32-7 4.9 5.1 5.2 12 Sandy clay 7.8 5.9 1.6 1.4 13 Sandy peat soil 26.6 7.0 7.9 8.5 The corresponding chloride contents in soil are five times greater and range from 3.0 to 65 p.p.m.of difference (P > 0.5) between the results for the electrode and the colorimetric methods. Duplicate analyses of twenty-one water samples in the range 0.5 to 18.8p.p.m., to 30ml of each of which had been added 3 ml of a solution that was 5 M and 0.3 M with respect to ammonium nitrate and nitric acid, respectively, showed no indication of difference between the results for the two methods (Table 11). TABLE I1 DETERMINATION OF CHLORIDE IN WATER SAMPLES Sample No. 1 2 3 4 6 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 PH 6.0 6.3 6.5 6-4 6.2 6.9 6-2 6.4 6.3 6.0 5.3 4.0 6.2 6.5 6.0 5.5 7-0 5.1 7.0 4.6 5.8 Specific conductance/ $3 cm-l 40 42 83 95 75 63 191 130 35 30 84 87 39 28 30 19 25 23 23 30 54 Chloride content/mg l-l, found by- I A \ electrode method AutoAnalyzer method Duplicate results Mean Duplicate results Mean & 1.9,34 2-5 2-4,2-3 2-4 1.2,2*6 1.9 1.8.1-9 1.9 8.1,8.5 8.3 7*6,7*6 7.6 10*3,10*3 10.3 9-6,9-5 9.6 8*5,8*9 8.7 6*9,6-9 6.9 5*4,5.4 5.4 5.9,5.8 5.9 17-0,18.8 17-9 16*8,16*8 16.8 11~1,12.0 11.6 10-7,10.7 10.7 2*6,2*3 2.5 1*9,1-8 1-9 0*5,1.9 1.2 1.6,1-5 1.6 0*7,1.7 1.2 1.3,1.3 1.3 16-0,16.2 16.1 17-5,17*5 17.5 1-2,2-7 2.0 2-3,2*3 2.3 2*4,2+4 2.4 2.4,2*4 2.4 2.8,2*8 2.8 2*5,2*5 2.5 2*0,1*7 1.9 1*4,1.4 1.4 2*6,2-2 2.4 1.9,2-0 2.0 1*4,1-1 1.3 1-6,1*6 1.6 1*2,1-9 1.6 1.6,l.g 1.8 1*6,2-2 1.9 2*7,2*8 2.8 4.8.4-9 4.9 4-7,4.7 4.7 SENSITIVITY AND PRECISION OF THE METHOD- As the slope of the graph of millivolts versus concentration is not constant below a concentration of 10 p.p.m.(Fig. l), the chloride content cannot be read directly from the meter. The chloride concentration can then be measured down to 0.5 p.p.m. At this level, however, the reproducibility is poor. In duplicate analyses of thirteen soil extracts in the range A calibration graph must therefore be drawn for the lower concentrations.June, 19731 EXTRACTS AND WATER BY MEANS OF A CHLORIDE-SELECTIVE ELECTRODE 415 0.6 to 11 mg 1-1 of chloride, the coefficient of variation was 4.5 per cent. For thirteen water samples with a relatively low content of chloride (0.5 to 3 mg F), the coefficient of variation was as high as 31 per cent. For higher concentrations (443 to 18.8 mg 1-1) it was found to be about 5 per cent. I I I 1 I 1 1 0 5 10 15 2c Chloride concentration, p.p.m. Fig. I. Standard graph The results indicate that for low chloride concentrations (below 3 mg 1-1 of extract) it is better to use a colorimetric AutoAnalyzer method than the electrode method if accurate measurements are needed. Above this concentration, the chloride electrode seems to be satisfactory. The chloride content in the range 10 to 100mg1-1 can be determined by direct read-out on the instrument. REFERENCES 1. 2. 3. 4. 5 . Bremner, J. M., in Black, C. A., Editor, “Methods of Soil Analysis, Part 2, Agronomy, 9,” American Davey, B. G., and Bembrick, M. J., Proc. Soil Sci. Soc. Amer., 1969, 33, 386. Technicon AutoAnalyzer Methodology, Chloride IIa Method, Technicon Controls Inc., Chauncey, O’Brien, J. E., Wastes Engng, 1962, 670. Orion Instruction Manual for Halide Electrodes, Orion Research Incorporated, Cambridge, Mass., Received October 30th, 1972 Accepted January 9tk 1973 Society of Agronomy, Madison, Wisc., 1965, p. 947. N.Y., 1960. 1967,

