摘要:

ANALYST AUGUST 1986 VOL. 111 901. Optimisation in Inductively Coupled Plasma Mass Spectrometry Stephen E. Long and Roger M. Brown Environmental and Medical Sciences B55 7 AERE Harwell Oxfordshire OX I I ORA UK The performance of a commercial inductively coupled plasma mass spectrometer has been investigated as a function of a number of parameters associated with sample introduction and plasma operation. The forward r.f. power to the plasma and the nebuliser argon pressure are the two that have the greatest effect. For a variety of elements similar optimum settings for these two variables are found. Optimising the signal due to M+ generally minimises the signal due to M*+ or MO+. Substantial differences between the optima for these variables in ICP-MS compared with ICP-ES are found.Keywords Optimisation; inductively coupled plasma mass spectrometry Inductively coupled plasma mass spectrometry (ICP-MS) is a relatively new technique'-3 based on the hybridisation of a conventional ICP optical emission source and a quadrupole mass spectrometer. The ions produced in the plasma at atmospheric pressure are analysed as a function of their mass to charge ratio rather than by optical emission from the excited state. One of the major benefits of this mode of detection of species in the ICP is improved sensitivity. The ICP is a very efficient ionisation source. However the probability of radiative transfer between energy levels is often low such that for some elements the sensitivity is poor. It therefore can be advantageous to separate physically and count the ions generated.ICP-MS detection limits in solution are generally one to two orders of magnitude better than those with optical techniques,2 with more marked improvements for the heavy metals such as lead. ICP-MS has only recently become commercially available, and therefore there is little information in the literature or integrated experience concerning such aspects as the nature and extent of interferences (both spectral and matrix) analyte transport phenomena through the interface or technique optimisation procedures. Clearly these issues are of impor-tance if accurate and precise data are to be obtained with this technique. In this study we have investigated the effect of varying one at a time the gas flows to the plasma the power coupled to the plasma and the solution uptake rate on the magnitude and behaviour of the signal generated by the analyte the background signal and element oxide and doubly charged species signals.Gray and Date4 have made a preliminary study of these although their data relate to a prototype ICP-MS system. Horlick et al. 5 have also published a preliminary study of the effect of plasma operating parameters in several analyte ion signals. Their work relates to a different ICP-MS production system. They concluded that the two important parameters were aerosol flow-rate and applied power. Most of this work was conducted at a sampling depth of 31 mm which is greater than that used in the study reported here (10 mm). However, they also indicated that sampling depth was not critical on their instrument for the range 17-25 mm.Experimental Apparatus The instrument used in these studies was a commercial ICP-MS system (PlasmaQuad VG Isotopes Winsford, Cheshire UK). The mass spectrometer unit is based on a conventional VG 12-12 quadrupole assembly with a combi-nation of two oil diffusion pumps and two rotary pumps for evacuating the interface and mass spectrometer chambers. The plasma module consists of a horizontally configured torch and matching box (RFA Eastbourne Sussex UK Model BTP 1500) linked to a 27.16-MHz. crystal controlled gener-ator (Plasmatherm Kresson NJ USA Model HFP 1500F). This plasma generator is not the unit fitted as standard to other VG production instruments. The torch is of a Fassel design, with the sample being introduced via a Scott-type double-pass spray chamber.A concentric nebuliser (Meinhard Type TR30-A3) is used in preference to the Jarrell-Ash cross-flow nebuliser supplied as standard with the instrument. The reasons for this change are detailed elsewhere.6 The nebuliser uptake rate is controlled by means of a Minipuls 2 peristaltic pump (Gilson Instruments France) and for these studies was maintained at 0.4 ml min-1. Ions generated within the plasma are extracted through an interface assembly equipped with focusing lens elements and a photon stop to prevent optical interference at the detector. The major component of the interface is the sampling cone, which for these studies is machined from copper. An inlet orifice of 0.45 mm is drilled through the centre of the cone to permit extraction of the ions into the mass spectrometer.This sampling cone has recently been superseded by a titanium nitride-coated nickel cone with a 0.75 mm orifice. The data acquisition system consists of a Channeltron-type electron multiplier (DeTech Brookfield MA USA Model 401) and a multi-channel analyser (Tracor Northern Middle-ton WI USA Model TN1710). An IEEE 488 data bus allows the transfer of data to a central control computer (Olivetti M20) which is also used for data storage and manipulation and for maintaining critical instrument functions. For the experiments reported here the following standard instrument settings were chosen and were kept constant except for the one variable being investigated: Forward power .. . . . . . . . . 1.3 kW Reflected power . . . . . . . <25 W Nebuliser pressure . . . . . . . . 20 lb in-2 (equal to 850 ml min-1 flow) Coolant flow-rate . . . . . . . 13 1 min-1 Auxiliary flow-rate . . . . . . . . 0.4 1 min-1 Solution uptake rate . . . . . . . . 0.4 ml min-I Ion lens settings . . . . . . . . Optimised for (pumped) ZosPb then kept constant The nebuliser flow-rate was measured using a soap-film bubble meter. The argon flow through the nebuliser increases linearly with increasing pressure over the range 14-28 lb in-2, according to the equation F = 50 + 40 P where F is the nebuliser flow-rate (ml min-1) and P is the operating pressure (lb in-2) 902 ANALYST AUGUST 1986 VOL. 111 100 1 Reagents The reagents used were prepared from Specpure materials (Johnson Matthey).Where necessary. standards were pre-pared by serial dilution with de-ionised water. All solutions were prepared in 1 % V/V nitric acid from Fisons Primar-grade reagent. ICP Emission Studies For the emission studies reported here a Plasmatherm 2500 source was used together with a Spex 1-m high-resolution scanning monochromator operated in second order. This instrument has been described in detail elsewhere.’ The sample introduction system used was of the same specification as that reported for the ICP-MS system. Results and Discussion Optimisation of the plasma for optical emission spectrometry has been the subject of other studies.8.9 The major variables that influence the performance are the nebuliser coolant and auxiliary gas flows to the torch the power coupled to the plasma and the observation height of the emitted radiation.It is of course important to decide on the parameter that is to be optimised. In ICP-ES this is usually the ratio of the emission signal to the emission background or noise on the background. Continuum background emission is however often more critically dependent on a plasma variable than is the emission signal from the analyte so that the optimised case is a balance between the two. In ICP-MS the situation is different in that the plasma must be exclusively optimised for the production of a large population of ions and efficient transport of these ions through the interface with minimum inteferences. Continuum background emission from the plasma is not measured directly but will make a minor contribution to the mass spectral background from scattering processes.It is expected that for ICP-MS those variables which primarily affect the properties of the energy input region of the plasma will be most important. Graylo has reported that changes in the nebuliser gas flow will alter the ion energies leading to changes in signal. In this study the sampling position was kept constant owing to mechanical limitations. However Gray and Date4 have studied the effect of sampling depth on a similar system and recommended a value of 10 mm for normal operation which is the distance used in this study. Effect of Gas Flows The effect on the signal from 208Pb of varying the coolant and auxiliary gas flows is shown in Figs.1 and 2 respectively. In these and all other figures the points plotted are the means of five measurements. The relative standard deviations are better than 10% in all instances. The vertical axes are in units of thousands of counts per second. In common with ICP-ES, the coolant flow-rate does not significantly affect the signal across the range normally used for sustaining the plasma discharge. The auxiliary gas flow produces a flat signal response over the range 0-0.6 1 min-1 after which the signal decreases more rapidly. A low flow-rate just sufficient to keep the plasma from heating the torch injector tube is therefore advisable for maintaining stability. Effect of Nebuliser Pressure A plot of the effect of varying the nebuliser pressure while the r.f.generator forward power is held constant on the signal from several ions is presented in Fig 3. These ions were 1 I I I I I I 10 11 12 13 14 15 16 70 9 Coolant flow/l min-’ Fig. 1. 1.3 kW; 0.1 pg ml-1 2OSPb Effect of coolant gas flow-rate on signal in ICP-MS. Power, 80 I 0 0.2 0.4 0.6 0.8 1 .o 1. 40 Auxiliary gas flow/l min-1 Fig. 2. 1.3 kW; 0.1 yg ml-’ 20SPb Effect of auxiliary gas flow-rate on signal in ICP-MS. Power Nebuliser pressure/l b in - 2 Fig. 3. kW; 0.1 pg ml-1 A Pb; B Cs; C Ba; D Zn; E Sm; and F B Effect of nebuliser pressure on ICP-MS signal. Power. 1.3 chosen to represent a cross-section of mass first and second ionisation potentials and chemistries. The response curves show a critical dependence of mass spectral count rate on the operating pressure with an optimum in the region of 18-22 Ib in-2.This would be predicted because a variation of ga ANALYST AUGUST 1986 VOL. 111 90 80 .g 70 P 5 6 0 -2 o 50 n r 40-+-U 0 Y 0 C 0 30-20 10 903 ------flow-rate into the injector channel of the ICP can substantially alter the temperature and hence the ion population.11 It is also expected that there would be a balance between decreasing temperature as the pressure is increased and an increasidg mass transport rate of the analyte. The effect is however, more pronounced than in ICP-ES. The peak shape is noteworthy being distinctly Gaussian in shape with a slightly rounded flat top. This implies that for a Meinhard nebuliser small fluctuations in nebuliser gas pressure (and hence gas flow) cause only small changes in the M+ signal.This behaviour may not be shown by other types of loo r---n l I I I I I I 12 16 20 24 28 32 36 40 Nebuliser pressure/lb Fig. 4. Effect of nebuliser pressure on ICP emission signal. Power, 1.3 kW; entrance slit 50 pm; exit slit 10 pm; viewing height 16 mm above load coil. A Cu I 324.75 nm; and B Cu 11 224.70 nm nebuliser; indeed the manufacturer is recommending that a mass flow controller be fitted to the nebuliser gas flow line when using the standard cross-flow nebuliser. The short-term precision on our instrument fitted with a Meinhard nebuliser is generally better than 1% on signals in excess of 100 000 counts s-1. There is no correlation between the first ionisation potential and the optimum pressure for the six elements investigated.Operating at 19-20 Ib in-2 allows multi-element analysis within 10% of the optimum sensitivity. A comparative study of the effect of nebuliser pressure on the ICP emission signal to background for the copper atom line at 324.75 nm and the copper ion line at 224.70 nm is presented in Fig. 4. Copper was chosen as it exhibits typical emission properties. It is clear that in both instances the signal to background ratio has a much flatter response to variations in pressure than that in ICP-MS. Effect of Power Fig. 5 illustrates the effect of power applied to the plasma on the count rate obtained from 2ogPb 152Sm and 1lB. In this experiment the nebuliser was operated at its apparent optimum pressure of 19 Ib in-2.The count rate increases as the applied power is increased with a maximum in the range 1.2-1.4 kW. Above 1.4 kW the signal apparently decreases, which could be taken to imply that power in excess of 1.4 kW would not be necessary. Gray and Date4 have also observed this effect on 115In concluding that there was a maximum count rate at 1.6 kW and an optimum signal to noise ratio at 1.4 kW. Their work relates to their prototype system. Fig. 6 represents a two critical parameter map of instrument response as a function of nebuliser operating pressure and applied power. Clearly as the power is increased the optimum nebuliser pressure also increases. This behaviour was also observed by Horlick et al.5 For the equipment installed in this laboratory the maximum practical output power of the generator for routine use was found to be approximately 1.35 kW.It was therefore not possible to extend the study to higher powers than this. Gray and Date4 rightly pointed out, however that at very high powers (>1.4 kW) there would be excessive heat dissipation in the interfacial