ISSN:0003-2654    

DOI:10.1039/AN9739800412

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

416 Analyst, June, 1973, Vol. 98, #p. 416-425 Precise Coulometric Determination of Acids in Cells Without Liquid Junction Part 111.” Determination of the Silver Error by Amperostatic Anodic Strippingt BY E. BISHOP AND M. RILEY1 (Chemistry Department, University of Exeler, Stocker Road, Exeter, EX4 4QD) The plating and stripping of silver on platinum electrodes have been examined in the context of the determination of the silver error, which arises from the solubility of silver bromide when the deposition of bromide on a silver anode is used as the auxiliary reaction in the coulometric assay of acids. In order to calibrate the stripping method, it is necessary to plate known amounts of silver quantitatively on to platinum-gauze electrodes. Low recoveries are obtained when the platinum is not fully reduced.The oxidation and reduction of platinum electrode surfaces have been briefly examined, and it is demonstrated that oxide forms on an electrode when its potential is allowed to rise beyond 0.8 V, the termination potential in silver stripping. For calibration purposes, plating and stripping in a 0.1 M solution of silver nitrate in 0.1 M perchloric acid was first investigated. Amperostatic and potentiostatic reduction of the platinum electrode are shown to be ineffective, but chemical reduction leads to excellent plating and recoveries, provided great care is taken completely to remove all traces of reductant. Calibration being satisfactory, stripping in 0.1 M perchloric acid, as in an actual acidimetric assay, has been examined and shown to give excellent recoveries.The anodic stripping curves show an extended second wave, which is identified as arising from reduction of oxygen to hydrogen peroxide at the auxiliary stripping electrode, particularly when the latter becomes plated with silver. The hydrogen peroxide is oxidised at the stripping electrode, and the process is cyclic. IN the coulometric determination of acids, the use of a silver auxiliary anode on which bromide is deposited has been canvassed.1 The slight, but significant, solubility of silver bromide in the electrolyte gives rise to a “silver error” by deposition of silver on the working platinum cathode either by direct electro-reduction or by reduction of silver ion by hydrogen atoms in the compact layer (the combination of which, to give hydrogen molecules, is the rate-determining step of the main cathodic reaction). Additionally, any precipitate of silver bromide in the bulk of the solution is liable to be caught on the cathode and there reduced, although the experimental conditions are chosen so as to avoid the formation of such precipitates. The error incurred is about 1 to 2 C in a total of 5000 C, and a rapid and convenient method is needed for the determination of about 1 mg of silver deposited on a 125-cm2 platinum-gauze electrode.If this amount can be determined with an accuracy of 1 per cent., it will then represent an over-all accuracy of about 2 to 4 p.p.m. in the deter- mination of 0.05 mol of a monobasic acid. Amperostatic anodic stripping proved simple and rapid,l but is of unknown accuracy.In order to assess the accuracy a method must be discovered for quantitatively plating known amounts of silver on to platinum for calibration of the stripping process. Preliminary experiments showed that while the potential rise at completion of the stripping reaction was satisfactorily sharp, the accuracy was poor, recoveries of silver being only 70 per cent. This finding led to the investigation of plating of silver on to platinum, of oxidation of platinum surfaces, and therefore of the pre-conditioning of the platinum electrode. Mechanical stripping, i.e., loss of some of the loose deposit, was suggested by Lord, O’Neill and Rogers2 as the cause of low recoveries. Nisbet and Bard3 obtained 100 per cent. recovery on a platinised platinum electrode, but low recoveries on an oxidised * For particulars of Parts I and I1 of this series, see reference list, p.425. t Presented a t the Second SAC Conference, Nottingham, 1968. 0 SAC and the authors. For Part IV, see p. 426. Present address : Electronic Instruments Limited, Hanworth Lane, Chertsey, Surrey.BISHOP AND RILEY 41 7 electrode. A platinised surface was therefore prepared by a.c. electrolysis in 1 M perchloric acid, and the electrode was finally reduced almost until hydrogen was evolved. Silver was plated on to and then stripped from this electrode amperostatically, and the plating - stripping cycle was repeated several times with anodisation to + l o 6 V each time; progressively shorter stripping times were found for the same plating time.Nisbet and Bard3 reported similar results, which were claimed by them to show that some of the silver is retained on an oxidised platinum surface and can be removed only after reduction of the underlying oxide. Bixler and Bruckenstein4 claimed 100 per cent. recovery when a reduced platinum electrode was used. They confirmed that repetition of the plating - stripping cycle with anodisation to +1-6 V gave short recovery times, but could find no evidence of retention of silver on the electrode. They suggested that the apparent loss of silver arose from partial reduction of the oxidised platinum surface during plating, together with mechanical loss. EXPERIMENTAL The apparatus and reagents used have been described previously.G The cell and circuit used in the present study are shown in Fig.1. The milliammeter shown in series with the working electrodes was used to set the current to the nearest 10pA, but in the later work was replaced with a 10-l2 standard four-terminal resistor and the current set to the nearest 1 pA by using the P3 potentiometer. Titanium(1V) sulphate was prepared by heating 10 g of titanium dioxide with 20 ml of concentrated sulphuric acid for 30 minutes, cooling, diluting the solution with 50 ml of cold water, allowing it to stand for several days and filtering it through a Whatman No. 42 filter-paper. All electrode potentials are given with reference to the standard hydrogen electrode (S.H.E.). Clock r"l I OACS 1- P-3 T I ( t Reversing switch Plat i n urn-wire Fig. 1. Coulometric cell and circuit for the study of anodic stripping of silver from platinum gauze (OACS = operational ampli- fier constant-current source6) RESULTS AND COMMENTS Preliminary plating experiments on untreated platinum-gauze electrodes gave well defined stripping waves, but recoveries were between 50 and 85 per cent.A second, poorly defined wave at +1.0 V is ascribed to oxidation of the platinum. The gauze electrode was reduced at 1 mA for 100 s in nitrogen-purged 0.1 M perchloric acid, and was allowed to stand in this solution for about 10 minutes until its potential became stabilised, before being418 BISHOP AND RILEY : PRECISE COULOMETRIC DETERMINATION [ArtdySt, VOl. 98 transferred to the plating - stripping medium consisting of a 0.1 M solution of silver nitrate in 0.1 M perchloric acid.Two cycles of plating and stripping are shown in Fig. 2. The amount of silver plated in each cycle was constant at 400 mC and the respective recoveries were 87 and 73 per cent. The beginning of the second stripping wave can be seen in each instance and, furthermore, in the second plating step the potential takes some time to reach the steady value indicative of silver deposition, which is taken to indicate at least partial reduction of the oxidised surface produced at the tail of the first stripping step. Decrease in stripping time on repetitive cycling has been observed before.8~~ Clearly, oxide formation is involved and requires investigation. Potential - Potential versus S.H.E. - +- 0-6 V i Start - +1*ov \ - + 1.4 v Fig. 2. Consecutive silver plating and strip- Fig.3. Three cycles of oxidation and reduction ping cycles in a 0.1 M solution of silver nitrate in of a 126-cma platinum-gauze electrode at 2 mA in 0.1 M perchloric acid. The short vertical lines oxygen-free 0.1 M perchloric acid indicate points where the current was switched on or off or reversed Oxidation and reduction of platinum after the amperostatic cathodic pre-treatment was examined in nitrogen-purged 0.1 M perchloric acid at total currents of 0.5 to 2 mA. Fig. 3 shows the behaviour of an electrode, previously reduced at 2 mA for 200 s, during three cycles a t 2 mA between 1.4 and 0.5 V. Electrode oxidation and reduction start at +1.0 and + 0.9 V, respectively, although the former is ill defined. The time taken for the potential to rise from 0.5 to 1.4 V decreases with continued cycling although the cathodic step remains essentially constant a t 45 s.Similar behaviour was observed by Feldberg, Enke and BrickerJs who found that the ratio of anodisation to cathodisation times started at 2 : 1 and fell eventu- ally to 1 : 1. In the present work, the initial ratio was about 2: 1 but, after six or seven cycles, became constant at 1.2: 1. Anodisation to potentials greater or less than 1.4 V produced longer or shorter cathodisation times, respectively, and the quantity of electricity required for cathodisation was virtually independent of the current ; for example, halving the current produced an increase in cathodisation time by a factor of about 2.2. A mean value was found for the reduction over a variety of conditions of 0.74 mC cm-2, assuming an electrode area of 125 cm2. Kolthoff and Tanaka' reported a value of 0.92 mC cm-2 for anodic or chemical oxidation.Several chronopotentiometric determinations of oxide have been reported and are summarised in Table I. TABLE I CHRONOPOTENTIOMETRIC OXIDATION AND REDUCTION OF PLATINUM IN OXYGEN-FREE 1.0 M SULPHURIC ACID Transition Electrode Current Quantity of Reference time measured area/cm8 density/mA cm-a electricity/mC cm-8 8 Anodic 2 70-300 0.8 9 Cathodic 6 100 0.98 10 Anodic 0.25 60 0.94 11 Cathodic 0.33 180 0.6 The current densities were much higher than the 8 pA cm-2 used here, but the coverage E. Bishop and B. Cooksey (unpublished work) have values are all of similar magnitude.June, 19731 OF ACIDS I N CELLS WITHOUT LIQUID JUNCTION.PART I11 419 shown that at potentials above 1.0 V the formation of molecular oxygen is in competition with oxide film formation, and attempted to develop growth laws. The oxidation and reduction of platinum are dealt with in detail elsewhere12Js; for the present purpose, it is clear that pre-conditioning of the electrode is necessary for quantitative plating of silver, and that oxidation of the electrode surface occurs when stripping is taken beyond 0.8 V. PLATING AND STRIPPING OF SILVER IN SILVER NITRATE SOLUTIONS- For the efficient quantitative plating of silver, a completely reduced platinum surface is clearly necessary. Three basic types of conditioning, amperostatic, potentiostatic and chemical, have been examined for this purpose.Plating and stripping were carried out in 0.1 M solutions of silver nitrate in 0.1 M perchloric acid. Between experiments, the platinum auxiliary electrode was always cleaned with 1 + 1 nitric acid so as to remove any silver deposited on it during the stripping step, which normally followed the plating step by reversal of the current. A mperostatic cathodic reduction-The pre-treatment was performed in de-oxygenated 0.1 M perchloric acid at 2 to 500 mA. Before reduction, the platinum-gauze electrode was always in an oxidised condition that resulted from a previous anodic stripping run, from prior anodic oxidation to a potential higher than 1.4 V in de-oxygenated 0.1 M perchloric acid, or from immersion, together with the auxiliary electrode, in 1 + 1 nitric acid during the cleaning of the latter.The reduction was followed by treatment to remove any hydrogen remaining on the electrode surface. This treatment was carried out in the same solution and comprised either (a) allowing the electrode to stand for 2 to 20 minutes in the stirred pre-treatment solution, or (b) short-circuiting it to a S.C.E. , connected via an intervening potassium sulphate salt bridge, for 5 to 50 minutes. The electrodes were then rapidly transferred to the silver solution. When the cathodic pre-treatment was of short duration, so that the electrode potential did not become low enough for the evolution of hydrogen, the stripping curves were of the same shape as in Fig. 2, but the over-all recovery was only 85 per cent. When the pre-treatment was more drastic and involved the evolution of hydrogen, and the electrodes were given treatment (a) above, then the stripping curves often showed an additional step just prior to the potential rise, which indicated completion of stripping.Measurement of the stripping time up to a potential of +Om8 V gave recoveries greater than 100 per cent. Fig. 4 shows two such curves at a stripping current of 2 mA for an electrode pre- treated by oxidation and reduction at 2 mA. Curve A was obtained with an electrode re- duced to -0.048 V and allowed to stand for 3 minutes, and curve B with an electrode reduced to -0.072 V and allowed to stand for 2 minutes; the recoveries were 111 and 119 per cent., respectively. Compared with the normal stripping step as shown in Fig. 2, the distortion caused by oxidation of residual hydrogen is apparent, and the hydrogen is evidently underneath the silver plate.Reproducibility was poor, but, in general, the longer the time the electrode was left after pre-treatment the lower was the recovery of silver and the more Potential versus S.H.E. -+ -+ -+ Fig. 4. Distortion of anodic stripping curves due t o the presence of residual hydrogen on the electrode after cathodic pre-treatment and under the silver plate420 BISHOP AND RILEY : PRECISE COULOMETRIC DETERMINATION [Analyst, Vol. 98 closely the shape of the curve approached that in Fig. 2. Recoveries of 95 to 105 per cent. were obtained with electrodes left for 20 minutes and the stripping curve still showed a slower potential rise than was normal at the completion of stripping.Short-circuiting the electrode to the S.C.E. appeared effectively to remove the hydrogen and gave normal stripping curves, but the recovery averaged only 90 per cent. A similar recovery was obtained after using Bixler and Bruckenstein’s method of pre-treatment A.* De-oxygenation of the solution for plating and stripping had no detectable influence on the shape of the plating curve or on the recovery of silver. Potentiostatic cathodic reduction-Potentiostatic reductions were performed at command potentials of +0.6, +0.5 and +0-25 V in nitrogen-purged 0.1 M perchloric acid until the current became essentially constant. The platinum gauze was always in an oxidised form before treatment, as in the amperostatic method. The electrodes were then rapidly trans- ferred into the silver nitrate solution.Currents in the reduction started at 1 to 4 mA and decreased rapidly, to one tenth of the initial value in 10 minutes, becoming constant at the two higher command potentials after 30 minutes and decreasing only slowly at +0-25 V; this condition produced the largest residual current, of the order of 0.1 mA. Reduction times from 35 minutes to 3 hours were used, but no significant decrease in current occurred after the first 60 minutes. Silver recoveries as high as 95 per cent. were attained, but the mean re- covery in fourteen runs was only 90 per cent., and there was no appreciable difference between the effects of the three pre-treatment potentials on recoveries. De-oxygenation ofthe plating - stripping solution was without influence.Chemical reduction-Reduction of oxidised platinum by immersing it in a solution of a reductant has been carried out several times. Kolthoff and Tanaka7 used a 5-minute immersion in a 0.01 M solution of iron(I1) sulphate in 0.05 M sulphuric acid, but found that a 0.01 M solution of arsenic(II1) in 1 M hydrochloric acid had no effect in 60 minutes. Ross and Shain14 used a 10-minute immersion in 0.1 M iron(I1) perchlorate solution, Anson and Linganes found an acidic iodide solution to be effective and Anson15 used a 1-minute immersion in a 0.2 M solution of iron(I1) sulphate in 1.0 M sulphuric acid. Trials with immersion of 2 to 10 minutes in 0-2 to 0.5 M solutions of iron(I1) sulphate in 1.0 M sulphuric acid gave recoveries of silver between 90 and 99 per cent.Recoveries improved with increasing im- mersion times, and it became clear that a period of over 15 minutes was required for the platinum-gauze electrode to be reliably reduced a€ ter anodic oxidation or immersion in 1 + 1 nitric acid, and even longer times after cleaning in aqua regia. I t also became clear that the potential displayed by the electrode when immersed in the plating solution gave a good indication of the efficiency of pre-treatments. A fully reduced electrode took up a potential of about 0.75 V, while incompletely reduced electrodes showed potentials between 0.85 and 0.9 V. This was confirmed by immersing two closely similar platinum-wire elec- trodes, one anodically oxidised and the other cathodically reduced, in the plating solutions.The potentials became stabilised at 0.90 V for the oxidised and 0.74 V for the reduced elec- trode, and when a clean silver-wire electrode and a platinum-wire electrode previously reduced for 30 minutes in the iron(I1) solution were placed in the same plating solution, both im- mediately displayed the same stable potential of 0.72 V. This potential was also taken up by platinum-wire electrodes that had been treated with aqua regia, 1 + 1 nitric acid or chromic acid and by immersion, after washing, in the reduction solution for only 1 minute. The platinum-gauze electrode required prolonged immersion because the woven structure presented many hundreds of points of contact in the mesh, which created crevices that were relatively slowly accessible to the solution.Contamination of the silver nitrate solution could arise if the gauze electrode was not completely washed before transfer from the reduction bath. This effect was examined first by adding 0.1 ml of reduction solution that had been diluted 1 + 9 with water to the plating solution just before starting a plating - stripping cycle. This treatment gave a recovery of about 115 per cent. and the stripping curve was much more rounded in the end-point region, while the time required for the potential to rise from +0.8 to +1.4 V was about 30 per cent. longer than usual. Secondly, a reduced electrode that had not been washed was plated at 2 mA for 100 s and immediately stripped at 2 mA. The initial stripping potential was 0.72 V and, after 1900 s, had decreased by only 5 mV.At this stage the current was increased to 5 mA, and after a further 180 s a fairly rapid rise in potential to 0.77 V occurred, followed by a gradual rise over 800 s to 0.8 V. When prolonged washing of the gauze electrodeJune, 19731 OF ACIDS I N CELLS WITHOUT LIQUID JUNCTION. PART 111 421 was performed, by allowing it to stand, after initial thorough rinsing, in three successive batches of 200 ml of vigorously stirred water with thorough washing and draining between each washing, 100 per cent. recoveries of silver were obtained. In addition to the initial potential of the electrode in the plating solution, the shape of the plating curve is also diagnostic of the extent of reduction of the electrodes. This is evident on comparison of Fig.2, which shows incomplete reduction, with Fig. 5 (curve A), which shows a complete plating - stripping cycle at 2 mA in a 0.1 M solution of silver nitrate in 0.1 M perchloric acid with a chemically reduced electrode. The short vertical lines in all of the chronopotentiograms indicate points at which the current was switched on or off. The plating curve in Fig. 5 (curve A) is smooth, while those in Fig. 2 show small negative-going peaks at the beginning of plating. Nisbet and Bard3 observed a similar phenomenon during plating on to incompletely reduced electrodes, Once again de-oxygenation of the plating solution had no influence on recoveries of silver. Fig. 5. Plating and stripping of silver (curve A) and stripping of a silver-free electrode (curve B) by using fully reduced platinum-gauze electrodes PLATING - STRIPPING EFFICIENCIES IN A 0.1 M SOLUTION OF SILVER NITRATE IN 0.1 M As a result of the foregoing investigations, a standard pretreatment procedure for the gauze electrodes was adopted.A minimum immersion time of 20 minutes for oxidised electrodes, or 3 hours, but preferably overnight, for electrodes treated with aqua regia, was used in a stirred 0-2 M solution of iron(I1) sulphate in 1.0 M sulphuric acid. After being thoroughly washed, the electrodes were immersed in clean water for at least 5 minutes. Pre-treated platinum gauze was dried in warm air, weighed, re-treated and plated with about 3 mg of silver. After thoroughly washing it, the electrode was again dried in warm air and re-weighed. The increase in mass was well within 1 per cent.of that calculated from the quantity of electricity passed, thus confirming the quantitativeness of the plating process. Plating and stripping were performed at currents of 2 or 5 mA, the crystal clock was automatically triggered on starting and stopping the plating and on starting the stripping, and was switched off manually when the stripping potential reached 0-8 V. Stripping was started 15 to 30 s after plating so that accurate times could be noted. A series of twenty-two cycles gave the results shown in Table 11. The precision improves as the amount of silver increases. The 95 per cent. confidence limits are &l-6 per cent. for 0.2 mg of silver and 1.0 per cent. for larger amounts. The uncorrected recoveries calculated from the stripping time to 0.8 V exceed 100 per cent.for amounts of 0.4 mg of silver or less, and increase as the amount of silver and the stripping current decrease. The stripping time for a clean, unplated electrode (curve B, Fig. 5) was 3.5 5 0-5 s at 2 mA. If the recoveries are corrected for this effect, they become 99.4 per cent. for 0.2 mg of silver and 99.5 per cent. for 0-4 mg of silver. This “blank” stripping time could be due to a small amount of oxidation of the electrode at 0.75 to 0.8 V, or to the chemical deposition of a partial monolayer of silver on the electrode. Even assuming the extreme value of 100 pF cm-2 for the double-layer capacitance, PERCHLORIC ACID-422 BISHOP AND RILEY : PRECISE COULOMETRIC DETERMINATION [Analyst, VOl. 98 only 0.3 s at 2 mA would be required for charging over the potential range involved.Allen and Hickling16 claimed that platinum was coated with silver on immersing it in 1 M silver nitrate solutions after cathodic reduction. They detected silver colorimetrically after dis- solution from a 100-cm2 platinum foil, and also observed a step in the stripping curve of a pre-reduced 1-cm2 platinum electrode between 0.7 and 0.8 V, corresponding to about 3.5 mC of electricity. On the assumption of a roughness factor of 2, they calculated that a layer of silver seven atoms thick was formed on immersion of platinum in silver nitrate solution, which in the present instance would represent about 0.4 mg of silver, whereas the “blank” stripping time corresponds to 7 pg of silver. It is likely that the deposit, and the stripping step observed by Allen and Hickling, arose from residual hydrogen from the cathodic reduction; no steps taken to remove this deposit were described.TABLE I1 AMPEROSTATIC ANODIC STRIPPING OF SILVER IN A 0.1 M SOLUTION OF SILVER NITRATE I N 0.1 M PERCHLORIC ACID Stripping Approximate mass of Number Mean recovery, Relative standard current/mA silver plated/mg of tests per cent. deviation, per cent. 2 0.2 10 102.9 0.7 2 0.4 6 101.2 0.4 6 0.5 to 1.1 6 100.1 0.4 ANODIC STRIPPING OF SILVER IN PERCHLORIC ACID SOLUTION- Having obtained a satisfactory recovery in silver nitrate solutions, it was then necessary to check the recoveries under the conditions appertaining to “silver error” determinations at the end of an acid assay. After plating in the silver nitrate - perchloric acid solution, the platinum-gauze electrode was removed from the plating solution, drained, carefully and thoroughly washed with water, and allowed to stand in about 100ml of water for several minutes before being transferred to a second coulometric cell, as in Fig.1, containing 0.1 M perchloric acid. It was then stripped at 2 or 5mA, the clock being automatically started and manually stopped at a stripping potential of 0.8 V. The elapsed time between completion of plating and starting of stripping was usually about 10 minutes. Craig, Law and Hamer17 have demonstrated that the rate of dissolution of finely divided silver in perchloric acid solutions is so slow that it can be neglected with confidence. The results of thirteen experiments involving 0-2 to 1.2 mg of silver are summarised in Table 111, and again show an improvement in precision as the amount of silver and the stripping current increase.TABLE IT1 AMPEROSTATIC STRIPPING OF SILVER INTO 0.1 M PERCHLORIC ACID Stripping Approximate mass of Number Mean recovery, Relative standard current/mA silver plated/mg of tests per cent. deviation, per cent. 2 0.2 to 1.2 9 100.9 1.0 5 0.8 to 1.1 4 100.3 0.7 The 95 per cent. confidence limits were 5 2 . 