expansion stage, which might lead to operational problems. Effect of Solution Uptake Rate The solution uptake rate is largely determined by the pressure drop at the nebuliser end of the uptake tube. If a pump is included in the system it is possible to vary the uptake 0.9 1.0 1.1 1.2 1.3 1.4 Forward power/kW Fig. 5. Effect of power on signal in ICP-MS. Concentration 0.1 pg r n - I ; nebuliser pressure 19 Ib in-'.A 20xPb; B 152Sm; and C. I1B Nebuliser pressurellb in-2 Fig. 6. Two-dimensional plasma parameter map for ICP-MS. Concentration 0.1 pg ml-1 ZOXPb. Power A 1.4 kW; B 1.3 kW; C. 1.2 kW; and D 1.1 k 904 ANALYST AUGUST 1986 VOL. 111 rate as desired. This is useful from the points of view both of handling small solution volumes and of maintaining the uptake rate at a constant value when solutions with high dissolved solids or high viscosity are to be aspirated. Browner et a1.12 found that for ICP-ES there is little change in the emission signal for a wide range of solution uptake rates. They proposed that the increase in nebuliser mass transport as the flow-rate is increased is balanced by a lower nebuliser transport efficiency so that the net mass transport rate to the plasma is roughly constant.The effect of varying the solution uptake rate in the range 0.15-1.4 ml min-1 for the ICP-MS system is shown in Fig. 7. Above 0.6 ml min-1 there is little variation in the count rate. Below this value the signal reaches a peak at 0.4 ml min-1, with a decreased signal at lower flow-rates. The optimum flow-rate for the concentric nebuliser was chosen to be 0.4 ml min-1. The signal to background ratio was also a maximum at this value. These data suggest that there is little point in operating the nebuliser at its natural uptake rate (2.2 ml min-1) because larger solution volumes are required with no attendant increase in detection capability. The routine use of a pump is therefore recommended.Effect of Plasma Variables on Background An important difference between ICP-ES and ICP-MS is that the background in the former originates from continuous optical emission in a line of sight from the plasma whereas in the latter it appears to originate mostly from scattered photons or stray ions in the mass spectrometer. The effect of applied power on the background in the middle of the mass spectrum at 130 a.m.u. is shown in Fig. 8. There was also found to be little variation in the absolute magnitude of the background for the range 80-240 a.m.u. A linear response to applied power is apparent. If the signal to background ratio (Table 1) is calculated using data for 208Pb it is apparent that there is much benefit in the use of high powers in order to achieve low detection limits.This contrasts strongly with a comparable study using ICP-ES (Table 2) in which the signal to background ratios decrease as the power is increased. This may be attributed to the stronger dependence of background on applied power than the analytical signal. A plot of the variation of background signal with nebuliser pressure is shown in Fig. 9. The background signal decreases as the nebuliser pressure is increased. Presumably at high operating pressures the increased gas flow-rate lowers the temperature of the plasma and hence the number of scattered photons that are detected within the mass spectrometer. Using data obtained in this work the optimum signal to background ratio at 1.3 kW would be obtained at a nebuliser pressure of 19-20 lb in-2 compared with the 20-24 lb in-2 found in our emission studies (Fig.4). Effect of Plasma Variables on Oxide and Doubly Charged Ions Three reasons for the fall in signal from a singly charged atom ion (M+) as the nebuliser pressure or r.f. forward power move away from optimum can be postulated. Firstly it may be caused by a reduction in ion formation as a result of less efficient energy transfer from the r.f. load coil via the plasma, to the analyte. Secondly it could be due to the formation of neutrally charged species which will not be detected. Thirdly, the reduction in M+ species could be a result of the formation of other ions such as MO+ and M2+. These ions are frequently encountered in elemental mass spectrometry and can be the principal species causing spectral interference.Ions such as these are also observed in ICP-MS. The M+/MO+ and M+/M2+ ratios for caesium samarium, barium and lead with varying nebuiliser pressure or varying forward power are shown in Figs. 10 11 and 12 respectively. loo 77 20 0 "L 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Flow ratelm1 min-' Fig. 7. nebuliser operating pressure 17 lb in-2; 0.1 pg ml-1 zOxPb Effect of solution uptake rate in ICP-MS. Power 1.3 kW; 120 100 VI a al $ 80 h 2 60 0 c. c. 40 20 1 .o 1.1 1.2 1.3 Power/kW Fig. 8. Effect of power on background signal in ICP-MS. Mass 130 Table 1. Effect of power on the signal to background ratio in ICP-MS. Isotope 20sPb; total lead concentration 100 pg 1-I; nebuliscr pressure 16 lb in-2 (690 ml min-1) Signal (S)/ Background (B)/ PowerlkW counts s- counts s-l SIB 1.1 20 000 52 380 1.2 43 000 80 540 1.3 113 000 113 1000 Table 2.Effect of power on signal to background ratio in ICP-ES (arbitrary units). Data obtained for lead 11 220.359 nm 10 pg ml-I. Entrance slit 50 pm; exit slit 10 pm; viewing height 16 mm. Background measured on blank at line centre Signal Background Power/kW (S) (B) S/B 1.1 6.0 0.45 13.3 1.2 8.3 0.70 11.9 1.3 9.3 0.90 10.3 The optimum nebuliser pressure is again in the range 16-20 Ib in-2 with M+/M2+ ratios in excess of 100 (i.e. M2+ is 1% or less of M+) and M/MO+ ratios in excess of 300. With caesium, the Cs2+ and CsO+ levels are both less than 0.1% at the optimum pressure. As the pressure moves off-optimum the ratios fall principally as a result of a reduction in the M+ population rather than any significant rise in M2+ or MO+.Whereas for both M+/M2+ and M+/MO+ ratios an optimum is observed in the range 18-20 Ib in-2 for caesium samarium and lead the ratios for barium continue to rise as the pressure falls. This indicates that the lower the pressure the better ANALYST AUGUST 1986 VOL. 111 905 loo( 80 -7 v) u a, n . 4-60- c 3 0 0 40 -1 24 20 12 14 16 18 20 22 Pressurellb in-2 Fig. 9. Power 1.3 kW; mass 130 Effect of nebuliser pressure on background signal in ICP-MS. 2400 2000 1600 0 r_ 3 1200 w-0 m .- c a 800 400 0 12 14 16 18 20 22 24 26 Nebuliser pressure/lb inp2 Fig. 10. Ba+/BaO+; B Pb+/PbO+; C Cs+/CsO+; and D Sm+/SmO+ M+/MO+ ratio as a function of nebuliser pressure.A, However as both ratios for barium are greater than 1000 at 16 lb in-2 pressures below this level would not be required in routine analysis. Generally both ratios rise for barium samarium and lead as the forward r.f. power increases the notable exception being the Pb+/PbO+ ratio which peaks at 1.2 kW largely owing to the fall in the absolute Pb+ signal at high forward power (Fig. 5). The low Sm+/Sm2+ and Ba+/Ba2+ ratios at 1.3 kW can be offset to some extent by reducing the nebuliser pressure from the 20 lb in-2 used in this study to 18 lb in-2 (Fig. 11). In general the data presented here are consistent with those published by Horlick et al.5 However they found that for elements with a low second ionisation potential (Sr Ba) the M+/M2+ ratio falls as the nebuliser pressure and applied power to the plasma are increased.This is opposite to the behaviour observed in this study. The reasons for this are not immediately explainable although the two instrumental systems and conditions used are very different. It is clear that 1200 800 0 .- 4-a 400 16 18 20 22 24 Nebuliser pressurellb in-* Fig. 11. Ba+/Ba2+ ( x 10); B Pb+/Pb2+; C Cs+/Cs2+; and D Sm+/Sm2+ M+/MZi ratio as a function of nebuliser pressure. A, 1200 1000 800 0 .-600 a 400 200 -PowerlkW Fig. 12. M+/M2+ and M+/MO+ ratios as functions of applied power. A Ba+/Ba2+ ( x 10); B Ba+/BaO+; C Pb+/Pb2+; D Pb+/PbO+; E, Sm+/Sm2+ ( ~ 1 0 ) ; F Sm+/SmO+; and G Sm+/Sm2+ further work is required in order to consolidate a general model for this type of behaviour that is applicable to all systems.Quadrupole Optimisation For the equipment used in these studies several lens and extraction voltage supplies were available for optimising the transport of ions into the quadrupole of the mass spec 906 ANALYST AUGUST 1986 VOL. 111 trometer. Whereas variations in the count rate as a result of altering the plasma variables could be restored to some extent by altering these settings the general trends and values of the optima were unaffected by control of these voltages. In practical analysis of course the analyst does not wish to re-optimise the lens voltages for each sample. Conclusions We have investigated the variable parameters associated with sample introduction and plasma operation in order to obtain the optimum signal signal to background and M+/M2+ and M+/MO+ ratios.For a number of elements covering a wide range of atomic mass first and second ionisation potentials and chemistries very similar optimum conditions have been found. A nebuliser argon pressure in the range 18-22 lb in-2 produces the optimum signal with the Meinhard nebuliser used. Small pressure fluctuations about the optimum cause a small variation in the absolute signal. Pressures in this range also maximise the M+/M2+ and M+/MO+ ratios although operation at the lower end of this range improves the M+/M2+ ratio for elements such as barium with a low second ionisation potential. In general high forward r.f.powers are preferred to low powers Moving off-optimum in either nebuliser pressure or forward power generally results in a fall in the M+ signal, rather than a significant rise in the M*+ or MO+ signals (there are however notable exceptions). The data presented here and those published by Horlick et al.,5 are generally consistent despite the application of different ICP-MS systems with significantly different interface designs. However a notable exception to this is the behaviour of M+/M2+ ratios with varying nebuliser pressure or applied power. For the hardware system considered here the optimum conditions may be summarised as follows: Forward r.f. power . . . 1.3 kW Solution uptake rate (pumped) Auxiliary gas flow-rate . . . . 0.4lmin-1 Coolant gas flow-rate .. . . 13 1 min-1 Nebuliser pressure . . . . . 18 lb in-2 (770 ml min-1) These conditions offer a meaningful compromise to allow sensitive multi-element analysis with minimum spectral inter-ference from MI+ or MO+ ions. If optimum sensitivity is required and spectral interference is not important then a nebuliser pressure of 20 lb in-2 may be more appropriate. A high degree of stability of these five variables is required in order to give stable absolute signals and stable M+/M2+ and M+/MO+ ratios. This is especially true of the forward r.f. power and the nebuliser pressure. Although a mass flow controller on the nebuliser argon (using a Meinhard nebuliser) may not be required to produce a stable M+ signal it may be needed to stabilise the M+/M2+ and M+/MO+ ratios for some elements.0.4 ml min-1 If the optima listed above are compared with those usually encountered in ICP-ES there are some obvious differences. A knowledge of ICP-ES operation is beneficial but may not be directly transferable to ICP-MS. Addendum It has recently been brought to our attention that the data presented in this paper relating to oxide and doubly charged ion ratios have recently been quoted at a conferencei3 suggesting that they represent the optimum obtainable with a production PlasmaQuad. This conclusion should not be drawn from our data. We would like to reiterate that the purpose of this study was to ascertain the optimum plasma parameters and to note the effect of changing these operating conditions. The study was not designed to maximise the absolute value of ion count rate or ion-intensity ratios.We also wish to point out that the above data were obtained in mid-1985 on the third PlasmaQuad produced and under-stand that VG Isotopes have recently announced changes to the design of their interface that substantially reduce the relative levels M2+ and MO+ ions. The work described in this paper was undertaken as part of the Underlying Research Programme of the UKAEA. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. References Gray A. L. and Date A. R. Int. J. Mass Spectrom. Ion Phys. 1983 46 7. Date A. R . and Gray A. L. Analyst 1983 108 159. Houk R. S. Svec H. J . and Fassel V. A. Appl. Spectrosc., 1981 35 380. Gray A. L. and Date. A. R. Analyst 1983. 108 1033. Horlick G. Tan S. H. Vaughan M. A. and Rose C. A., Spectrochim. Acta Part B 1985 40 1555. Long S. E. and Brown R. M. Report AERE R-11863 HM Stationery Office London 1985 in the press. Pickford C. J. and White G. F. Report AERE R-10833 HM Stationery Office London 1983. Ebdon L. Cave M. R. and Mowthorpe D. J . Anal. Chim. Acta 1980 115 179. Moore G. L. Humphries-Cuff P. J. and Watson A. E . , Spectrochim. Acta Part B 1984 39 915. Gray A. L. paper presented at the European Winter Conference Leysin Switzerland 1985. Long S. E. unpublished work. Browner R. F. Boorn A. W. and Smith D. D. Anal. Chem. 1982 54 1411. Douglas D. Paper presented at Groupernent pour 1’Avance-rnent des MCthodes Spectroscopiques (GAMS) Paris May 1986. Paper A51343 Received September 24th 1985 Accepted March 17th 198

ISSN:0003-2654    

DOI:10.1039/AN9861100901

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, AUGUST 1986, VOL. 111 907 Spectrofluorimetric Determination of Titanium with 2-Met hyl-5- hydroxy-7-met hoxyisof lavone Takushi Ito* College of Engineering, Shizuoka University, Johoku, Hamamatsu, 432, Japan Yoshiaki Tsubomatsu, Tetsuo Suzuki and Akira Murata Facu It y of Engineering, Sh izuo ka University, Jo h o ku, Ham am a tsu, 432, Japan Titanium reacts with hydroxy and methoxy derivatives of 5-hydroxyflavone and 5-hydroxyisoflavone to form water-insolu ble complexes that can be extracted into carbon tetrachloride. The absorptive and fluorescent characteristics of the titanium complexes of these derivatives have been studied. The best fluorescent reagent for the determination of titanium is 2-methyl-5-hydroxy-7-methoxyisoflavone. The maximum wavelengths of the excitation and emission spectra of the titanium complex in carbon tetrachloride are 390 and 520 nm, respectively.Titanium can be determined in the range 0.05-1.3 pg ml-1 when extracted from the solution at pH 7.0-7.5 into carbon tetrachloride. Keywords: Titanium determination; fluorescence analysis; 2-meth yl-5-h ydroxy-7-methoxyisoflavone; 5-h ydroxyflavone derivatives; 5-h ydroxyisoflavone derivatives Various spectrophotometric methods for the determination of titanium have been reported,' but only a few of these methods are spectrofluorimetric,*-7 as complexes of transition metals with organic reagents generally do not fluoresce. The only organic reagents that form a fluorescent complex with titanium are benzal-2-amino-3-nitrilo-4,5-diphenylfuran, dimethoxybenzal-2-amino-3-nitrilo-4,5-diphenylfuran,~ 2- methyl-3-ethyl-5-hydroxychr~mone~ and biacetyl monoxime nicotinylhydrazone.6.7 We have previously reported the spec- trofluorimetric determination of titanium with 2-methyl-3- ethyl-5-hydroxychromone as a complexing reagent.4 In that investigation, we found that the fluorescence intensity of the titanium complexes varied markedly with the insertion of some substituents at different positions on the 5-hydroxychro- mone. This paper describes the fluorescent reactions of titanium with alkyl and methoxy derivatives of 5-hydroxyflav- one and 5-hydroxyisoflavone, which are a 2-phenyl and a 3-phenyl derivative of 5-hydroxychromone, respectively, and the spectrofluorimetric determination of titanium with 2-methyl-5-hydroxy-7-methoxyisoflavone. Experimental Reagents A mass of 0.143 g of titanium dioxide (Johnson Matthey Chemicals, Specpure) was dissolved by heating with 16 g of ammonium sulphate and 42 ml of concentrated sulphuric acid.After cooling, the resulting solution was diluted to 500 ml with water. A working solution of 0.05 M sulphuric acid was prepared from a standard solution. 2-Methyl-5-hydroxy-7-methoxyisofl avone and the other 5-hydroxyisoflavone and 5hydroxyflavone derivatives listed in Table 3 were synthesised as described earlier.839 These reagents were used as methanolic solutions. All other chemicals were of analytical-reagent grade. Apparatus Fluorescence spectra and fluorescence intensities were obtained with a Hitachi Model 650-10s spectrofluorimeter fitted with a xenon lamp. Fluorescence spectra were not corrected.An aqueous solution (0.15 pg ml-1) of sodium * To whom correspondence should be addressed. fluorescein was used to adjust the sensitivity of the spectro- fluorimeter. Absorption spectra were recorded with a Hitachi Model 124 spectrophotometer and absorbance measurements were obtained with a Hitachi Model 101 spectrophotometer. Quartz cells (10 x 10 x 45 mm) were used for all measurements. A Toadenpa Model HM-1OB pH meter was used for pH measurements. Procedure To a sample solution containing 0.5-13 pg of titanium, add 8 ml of methanolic solution (5 x 10-3 M) of 2-methyl-5-hydroxy- 7-methoxyisoflavone, 4.5 ml of methanol (the final methanol content should be 50% VIV), 2.5 ml of 1 M ammonium acetate solution and a sufficient amount of 1 M ammonia solution to adjust the pH to 7.0-7.5.Dilute the mixture to 25 mi with water. After about 1 h, extract the titanium complex with 10 ml of carbon tetrachloride by shaking vigorously for 2 min. Separate the organic phase and dry over sodium sulphate. Irradiate the solution with 410 nm radiation and measure the fluorescence intensity at 520 nm. Results and Discussion Titanium reacts with 2-methyl-5-hydroxy-7-methoxyisoflav- one to form a water-insoluble complex that is soluble in aqueous solutions containing not less than 80% V/Vmethanol, 80% V/V methyl Cellosolve or 70% V/V dimethylformamide, although these solutions scarcely fluoresce. The complex is insoluble in aqueous solutions containing 80% V/V dimethyl sulphoxide, or dispersing agents such as Triton X-100, poly(viny1 alcohol), Tween 80 (Wako) and Brij 35 (Wako).However, a titanium complex extracted into organic solvents exhibits fluorescence. The fluorescence intensities of the complex and the reagent blank extracted into organic solvents are shown in Table 1. Carbon tetrachloride is seen to be the most suitable solvent. Some types of organic solvents, especially hexane and cyclohexane, gave an insoluble reagent at the phase boundary. We also studied the influence of the addition of ion-association reagents of different natures (cationic. zephiramine and Brilliant Green; anionic, sodium lauryl sulphate and erythrosine) on the fluorescence intensity, but they had no effect on the fluorescence intensity of the complex after extraction.908 .- c m - $ 1 0 - ANALYST, AUGUST 1986, VOL.111 - Table 1. Effect of extraction solvents. Titanium, 2 X 10-7 M ; reagent, 4 X 10-5 M; pH, 7.5; solvent, 10 ml Relative fluorescence intensity Solvent Carbon tetrachloride Xylene . . . . . . Cyclohexane . . . . Chlorobenzene . . Toluene . . . . . . Benzene . . . . . . Diisopropyl ether . . Hexane . . . . . . Chloroform . . . . 4-methylpent an-2-one Titanium complex . . 54.3 . . 34.8 . . 33.0 . . 26.5 . . 21.1 . . 20.8 . . 9.2 . . 9.0 . . 3.8 . . 0.3 Reagent 5.7 4.3 4.8 4.8 5.6 4.8 5.0 3.0 5.5 5.0 I I 0' I I I I 350 400 450 500 550 600 Wavelengthhrn Fig. 1. Fluorescence spectra of the titanium complex of 2-methyl-5- hydroxy-7-methoxyisoflavone in carbon tetrachloride. A and B are excitation spectra of (A) titanium complex and (B) reagent; C and D are corresponding emission spectra Fluorescence Spectra The excitation and emission spectra of the titanium complex and the reagent in carbon tetrachloride are shown in Fig.1. The excitation and emission spectra have maxima at 390 and 520 nm, respectively. An excitation wavelength of 410 nm was selected for further study, as the reagent strongly absorbs exciting light below 400 nm. Effect of Reaction Variables The effect of the pH of the aqueous phase on the extraction of the titanium complex into carbon tetrachloride is shown in Fig. 2. The maximum constant fluorescence intensity is obtained in the pH range 7.0-8.5, although the fluorescence intensity of the reagent is constant over a wide pH range. In order to prevent the hydrolysis of other metallic ions, a pH of 7.0-7.5 is recommended for the determination of titanium.The effect of the concentration of the reagent was exam- ined. The maximum fluorescence intensity is obtained in a 4 x 10-4-16 x M reagent solution. Higher reagent concentrations cause a decrease in the fluorescence intensity, probably because of an inner filter effect. Taking into account the consumption of the reagent with co-existing ions, 8 ml of 5 X 10-3 M reagent solution were added. The reagent must be dissolved in organic solvents as it is insoluble in water. Methanol, ethanol, acetone and dioxane can easily dissolve the reagent. The titanium complex in the aqueous solution containing acetone or dioxane is hardly 1 a - . a - - - - - .B 4 5 6 7 8 9 1 0 0 I 1 1 I I I PH Fig.2. Effect of pH on the extraction of the titanium complex of 2-methyl-5-hydroxy-7-methoxyisoflavone into carbon tetrachloride. A, Titanium complex (2 X mol Ti); B, reagent (4 X 10-5 mol) Table 2. Effects of foreign ions on the determination of 4.79 pg of titanium Tolerance limit/kg Ion 1 10 100 Phosphate, EDTA, Be, Al, Sc, Ga, Y, Zr, Hf, Cr(III), Citrate, tartrate, In, Ce(III), Sn(IV), Sb(III), Bi(III), F-, oxalate, Cu(II), Zn, La, Pb, As(V), Co(II), Fe(II1) Mn(I1) Ni(II), Pd, Pt(IV) 1000 Mg, Ca, V(V), Mo(V1) 10000 so42- 100000 N03-, C1- extracted into carbon tetrachloride. On the other hand, the titanium complex in a methanolic or an ethanolic aqueous solution is extracted into carbon tetrachloride and fluoresces. A study of the effect of the methanol content of the aqueous phase on the extraction of the complex showed that the optimum content is 50% V/V.Contents lower than about 40% give an insoluble reagent at the phase boundary, whereas contents above 60% cause a decrease in the fluorescence intensity, probably because of a reduced extraction efficiency. The effect of the ethanol content on the extraction of the complex showed the same tendency as that of methanol but the fluorescence intensity of the titanium complex extracted from an ethanolic solution is only about 60% of that from a methanolic aqueous solution. The effect of the standing time before extraction on the fluorescence intensity was examined. The fluorescence inten- sity increases gradually up to 30 min, after which it remains constant up to 120 min.The titanium complex is readily extracted into carbon tetrachloride by shaking for 2 min, and the fluorescence intensity remains constant for at least 120 min after extraction. Calibration Graph Under the recommended conditions, the calibration graph is linear over the range 0.5-13 pg of titanium per 10 ml of carbon tetrachloride, The coefficient of variation obtained from six measurements of 4.79 pg of titanium is 1.8%. The sensitivity of the proposed method is approximately the same as that of the 2-methyl-3-ethyl-5-hydroxychromone method reported earlier,4 and the linear range of the calibration graph of the proposed method is wider than that of the former method. Further, the sensitivity of this method is approximately theANALYST, AUGUST 1986, VOL.111 909 Table 3. Absorptive and fluorescent characteristics of titanium complexes in carbon tetrachloride Reagent Absorption Fluorescence spectra spectra E x 10Y h,,,,/nm 1 mol-1 cm-1 h,,./nm h,In./nm Flavone 7.7 420 555 7.2 410 555 8.8 410 555 6.8 410 550 5-Hydroxy- . . . . . . . . . . . . 410 3-Methyl-5-hydroxy- . . . . . . . . 420 3-Ethyl-5-hydroxy- . . . . . . . . . . 400 5-Hydroxy-7-methoxy- . . . . . . . . 400 Is 9oflavone 5-Hydroxy- . . . . . . . . 2-Methyl-5-hydroxy- . . . . 2-Ethyl-5-hydroxy . . . . . . 2-Me thyl-5-hydroxy-7-me thoxy- 2-Ethyl-5-hydroxy-7-methoxy- . . . . . . . . . . . . . . . . . . . . . . 395 392 392 380 380 2.3 9.5 9.7 7.5 9.8 405 400 402 390 385 534 532 530 520 490 Fluorescence intensity* Complex Reagent 2 12 1 13 2 7 7 13 3 82 100 88 60 13 32 33 15 187 * Relative to the 2-ethyl-5-hydroxyisoflavone. same as that of the other spectrofluorimetric methods for the determination of titanium.2 3 ~ 6 Effect of Foreign Ions The effect of foreign ions on the determination of 4.79 pg of titanium is summarised in Table 2. Phosphate, EDTA, citrate and tartrate cause serious negative errors. The positive interference of beryllium, indium, tin and antimony can be attributed to the fact that these elements also form extractable fluorescent complexes. Various cations give negative errors, probably by forming non-fluorescing complexes or by adsorb- ing the titanium complex on colloidal hydrolysis products. Accordingly, for practical analyses, preliminary separations will be essential.Composition of Complex Spectrophotometric and spectrofluorimetric continuous varia- tions methods were attempted in carbon tetrachloride, but the molar ratio could not be determined because at low concen- trations the signals were extremely faint and at high concen- trations the reagent or the hydrolysis product of titanium was deposited at the phase boundary. The composition of the titanium complex in carbon tetrachloride was examined in the following way. The titanium concentration was kept constant at 8 x 10-6 M and the reagent concentration was varied from 1 x 10-4 to 16 x 10-4 M. A series of solutions containing titanium and the reagent were prepared at pH 7.0, and the absorbances of the titanium complex extracted into carbon tetrachloride were measured. The molar ratio of titanium to ligand was found to be 1 : 2 by the methods of Bent and French10 and Benesi and Hildebrand.” Comparison of Substituent Effects Titanium reacts with 5-hydroxyflavone, 5-hydroxyisoflavone and their derivatives to form water-insoluble complexes, which are extracted into carbon tetrachloride.The wavelengths of maximum absorption, the molar absorptivities and the maximum excitation and emission wavelengths of the titanium complexes at pH 7.5 are summarised in Table 3, together with the relative fluorescence intensities measured at the maximum wavelengths of the relevant spectrum. Titanium complexes of 5-hydroxyisoflavone, 5-hydroxyflavone and their derivatives show a very weak fluorescence, but the insertion of an alkyl group (methyl < ethyl) at the 2-position of 5-hydroxyisoflavone increases the fluorescence consider- ably.As reported earlier,4 the insertion of an alkyl group (methyl < ethyl) at the 2-position of 5-hydroxychromone also increases the fluorescence intensity of the titanium complex. The insertion of a methoxy group at the 7-position of 2-methyl-5-hydroxyisoflavone slightly increases the fluor- escence, whereas the insertion of a methoxy group at the 7-position of 2-ethyl-5-hydroxyisoflav~ne decreases the fluor- escence. The effect of an insertion of a methoxy group at only the 7-position of 5-hydroxyisoflavone is still uncertain. 2-Methyl-5-hydroxy-7-methoxyisoflavone was selected for the detailed study described above, taking into account the fluorescence intensity of the reagent. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. References Snell, F. D., “Photometric and Fluorimetric Methods of Analysis, Part 2,” Wiley, New York, 1978, p. 1077. Titkov, Yu. B . , Ukr. Khim. Zh., 1969, 35, 887. Tashkhodzhaev, A. T., Zel’tser, L. E., Sabirova, T., and Morozova, L. A., Dokl. Akad. Nauk SSSR, 1976,32. Ito, T., and Murata, A., Anal. Chim. Acta, 1980, 113, 343. Luque de Castro, M. D., and Valcarcel, M., Talanta, 1980,27, 645. Cejas, M. A., Gomez-Hens, A. , and Valcarcel, M., Anal. Chim. Acta, 1984, 158, 287. Rubio, S . , Gomez-Hens, A., and Valcarcel, M., Anal. Chem., 1985,57, 1101. Murata, A., Tominaga, M., Inoue, H., and Suzuki, T., Bunseki Kagaku, 1973,22, 179. Murata, A., Torninaga, M., and Suzuki, T., Bunseki Kagaku, 1974, 23, 1349. Bent, H. E . , and French, C. L., 1. Am. Chem. SOC., 1941,63, 568. Benesi, H. A., and Hilderbrand, J. H., J . Am. Chem. SOC., 1949, 71,2703. Paper A6145 Received February 13th, 1986 Accepted March 6th, 1986