3 per cent. at 2 mA and k 2 . 2 per cent. at 5 mA. Recoveries are again slightly high, but the bias decreases as the amount of silver increases, and is insignificant for amounts in the region of 1 mg. When silver-free, reduced electrodes were stripped, it was found that their initial potential in 0.1 M perchloric acid was about +0.9 V, and so there is no blank stripping time under these conditions.Typical stripping curves for about 0.4 mg of silver and for a reduced electrode at 2 mA are shown in Fig. 6. The potentials displayed by a clean silver-wire electrode and a reduced platinum-wire elec- trode in 0.1 M perchloric acid were approximately 0.43 and 0.98 V, respectively. THE SECOND ANODIC WAVE- It was observed during silver error determinations1 that the second wave in the stripping curve associated with oxidation of platinum was abnormally long. Fig. 7 shows a typical curve for a silver error determination at 2 mA; the rise time from 0.8 to 1.4 V corresponds to about 500mC in this instance, and values ranging from 200 to 700mC are commonly obtained.The silver deposit in these experiments was produced under different conditions and in a different medium from the calibration plating described above. Values of aboutJune, 19731 OF ACIDS I N CELLS WITHOUT LIQUID JUNCTION. PART I11 423 110 mC were found in the preliminary work described in this paper, but pertained to partly reduced electrodes; values for the second wave on stripping in silver nitrate solutions at 2mA were 165 to 170mC for both silver-plated and silver-free electrodes. For stripping into perchloric acid, as described in the preceding section, similar extended second waves appeared, Although variable, the results for the corresponding quantities of electricity were 180 to 380mC, and longer stripping times were associated with longer second waves.Potential versus S.H.E. - + 0-5 V -+ 0-8 -+ 1.1 --+ 1.4 Potential versus S. H . E . + 0.6 V- + 1.ov- + 1.4 V- Fig. 6. Anodic stripping a t 2 mA in 0.1 M perchloric acid of A, silver- plated platinum; and B, silver-free platinum Fig. 7. Extended second wave obtained during a “silver error” determination a t 2 mA These second waves are pertinent to the ensuing pre-treatment of electrodes, and they warranted exploration. It is postulated that they are due to the anodic oxidation of hydrogen peroxide introduced into the solution by the reduction of dissolved oxygen at the auxiliary electrode during the stripping process. Laitinen and Kolthoff l8 claimed that hydrogen peroxide was a product of reduction of dissolved oxygen at a platinum electrode, 0, + 2H+ + 2e + H,O, but later studie~l~-2~ have shown that this is so only under certain conditions of the state of the electrode surface.Peters and Mitchel121 demonstrated spectrophotometrically that significant amounts of hydrogen peroxide appeared in the bulk of the electrolyte when reduction was performed at an “aged” electrode, which was produced by allowing a freshly reduced electrode to stand in oxygen-saturated electrolyte for over 24 hours, but could detect no hydrogen peroxide when the reduction was performed with freshly reduced or pre-oxidised electrodes. They supposed that hydrogen peroxide was also produced at a freshly reduced electrode, but immediately disproportionated to water and oxygen at the “active” platinum surface, and that as the electrode “aged” it lost the ability to catalyse the disproportionation.In the present instance, the pre-reduced auxiliary electrode will be oxidised during plating then at least partially reduced during the first part of the stripping step, and then, concurrently with the reduction of dissolved oxygen, some of the silver stripped from the anode will be deposited on it in increasing amount as the stripping time increases. Kolthoff and Laitinen22 found that dissolved oxygen was reduced at silver at more positive potentials (less charge- transfer overpotential) than at platinum. It did not seem that the condition of the auxiliary electrode was suitable for hydrogen peroxide formation in the present work, and so reduction of dissolved oxygen at silver electrodes was investigated.Deliberate addition of trace amounts of hydrogen peroxide, either during the stripping of silver in 0.1 M perchloric acid, or just prior to stripping a silver- free reduced electrode, produced extended second waves. This observation agrees with that hydrogen peroxide is oxidised at an unoxidised platinum electrode, but424 BISHOP AND RILEY: PRECISE COULOMETRIC DETERMINATION [Analyst, VOl. 98 only qualitative correlation could be made between the amount of hydrogen peroxide added and the increase in duration of the second wave. When oxygen-free 0.1 M perchloric acid was used for silver stripping no extension of the second wave occurred, even with long stripping times, and the recovery of silver was unchanged. When the auxiliary electrode was pre- plated with silver and the solution contained oxygen, significantly extended second waves appeared, even with short stripping times, and hydrogen peroxide was detected spectrophoto- metrically in the cell electrolyte.In two examples with silver stripping times of 120 and 500 s, the second waves were equivalent to 340 and 550 mC, respectively. The peroxodisulphatotitanic(1V) acid method was used to determine hydrogen peroxide. The absorption spectrum of the complex showed a broad peak with A, = 410 nm. A linear Beer’s law graph was used as a calibration graph ; the molar absorptivity of the complex at 410 nm was 40 1 mol-l mm-l. Blanks of freshly prepared 0.1 M perchloric acid electrolyte were used. A 10-ml sample of the cell electrolyte, after about 80 per cent.of the silver had been stripped, was treated with 2 ml of the titanium(1V) reagent and the absorbance at 410 nm was measured. The amount of hydrogen peroxide found in the electrolyte was of the order of low5 mol l-l, but could be only qualitatively correlated with the extension of the second wave. The reaction is probably cyclic. DISCUSSION Clearly, the efficiency of plating silver on incompletely reduced platinum electrodes is significantly below 100 per cent., and the loss is due to the concurrent reduction of remanent oxide, The oxidation and reduction of platinum show behaviour in accord with other work, but the interpretation is still contentious and is dealt with elsewhere. Insufficient attention seems to have been paid to the competitive formation of molecular oxygen in the oxidation half-cycle. Failure of the electro-reduction methods to give fully reduced active electrodes for plating is puzzling, but is tentatively ascribed to electro-sorption of impurities of which minute traces exist in sulphuric acid12 or insufficient prior oxidation, or both.Later work suggests that repeated anodic - cathodic cycling is necessary in preparing a fully active electrode. The iron(I1) reduction method, when correctly applied, gives electrodes at which good plating efficiency is attained, but very thorough washing of the electrode is essential, otherwise cyclic oxidation and reduction of iron(I1) and iron(II1) at the two electrodes occurs and no silver is plated on the electrode. On the basis of plating calibrations involving the use of electrodes pre-conditioned by chemical reduction with iron( II), silver stripping recoveries both in silver nitrate solutions and in pure perchloric acid solutions are excellent for 0.2 to 1.2 mg of silver and are of a precision entirely adequate for the determination of silver errors.The blank found in silver nitrate solutions but absent in perchloric acid shows that about one tenth of a monolayer of silver is deposited chemically on platinum immersed in a silver solution, and this finding is supported by the potentials of silver and platinum electrodes in these media, Hydrogen peroxide is produced at the auxiliary electrode by reduction of dissolved oxygen during the stripping of silver into perchloric acid, and gives rise to an extended second wave in the stripping chrono- potentiogram.The deposition of stripped silver on the auxiliary electrode favours this process because of the lower charge-transfer overpotential of reduction of oxygen on this metal. The phenomenon is absent in de-oxygenated media, and in any event does not affect the recovery of silver; it does, however, add emphasis to the importance of adequate pre- treatment of platinum electrodes on to which silver is to be plated. CONCLUSIONS The present study has shown that silver present on a platinum-gauze electrode can be determined by amperostatic anodic stripping with an accuracy and precision that are adequate for the required “silver error” determination, and that such an electrode can be washed and transferred into the stripping solution without significant loss of silver.The amount of silver present on the working electrode at the end of an acid assay that involves the passage of 5000 C can be determined simply and rapidly by this method. Extended second waves (which do not affect the silver determination) on the stripping curves arise from the oxidation of hydrogen peroxide produced at the auxiliary electrode or oxidation of residual hydrogen on the electrode surface following the acid determination, or both.June, 19731 OF ACIDS I N CELLS WITHOUT LIQUID JUNCTION. PART 111 425 One of us (M.R.) is deeply indebted to the Charitable and Educational Trust of the Worshipful Company of Scientific Instrument Makers for financial support in the form of a Research Studentship. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. REFERENCES Bishop, E., and Riley, M., Analyst, 1973, 98, 313. Lord, S. S., O’Neill, R. C., and Rogers, L. B., Analyt. Chem., 1952, 24, 209. Nisbet, A. R., and Bard, A. J., J . Electroanalyt. Chem., 1963, 6, 332. Bixler, J. W., and Bruckenstein, S., Analyt. Chem., 1965, 37, 791. Bishop, E., and Riley, M., Analyst, 1973, 98, 305. Feldberg, S. W., Enke, C. G., and Bricker, C. E., J . Electrochem. Soc., 1963, 110, 526. Kolthoff, I. M., and Tanaka, N., Analyt. Chem., 1954, 26, 632. Anson, F. C., and Lingane, J. J., J . Amer. Chem. Soc., 1967, 79, 1016. ~- , Ibid., 1957, 79, 4901. Lingane, J. J., J . Electroanalyt. Chem., 1960, 1, 379. Peters, D. G., and Lingane, J. J., Ibid., 1962, 4, 193. Hitchcock, P. H., Ph.D. Thesis, University of Exeter, 1969. Bishop, E., “Coulometric Analysis,” Volume IID of Wilson, C. I,., and Wilson, D. W., Editors, Ross, J. W., and Shain, I., Analyt. Chem., 1956, 28, 648. Anson, F. C., Ibid., 1961, 33, 934. Allen, P. L., and Hickling, A., Analytica Chin$. Acta, 1954, 11, 467. Craig, D. N., Law, C. A., and Hamer, W. J., J . Res. Natn. Bur. Stand., 1960, 64A, 127. Laitinen, H. A., and Kolthoff, I. M., J . Phys. Chem., 1941, 45, 1061. Lingane, J. J., J . Electroanalyt. Chem., 1961, 2, 296. Sawyer, D. V., and Interrante, L. V., Ibid., 1961, 2, 310. Peters, D. G., and Mitchell, R. A., Ibid., 1965, 10, 306. Kolthoff, I. M., and Laitinen, H. A., Science, N.Y., 1940, 92, 160. Hickling, A., and Wilson, W. €I., J . Electrochem. SOC., 1951, 98, 425. Giner, J., 2. Elektrochem., 1960, 64, 491. Lingane, J. J., and Lingane, P. J., J . Electroanalyt. Chem., 1963, 5, 411. Liang, C. C., and Juliard, A. L., Ibid., 1965, 9, 390. NOTE-References 1 and 5 are to Parts I1 and I, respectively, of this series. “Comprehensive Analytical Chemistry,” Elsevier, Amsterdam, 1973, pp. 243-263. Received December 18th, 1972 Accepted January 19th, 1973