ISSN:0003-2654    

DOI:10.1039/AN9861100907

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, AUGUST 1986, VOL. 111 91 1 Micro-determination and Separation of Molybdenum Using a Liquid Ion Exchanger Sobhana K. Menon and Yadvendra K. Agrawal Analytical Laboratories, Pharmacy Department, Faculty of Technology and Engineering, M.S. University of Baroda, Baroda, India A sensitive and selective method for the micro-determination of MOW) has been developed, involving selective extraction of the yellow MOM) - N-phenylbenzohydroxamic acid - SCN- complex with a liquid ion exchanger, Aliquat 336, from a medium of 2-5 M hydrochloric acid. The molar absorptivity of the complex is 7.2 x lo4 I mol-1 cm-1 at 350 nm and the colour system obeys Beer's law in the range 0.34-10 p.p.m. of Mo(VI). The extracted metal can be quantitatively back-extracted with 0.3 M ammonia solution, thus rendering the method applicable for the concentration, determination and separation of MOW) in samples containing very low levels of the metal. Molybdenum has been determined in standard steel samples and in a plant sample. Keywords: Molybdenum determination; spectrophotometry; liquid ion exchanger; steel analysis; plant ana I ysis The extraction of coloured species of a metal ion with a liquid ion exchanger enhances the extractability and sensitivity considerably1.2 by the formation of a complex coloured species within the ion exchanger phase, and also permits the direct detection of the metal and its pre-concentration, separation and determination.This method obviates the need for a stripping step to complete the analysis and provides a means of separation when more than one metal is extracted, by stripping one metal while retaining the other for determi- nation in the organic phase.Quantitative recovery of the extracted metal can be effected by choosing a suitable stripping agent. A survey of the literature of the last decade indicated that little work has been reported on the extraction of Mo(V1) with liquid ion exchangers and most of the extractions were from mineral acid media, which are less selective.3-9 Recently an extraction with 8-hydroxyquinoline-5-sulphonate has been reported, which provides for direct spectrophotometric deter- mination in the organic phase.10 Hydroxamic acids are excellent reagents, used for the sensitive and selective determination of metals.11,12 A few hydroxamic acids have been reported for the determination of Mo(VI),13-18 mostly from acidic media.The yellow complex of Mo(V1) with N-phenylbenzohydroxamic acid (PBHA)19 was found to be extracted instantaneously into Aliquat 336, with a considerable enhancement of the selectivity and sensitivity, which can be applied to the determination of microgram amounts of Mo(V1) and to its separation from closely associated metals. Experimental Chemicals and Reagents All chemicals were of AnalaR or general-reagent grade from BDH Chemicals and Merck, unless specified otherwise. N-Phenylbenzohydroxamic acid (PBHA). PBHA was syn- thesised by the method of Priyadarshini and Tandon20 and its purity was checked from its melting-point and IR and UV spectra. A 1% solution of PBHA in ethanol was prepared.Molybdenum( VI) standard solution. A stock solution of Mo(V1) was prepared by dissolving the required amount of ammonium molybdate tetrahydrate in doubly distilled water. The solution was standardised gravimetrically with 8-hydroxy- quinoline21 and the metal content was found to be 6.900 mg ml-1. The solution was diluted as required. Ion Exchangers Amberlite LA-1 [(N-dodecyltrialky1)amine (Rohm & Haas, Philadelphia, PA, USA)], Aliquat 336 [tricaprylmethylammo- nium chloride (Fluka, Buchs, Switzerland)] and trioctylamine (Fluka), dissolved in suitable diluents in varying proportions, were used. Apparatus A VSU2-P spectrophotometer (Carl Zeiss, Jena, GDR) with matched quartz cells was used for spectral measurements. Procedure A sample solution containing 5-200 pg of the metal was taken in a 60-ml separating funnel and concentrated hydrochloric acid and water were added so that the acidity of a total volume of 15 ml of the aqueous phase was between 2 and 5 M.A 2-ml volume of 1% PBHA solution was added and mixed well and then 2.5 ml of 5% ammonium thiocyanate solution were added. The mixture was shaken gently with 15 ml of a 3% solution of Aliquat 336 in toluene for about 1 min. The phases were allowed to separate and the organic extract was dried over anhydrous sodium sulphate and transferred into a 25-1111 calibrated flask. To ensure the complete recovery of Mo(VI), the extraction was repeated with 5 ml of the extracting solvent, dried over sodium sulphate and finally the combined extracts were diluted to volume with the solvent.The absorbance was measured at 350 nm against a reagent blank. For the recovery of Mo(VI), the extracted metal was back-extracted from the organic phase by shaking with 10 ml of 0.3 M ammonia solution for 2 min. The two phases were allowed to settle and the aqueous phase was withdrawn carefully. Results and Discussion The yellow Mo(V1) - PBHA compiex is instantaneously extracted into a 3% solution of Aliquat 336 in toluene from a medium of 2-5 M hydrochloric acid. The extracted species gives a general absorption spectrum without any particular maximum, but the absorbance values from 380 to 390 nm are reproducible and spectral measurements can be made in this range. Addition of thiocyanate favours the extraction through a synergic effect and increases the molar absorptivity con-912 ANALYST, AUGUST 1986, VOL.111 siderably. The extracted species shows a sharp absorption maximum at 350 nm. The blank does not absorb appreciably at this wavelength. Acids and Acidity The Mo(V1) complex was extracted from hydrochloric, sulphuric and perchloric acids. Extraction was incomplete in sulphuric and perchloric acid media, whereas extraction with 2-5 M hydrochloric acid was quantitative and complete. At lower hydrochloric acid concentrations (< 1 M) the intensity was less but increased steadily as the hydrochloric acid concentration increased to a maximum between 1.5 and 5.5 M, and hence the range 2-5 M was adopted for the extraction measurements. Reagent Concentration Extraction with various concentrations of the reagent showed that 1-2 ml of 1% PBHA solution was adequate for the quantitative extraction of Mo(V1).The absorbance of the Mo(V1) complex was constant with the use of excess of the reagent. Ammonium Thiocyanate Concentration The use of ammonium thiocyanate favours the extraction and increases the intensity of the colour. For maximum extraction and colour intensity, 2-3-ml of a 5% solution of ammonium thiocyanate was found to be sufficient. A large excess of ammonium thiocyanate decreases the colour intensity. Potas- sium thiocyanate also gives the same results and hence can be used for extraction. Aliquat 336 Concentration The optimum concentration of Aliquat 336 was studied by varying the concentration from 1 to 10% in toluene. The extraction was quantitative from 2% and remained constant up to 6%.A tendency to form an emulsion was observed at higher concentrations of the ion exchanger. A 3% solution of Aliquat 336 was adopted for extraction. Diluents Mo(V1) was extracted with 3 and 5% Aliquat solution in various diluents. Equilibration was effected by maintaining the ratio of organic to aqueous phase at 1 : 1 and the percentage extraction, E , was calculated in each instance (Table 1). The extraction was very poor with chloroform and isobutyl methyl ketone and carbon tetrachloride, hexane and xylene gave incomplete extraction. The extraction was com- plete and quantitative with benzene and toluene and a clean separation was obtained. As benzene is highly toxic, toluene was used as the diluent in subsequent work.Table 1. Effect of various diluents on the extraction (YO) of Mo(V1) with Aliquat 336 Aliquat 336 concentration, % Diluent Toluene . . . . . . Benzene . . . . . . Xylene . . . . . . Carbon tetrachloride Hexane . . . . . . Chloroform . , . . Isobutyl methyl ketone 3 . . . . 99.9 . . . . 99.8 . . . . 90.9 . . . . 86.3 . . . . 81.8 . . . . 58.0 . . . . 55.4 5 99.9 99.8 90.9 87.4 84.0 59.5 58.0 Type of Liquid Anion Exchanger The extraction of Mo(V1) was carried out with three extrac- tants in various diluents (Table 2). Aliquat 336 in toluene was found to be the best for extraction. Equilibration Time and Stability The time of shaking was varied from 30 s to 5 min. The results showed that the extraction is quantitative within 30 s. The complex extracted under optimum conditions is stable for several days.Optical Properties The colour system obeyed Beer’s law from 0.34 to 10 p.p.m. of Mo(V1) at 350 nm and the optimum range (Ringbom plot) was 0.2-8.5 p.p.m. The molar absorptivities are 4.3 X 104 1 mol-1 cm-1 (without thiocyanate) and 7.2 X lo4 1 mol-1 cm-1 (with thiocyanate), whereas the molar absorp- tivity of the Mo(V1) - PBHA complex extracted into isoamyl alcohol is 3.2 x 103 1 mol-1 cm-1.19 Composition of the Complex The composition of the Mo(V1) mixed ligand complex was studied by the slope ratio method,22 i.e., by plotting the logarithm of the distribution coefficient of the metal, log D M , against the logarithm of the ligand concentration, log (ligand). The extraction was carried out by taking a fixed amount of Mo(V1) in the presence of ( a r a constant amount of PBHA and Aliquat 336 and varying the concentration of thiocyanate, (b) a constant amount of thiocyanate and Aliquat 336 and varying the concentration of PBHA and (c) a constant amount of PBHA and thiocyanate and varying the concentration of Aliquat 336.In all three instances straight lines were obtained, with slopes of 1.9, 2 and 1.75, respectively, which indicates that the composition of the complex is Mo(V1) - PBHA - SCN- - Aliquat 336 (1 : 2 : 2 : 2). The possible mechanism of the extraction is as follows: M00z2+ + 2 {(C6H5)2CONHOH} + 2 C1- ----+ [M002{ (C6H5)2CONH0}2C12]2- + 2H+ [M002{ (C6H5)2CONH0}2C12]2- +2 SCN- - [MOO,{ (C6H5)2CONH0}2(SCN)2]2- + 2 C1- - [R4N+]2 [MOO,{ (C,jH5)2CONH0}2(SCN)2]2- + 2 C1- [MOO,{ (C6H5)2CONH0}2(SCN)2]2- + 2 R4NfCl- where R4N+ represents the cationic part of the liquid anion exchanger.Hence the composition of the extracted species is [(R4Nf)2M002 { ( C,jH5)2CONH0}2(SCN)22-]. Back-extraction For the recovery of the Mo(V1) from the liquid ion exchanger phase, the extracted metal was back-extracted using a suitable reagent. To choose the back-extracting agent, Mo(V1) extrac- ted into the organic phase was stripped with 10 ml of varying Table 2. Effect of different liquid anion exchangers on the extraction of Mo(V1) Liquid anion exchanger Diluent Aliquat336(3%) . . . . . . Toluene Amberlite LA-1 (3%) . . . . Toluene Trioctylamine(3%) . . . . Toluene Xylene Chloroform Xylene Chloroform Xylene Chloroform Extraction, YO 99.9 90.9 58.0 86.4 72.8 56.0 82.0 64.2 39.0ANALYST, AUGUST 1986, VOL.111 913 Table 3. Effect of diverse ions. Amount of Mo taken = 86.25 pg per 25 ml Table 4. Analysis of steel samples Foreign ion Ag+ . . . . . . Be2+ . . . . . . Mg2+ . . . . . . Ca2+ . . . . . . Ba2+ . . . . . . Fe2+ . . . . . . Fe3+ . . . . . . Pb2+ . . . . . . a*+ . . . . . . AS3+ . . . . . . Bi3+ . . . . . . co2+ . . . . . . cu2+ . . . . . . Hg2+ . . . . . . Ni2+ . . . . . . Zn2+ . . . . . . Mn2+ . . . . . . cr3+ . . . . . . A13+ . . . . . . v s + Ti,+ Zr4+ U6+ . . . . . . . . . . . . . . . . . . . . . . . . w . . . . . . F- . . . . . . c1- . . . . . . Br- . . . . . . I- . . . . CH3C00- . . . . Cit3- . . . . . . so42- . . . . . . po43- . . . . . . Added as BeS0, MgS04.7Hz0 BaCI, .2H20 Fe(S04)3(NH,)2S0,.24H20 Pb(N03)2 3CdS04.8H20 AS203 Bi(N03)3 .5H20 COCI, CuS04.5HzO HgCl2 Ca(N03)2 Fe2(S04):, (NH4)3S04.24H20 NiCI,.6H20 ZnS0, .7H20 MnS04. 4H20 CrC13 AICI3.7H20 NH4V03 Ti02 Zr(N03),. 5H20 UO,(CH,C00)2 Na2W04 NaF NaCl NaBr NaI CH3COONa Citric acid Na,S04 Na3P04 * Reduced with ascorbic acid. t Reduced with FeSO,. f Masked with NaF. Tolerance limit/mg 25 30 30 30 30 30 20* 10 25 30 10 30 30 30 25 10 20 30 30 25 t 25f 10 10 15 5 50 50 20 50 30 50 10 concentrations (0.05-5 M) of sulphuric acid, hydrochloric acid, nitric acid, sodium chloride, sodium carbonate, sodium hydroxide and ammonia solution. The back-extracted metal was determined spectrophotometrically with Tiron.23 After adjusting the pH of the aqueous extract to 7.0 (potassium dihydrogen orthophosphate - sodium hydroxide buffer), 2 ml of a 2% aqueous solution of Tiron were added.The solution was diluted to 25 ml and the absorbance measured at 390 nm against a reagent blank. The amount of Mo(V1) was calculated from a calibration graph. The back-extraction was very poor and incomplete with all concentrations of mineral acid and also with solutions of sodium chloride and sodium carbonate. However, quantitative recovery was possible with sodium hydroxide and ammonia solutions at concentrations above 0.3 M. Effect of Diverse Ions Mo(V1) (86.25 pg) was extracted and separated in the presence of large number of different ions (Table 3). The tolerance limit was set as the amount of foreign ion causing a +2% error in the determination of Mo(V1). Moderate amounts of various metal ions associated with Mo(V1) were tolerated and also most anions.Ti(1V) and V(V) interefered seriously. Ti(1V) can be masked with 2 ml of 1% sodium fluoride solution and the interference due to V(V) can be eliminated by reducing it to V(IV) with 2 ml of 1% iron(I1) sulphate solution. V(V) is preferentially reduced to V(1V) without affecting the Mo(V1) because their standard oxidation potentials (E") differ widely [-0.25 V for V(V) - V(1V) and 0.01 V for Mo(V1) - Mo(V)]. Alternatively, Ti(1V) and V(V) Mo Mo Relative present, found,* error, Sample Composition, % % Y O YO BS1501-261 . . C0.14,Mn0.55, 0.55 0.56 1.8 Ni 0.30, Cr 0.95, Mo 0.55 Ni 0.40, Cr 1.25, Mo 0.55 Ni 0.30, Cr 2.25, Mo 1.05 BS1503-621 . . C0.13,Mn0.55, 0.55 0.56 1.8 BS 1501-622 . . C0.14,Mn0.60, 1.05 1.02 2.8 * Average of six determinations. can be preferentially eluted from the organic phase with 10 ml of 4 M sulphuric acid and 0.2 M sodium acetate solution, respectively, without affecting the Mo(V1) complex in the organic phase.The interference due to Fe(II1) can be removed by reduction with ascorbic acid. W(VI) shows no interference and so Mo(V1) can be separated and determined in the presence of any amounts of W(V1) up to 15 mg. Hence the proposed method permits the separation of Mo(V1) from W(VI), Ti(IV), V(V), Cr(III), Mn(II), Fe(III), Cu(II), Ni(I1) and Zn(II), which are generally associated with the metal in alloys, steels and environmental samples. Analysis of Steel Samples Steel samples (0.5 g) were dissolved in hydrochloric acid - nitric acid (2 + 1) containing a few drops of hydrofluoric acid.The solution was evaporated to dryness and the residue was dissolved in 5 ml of concentrated hydrochloric acid and diluted to 100 ml. Suitable aliquots were taken and analysed for Mo(V1) by the general procedure. The results are given in Table 4. Analysis of Plant Sample An ergot was ashed and 5 g of ash were digested under a reflux condenser with a mixture of hydrochloric acid and nitric acid and finally centrifuged. The residue was digested with perchloric acid and hydrochloric acid and centrifuged. The two mother liquors were mixed and diluted to 25 ml with 0.1 M hydrochloric acid. This solution was used for the determina- tion of Mo(V1). The Mo(V1) content obtained was 0.0050 mg per 100 g of dried sample. This result can be compared with that obtained by atomic absorption spectrometry, viz., 0.0052 mg per 100 g of dried sample.Conclusion The proposed method is suitable for the pre-concentration, separation and micro-determination of Mo(V1) in steel and plant samples. One of the authors (S. K. M.) is grateful to the UGC, New Delhi, for the award of a research associateship. References 1. Menon, S. K . , and Agrawal, Y. K., Analyst, 1984, 109, 27. 2. Menon, S. K . , and Agrawal, Y. K., Analyst, 1986, 111, 335. 3. Tseryuta, Yu. S., Bagreev, V. V., Agrinskaya, N. A., Gushchin, N. V., Basov, A. S . , and Solotov, Yu. A., Zh. Anal. Khim., 1973, 28, 946. Vieux, A. S., Rutagengwa, N., and Mpeti, N., Analusis, 1976, 4, 134. Vieux, A. S., Rutagengwa, N., and Noki, V., Inorg. Chem., 1976, 15, 722. 4. 5.914 ANALYST, AUGUST 1986, VOL. 11 1 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. Ray, U. S., and Modak, S . , J. Indian Chem. SOC., 1982, 59, 911. Kim, C . H., and Park, M. S . , Haksul Yoguchi-Chungnam Tachakkyo, Chayon Kwahak Yonguso, 1978, 6, 61. Ponomareva, A. A., Tr. Novocherk. Politekh. Inst. , 1972,266, 103. Rao, R. R., and Khopkar, S. M., Analyst, 1983, 108, 346. Sugawara, M., Uto, M., and Kambara, T., Bull. Chem. SOC. Jpn., 1983, 56, 3179. Agrawal, Y. K., Rev. Anal. Chem., 1980, 5, 3. Agrawal, Y. K . , and Patel, S. A., Rev. Anal. Chem., 1980, 4, 237. Rowland, R., and Meloan, C. E., Anal. Chem., 1964,10,1977. Dutta, R. L., J. Indian Chem. Soc., 1959, 36, 285. Agrawal, Y. K., J. Indian Chem. SOC. , 1977, 54, 451. Alimarin, I. P., and Borzenkova, N. P., Vest. Mosk. Gos. Univ., Ser. Khim., 1969,24, 65. 17. 18. 19. 20. 21. 22. 23. Abbasi, S. A., Sep. Sci., 1976, 11, 293. Agrawal, Y. K . , and Jain, R. K., Croat. Chem. Acta, 1981,54, 249. Patel, S. A., PhD Thesis, M.S. University of Baroda, 1980. Priyadarshini, U., and Tandon, S. G., J. Chem. Eng. Data, 1967, 12, 143. Vogel, A. I., “A Textbook of Quantitative Inorganic Analy- sis,” Longman Green, London, 1968, p. 509. Tomazic, B. B., and O’Laughlin, W. J.,Anal. Chem., 1973,45, 1519. Will, F., and Yoe, J. H., Anal. Chim. Acta, 1953, 8, 546. Paper A5141 6 Received November 12th, 1985 Accepted March 3rd, 1986

ISSN:0003-2654    

DOI:10.1039/AN9861100911

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, AUGUST 1986, VOL. 111 91 5 Spectrophotometric and Spectrofluorimetric Determination of Cycloserine with p-Benzoquinone Laila El-Sayed, Zeinab H. Mohamed and Abdel-Aziz M . Wahbi Pharmaceutical Chemistry Department, College of Pharmacy, King Saud University, P. 0. Box 2457, Riyadh- 11451, Saudi Arabia Spectrophotometric and spectrofluorimetric methods are described for the determination of cycloserine. Both methods are based on the reaction of cycloserine in aqueous solution of pH 8.2 with p-benzoquinone. A fluorescent compound with an absorption maximum (and also an excitation maximum) at 381 nm is obtained. The complex has an apparent molar absorptivity of 4.76 x lo3 I mot-1 cm-1 and Beer's law is obeyed over the range 4-20 pg ml-1. The emission maximum of the complex is at 502 nm.The fluorescence intensity is linearly related to concentration over the range 0.04-0.2 pg ml-1. When applied to capsules labelled to contain 250 mg, the proposed methods gave mean recoveries of 100.97 k 1.14 and 99.8 k 1.13%, respectively. Keywords: Cycloserine determination; p- benzoquinone; spectrophotometry; spectrofluorimetry Cycloserine is an antimycobacterial agent with antitubercular activity. The L-isomer is ten times more active than the D-isomer. Several methods have been reported for its deter- mination , including titrimetric,l colorimetric,' spectrophoto- metric2 and thermogravimetric methods.3 H2N H*N\;co XJ0" I II Cycloser i n e Reaction of aliphatic primary and secondary amines with p-benzoquinone has been used as the basis for the spectro- photometric determination of a considerable number of amine-containing medicinal compounds.4-9 Cycloserine con- tains an ionisable primary amine group with pK, 7.4,1° which can react with p-benzoquinone to form a complex. This paper discusses this reaction and its use for the spectrophotometric and spectrofluorimetric determination of cycloserine in phar- maceu tical preparations. Experimental Reagents and Solutions p-Benzoquinone solution, 0.6% in 95% ethanol.Freshly prepared. Cycloserine solution, 0.02% mW. This solution must be freshly prepared and used within 1 h. Buffer solution.11 Prepared by mixing 58.0 ml of 0.05 M sodium tetraborate solution and 42 ml of 0.1 M hydrochloric acid to give 100 ml of buffer (pH 8.2). Apparatus A Varian DMS 90 double-beam spectrophotometer with 1-cm quartz cuvettes and a Perkin-Elmer 552 spectrofluorimeter with 1-cm quartz cuvettes and fitted with a Model 057 X - Y recorder and X - Y interface were used.General Procedures Spectrophotometric method Pipette 1-5-ml portions of standard cycloserine solution into 20-ml test-tubes. Add exactly 1 ml of p-benzoquinone solution, 2 ml of buffer solution (pH 8.2) and distilled water to 10 ml. Mix well and heat in a boiling water-bath for 30 min. Allow to cool for 2 min, then transfer into 50-ml calibrated flasks using buffer solution. Dilute to volume with buffer solution and set aside for 15 min. Prepare a reagent blank and measure the absorbance against the blank in 1-cm cells at 381 nm. Spectrofluorimetric method Into a series of 100-ml calibrated flasks, pipette 1 ml of each of the final solutions prepared for the spectrophotometric method above.Dilute to the mark with distilled water and mix well. Measure the fluorescence intensity at 502 nm using 381 nm as the excitation wavelength. Assay of Capsules Empty the contents of ten capsules into a weighing bottle, weigh accurately and mix well. Extract an accurately weighed amount of the powder equivalent to about 20 mg of cycloserine with four successive 20-ml portions of distilled water into a 100-ml calibrated flask. Dilute to volume with distilled water. Proceed as described under General Pro- cedures. Calculate the concentration of cycloserine from a calibration graph of absorbance or fluorescence intensity versus concentration.Results and Discussion Cycloserine in aqueous, freshly prepared solution occurs in keto - enol equilibrium (I and II).12?13 It forms a dipolar ion (111), which on standing dimerises to 2,5-bis(aminoxyrnethyl)- 3,6-diketopiperazine (IV). 0 Ill IV The amino group in the cycloserine molecule was found to be essential for its antibacterial activity.14 It was proposed that cycloserine forms a Schiff base by condensation with pyridoxal phosphate, the coenzyme of alanine racemase, which is thought to be the site for cycloserine inhibition of cell wall biosynthesis that is the basis of its antibacterial action.15916 ANALYST, AUGUST 1986, VOL. 111 Similar speculation on the reaction of cycloserine with p-benzoquinone suggests that the free primary amine of the former condenses with the carbonyl group of the latter to form a Schiff base (V).0 NMR spectral analysis of the isolated reaction product shows no evidence of substitution on the quinone ring. The four aromatic protons of p-benzoquinone give a singlet positioned at 6 6.52 p.p.m., which is ca. 0.3 p.p.m. upfield of the position of the same protons in the p-benzoquinone spectrum in the same solvent (deuteriated dimethyl sulph- oxide). This supports the suggestion that the reaction of cycloserine with p-benzoquinone results in the formation of a Schiff base. The upfield shift of the aromatic proton signal observed in the spectrum of the product compared with that of p-benzoquinone could be attributed to the difference in electronegativities and consequently the shielding effect of oxygen relative to the nitrogen atom in the complex V.The IR spectrum of V does not contain the overtone band at 2200 cm-1 characteristic of the -NH3+ zwittterion that is present in the cycloserine spectrum. The hydroxamate anion band extending in the range 1400-1500 cm-1 had also disappeared in the IR spectrum of V; the broad band extending from 3400 to 2500 cm-1 is possibly due to the hydrogen-bonded OH stretching vibration, and the OH bending vibration is seen as a sharp, strong peak at 1510 cm-1. Although there is no positive IR evidence for the formation of - E N , as the C=O band overlaps with the -C=N band region, the absence of an -NH or -NH2 group characteristic stretching is indicated. The absorption characteristics of the complex formed were found to be dependent on pH.At pH < 7.0 no complex formation was observed, probably owing to decomposition in acidic medium. At pH 7.0, the complex formed has an absorption maximum at 335 nm; on increasing the pH to 8.2 there is a bathochromic shift of h,,,, to 381 nm. No appreciable change in absorptivity was observed in the pH range 7.0-8.2, whereas at higher pH there was a considerable decrease in the intensity at the A, (381 nm). Absorption spectra of the product for pH values between 7 and 10 are shown in Fig. 1. pH 8.2 was selected as the optimum pH for this study. The aqueous cycloserine solution, on standing for more than 1 h, gave lower and inconsistent results, probably owing to the formation of the dimeric form IV, which cannot react with p-benzoquinone.For this reason, the cycloserine solution should be used within 1 h of its preparation. The optimum concentration of p-benzoquinone for the reaction was investigated in the range 0.1-1% mlV. Fig. 2 shows that on increasing the p-benzoquinone concentration beyond 0.5% there was no increase in the absorbance at 381 nm. Accordingly, a concentration of 0.6% mlV was used in subsequent work. The volume of p-benzoquinone solution, the buffer solution and the heating time given under General Procedure were optimised to achieve high sensitivity and high stability. Under the optimum experimental conditions, the absorbance of the complex at 381 nm was found to be stable for several hours. The complex was found to possess strong fluorescence at 502 nm when excited at 381 nm (Fig.3). Beer's law was obeyed over the concentration range 4-20 yg ml-1 for the spectro- photometric method and 0.04-0.2 yg ml-1 for the spectroflu- 250 290 330 370 410 450 500 0 A h m Fig. 1. Effect of pH on the reaction product of 0.0014% mlV cycloserine withp-benzoquinone: A, pH 7.0: B, pH7.4; C, pH 8.0; D, pH 8.4; E, pH 9.0; and F, pH 10.0 0.5 r 1 0.4 1 E 0.3 - 0 c - q 0.2 - 0.1 - 0 0.2 0.3 0.4 0.5 0.6 0.7 p-Benzoquinone concentration, Ol0 m/V Fig. 2. 381 nrn for 0.001% mlV cycloserine solution at pH 8.2 Effect of p-benzoquinone concentration on absorbance at loo I h!nm Fig. 3. Fluorescence excitation (A) and emission (B) spectra for cycloserine reaction product with p-benzoquinone and blank emission spectrum (C) orimetric method. The apparent molar absorptivity was 4.76 x 103 1 mol-1 cm-1.The linearity between absorbance at 381 nm ( A ) and concentration ( c , mg per 100 ml) was found to be expressed by the equation A = 0.005 + 0.466c, with a correlation coefficient of 0.9996. The linear graph of relative fluorescence intensity to concentration was found to pass through the origin, with a correlation coefficient of 0.9999.917 ANALYST, AUGUST 1986, VOL. 111 Table 1. Spectrophotometric and spectrofluorimetric determination of cycloserine in capsules with p-benzoquinone. Capsules labelled to contain 250 mg each Found, * Spectrophotometry Spectrofluorimetry (Amax, = 381 nm) (Aex. = 381, kern. = 502 nm) BP method 99.8 k 1.23 (4) Mean k s.d. ( n ) . . 100.9 k 1.14 (4) 99.8 f 1.13 (4) t 1.37 (2.447)t .. . . . . . . F . . . . . . . . 1.02 (9.28)t * Expressed as a percentage of the labelled amount. t Theoretical values at P = 0.05. Table 2. Recovery of cycloserine added to capsules using the spectrophotometric and spectrofluorimetric methods Spectrophotometry Spectrofluorimetr y Expt. No. Addedlmg* Recovery, YO Added/mg* Recovery, YO 1 29.1 99.1 21.7 97.3 2 29.1 102.1 21.7 98.6 3 19.7 99.7 19.7 99.0 4 19.7 101.0 19.7 99.0 Mean k s.d. 100.6 5 1.59 98.5 f 0.81 * Added to powdered capsule equivalent to 20 mg of cycloserine. The two proposed methods were applied to the determina- tion of cycloserine in capsules labelled to contain 250 mg of active ingredient. The spectrophotometric and spectroflu- orimetric methods gave mean percentages (k standard deviation) of the labelled amount of 100.9 k 1.14 and 99.8 k 1.13%, respectively (Table 1).According to the variance ratio test (F-test), the calculated F value was found to be 1.02 for n1 = 4 and n2 = 4; the theoretical value is 9.28 ( P = 0.05). This means that there was no significant difference between the spectrophotometric and spectrofluorimetric methods with respect to precision. The calculated t was found to be 1.37; the theoretical value at 6 degrees of freedom (P = 0.05) is 2.447. There was no significant difference between the two means with respect to accuracy. The official BP spectrophotometric method, based on reaction with sodium nitroprusside, gave a mean result of 99.8 k 1.23%, which agrees well with the results of our spectrophotometric and spectrofluorimetric methods (Table 1).The latter method is 100 times more sensitive than the spectrophotometric method. Recovery experiments on cycloserine added to the powdered capsule were carried out using the two methods. Table 2 shows that the mean recoveries were 100.6 k 1.6 and 98.6 k 0.81% using the spectrophotometric and spectroflu- orimetric methods, respectively. The latter method is suffi- ciently sensitive to permit the determination of cycloserine in biological fluids after a suitable clean-up procedure. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. References “British Pharmacopoeia 1980,” HM Stationery Office, London, 1980, pp. 134 and 528. Wahbi, A. M., Mohamed, M. E., Abounassif, M., and Gad-Kariem, E., Anal. Lett., 1985, 18B, 261. Harned, R. L., Hidy, P. H., and Baro, E. K., Antibiot. Chemother., 1955,5,204. Cavett, J. W., and Heotis, J. P., J . Assoc. Off. Agric. Chem., 1955,41, 323. Beckmann, H. F., and Feldman, L., J. Agric. Food Chem., 1960, 8, 227. Loucks, M. F., and Nauer, L., Jr., J . Assoc. 08. Anal. Chem., 1967, 50, 268. Benson, G. A., and Spillane, W. J., Anal. Chem., 1976, 48, 2149. Korany, M. A., Wahbi, A. M., and Abdel-Hay, M. H., J . Pharm. Biomed. Anal., 1984, 2, 537. Wahbi, A. M., Abounassif, M. A., and Gad-Kariem, E. A., Talanta, 1986, 33, 179. Lamb, J. W., in Florey, K., Editor, “Analytical Profiles of Drug Substances,” Volume 1, Academic Press, New York, 1976, p. 53. “Documenta Geigy, Scientific Tables,” Ciba-Geigy, Bade, 1984, Part 3, p. 59. Kuehl, F. A., Jr., J . Am. Chem. SOC., 1955, 77,2344. Hidy, P. H., and Hodge, E. B., J. Am. Chem. SOC., 1955,77, 2345. Foye, W., “Principles of Medicinal Chemistry,” Second Edi- tion, Henry Kimpton, London, 1981, p. 746. Raudo, R. R., Biochem. Pharmacol., 1975,24, 1153. Paper A51406 Received November 6th, 1985 Accepted February 12th, 1986