ISSN:0003-2654    

DOI:10.1039/AN9739800416

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

426 Analyst, June, 1973, Vol. 98, pp. 426-431 Precise Coulometric Determination of Acids in Cells Without Liquid Junction Part IV.* The Assay of Primary Standard Sulphamic Acidt BY E. BISHOP AND M. RILEY: (Chemistry Department, University of Exeter, Stocker Road, Exeter, EX4 4QD) The precise (1 to 2 p.p.m.) location of the end-point in the pre-titration of the supporting electrolyte and in the titration of sulphamic acid has been examined, and d.c. differential electrolytic potentiometry gives excellent results. The methods and simple apparatus previously described in Parts I, I1 and I11 are applied to the assay of primary standard sulphamic acid previously collaboratively assayed by mass titrimetry, and give results with an accuracy and precision of 100 p.p.m., and a 95 per cent.confidence level of 0.02 per cent., but with a positive bias of 0.014 per cent., the reasons for which are canvassed. The method is of high merit; it is simple, fast and direct, and is capable of further refinement. EARLIER work has established that current sources of adequate stability and a timing device of adequate accuracy are available for high-precision amperostatic acidimetryl ; that suitable conditions can be chosen that permit the use of a level of deposition of more than 5000 C of bromide on a silver anode as the auxiliary reaction in the same compartment in which the cathodic determination of the acid is conducted2; and that the error arising from the solubility of silver bromide in the cell electrolyte and consequent deposition of silver on the working cathode can be evaluated with adequate accuracy and preci~ion.~ There remained for investi- gation first, a means of locating the end-point of the reaction with adequate accuracy and precision, and secondly, the testing of the whole method by the assay of an independently standardised primary standard acid. For the latter, sulphamic acid, which was collaboratively assayed by the Analytical Standards Sub-committee of the Analytical Methods Committee of the Society for Analytical Chemistry and recommended and accepted as an international primary was selected.For end-point location, d.c. differential electrolytic potentiometry has an adequate reserve of sensitivity.s EXPERIMENTAL The equipment used has been described previous1y.l The cell used in the final deter- minations is shown in Fig.1. The platinum-gauze cathode was the larger Model 72020, The silver anode comprised two electrodes, one each of sizes A and B, of total area 275 to 315 cm2. The twin antimony differential electrolytic potentiometric electrodes were mounted centrally so as to be out of the generating current field, and were held in a rubber bung, which was replaced with a plain bung when the indicating electrodes were removed from the cell. Pure oxygen, humidified by passage through a water bubbler, was passed over the surface of the solution, and served to exclude carbon dioxide and to oxygenate the solution for the benefit of the antimony electrodes.’ Excess of oxygen escaped through an exit tube that was provided with a spray trap. The platinum-gauze electrode was initially cleaned by immersion for 1 to 2 minutes in freshly prepared aqua regia, followed by very thorough washing with and storage in water.Subsequently, after “silver error” determinations, it was immersed in a 0.2 M solution of iron(I1) sulphate in 1.0 M sulphuric acid for at least 30 minutes before further use.3 Other electrodes were treated as described previously.1 The circuit, in which the cell is indicated in plan view, is shown in Fig. 2. The constant-current sources were run continuously, being switched to dummy loads when not in use. The selector switch on the P3 potentiometer was used to select the standard resistor that was required when power supplies were exchanged. Switching in power supplies to the cell automatically triggered the clock, and switching over to the dummy load stopped the clock.For particulars of Parts I, I1 and I11 of this series, see reference list, p. 431. t Presented at the Second SAC Conference, Nottingham, 1968. @ SAC and the authors. Present address: Electronic Instruments Limited, Hanworth Lane, Chertscy, Surrey.BISHOP AND RILEY 427 Twin antimony DEP electrodes Platinum -gauze cat hod e Fig. 1. Coulometric cell for acid determina- tion (DEP denotes differential electrolytic poten- tiometric) SULPHAMIC ACID- by the Analytical Standards Sub-committee. with reference to atomic-mass grade silver at 100.001 per cent. purity. dried under vacuum over fresh magnesium perchlorate for at least 24 hours before use. WEIGHING AND TRANSFER OF SAMPLES- Catch-weights of 4 to 5 g of dry sulphamic acid were weighed in small glass weighing bottles that had outside-fitting lids, with a hole drilled in each ground face so that pressure could be equalised by turning the lid to register the holes.Preliminary rough weighings to the nearest 1 mg were made on a CL3 balance, and the accurate weighings made on the Samples were taken from a 100-g batch of sulphamic acid provided in a sealed container The sample had been collaboratively assayed The samples were Fig. 2. Circuit for coulometric acidi- metry: c, plan view of coulometric cell; PSU (power supply unit), 2-A source AS 1411 ; OACS, operational amplifier constant- current source ; P, P3 potentiometer ; T, crys- tal clock; M, 39A pH meter; Vs, DEP voltage source; R,, 0.6-SZ standard resistor (two 1-SZ standards in parallel) ; Rz.104 standard resistor; R,, dummy load resistor, about 0.6 a; R,, 1042 dummy load resistor; and RB, DEP ballast resistor428 [Analyst, Vol. 98 special atomic-mass balance that had a standard deviation of 1.3 pg on 100 g and was provided with both 5 and 0-5-mg riders. Once the riders had been located in the correct notch, the balance was left for 30 minutes, then released and the swings were observed by telescope from a distance of 20 feet. The sequence of measurements were zero, weighing bottle PLUS sample, sensitivity, zero, empty weighing bottle, sensitivity, zero. Buoyancy corrections were made by using 8.0 g ~ m - ~ for the density of the weights and 2.126 g ~ m - ~ for the density of sulphamic acid.The sample was transferred into the cell, which had already been filled with pre-titrated electrolyte, with oxygen passing through it, via a small glass funnel inserted in the lid. Most of the crystals passed straight through into the cell. The lid of the weighing bottle was replaced and the bottle re-weighed. Before adding the sample to the cell, 20 ml of the neutral electrolyte were withdrawn into a hypodermic syringe, and this solution was used to wash the funnel thoroughly, which was then removed and the hole closed with a bung. END-POINT LOCATION- A conventional differential electrolytic potentiometric circuit8 was used as indicated in Fig. 2. The source voltage was 240 V and the ballast resistance was 960 MQ, giving a dif- ferentiating current density of about 4 pA cm-2.Differential potentials were measured on the 39A pH meter and recorded on a 10-mV recorder. The electrodes were inserted for pre-titration, then removed and replaced within 1 C of the end-point. Preliminary appraisal of the response was made by titrating 7 to 10mg of sulphamic acid at concentrations of 2.5 to 3-5 X M at currents of 10 mA in 0.03 M potassium bromide solution. Satisfactory differential peaks, 50 to 80 mV in amplitude, were obtained at differentiating current densities of 2.2 to 6.6 pA cm-2 and ballast loads of 3-6 x 1O1O to 1.1 x 1011 VQ, but with continuous generation the results showed a positive bias of 1 to 2 per cent. Bishop and Short6 determined similar amounts of perchloric acid by continuous generation at 2 to 10 mA but found a bias that was not greater than 0.2 per cent.The cause of the bias (electrolyte retention in the meshes of the electrode) was not immediately apparent, but it was concluded that incremental generation, with automatic time integration with the crystal clock, would be more apposite in high-precision work. With incremental generation, relatively long times were required for the differential potential to become stabilised near the end-point. Times of 4 to 5 minutes were usual in pre-titration (see below) and 6 to 10 minutes during the final end-point determination. The positive-going drifts were not caused by the electrodes, because glass indicator electrodes referenced to a saturated mercury( I) sulphate electrode showed similar drifts to higher pH values. The drift is due to slow diffusion of hydroxyl ions from the interstices of the gauze electrode into the bulk of the solution.Eckfeldt and Shafferg observed similar drifts in unbuffered solutions when using an electrode constructed in the form of a mesh of platinum strips, and found that the drift was eliminated when smooth platinum wire was used. Satisfactory end-points were achieved in sulphamic acid determinations by using incremental generation in the vicinity of the end-point, although the long equilibration times between increments made the procedure rather tedious. Differential potentials were con- sidered to be stable when a drift of less than 0.2 rnV min-l was recorded. A typical end-point determination is shown in Fig. 3 for incremental generation at 10mA. Exact end-points could be extracted from such curves geometrically,6 but could usually be estimated visually with sufficient accuracy; an error of 1 s in the end-point location corresponded to an over-all error of 2 p.p.m.in the determination. Curves for the pre-titration were identical, although peak heights were greater by about 10mV. Pre-titration of the cell electrolyte, before addition of the sample, was carried out at a current of 5 mA so as to correct for any residual carbon dioxide in the water or acidic impurities in the potassium bromide. This pre-titration required about 0.05 C, corresponding to a residual carbon dioxide concentration of 1.7 x M, assuming this to be the only impurity present, and is in good agreement with the earlier value6 of 1.6 x 10-6 M. PROCEDURE FOR SULPHAMIC ACID ASSAY- From the approximate mass of the sample, the theoretical amount of potassium bromide was calculated, and this amount plus an excess of 1.0 to 1.1 g was weighed.The power supply to the crystal clock was switched on at this stage: as the circuit is entirely of the low-power