ISSN:0003-2654    

DOI:10.1039/AN9861100915

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, AUGUST 1986, VOL. 111 919 Use of Cerium(lV) Sulphate in the Spectrophotometric Determination of Paracetamol in Pharmaceutical Preparations Salah M. Sultan, lbrahim 2. Alzamil, Abdullah M. Aziz Alrahman, Saad A. Altamrah and Yousif Asha Chemistry Department, College of Science, King Saud University, Riyadh 11451, P.O. Box 2455, Saudi Arabia An accurate spectrophotometric method is proposed for the determination of paracetamol. Cerium(lV) sulphate is used to oxidise paracetamol in 5 M H2S04 to p-benzoquinone, which is then determined at 410 nm. The method has been successfully applied to the analysis of commercial pharmaceutical preparations and the results have been statistically compared with those obtained by the official (BP) method. Keywords: Pharmaceutical preparations; cerium(lV) sulphate; paracetamol determination; sulphuric acid; spectrophotometry Paracetamol (N-acetyl-p-aminophenol) is well known as an analgesic anti-pyretic drug.Many methods for its determi- nation have been described, including titrimetry,1-7 chromato- graphy,%z4 electrochemistry25-27 and spectrophotometry.28-41 In the official method (BP),l paracetamol is determined titrimetrically with cerium(1V) in acidic media using 1 ,lo- phenanthroline - iron(I1) complex (ferroin) to determine the end-point. The titration is performed in ice. In the present method, cerium(1V) was reacted for 90 min with paracetamol in concentrated sulphuric acid in a water- bath maintained at 80 "C. The final product of the oxidation of paracetamol, p-benzoquinone, was spectrophotometrically determined at the wavelength of its maximum absorption (410 nm).This method was applied to the determination of paracetamol in drugs used for the treatment of coughs, colds and influenza. Experimental Apparatus A Beckman Model 35 spectrophotometer connected to a Beckman Model 24-25 ACC recorder was used for all absorbance measurements. Matched sets of W 210/UU 10.00 mm cells were used throughout. Reagents and Samples Quartz-processed high-puri ty distilled water was used throughout. All chemicals and reagents were of analytical- reagent or pharmaceutical grade. Paracetamol (pure form). This was prepared in our labora- tory by acetylation of p-aminophenol.42 The purity of the crystalline powder was checked by melting-point, IR and NMR techniques.The stock solution from this compound (1 mg ml-1) was prepared by dissolving exactly about 1 g in warm water, stirring for 10 min and diluting to 1 1 in a calibrated flask after cooling. The absolute purity of this material was 99.81% as determined by the official method' and it complied with the BP specifications. Paracetamol (tablets). Ten tablets were weighed and ground into a fine powder. A mass of powder containing about 500 mg of paracetamol was weighed accurately, mixed with about 150 ml of water, warmed and stirred for 10 min. This was then filtered through a Whatman No. 41 filter-paper, washed with water and then the filtrate plus washings were diluted to 500 ml with water in a calibrated flask after cooling. Paracetamol (capsules). The contents of 10 capsules were carefully mixed and weighed.An amount equivalent to 500 mg of paracetamol was accurately weighed and dissolved in 150 ml of water, warmed and stirred for 10 min. The insoluble mass was filtered through a Whatman No. 41 filter-paper and washed with water. The filtrate and washings were transferred into a 500-ml calibrated flask and diluted to volume after cooling to room temperature. Ceriurn(ZV) solution. A stock solution of Ce(S04)2 (20 mg ml-1) was prepared in 10 M sulphuric acid. Sulphuric acid. A stock solution of 10 M was prepared in the usual way. Procedure A 4.00-ml volume of cerium(1V) sulphate solution was placed in a 50-ml calibrated flask to which 21.00 ml of sulphuric acid stock solution and the appropriate amount of paracetamol solution were added.The flask and contents were swirled, placed in a water-bath at a temperature maintained at 80 "C for 90 min, cooled under a tap and then diluted to the mark with water. The coloured product was determined at 410 nm against a reagent blank. Results and Discussion Kinetics This method was based on the oxidation reaction of para- cetamol with cerium(IV) in sulphuric acid media. The brown - red species determined at a A,,,, of 410 nm was believed to be p-benzoquinone, the final product of paracetamol oxidation. This maximum was the only one obtained for the absorption spectra run in the range 300-750 nm. It was observed that the reaction was dependent on the concentration of sulphuric acid, and that the reaction only takes place at acid concentra- tions greater than 2 M.The rate of the reaction accelerated as the acid concentration was increased and slowed as the concentration decreased. At higher concentrations of sul- phuric acid the redox potential of Ce(S04)2 is such that it can be oxidised. This indicates that the de-acetylation of paracet- amol top-aminophenol is the rate-determining step. p-Amino- phenol is then further oxidised with Ce(1V) to p-benzoquin- one. This mechanism is similar to that recently proposed2 for the oxidation of paracetamol using iron(II1) as a one-electron oxidant in 8 M H2S04. OH OH920 ANALYST, AUGUST 1986, VOL. 111 Table 1. Results of the determination of paracetamol in pure form and in pharmaceutical preparations compared with the official method. * Nos. 2-11 were in tablet form and 12 in capsule form Recovery* 5 standard deviation, YO Drug proprietary name No.(supplier) Composition/mg Proposed method 1 2 3 4 5 6 7 8 9 10 11 12 Paracetamol (laboratory made) 99.81% paracetamol (absolute) 100.83 k 0.27 Panadol (Winthrop, UK) 500 paracetamol 99.84 k 0.53 Paracetarnol (Smith, France) 500 paracetamol 99.84 k 0.53 Revanin (Arab Pharmaceutical Manuf. Co., Jordan) 500 paracetamol 98.98 t 0.16 Sinutab (Warner, UK) 300 paracetamol 98.58 t 0.62 25 phenylpropanolamine. HCI 22 phenyltoloxamine Revacod (Arab Pharmaceutical 500 paracetamol 99.43 k 0.30 Manuf. Co., Jordan) Switzerland) 200 paracetamol 100.21 k 0.91 10 codeine phosphate Triatussic (Wander, 5 mepyramine. HCl 4.2 pheniramine. HCl 20 noscapine 13 caffeine 90 terpin hydrate 12.5 phenylpropanolamine.HCl Trimedil (Zyma, Switzerland) 100 paracetamol 0.25 dimetindene maleate 15 o-(P-hydroxyethy1)- rutoside 40 ascorbic acid 1.25 phenylephrine.HC1 Efferalgan (Upsa, France) 330 paracetamol 200 ascorbic acid Prontopyrin (Mack, FRG) 200 paracetamol 250 aluminium acetylsalicylate 50 caffeine Veganin (Warner, UK) 250 paracetamol 250 acetylsalicylic acid 6.8 codeine phosphate Neopyrin-N (Knoll, FRG) 200 paracetamol 200 ethenzamide 10.72 isometheptene mucate 14.28 octamylamine mucate Official method (Bp)l 100.80 k 0.22 100.50 k 0.18 100.11 k 0.10 99.11 k 0.17 99.02 2 0.14 99.65 t 0.11 100.28 t 0.25 185.51 k 0.72 134.33 t 0.23 133.5 k 0.49 140.48 k 0.15 110.58 t 0.56 - 123.89 k 0.49 - 98.64 ? 0.61 98.32 t 0.14 * Average of five determinations, assay as a percentage of label claim.t For five determinations at different concentrations in the range 4&160 pg ml- I . $ Theoretical value 2.78 (p = 0.05). Correlationt coefficient * 2 fcalc.$ 0.28 1 . 0000 1 .0002 1.04 1.0006 0.99 1.0010 0.67 0.9993 0.88 1 .oo 0.87 0.99 0.69 1 .0008 1.28 The reaction rate accelerated at elevated temperatures; 80 "C was found to be suitable to study the interference of other drug constituents without fear of chemical changes. When the reactants in 5 M H2S04 were left in the water-bath at 80 "C, the absorbance of the product at 410 nm increased with time. The reaction product attained maximum absorb- ance within 75 min and was found to be stable for 24 h at room temperature. Measurements taken earlier than 75 min were meaningless and rendered inaccurate results, and hence 90 min was found to be suitable for all experiments.Analytical Data Beer's law was valid over the concentration range between 30 and 160 pg ml-1. Although the range was small, the results were accurate and were not affected at common dose levels. The linearity of concentration versus absorbance was calcu- lated using a computer and the correlation coefficient was found to be 0.9967 with an intercept of 0.0781. The molar absorptivity, calculated from the slope, was found to be 811.21 1 mol-1 cm-1. Precision The method was applied to the determination of paracetamol in its pure form and in a number of pharmaceutical prepara- tions. The results obtained (Table 1) seemed to be highly accurate and the method was successfully applied to the determination of paracetamol in the presence of excipients such as lactose, glucose and starch, which are usually added during the preparation of tablets and capsules.The presence of codeines, caffeines, noscapine, terpins, phenylephrines, phenylpropanolamine, phenyltoloxamine, mepyramine, pheniramine, dimethindene maleate, rutosides, isometh- eptene mucate, octamylamine mucate and ethenzamide together with paracetamol in drug formulations did not interfere with the results. However, the presence of ascorbic acid and acetylsalicylic acid in Efferalgan and Trimedil tablets and in Prontopyrin and Veganin tablets, respectively, showed positive interferences. In the determination of paracetamol in Prontopyrin and Veganin tablets by the official method,' the interference caused by acetylsalicylates made it impossible to determine the end-point of the titration.ANALYST, AUGUST 1986, VOL.11 1 The results of this method were compared with the results obtained by the official method’ using the batch of samples presented in Table 1. The calculated t-values at the 95% confidence level did not exceed the theoretical value, indicat- ing no significant difference between the two methods. However, this method is regarded as being superior to the official method’ because it can be used for the determination of trace amounts of paracetamol whereas the official method’ is titrimetric and cannot be applied accurately to trace analysis. We thank Mr. Ahmed M. A. Yousif for preparing the pure paracetamol sample. We are grateful to the Research Centre of the College of Science, King Saud University, for support- ing this research.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. References “British Pharmacopeia 1980,” HM Stationery Office, London, 1980, p. 326. Srivastava, M. K., Ahmed, S . , Singh, D., and Shukla, I. C., Analyst, 1985, 110, 735. Mohmoud, S. A., and Abdel, S. F., Pharmazie, 1979,34,569. Kalinowska, Z. E., and Hasztar, H., Farm. Pol., 1976,23,447. Walash, M. I . , Agarwal, S. P., and Martin, M. I., Can. J . Pharm. Sci., 1972, 7, 123. Agarwal, S. P., and Walash, M. I., Indian J. Pharm., 1974,36, 47. Inamdar, M. C., Saboo, J. C., Kamdar, C. N., and Sanghavi, N. M., Indian J . Pharm., 1973, 35, 187. Carnevale, L., J. Pharm. Sci., 1983, 72, 196. Beckett, A. H., and Wilkinson, G.R., J . Pharm. Pharmacol., Suppl., 1965, 17, 104s. Wallo, E. W., and d’Adamo, A., J. Pharm. Sci., 1982, 71, 1115. Brochmann-Hanssen, E., and Svendsen, A. B., J. Pharrn. Sci., 1962,51, 1096. Barnhart, J. W., and Caldwell, W. J., J . Chromatogr., 1977, 130, 243. Plakogiannis, F. M., and Saad, A. M., J . Pharrn. Sci., 1977,66, 604. Chao, M. K., Holcomb, I. J., and Fusari, S. A., J . Pharm. Sci., 1979, 68, 1463. Kubiak, E. J., and Munson, J. W., J . Pharm. Sci., 1980, 69, 152. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 921 Menyharth, A., Mahn, F. P., and Heveran, J. E., J . Pharm. Sci., 1974, 63, 430. Sena, F. J., Piechocki, J. T., and Li, K. L., J. Pharrn. Sci., 1979, 68, 1465. Soni, S. K., and Dugar, S. M., J . Forensic Sci., 1979, 24, 437. Das Gupta, V., J.Pharm. Sci., 1980, 69, 110. Burke, D., and Sokoloff, H., J. Pharm. Sci., 1980, 69, 138. McSharry, W. O., and Savage, I. V. E., J. Pharm. Sci., 1980, 69, 212. KO, C. Y., Marziani, F. C., and Janicki, C. A., J . Pharm. Sci., 1980, 69, 1081. Townley, E., and Roden, P., J. Pharm. Sci., 1980, 69, 523. Munson, J. W., and Kubiak, E. J., Anal. Lett., 1980, 13(B8), 705. Miner, D. J., Rice, J. R., Riggin, R. M., and Kissinger, P. T., Anal. Chem., 1981, 53, 2258. Jollow, D. J., Thorgeirsson, S. S . , Potter, W. Z., Hashimito, 27. 28. 29. 30. 31. 32. 33. 34. 35 * 36. 37. 38. 39. 40. 41. 42. M., and Mitchell, J-R., Pharmacology, 1974, 12, 251. Miner, D. J., and Kissinger, P. T., Biochem. Pharmacol., 1979, 28, 3285. Verma, K. K., Gulati, A. K., Palod, S . , andTyagi, P., Analyst, 1985, 109, 735. D’Souza, A. A., and Shenoy, K. G., Can. J . Pharm. Sci., 1974, 36, 47. Kalatzis, E., and Zarbi, I., J. Pharm. Sci., 1976, 65, 71. Inamdar, M. C., Gore, M. S . , and Bhide, R. V., Indian J. Pharm., 1974, 36, 7. Ellock, C. T., and Fogg, A. G., Analyst, 1975, 100, 16. Davis, D. R., Fogg, A. G., Burns, D. T., and Wragg, J. S . , Analyst, 1974, 99, 12. Belal, S. F., Elsayed, M. A. H., Elwalily, A., and Abdine, H., Analyst, 1979, 104, 919. Elsayed, M. A. H., Belal, S . F., Elwalily, A., and Abdine, H., Analyst, 1979, 104, 620. Wallace, J., Anal. Chem., 1967, 39, 531. Elsayed, N., Belal, S. F., Abdel, F. M., and Abdine, N., J. Assoc. Off. Anal. Chem., 1979, 62, 549. Plakogianis, F. M., and Saad, A. M., J. Pharm. Sci., 1975,64, 1547. Mouton, M., and Masson, M., A n n . Pharm. Fr., 1960,18,759. Welch, R. M., and Conney, A. H., Clin. Chem., 1965, 11, 1064. Vaughn, J. B., J. Pharm. Sci., 1969, 58, 469. Vogel, A. I . , “Elementary Practical Organic Chemistry, Part 1,” Longmans, London, 1958, p. 242. Paper A6123 Received January 23rd, 1986 Accepted February 24th, 1986