solid-state type, the clock reaches thermal equilibrium quickly and does not need BISHOP AND RILEY : PRECISE COULOMETRIC DETERMINATIONJune, 19731 OF ACIDS I N CELLS WITHOUT LIQUID JUNCTION. PART IV 429 0 0 0 0 0 0 410 430 450 Generation time a t 10 mA/s Fig. 3. Typical end-point location graph to be run continuously. The pre-conditioned cell was rinsed with carbon dioxide free water, the potassium bromide and magnetic follower were placed in it, the lid and electrode assembly fitted and the flow of oxygen was started.About 300ml of carbon dioxide free water were dispensed from the reservoir into the cell through a spare hole in the lid, which was then closed with a bung and the stirrer switched on. The operational amplifier constant- current source was set to give an output of 5 mA, and the potential drop across the standard resistor and the temperature of the latter were noted. Pre-titration was performed by generation in increments of 2 to 4 s, allowing the differential potential to become stabilised between increments. The values of the cumulative generation time and the differential poten- tial were noted after each increment, and generation was continued until the differential potential showed a marked fall.The quantity of electricity passed after the differential elec- trolytic potentiometric peak was calculated. The antimony indicator electrodes were removed and washed with and stored in water, the hole in the lid being then closed with a bung. The 2-A current source with output settings of 2.0 A and 39.9 V (the maximum setting) was monitored, and the sulphamic acid sample was transferred into the cell. The crystal clock was re-set to zero. The potential drop across the standard resistors and their tem- peratures were noted and generation at 2 A was started. The theoretical quantity of elec- tricity, QT, required for neutralisation was calculated, and generation at 2 A was continued to within 50 C of QT. The lid and walls of the cell, the electrode stems and the spray trap were carefully washed down, which was effected by fitting a 20-ml hypodermic syringe with a suitably bent needle, opening the spare hole in the lid, withdrawing 10 to 15ml of cell electrolyte into the syringe and using this solution for the washing down, further portions of electrolyte being withdrawn if necessary. Use of the bent needle enabled all of the washings to be carried out without removing the lid.Generation at 2 A was continued to within 5 C of QT, when the 2-A current source was disconnected, the generation time at 2 A was noted and the washing operations were repeated. Measurements of the current flowing were made at 2-minute intervals during electrolysis a t 2 A. The operational amplifier source, set to 10-mA output, was then monitored, the crystal clock re-set to zero and generation continued to within 1 C of QT.After a further washing operation, the antimony indicator electrodes were rinsed with carbon dioxide free water, inserted in the cell and the differential electrolytic potentiometric circuit activated. Genera- tion at 10 mA was continued in 10 to 20-s increments until the differential potential showed a significant increase, when a final washing down was performed. The final part of the generation was conducted in 4 to 5-s increments, noting the time and stabilised differential potentials after each increment, and was continued until four or five points after the end-point had been passed. The platinum-gauze electrode was then carefully removed from the cell, gently washed with water and drained several times, and immersed in water for 5 to 10 minutes.It was then mounted in a stripping cell (Fig. 1 in Part 1113) containing 0.1 M perchloric acid electrolyte, and the 39A pH meter was connected between the gauze electrode and the saturated mer- cury(1) sulphate electrode. The operational amplifier source was set to 5 mA, the crystal430 [finahst, vol. 98 clock re-set and the silver stripped from the gauze electrode as previously described,3 the time taken to reach a potential of +O-S V being measured by stopping the clock manually at this stage. CALCULATION OF RESULTS- The end-point was located graphically as in Fig. 3, for generation time at 10 mA, and the total number of coulombs used in the acid determination, QF, was calculated as follows- QF = QD + QE - Qs where QD is the number of coulombs passed at 2 A, calculated from the mean of the measured values of the current, QE is the number of coulombs passed at 10 mA in reaching the end-point PZZH any excess of generation at 5 mA in the pre-titration, and Qs is the number of coulombs required for the anodic stripping of silver.The result of the coulometric assay was therefore the ratio of QF to QT, expressed as a percentage. BISHOP AND RILEY : PRECISE COULOMETRIC DETERMINATION RESULTS AND DISCUSSION The results obtained on one batch of sulphamic acid, comprising eight assays of 4 to 5-g portions, are shown in Table I and are typical. The values of QT were calulated, to the nearest 0.01 C, by using the recommended value of the Faraday constantlo of 96 486.70 & 0-5 A s mol-1 and a relative molecular mass of 97.093 for sulphamic acid.Evidently, the method is capable of high precision: the 95 per cent. confidence limits are &Om02 per cent. The two results, 99.999 and 100.028, attenuate the precision rather severely, but cannot be excluded because they are, although only just, statistically significant. TABLE I COULOMETRIC ASSAY OF BATCH 3 SULPHAMIC ACID Sample masslg 4.850 03 4.496 97 4-702 20 6.138 46 4.325 46 4.320 23 4.683 28 4.895 63 Quantity of electricity h r \ Q D l C Q E l C QslC 4815.25 7.422 2.543 4468.38 4.645 2- 194 4672.73 3.131 2.246 5 104.42 5.675 2.637 4297.32 4.432 2.101 4289.51 6.3 16 2.173 4654.33 2.968 2.400 4864.78 3.863 2.506 Mean .. .. .. .. Relative standard deviation , .Assay, per cent. 100.008 99.999 100.01 6 100*021 100.028 100~009 100.0 18 100.022 100.0 15 0.009 per cent. It is clear that the results show a positive bias, being 0.014 per cent. higher than in the mass titrimetric assay, which in one respect is fortunate in that it shows that cancellation of errors had not occurred: it is probable that several factors contribute to the bias. A bias in the measurement of the 2 A current was adumbrated earlier,l and it is possible that the uncertainty of &SO p.p.m. in the values of the two 1-R standard resistors used for measure- ment of high currents could account for one third of the bias. The possibility that the measured silver errors, whose constancy reflects the fact that the total times occupied by the acid determinations were similar in all instances, were low because of some mechanical loss of silver from the gauze electrode cannot be entirely discounted, although such a loss might be expected to show larger variations in the value of Qs.It is not improbable that anodic reactions other than halide deposition could occur, which could be significant particularly towards the end of the determination when the pH is rising rapidly. These reactions might include the anodic directions of iC AgOH + e + Ag + OH- i a 0, + 4H,0+ + 4e + 6H,O 0, + 2H,O + 4e + 40H- iC i a i o i,June, 19731 OF ACIDS I N CELLS WITHOUT LIQUID JUNCTION. PART IV 431 and, although the solution is oxygenated it is not purged, so that some molecular hydrogen in solution may give a small anodic current due to its re-oxidation- i c 2H30+ + 2e + H, + 2H,O i* It is also possible that 1 or 2 p.p.m.of the anodic current produces bromine from the bromide, and the bromine would hydrolyse at pH above 5, thus producing hydrogen ions. The drifting differential potentials near the end-point made the finish of the determina- tions rather tedious, so that a complete determination required about 5 hours, although this is much quicker than the titrimetric assay, which takes 3 to 5 days. That the drifting is caused by diffusion of hydroxyl ions from the interstices of the platinum gauze is in agreement with previous finding^.^ It also explains the difference in equilibration times between pre- titration and final end-point location, because the concentration of residual hydroxyl ions on the electrode would be expected to be much larger near the end of the reaction.A coiled platinum rod or perforated heavy gauge sheet would be better than gauze, and could reduce the experimental time by 90 minutes. The measured values of the large current were found to change only very slowly over the 35 to 40 minutes of generation, and the maximum deviation was only 20 p.p.m. The cell resistance was less than 0.5 Q initially and rose to 2 Q at the end of the high-current generation period: the temperature of the cell electrolyte rose by 5 to 6 "C from the initial 20 "C. CONCLUSIONS Results obtained in the assay of primary standard sulphamic acid show that a precision and accuracy in the region of 100 p.p.m. can be achieved with the simple cell and equipment described. Refinement of the method, by using better current measuring equipment, smooth platinum instead of gauze and purging of the electrolyte with 1 + 6 oxygen - nitrogen or oxygen - argon mixture, could enhance the precision and reduce the time required. Operation in the differential mode, with two cells in series, for the intercomparison of chemical standards would certainly yield improved precision, and would give the method even more marked advantages over the conventional methods with multicompartment cells. One of us (M.R.) is deeply indebted to the Charitable and Educational Trust of the Worshipful Company of Scientific Instrument Makers for financial support in the form of a Research Studentship. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. REFERENCES Bishop, E., and Riley, M., Analyst, 1973, 98, 305. 8 , Ibid., 1973, 98, 313. , Ibid., 1973, 98, 416. -- 3 -- Analytical Methods Committee, Ibid., 1967, 92, 587. International Union of Pure and Applied Chemistry, Analytical Division, Pure AppZ. Chem., 1969, Bishop, E., and Short, G. D., Analyst, 1964, 89, 687. Short, G. D., and Bishop, E., Analyt. Chem., 1965, 37, 962. Bishop, E., and Short, G. D., AnaEyst, 1962, 87, 467. Eclrfeldt, E. L., and Shaffer, E. W., Analyt. Chem., 1965, 37, 1634. Taylor, B. N., Parker, W. H., and Langenberg, D. N., Rev. Mod. Phys., 1969, 41, 376. NOTE-References 1, 2 and 3 are to Parts I, I1 and 111, respectively, of this series. 18, 443. Received December 18th. 1972 Accepted January 19tk, 1973