ISSN:0003-2654    

DOI:10.1039/AN9861100919

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, AUGUST 1986, VOL. 111 923 Ultraviolet Spectrophotometry for the Determination of Benperidol by a Difference Procedure and for the Elucidation of its Polymorphic Structures* Ahmed E. H. Gassim, Pamela Girgis Taklat and Kenneth C. James Welsh School of Pharmacy, University of Wales Institute of Science and Technology, P.O. Box 13, Cardiff CF13XF, UK Benperidol contains two chromophores separated by a non-conjugated system. Its ultraviolet absorption spectrum, in acidic or neutral aqueous solution, shows maxima due to the p-fluoroacetophenone chromophore (at 249 nm) and to the benzimidazolone chromophore (at 232 and 279 nm). In alkali, owing to ionisation, there is a bathochromic shift of the benzimidazolone absorption, and peaks occur at 220,248 and 288 nm. An apparent hyperchromic effect at 248 nm is caused by the overlapping of peaks from the two chromophores.A procedure is proposed for determining benperidol in tablets by measuring the difference in absorbance at 292 nm of solutions prepared in alkali and in acid. This can eliminate background interference, and hence avoid preliminary purification steps. Infrared spectra indicate that benperidol can act as a zwitterion. Ultraviolet absorption spectra of benperidol in organic solvents show marked bathochromic shifts of the longer wavelength benzimidazolone maximum, compared with spectra in acidic or neutral aqueous solution. In ethers, the shift is of the same magnitude as that produced in aqueous alkali and probably indicates hydrogen bonding with the enol form. Alcohols can stabilise the zwitterion by hydrogen bonding with the oxygen of the-N=C-0- grouping, which would explain the formation of a polymorph with no amide absorption on recrystallisation from isopropanol.Keywords: Benperidol determination; pharmaceutical preparations; difference spectrophotometric assay; tautomerism; polymorphic zwitterion Benperidol (IA) belongs to the butyrophenone group of major tranquillisers, which also includes haloperidol (11). Pharma- copoeial assay methods1.2 for dosage forms containing the latter compound depend on preliminary extraction procedures to eliminate background interference, followed by the measurement of ultraviolet absorption. Thin-layer chromato- graphy, to limit the presence of related impurities or degra- dation products, is an additional requirement of the British Pharmacopoeia.Benperidol has no official monographs. Its purity determi- nation by thin-layer chromatography has been reported,3 but its assay was found to require investigation. A difference spectrophotometric procedure has therefore been developed from our earlier observation3 that the ultraviolet absorption of benperidol shows bathochromic, hyperchromic shifts on changing from an acidic or neutral to an alkaline pH. These shifts can be ascribed to ionisation of the benzimidazolone chromophore.4 The method avoids the tedious separations needed for haloperidol preparations and can be applied directly to benperidol tablets. Benperidol is known to exist in different polymorphic forms,5 whereas no evidence has been found of polymorphism in haloperidol.6 The underlying molecular changes associated with the polymorphism of benperidol can be explained from infrared spectral data.7 A study of the ultraviolet absorption spectra of benperidol in a number of organic solvents has produced results that support the suppositions already made.Experimental Reagents Analytical-reagent grade or spectroscopic-reagent grade materials were used. Benperidol. Melting range 172-174 "C; loss of mass on drying to constant mass in vacuo at 60 "C, not more than 0.04%. The sample was kindly supplied as a gift by Janssen Pharmaceutical Limited. * Presented at SAC 86, the 7th SAC International Conference on Analytical Chemistry, Bristol. UK, 20-26 July 1986. t To whom correspondence should be addressed.Benperidol stock standard solution, approximately 0.1 % Water. De-mineralised using an Elgostat de-mineraliser. mlV in methanol. Accurately prepared. Apparatus Ultraviolet absorption spectra were determined with a Pye Unicam SP 1800 spectrophotometer. Quantitative measure- ments of absorbance made at fixed wavelengths were obtained using a Pye Unicam SP 500 Series 2 spectrophotometer. All measurements were made using 1-cm silica cells. Procedure Determination of benperidol in tablets by ultraviolet difference spectrophotometry Weigh and powder 20 tablets. Extract an accurately weighed amount of the powder equivalent to about 0.75 mg of benperidol by shaking with 0.05 M sulphuric acid, making the total volume up to 10.0 ml. Centrifuge and withdraw two aliquots of the clear supernatant liquid, each of 4.0 ml.To one aliquot, add 1.4 ml of 1 M sodium hydroxide solution and dilute to 10.0 ml with water (test solution). Dilute the other aliquot to 10.0 ml with 0.05 M sulphuric acid (test reference solution). Measure the absorbance (AT) of the test solution at 292 nm against the test reference solution set to read zero absorbance. Dilute 3.0 ml of the benperidol stock standard solution to 10.0 ml with water and measure two aliquots of the solution, each of 1.0 ml. To one aliquot add 4.0 ml of 0.05 M sulphuric acid and 1.4 ml of 1 M sodium hydroxide solution and dilute to 10.0 ml with water (standard solution). Dilute the other aliquot to 10.0 ml with 0.05 M sulphuric acid (standard reference solution). Measure the absorbance (As) of the standard solution at 292 nm against the standard reference solution set to read zero absorbance.To 4.0 ml of 0.05 M sulphuric acid, add 1.4 ml of 1 M sodium hydroxide solution and dilute to 10.0 ml with water (reagent blank solution). Measure the absorbance (AB) of the reagent blank solution at 292 nm against 0.05 M sulphuric acid set to read zero absorbance.924 ANALYST, AUGUST 1986, VOL. 111 Calculate the content of C22H24FN302 in milligrams per tablet using the equation (AT - AB) x 7.5 x mean tablet mass (g) X stock standard concentration (YO m/V) (As - AB) x mass of powder (g) Note-The methanol used in preparing the benperidol stock standard solution should not interfere in the measurement of AB. Results and Discussion The ultraviolet absorption of benperidol (Fig.1) shows maxima in 0.1 M sodium hydroxide solution at 248 nm 452) and 288 nm 203). In 0.05 M sulphuric acid, it shows maxima at 232 nm (A::m 265), 249 nm (A::m 3.52) and 279 nm ( A : O t r n 207). The spectra in acidic solution are almost identical with the spectra determined in de-ionised water. Spectra determined similarly for benzimidazolone and p-fluoroacetophenone (Fig. 2) showed that only the benz- imidazolone spectrum was pH sensitive. Summation of the spectra obtained separately for the two compounds produced spectra closely similar to the spectra found for benperidol. The two chromophores are separated from each other in the benperidol molecule by a non-conjugated system and, as expected, they therefore exert their effects on light absorption independently of each other.Wavelengthhm Fig. 1. Ultraviolet absorption spectra for 1.50 mg-% mlV benperidol in (A) 0.1 M sodium hydroxide and (B) 0.05 M sulphuric acid Wavelength/nrn Fig. 2. Ultraviolet absorption spectra for 5 x 10-5 M benzimidazol- one in (A) 0.1 M sodium hydroxide and (B) 0.1 M hydrochloric acid and for 5 x 10-5 M p-fluoracetophenone in ( C ) 0.1 M sodium hydroxide and (D) 0.1 M hydrochloric acid Benperidol absorption maxima obtained in acidic or neutral solution can hence be assigned to the p-fluoroacetophenone fragment (249 nm) or to the benzimidazolone moiety (232 and 279 nm). In 0.1 M sodium hydroxide solution, the benzimid- azolone maxima shift to 220,248 and 288 nm, and the central peak overlaps the 249 nm p-fluoroacetophenone absorption peak.The proposed procedure for the analysis of tablets measures the difference in the (benzimidazolone) absorption of benperidol (AA) at 292 nm in alkali and in acid; the method will eliminate background interference provided that it is unaffected by pH changes. The wavelength selected was the maximum in the difference s ectrum. Thus, at 292 nm, the selected wavelength, the AAl cm value was 155, higher than the value of 95 obtained at 288 nm (the wavelength maximum in alkali). Calibration graphs of AA versus concentration were linear up to 4.0 mg-Yo mlV benperidol. The absorbance (AB) due to sodium hydroxide in the reagent blank solution was low (0.015 at 292 nm), but it was necessary to take it into account. The other reagents used in the assay, methanol and 0.05 M sulphuric acid, showed no absorbance at 292 nm when measured against de-ionised water as a blank.Results obtained for a commercial sample of benperidol tablets are shown in Table 1, together with results obtained by the US Pharmacopeia procedure2 for haloperidol tablets. The procedure was slightly modified. An accurately weighed amount of powdered tablets equivalent to about 1.5 mg of benperidol was shaken with 0.1 M sodium hydroxide solution (10 ml) and chloroform (25.0 ml), and 10.0 ml aliquots of the chloroform layer were each shaken with 20.0 ml of 0.05 M sulphuric acid. The acid layer ms diluted (a) by adding 7 ml of 0.5 M sodium hydroxide solution to 10.0 ml and diluting with water to 25.0 ml, and (b) by making 5.0 ml up to a volume of 10.0 ml with 0.05 M sulphuric acid.Absorbances were measured against the appropriate reagent blanks at 248 nm for the alkaline solution and 249 nm for the acidic solution. The content of C 2H24FN302 per tablet was calculated using the respective A{$m values of 452 and 352. Good precision was obtained in all the assays. Results obtained by the USP procedure were in agreement, irrespective of whether absorb- ance was measured in alkali (benzimidazolone plus p-flu- oroacetophenone moieties) or in acid Cp-fluoroacetophenone moiety), but were 3.2% lower than results obtained by the recommended difference procedure. This was attributed to loss of benperidol during the second extraction from chloro- form into acid. Results obtained by the USP procedure agreed with those obtained when the powdered tablets were shaken directly with chloroform and extracted into acid.Also, when a stock 0.1% mlV solution of benperidol was assayed after direct dilution in 0.05 M sulphuric acid, and then by the modified USP procedure after dilution in alkali, the mean recovery of benperidol obtained by repeating the extraction procedure six times was 96.7 k 0.7%, confirming the lower result obtained above. Results (Table 2) obtained for a tablet sample that had been stored in the laboratory at room temperature for more than 2 years indicate good stability for the drug in tablet form. The recovery of the drug by the extraction procedure was 94.4%, indicating a deterioration of about 2% associated with the p-fluoroacetophenone portion of the molecule.The proposed procedure appears to be a reliable indicator of benperidol content and is particularly suited for quality control purposes. Certain impurities or degradation products of benperidol, such as 1,3-dihydro-l-(4-piperidinyl)-2H- benzimidazol-2-one, could interfere in the assay and it is advisable that these should be controlled by thin-layer chromatography.3 Benzimidazolone is known to be capable of tautomerism and ionisation in alkali,4 and its presence in benperidol renders that compound amphoteric and able to exist in aqueous solution in an amide form (IA) or enol form (IB), depending on the pH. Infrared spectra of solutions of ?YoANALYST, AUGUST 1986, VOL. 111 925 Table 1. Precision of determination of benperidol in a tablet sample and comparison of results obtained using the recommended difference procedure and a modified USP procedure. Average mass per tablet = 0.0999 g.Labelled benperidol content = 0.25 mg. As - AB = 0.484. Stock standard concentration = 0.1042% m/V Difference procedure Modified USP procedure Powdermass/g AT - AB 0.3140 0.485 0.3004 0.447 0.3014 0.473 0.3045 0.474 0.3110 0.481 0.3109 0.476 0.3034 0.468 0.3100 0.482 0.3007 0.449 0.3030 0.470 Mean . . . . . . . . . . S.D. . . . , . . . , . . . . Relative standard deviation . , Statistical analysis: Student’s t . , . . . . , . Variance ratio , . . . . . . . Benperidol content per tablet/mg 0.249 0.240 0.253 0.251 0.250 0.247 0.249 0.251 0.241 0.250 0.248 0.004 1.6% Difference vs. modified USP 5.88’ 1.65 (NaOH) Renperidol Benperidol A248 nrn content per A249 nrn content per Powder mass/g (in NaOH) tabletlmg (in H2S04) tablet/mg 0.6062 0.520 0.533 0.5975 0.511 0.509 0.6070 0.530 0.525 0.6063 0.523 0.532 0.6084 0.528 0.525 0.237 0.243 0.236 0.235 0.241 0.239 0.238 0.242 0.240 0.238 0.239 0.003 1.3% 0.510 0.521 0.501 0.506 0.521 0.510 0.510 0.520 0.514 0.510 0.239 0.244 0.238 0.240 0.244 0.239 0.239 0.243 0.240 0.238 0.240 0.002 0.8% Difference vs.Modified USP (NaOH) modified USP vs. modified USP Critical (HS04) (H2S04) (P’ = 0.01) 5.14 1.34 2.88 1.79 1.08 5.35 Table 2. Determination of benperidol in a 2-year-old tablet sample using the recommended difference procedure and a modified USP procedure. Labelled benperidol content = 0.25 mg. Average mass per tablet = 0.0997 g. Stock standard concentration = 0.1042% m/V.As - AB = 0.484 Difference procedure Modified USP procedure Benperidol Renperidol Benperidol Powder content per Powder A248 nm content per A249 nm content per mass/g AT - AB tablet/mg masslg (in NaOH) tabletlmg (in H2S04) tablet/mg 0.3016 0.466 0.249 0.6012 0.51 1 0.235 0.500 0.236 0.515 0.237 0.495 0.234 0.3115 0.484 0.251 0.6033 0.513 0.235 0.500 0.235 0.514 0.235 0.510 0.240 Mean . . . . . . . . . . 0.250 0.236 0.236 R’ I R ’ = F R’ I VH CI R ’ I R‘ 1 Q R ‘ I 9 @c-o----H- OC*H5 N926 ANALYST, AUGUST 1986, VOL. 111 benperidol in non-aqueous solvents have indicated’ that benperidol can also exist as a zwitterion (IC) with a positively charged piperidyl nitrogen forming an intramolecular bond with the ketone group of the p-fluoroacetophenone moiety.To investigate this further, the ultraviolet absorption spectra of benperidol were determined in a number of organic solvents. The spectra obtained showed marked bathochromic shifts of the longer wavelength benzimidazolone peak, resem- bling the shift obtained in aqueous 0.1 M sodium hydroxide solution. Typical spectra are shown in Fig. 3. In methanol, isopropanol, hexane, heptane and diethyl ether the peaks occur at 282, 283, 284, 286 and 288 nm, respectively, compared with 288 nm in 0.1 M sodium hydroxide solution. It seems probable that in diethyl ether, as in 0.1 M sodium hydroxide solution, benperidol exists in the enol form (IB) stabilised by salt formation with sodium hydroxide or by hydrogen bonding in the ether (IV). R’ I (IV) The hydrogen bond is presumably not strong enough to stabilise the enol form in the solid state after recrystallisation from ethers, and the polymorphic form I is obtained after recrystallisation from diethyl ether, diisopropyl ether and 1,4-dioxane.This polymorph shows two infrared carbonyl stretching bands (ketone at 1685 cm-1 and amide at 1710 cm-1). It does not show the OH band at 3500 cm-l that is observed with a solution of benperidol in diethyl ether. In most other organic solvents, where the benzimidazolone peak appears in the range 282-284 nm, it appears very likely that the zwitterion is formed. Recrystallisation of benperidol from most solvents (including hexane, benzene, carbon tetrachloride, chlorobenzene, dichloromethane, tetrahydro- furan and dimethylformamide) does, however, still produce polymorphic form I.Alcohols, however, seem able to stabilise the zwitterion, probably owing to a hydrogen bonding effect. After recrystallisation from isopropanol, for example, ben- peridol is obtained as polymorphic form 11, which shows an infrared spectrum characterised by the absence of amide carbonyl and OH bands, and by a poorly defined N-H stretching band. With ethanol as solvent, the bonding effect is sufficiently strong to allow recrystallisation as an ethanolate The ketone carbonyl stretching band shifts from 1685 cm-1 in polymorph I to 1690 cm-1 in polymorph 11, and it is probable that the solid benperidol polymorph I1 exists as the (111). 5 Wavelengt hin rn Fig 3. (A) isopropanol, (B) diethyl ether and (C) hexane Ultraviolet absorption spectra for 5 X lo-’ M benperidol in zwitterion (IC) in an intramolecularly bonded state, whereas polymorph I is the amide (keto) form (IA). In heptane, the ultraviolet absorption maximum at 286 nm is closer to the 288 nm maximum found in diethyl ether. Azibi et aZ.5 have reported that a third polymorphic form can be obtained by recrystallisation from this solvent, but it could well be a mixture of forms I and 11, and this view is supported by the infrared spectral data. References 1. “British Pharmacopoeia 1980,” Volumes I and 11, HM Stationery Office, London 1980. 2. “United States Pharmacopeia, XXth Revision,” Mack, Easton, PA, 1980. 3. Girgis Takla, P., James, K. C., and Gassim, A. E. H., “Analytical Profiles of Drug Substances,” Volume 14, Academic Press, New York, 1985, p. 245. Efros, L. S . , and Eltsov, A. V., J . Gen. Chem. USSR, 1957,27, 755. Azibi, M., Draguet-Brughmans, M., and Bouche, R., Pharm. Acta Helv., 1982, 57, 182. Azibi, M., Draguet-Brughmans, M., and Bouche, R., J . Pharm. Sci., 1984, 73, 512. Gassim, A. E. H., Girgis Takla, P., and James, K. C., J. Pharm. Pharmacol., Suppl., 1985, 37, 126P. 4. 5. 6. 7. Paper A6/64 Received February 27th, I986 Accepted March 20th, 1986