ISSN:0003-2654    

DOI:10.1039/AN9739800426

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

432 Analyst, June, 1973, Vol. 98, $9. 432-442 A Quantitative Tunable Element-selective Detector for Gas Chromatography BY W. R. McLEAN, D. L. STANTON AND G. E. PENKETH (Imfierial Chemical Industries Limited, Petrochemicals Division, Billingham, Teesside, TS23 1 JB) A detector based on the atomic-emission spectra that result when organic compounds are decomposed in a low-pressure, microwave-sustained helium plasma is described. All of the non-metallic elements normally found in organic compounds can be sensitively and selectively detected in a linearly proportional and quantitative manner by means of conventional diffraction grating spectrometer equipment. A controlled amount of a scavenger gas is used to prevent carbon deposition inside the plasma tube. The chromato- graphic column outflow is split between the element-selective detector and a non-selective flame-ionisation detector.The latter acts as a reference for interpreting element-selective detector results and assists with the determina- tion of atomic ratios and the empirical formulae of organic compounds. THE gas-chromatographic detection of organic compounds by emission spectroscopy with a microwave-powered plasma was first reported by McCormack, Tong and Cookel and subsequently developed by Bache and Lisk2s3 and others4s5 The organic compounds emanat- ing from the gas chromatograph are fragmented in the high-energy plasma to produce emission spectra, which are then monitored with a suitable spectrophotometric detection system, In principle, the technique is applicable to a very wide range of elements, but most work to date has centred around the elements of interest in pesticide analysis, for example, sulphur, phosphorus and chlorine.Our objectives were to extend this range and to improve the quantitative characteristics of the system, and because of the latter objective we preferred to use a low-pressure helium plasma to produce the atomic-emission lines. The lower energy of an argon plasma is insufficient to prevent the association of atoms into pairs and this leads to the production of complex band-emission spectra, which reduces the spectral selectivity on atomic lines. The association of atoms into pairs also reduces the population of free atoms, and for quantitative work this effect adds a complicating mass-action influence to atomic emission and gives compound-specific effects.The higher energy of the helium plasma greatly inhibits formation of atom pairs, to the obvious benefit of the atomic-emission characteristics. In our early work, both the qualitative and the quantitative performance were hampered by persistent deposition of carbon on the inner walls of the plasma tube, but a dramatic improvement resulted when a small amount of air was continuously bled in. Subsequently, it was found that either oxygen or nitrogen would act as a carbon scavenger, and this discovery enabled either element to be included in the range of the technique by using the other as scavenger. APPARATUS- The apparatus is shown diagrammatically in Fig. 1. Gas chromatograph--We normally used a Pye 104 gas chromatograph but other makes have been used with equal success.Any column with low bleed characteristics can be used; the conditions for the chromatograms shown in this paper were- Column . . . . Temperature . . .. . . 112 "C (pre-heater 200 "C) Pressures .. .. . . Head of column, 170 kN m-2; tail of column, 35 kN m-2 Detector splitting ratio . . 1 : 1 Scavenger gas . . .. @ SAC and the authors. .. .. 3 m x 2.5 mm, packed with 10 per ccnt. Apiezon L on 00 to 80-mesh DCMS-treated Chromosorb W . . Oxygen (or nitrogen) : 0.1 to 1 per cent. of total plasma gas according to demandMCLEAN, STANTON AND PENKETH 433 Strip chart Strip chart recorder recorder I. 2. 3. 4. Microwave Nitrogen I Hydrogen Amplifier Amplifier Air (FID) I (MPD) Power Power supply handling (FID) (MPD) systems Sample - FID supply - Microwave cavity 5.Diffraction grating Plasma discharge Monochromator entrance slit Exit slit and photomultiplier 6. Scavenger gas supply 7. Air - hydrogen supply 8. Helium supply t o chromatograph Fig. 1. Block diagram of the apparatus blasma detector (MPDb-The Dlasma emission is contained in a thick-walled quartz capillary'tube of 10 mm 0.d. and 1 mm i.d., with an over-all length of 15 cm. This tube is mounted vertically in an assembly that also holds the microwave cavity in position. The pressure within the plasma is adjustable from an arbitrary 0.25 torr to higher pressures by means of a vacuum pressure regulator in the line connecting the plasma to the vacuum pump (Fig. 2). Microwave power, generated by a 0 to 200-W generator (Electro-Medical Supplies Limited, London) a t 2.45 GHz from 100 to 200 W, is supplied to the tuned 214L cavity via a reflected power meter and flexible wave guide.The tuning stubs on the cavity are adjusted until the reflected power is the minimum attainable. With this type of cavity, the power reflected can be made as low as 1 per cent. The plasma is initiated by means of a Tesla coil and gives an intense plasma discharge about 10 cm long. Gas su$+Zies-High-pressure sources of the following gases were used- Helium . . .. .. . . Grade A Carrier gas Nitrogen . . .. . . White spot grade Air . . .. .. . . Electrolytic grade : ~~~~~~~ ::$: detector support gases Oxygen . . .. .. Hydrogen . . .. Grade A helium was purified by passing it through a B.O.C.helium purifier, in which oxygen and nitrogen were removed by the hot titanium sponge and hydrocarbons were oxidised to carbon dioxide and water by hot copper(I1) oxide; these products were then removed in a modified 5A molecular sieve unit placed externally to the purifier and maintained at -80 "C. The scavenger gases were dried by passing them through a cold trap at -80 "C. Because of the relative amounts of each gas used, the purity of helium is approximately one hundred times more critical than that of the scavenger gases, so for the more common elements carbon, nitrogen, hydrogen and oxygen, the greater the spectral purity of the helium, the higher is the sensitivity. I t was realised at an early stage that the value of element-selective results was much enhanced by the simultaneous recording of non-selective results, e.g., those from a flame- ionisation detector.This realisation led us to develop an interface system that enabled a434 McLEAN, STANTON AND PENKETH: A QUANTITATIVE TUNABLE [Analyst, vol. 98 Sample Head pressure Tail pressure I I I Gas I chromatograph: I column I system I I I I I 7 I I I I I I 1 I I I I I I I I I i I Scavenger I L--- _ _ _ _ _ _ _ _ _ J bleed h L I MPD detector Plasma con t ro I Fig. 2. The chromatograph and the inter- facing system to two detectors gas-chromatographic system to be coupled with two or more detectors that were operated at atmospheric or lower pressures. The principle is illustrated in Fig. 2. The exit of the column system is maintained at a pressure in excess of that of the atmosphere by a separate supply of carrier gas; the extra supply then merges with the column exit flow and splits into equal parts across specially designed flow restrictors to the detectors.By allowing the total flow-rate to the detectors to be more than the optimum flow-rate through the gas-chromatographic column, no additional constraints are imposed upon the chromatography. 1-485.99 77 D-656. 7 . H-656.281 -486.1 33 rT. Ah = 0.134 nm Ah = 0.181 nm Fig. 3. Resolution of the hydro- gen - deuterium atomic-emission lines (wavelengths in nanometres)June, 19731 ELEMENT-SELECTIVE DETECTOR FOR GAS CHROMATOGRAPHY 435 Spectrophotometer-The monochromator used was a Hilger and Watts Monospek 1000 with a 102 x 102-mm diffraction grating of 1200 linesmm-l (blazed at 300-Onm) to give a dispersion of 0.88 nm mm-l. The original IP28 side-window photomultiplier was replaced with a similarly designed Hamamatsu R446 photomultiplier, which extended the optical range available from 190.0 nm to greater than 900.0 nm with an excellent degree of sensitivity. SPECTRAL CHARACTERISTICS- The spectroscopic system gave good line spectra for all of the elements examined with very little evidence of molecular band emission.As examples, Fig. 3 shows the resolution of the hydrogen and deuterium lines and Fig. 4 the resolution of the triplets of oxygen and nitrogen. The preferred wavelengths for the elements examined are shown in Table I. The technique should be equally applicable to boron and mercury' or, indeed, any element that can be introduced into the plasma.Energy levels involved in the emission of atomic spectra from non-metals are included in Table I and are illustrative of the high energy of the helium plasma. Nitrogen 746.879 I 744-256 742.388 I Oxygen 777.1 93 777.41 4 777.543 I I 740.0 750.0 760.0 770.0 780.0 Fig. 4. metres) Atomic-emission spectra for nitrogen and oxygen (wavelengths in nano- QUALITATIVE ELEMENT-SELECTIVE DETECTION- An artificial mixture containing most elements of interest was made up and run through the system to illustrate the element selectivity (Fig. 5 ) . To compare the selectivity for the various elements, n-heptane was used as a standard carbon and hydrogen containing com- pound, and in Table I1 the selectivity factor given is the ratio of the mass flow-rate of heptane to the mass flow-rate of the element required to give equal signals at the element-selective emission wavelengths.Inter-element effects are dealt with later. TABLE I ELEMENT-SELECTIVE EMISSION WAVELENGTHS AND EXCITATION ENERGIES Element C H D F c1 Br I S P N 0 He Emission wavelengthlnm 247.857 486.133 656.100 6854302 479.454 470.486 516.119 645.388 253.665 '746.879 777.193 587.662 Ionisation state 1 0 0 1 2 2 2 Energy levels/eV - E, + El 12.69 7.69 16.29 12.74 13-98 12.09 16.31 14-60 18.54 16-96 17.03 14.4 13.51 12.11 18.21 16.94 12.08 7.2 12-34 10.73 23.06 26-17 - -436 ,u MCLEAN, STANTON AND PENKETH: A QUANTITATIVE TUNABLE [Analyst, Vol. 98 14 ‘1 i; 1 I, Fig. 5 . Element-selective traces on a test sample for C(b), H(c), D ( d ) , O(e), N(f), F(g), CI(h), Br(i), I ( k ) and S(Z) versus a flame-ionisation detector reference (a).Emission wavelengths are: C, 247.867 nm; H, 656.281 nm; D, 656.100 nm; 0, 777.193 nm; N, 746.879 nm; I;, 686.602 nm; C1,1479.454 nm; Br, 470.486 nm;June, 19731 ELEMENT-SELECTIVE DETECTOR FOR GAS CHROMATOGRAPHY 437 11 7 Fig. 5, continued I, 516.119 nm; and S, 545.388 nm. Throughout, peaks are: 1, deuteroacetone; 2, nitroethane; 3, fluoro- benzene; 4, toluene; 5, n-butyl iodide; 6, n-nonane; 7, chlorocyclohexane ; 8, anisole; 9, diethyl disulphide; 10, octan-2-one ; 11, bromobenzene; 12, o-dichlorobenzene ; 13, o-bromotoluene ; 14, n-undecane438 MCLEAN, STANTON AND PENKETH: A QUANTITATIVE TUNABLE [Analyst, Vol. 98 TABLE I1 DETECTION LIMITS, SPECTRAL BACKGROUND LEVELS AND SELECTIVITY IN DETECTION Element C H D F c1 Br I S P N 0 Detection limit*/ng s-l 0.08 0-03 0.09 0.06 0.06 (0-06) 0.05 (0.05) 0.09 (0-05) 2.9 3.0 0.091 (0.02) - (0.009) Total background as element/ng s-l 0.8 1 0.22 0.17 0.091 0.46 0.72 0.86 1.1 113-0 98.0 - Selectivity ratio versus n-heptane* - - 880 2300 510 (44) 1300 (38) 400 (38) 390 (22) - (1000) - - * Figures in parentheses are values obtained by Bache and LiskZ for detection limits and selectivity ratio vemus phenanthrene.QUANTITATIVE ELEMENT-SELECTIVE DETECTION- Linearity of dual detection system-The flow to the two detectors maintained a constant splitting ratio irrespective of sample size. This property is illustrated in Fig. 6 and demon- strates the linear emission characteristics of the element detector relative to the flame- ionisat ion detector .150 E E 2 100 \ 0, a) JZ Y Q) P .- n 5 50 / I I I K / ' Acetone I FID peak heighthm 150 E -5 100 + J= 0) aJ t Y a L1 .- n 50 n 5 Benzene 0 I I 1 - 0 50 100 FID peak height/mrn Fig. 6. Linear response of the dual detection system on H ( a ) , C(b) and O(c) versws a Emission wavelengths are : flame-ionisation detector reference for acetone and benzene. C, 247.8 nm; H, 486.1 nm; and 0, 771.1 nm Sensitivity and dynamic range-The linear dynamic range is subject to an upper limit, which occurs when the concentration of a component is too large and either the scavenger is insufficient to prevent deposition of carbon or a quenching effect occurs, which perturbs the linear emission characteristics. At the lower end of the range, the limits of detection are subject to the values of the background signal levels at the element-selective wavelengths.Within these limits, the linear dynamic range for fluorine, for example, covers four decades. The best detectable limits and background levels so far obtained are shown in Table 11. The detectable limits given are twice the noise level observed on the background. The background is composed of photo-tube dark current, stray light and spectral contamination due to impurities, and the levels stated are calculated in terms of mass flow-rate for purposes of comparison.June, 19731 ELEMENT-SELECTIVE DETECTOR FOR GAS CHROMATOGRAPHY 439 Determination of atomic ratio and empirical formulae-With a manually tuned, single- channel spectrometer it is convenient to use the element-selective emission as a peak height ratio with the flame-ionisation detector signal.Effectively, this ratio is equivalent to the slopes of the graphs shown in Fig. 6. Let actual slopes be defined as- aC aH 8 0 a T D P a m # K D T ’ * 9 etc* For compound X, where FMpD - is the flow-rate (splitting) ratio to the two detectors, R, is the microwave plasma detector response per gram-atom of carbon, nc is the number of carbon atoms per molecule, n, is the number of gram-moles of compound X and (RF)X is the response factor per gram-mole of compound X on the flame-ionisation detector. FFID Similarly, and, on division- (E) =- n KO x 2 = constant x oxygen to carbon atomic ratio. x Kc nc 0.6 0.5 0.4 -1- 0.3 0 0 0.2 0.1 0 0.1 0.2 0.3 0.4 0.5 0 1 2 3 4 5 6 7 8 9 1 0 Atomic ratio carbon : oxygen Atomic ratio oxygen : carbon I Fig.7. Calibration graphs for oxygen to carbon and carbon to oxygen ratios The constant Ko/Kc is independent of the compound-specific response factor of the flame-ionisation detector and can be evaluated by reference to any known oxygen-containing compound. The independence of the emission signal from molecular properties is illustrated in Fig. 7, where the signal ratio - is a linear function of the known oxygen to carbon ratios in a variety of compounds. The inverse ratio, - , also expresses the carbon number if there is only one oxygen atom per molecule. Further results for hydrogen to carbon ratios, obtained by using ethylbenzene as a single reference standard, are shown in Table 111.Similar behaviour has been noted for the other elements detected selectively by this technique. ao aC X 8 0440 MCLEAN, STANTON AND PENKETH: A QUANTITATIVE TUNABLE [Analyst, Vol. 98 The highly selective detection of the elements carbon, hydrogen, deuterium, oxygen, nitrogen, fluorine, chlorine, bromine, iodine, sulphur and phosphorus in a linearly propor- tional and quantitative manner with good sensitivity has been a continuing feature of this work, and it is now possible to use the technique in order to obtain the empirical formulae of organic compounds separated by gas chromatography as a step towards the identification of compounds. Undoubtedly, the major factor leading to quantitative element-selective detection has been the use of a scavenger gas.As carbon is not appreciably volatile below 3500 "C (boiling-point 4200 "C) and silica melts at 1700 "C (boiling-point 2200 "C), it is to be expected that elemental carbon will plate out on the relatively cold walls. When organic compounds are pyrolysed in the plasma, the action of the scavenger is to hold the carbon in its volatile elemental state- DISCUSSION C +O---+CO . . . . . . * * (1) (ii) co - c + 0 . . .. - - (2) solid, ex pyrolysis gas (4 plasma gas energy frecatom when the necessary oxygen (or nitrogen) scavenging atoms are produced as a result of the plasma discharge. The bond strength of C-0 (256.7 kcal mol-1) is equivalent to 11.1 eV (C-N is equivalent to 7.8 eV) and virtually complete dissociation into separate atoms, as in equation (2), is readily achieved by the plasma energy available. An indication of the energy of the plasma is given in Table I, where energies of 12 to 19 eV are required in order to produce atomic emission from the non-metallic elements.In this work, oxygen (and nitrogen) scavenger gas levels in the plasma gas were kept in the 0.1 to 1.0 per cent. V/V range. Below 0.1 per cent., deposition of carbon was a problem. Above 1.0 per cent., deposition of carbon was not a problcrn, but the amount of carbonaceous material that could be tolerated could exceed that required to overload the linear atomic- emission characteristics of the plasma. This effect marks tlie upper limit of the dynamic range of tlie technique. TABLE I11 HYDROGEN TO CARBON ATOMIC RATIOS FOUND IN HYDROCARBONS H to C ratio found Theoretical H to C ratio Cyclopentane .. .. .. 1.990 2.000 Cyclohexane . . . . .. 2.020 2.000 Cyclooctane . . .. .. 2-023 2.000 Methylcyclohexane . , .. 2415 2.000 Dimethylcyclohexane . . .. 2.012 2.000 Trimethylcyclohexane . . .. 2.019 2.000 Isopropylcyclohexane . . .. 2.008 2.000 Cyclohexene . . .. .. 1.652 1-667 Pent-1 -ene .. .. .. 1.997 2.000 Hex- l-ene .. .. .. 2.052 2.000 Hept-3-ene .. .. .. 2.047 2.000 Oct-l-ene . . .. .. .. 2.027 2.000 Dec-l-ene . . .. .. .. 2.041 2.000 n-Hexane .. .. .. .. 2.347 2.333 n-Heptane.. .. .. .. 2-335 2.286 n-Octane . . * . .. * . 2.300 2.250 n-Nonane . . a . .. .. 2.266 2.222 n-Decane . . .. .. .. 2.249 2.200 n-Undecane . . .. .. 2.251 2.1 82 Benzene .. .. .. .. 0-982 1.000 Toluene .. .. .. .. 1.142 1.143 Ethylbenzene (reference standard) o-Xylene .. .. .. .. %3 %50 The lower end of the dynamic range is set by the noise level of the background signal at the various element-selective wavelengths. While some elements are detected more sensitively than others, in all instances the level of the background signal has a major effectJune, 19731 ELEMENT-SELECTIVE DETECTOR FOR GAS CHROMATOGRAPHY 441 on the sensitivity. This effect is illustrated by the data shown in Fig. 8 and shows good correlation between detection limits and background. It is significant that the least sensitive elements are oxygen and nitrogen, followed by hydrogen and carbon; attention to the following details can noticeably improve performance. The vacuum lines and joints should be tested very carefully for leaks, the gas-chromatographic columns should be pre-conditioned in situ, and the helium gas supply should be dried once more between the pressure regulators and the gas chromatograph by means of tubes containing phosphorus pentoxide.Also of great importance is the provision of very smooth and stable power supplies to the microwave generator and phototube. While the detection limits for some elements can be limited by contamination levels in the plasma, the background limitations on the less common elements are due to continuous radiation (plasma emission and stray light) phototube properties, e.g., dark current, spectral range and sensitivity, monochromator resolution and the optics of the light collection and filtration system. The shape of the emission source is optimised in the form of a cheaply replaceable, l-mm bore, thick walled, clear silica tube.10 t .- I-’ CI % ’ -a L)- I-’ .- 0.1 .- -I 0 4 1 0.001 Background Fig. 8. Effect of background signals on the limit of detection. (These data show that the limit of detection is more a function o f spectral background than element identity.) Basic units are gram- atoms x per second Interference effects are restricted to spurious spectral band emission when the plasma is overloaded but this is instantly recognisable from the magnitude of the flame-ionisation detector response. A chemical effect peculiar to fluorine and chlorine gives rise to phantom oxygen emission through the possible reaction scheme- plasma SiO, + F - SiF + 2 0 wall energy gas plasma - oxygen emission SiF - S + F - silica emission gas atomic This effect makes the detection of oxygen in polyfluoro and polychloro compounds very difficult by this technique. Equally, it inhibits the selective detection of silicon.An alterna- tive plasma tube material to silica would be useful in solving this particular problem.442 MCLEAN, STANTON AND PENKETH The original idea of adding a second, non-selective, detector to act as a reference for comparing the element-selective data has been extended to assist in the quantitative inter- pretation of the data into atomic ratios. Determination of the necessary atomic ratios for an evaluation of empirical formulae can be seen to be a very laborious process if the single- channel form of this technique described here is used. It is very much more efficient to use a multi-channel spectrometer with simultaneous detection of many elements and work is proceeding on the construction of an instrument of this type. Modern techniques of data handling could lend themselves to an automatic print-out of the empirical formulae of organic compounds eluted from a gas chromatograph. Because the amount of a compound is the sum of its atomic parts, a means of quantitative analysis is made available which is not compound-specific in its response and which does not require to be calibrated by use of the compounds being analysed. The authors thank Harry Fraser for his considerable practical contributions to this work. REFERENCES 1. 2. 3. 4. 6. 6. Imperial Chemical Industries Limited, British Patent Applications 20366/70, 41246/70 and 7. Received July 12th, 1972 Accepted January 24th, 1973 McCormack, A. J., Tong, S. S. C., and Cooke, W. D., Analyt. Chem., 1966, 37, 1470. Bache, C. A., and Lisk, D. J., Ibid., 1967, 39, 786. -,- , J . Ass. Off. Analyt. Chem., 1967, 50, 1246. Braman, R. S., and Dynako, A., Analyt. Chem., 1968, 40, 96. Dagnall, R. M., Pratt, S. J., West, T. S., and Deans, D. R., Talanta, 1970, 17, 1009. Bache, C. A., and Lisk, D. J., Avzalyt. Chem., 1971, 43, 960. 41960/70, 1970.

ISSN:0003-2654    

DOI:10.1039/AN9739800432

出版商:RSC    

年代:1973

数据来源: RSC