ISSN:0003-2654    

DOI:10.1039/AN9861100923

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, AUGUST 1986, VOL. 111 927 Automatic Spectrophotometric Determination of Amyloglucosidase Activity Using p-Nitrophenyl-a-D-glucopyranoside and a Flow Injection Analyser Kaj Andre Holm Enzyme Microbiological laboratory, NOVO Research Institute, DK 2880 Bagsvaerd, Denmark An automated flow injection analysis (FIA) method has been developed, using as a chromophore p-nitrophenyl-a-D-glucopyranoside, which is hydrolysed to glucose and p-nitrophenol by amyloglucosidase. The p-nitrophenol is determined spectrophotometrically at 400 nm as a measure of the enzyme activity. The general conditions of the analysis have been optimised. The analytical method is flexibile, as all methods for the analysis of enzymes using soluble p-nitrophenyl derivatives as chromophores may in principle be carried out with the FIA method described.The final result is a simple, automated method with high precision, sensitivity and speed. Keywords: Amyloglucosidase activity determination; spectrophotometry; p-nitrophenyl-a-D-gluco- pyranoside; automated flow injection analysis A general flow injection analysis (FIA) method was required for the determination of the activity of hydrolytic enzymes, using as chromophores various soluble p-nitrophenyl deriva- tives, which under the influence of the given hydrolytic enzyme split off the p-nitrophenol (yellow at alkaline pH) under standardised conditions. The amyloglucosidase hydrolysis of p-nitrophenyl-a-D-glucopyranoside is used here as a model system. The amyloglucosidase (exo-a-l,4-glucan glucohydrolase, E.C.3.2.1.3) is a glycoprotein which contains carbohydrate residues that are glycosidically linked through D-mannose to the hydroxy groups of serine and threonine in the polypeptide chain of the enzyme.1 The enzyme is produced by submerged fermentation of a strain of Aspergillus niger. The enzyme catalyses the stepwise hydrolysis of the a-l,4-linkages and also, at a slower rate, the a-1,6-bonds in liquefied starch, dextrins and oligosaccharides, by releasing single glucose units from the non-reducing end of the molecule. Most methods for the determination of the activity of amyloglucosidase described in the literature are based on the following principle. The maltose is hydrolysed by the enzyme to form glucose, which can then be enzymatically quanti- fied.293 The work reported here uses a simpler analytical principle in which the p-nitrophenol released by the hydrolysis can be determined spectrophotometrically after a pH change as follows.4-7 At pH 4.3, amyloglucosidase hydrolyses the colourless p-nitrophenyl-a-D-glucopyranoside (pNPG) form- ing glucose and p-nitrophenol (pNP).By adjusting the pH to basic values, the yellow colour resulting from p-nitrophenol can be determined spectrophotometrically at 400 nm as a measure of the enzyme activity. Experimental This method gives a relative measurement compared with a manually standardised amyloglucosidase standard.6 One NOVO amyloglucosidase unit (1 AGU) is the amount of enzyme which under given standardised conditions hydrolyses 1 p~ of maltose per minute at 25 "C and pH 4.3.The following standard conditions were used for the FIA method: chromophore, p-nitrophenyl-a-D-glucopyranoside, 2 g 1 -1; buffer, acetate, 0.1 M; incubation pH, 4.3; incubation temperature, 50 "C; incubation time, 20 s; colour reaction time 3 s; total analysis time 24 s. The flow injection analysis equipment used consisted of a sampler (Tecator 5007), a pump unit (Tecator 5020), a print unit (Alfacom Sprinter 40), a recorder (Servogor 120), a spectrophotometer (Shimadzu UV 120-02) and a cuvette (Helma QS 178-720). Reagents All chemicals were of analytical-reagent grade. Acetate buffer solution, pH 4.3, 0.1 M. The solution is prepared by dissolving 4.44 g of sodium acetate trihydrate and 3.75 ml of glacial acetic acid in 1 1 of de-mineralised water. p-Nitrophenyl-a-D-glucopyranoside solution, 2 g 1- (6.64 mmol 1-1).The reagent is prepared by dissolving 0.200 g of p-nitrophenyl-a-D-glucopyranoside in 100 ml of acetate buffer (pH 4.30, 0.1 M). The reagent is prepared freshly every day and must be protected against light. Sodium carbonate solution, 0.1 M, pH 11.6. This is prepared by dissolving 10.6 g of sodium carbonate in 1 1 of de- mineralised water. Amyloglucosidase standard solutions, 2.5, 5, 7.5 and 10 AGU ml-1 of NOVO enzyme in de-mineralised water. The solutions are stable for up to 1 month if kept frozen at - 15 "C in disposable centrifuge tubes. Procedure A diagram of the FIA procedure is shown in Fig. 1. The sample (30 pl) is injected into the carrier stream and the pNPG reagent is added. The incubation takes place at 50 "C, pH 4.3.Sampler De-mineralised C water R1 pNPG R2 Na2C03 0.53 ml min-1 Fig. 1. Flow diagram for the FIA amyloglucosidase method928 ANALYST, AUGUST 1986, VOL. 111 The enzyme reaction is stopped by adding sodium carbonate solution, resulting in a final pH of 10.1 , and the p-nitrophenol is then determined spectrophotometrically at 400 nm. Results The starting point of the present FIA method is an unpubli- shed manual analytical method,6 which uses the following conditions: 1 ml of the sample (0.5 AGU ml-I), 2 ml of pNPG (1 g 1-1) and an incubation time of 20 min (pH 4.3, 30 "C). The enzyme reaction is stopped by the addition of 3 ml of borax solution (0.1 M). The spectrophotometric determi- nation of p-nitrophenol is then immediately carried out at 400 nm.Incubation Temperature With the manual method, the incubation time is 20 min at 30 "C. With FIA, the incubation time is extremely short (only 20 s), i.e., it is the initial reaction rate of the enzyme that is registered. However, the degree of conversion of the p-nitro- phenyl-a-D-glucopyranoside at the short incubation time is smaller than that observed with the manual method. To compensate for this, the manual incubation temperature of 30 "C was raised to 50 "C, giving an increase in sensitivity for the detection of the enzyme reaction products of 300%. This is possible as the NOVO amyloglucosides enzyme is heat- resistant up to 55-60 "C at pH 4.3 . Incubation Time In an attempt to increase the sensitivity of the method, the incubation time was increased by using the "stopped flow" facility of the Tecator 5020.At a time ( T I ) after sample injection, when the sample was in the incubation coil, the flow was stopped at a previously fixed time ( T2). In the experiment the incubation time was increased from 20 s to 25, 30, 40 and 50 s total incubation time (Fig. 2). As was expected, a gain in sensitivity was obtained, but the calibration graph for enzyme was still rectilinear only up to an absorbance of 0.2 for 5 AGU mi-1 (T2 = 0). Incubation Buffer Acetate buffer (pH 4.3, 0.1 M) was used as an incubation buffer. In order to determine the p-nitrophenol by spectro- AGU ml-1 Fig. 2. to 30 s ( T2): Stopped flow assay. The incubation time is increased from 5 T ~ / S T,/s A 0 0 B 5 5 C 5 10 D 5 20 E 5 30 photometry, it is necessary to change the pH from 4.3 to a basic value.The molar absorptivity for p-nitrophenol is dependent on pH up to pH 10.8 The application of borate buffer (0.1 M), as indicated in the manual instructions, gave a pH of 9.1; however, when the borate buffer was replaced by a sodium carbonate buffer, a pH of 10.1 was recorded, which gave an additional gain in sensitivity of 5%. pNPG Reagent The specificity of the analytical method is dependent on the purity of the p-nitrophenyl-a-D-glucopyranoside reagent. Certain commercial products may be contaminated with p-nitrophenyl- (3-D-glucopyranoside . This is undesirable, because the (3-glucosidase will also be determined by the assay. It is possible to determine the purity of the p-nitro- phenyl-a-u-glucopyranoside in the following way: pure P-glu- cosidase may not give any reaction in a solution of p-nitro- phenyl-a-D-glucopyranoside.pNPG Concentration and Determination of K , Using the standardised conditions, the pNPG concentration was varied from 1 to 2 g 1-1 (Fig. 3); 2 g 1-1 gave the best response with regard to linearity and reproducibility. Using the Hanes plot, the Michaelis - Menten constant (K,) was determined to 6 mM pNPG 1-1. As can be seen from Fig. 3, the optimum sodium carbonate concentration is 0.1 M. p-Nitrophenol Spectral Curve The maximum absorption of the reaction end product, p- nitrophenol, was examined at pH 7 and 11.4 (Fig. 4). At neutral pH, a maximum absorption was seen at 310 nm ( A = 0.23) , but a pH shift to 11.4 gave a very sharp maximum at 400 nm ( A = 0.378).Establishment of a Sample Blank Value If the culture broths are very strongly coloured it is necessary to include a sample blank value. This may be done by changing the addition of reagent on FIA, so that the sodium carbonate reagent is added before the pNPG reagent. Sensitivity of Method With FIA, the enzyme can be determined from 0.5 to 10 AGU ml-1, whereas the corresponding level of activity that can be determined by the manual method6is only 0.1 to 1 AGU ml-1. 0.7 I I 0 2 4 6 8 10 AGU ml-1 Fig. 3. Experiments with variation in pNPG concentration: pNPG/g I- I Na2C03/h.I A 2 0.1 B 1 0.1 C 2 0.5ANALYST, 0.4 0.3 8 m $ 0.2 n a 0.1 4UGUST 1986, VOL. 111 300 350 400 450 5 Wavelengthhm 0 Fig. 4. 20 pmol I-’. pH: A, 11.4; and B, 7.0 Effect of pH on maximum absorption.pNP concentration, Speed of Analysis With a sampler coupled in, it is possible to analyse about 90 samples per hour. The present method has been used for 2 years in our laboratory with excellent results. 929 Discussion In principle, the present FIA method is very flexible as it can be used for the determination of various hydrolytic enzymes, which are able to cleave the p-nitrophenyl compounds. The described method is based on the amyloglucosidase - p-nitro- phenyl-a-D-glucopyranoside reaction, but has also, in our laboratory, been used for the determination of a-arabino- furanosidase with p-nitrophenyl-a-L-arabinofuranoside as the chromophore and a-galactosidase activity with p-nitrophenyl- a-D-galactopyranoside as the chromophore. 1. 2. 3. 4. 5 . 6. 7. 8. References Lineback, R., Carbohydr. Res., 1968, 7, 106. Bergmeyer, H. U., “Methoden der Enzymatischen Analyse,” 3. Aufl., Verlag Chemie, Weinheim, 1974, pp. 1241 and 1250. Holm, K. A., Anal. Chim. Acta, 1980, 117, 359. Katsuse, A. J., J. Biochem., 1939,30, 89. Halvorsen, H., and Ellias, L., Biochim. Biophys. Acta, 1958, 30, 28. Internal Analytical Method, NOVO, AF 115/2, BVR. 193, unpublished (available from NOVO Industri, Bagsvaerd, Denmark). Leisola, M., Ojamo, H., and Kauppinen, V., Enzyme Microb. Technol., 1980, 2, 121. Biochemica, Merck, Darmstadt, 197017. Paper A51384 Received October 28th 1985 Accepted February 19th 1986

ISSN:0003-2654    

DOI:10.1039/AN9861100927

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST AUGUST 1986 VOL. 111 93 1 Characterisation and Evaluation of the Use of Membrane Mimetic Agents to Amplify Chemiluminescence from the Lucigenin -Hydrogen Peroxide Reaction System Terrence E. Riehl,” Cheryl L. Malehornt and Willie L. Hinze Department of Chemistry Analytical Micellar Institute Wake Forest University P. 0. Box 7486, Winston-Salem NC 27109 USA The 10,1O‘-dimethyl-9,9’-biacridinium dinitrate (lucigenin) - hydrogen peroxide - N-methylacridone chemiluminescence (CL) system was characterised in different types of membrane mimetic agents and homogeneous solvents and the effects of various membrane mimetic agents on the lucigenin - hydrogen peroxide light reaction were assessed. Aqueous solutions of the surfactants hexadecyltrimethylammonium c h I o ri de 3-( N-do decyl - N N-d i m et h yl a m mo n io) p ropa n e- 1 -su I p ho nate PO I yoxyet h y I e ne( 23 )d odeca no I d i -octadecyldimethylammonium bromide and a- p- and y-cyclodextrin enhanced the CL intensity by factors of up to 3.4,2.5,4,1.6,2.3,14 and 12.6 respectively.A lowering of the CL emission was observed in the presence of all anionic surfactant media examined. The various experimental variables that influence the magnitude of the CL enhancements or reductions are briefly described and the advantages and limitations of these agents as CL enhancers are discussed. The alterations in the CL intensity are rationalised in terms of the effect of these different membrane mimetic agents on the rate of the reaction and/or excitation efficiency. Lastly the analytical implications and possible advantages of using such membrane mimetic agents in lucigenin ch em iI u m i nescent assays a re discussed.Keywords Enhanced lucigenin chemiluminescence; hydrogen peroxide; micelles; cyclodextrins; vesicles Aqueous solutions of different membrane mimetic agents (e.g. organised assemblies or ordered media such as cyclodextrins surfactant micelles and vesicles) have been shown to be useful in a variety of analytical applications.1-7 Their presence can favourably alter chemical and photo-physical pathways and rates modify equilibria alter the micro-environment around dissolved solutes enhance the water solubility of solutes and alter quantum efficiencies, among other effects.’ 8-10 Observations of enhanced chemiluminescence (CL) in solvent systems consisting of membrane mimetic agents have recently been investigated.11 The most widely studied of these CL reaction systems have been those involving 10,lO’-dimethyl-9,9’-biacridinium dinitrate (lucigenin) or related acridan-type compounds in which increases in CL quantum efficiency ranging from 4 to 130 times have been reported. l2-19 Physico-chemical studies have revealed that surfactant hexa-decyltrimethylammonium bromide (CTAB) micelles and dioctadecyldimethylammonium bromide (DODAB) ves-icleslz-15 and cyclodextrins” were effective in enhancing the CL efficiency observed from the lucigenin - hydrogen per-oxide reaction. No characterisation of the lucigenin -hydrogen peroxide reaction or evaluation of the relative effectiveness of these membrane mimetic agents on it have been reported.Such information is required before the use of membrane mimetic agents in analytical CL measurements can be considered and this has provided the impetus for further investigation of the lucigenin - hydrogen peroxide CL reaction system. This CL reaction is important analytically owing to its ability to quantitate either hydrogen peroxide lucigenin or acridan labelled compounds or certain enzymes or metal ions based on their catalytic or inhibitory effect on the reaction.ll. 2(k23 This paper describes the results of our characterisation of the lucigenin - hydrogen peroxide CL system in various mem-* Present address Department of Biology Washington University. t Present address Department of Chemistry Indiana University.St. Louis MO 63130 USA. Bloomington IN 47405 USA. brane mimetic agents and of our evaluation of such media as CL intensity amplification agents. This work aims to use the results of this study to improve the lucigenin CL assay for the determination of hydrogen peroxide lucigenin or peroxidase via the use of appropriate membrane mimetic agents. These applications will be described in subsequent ~ a p e r s . 2 ~ Experimental Apparatus Absorption spectra were determined with a Varian Cary 219 UV - visible spectrophotometer. Fluorescence data were obtained using an American Instruments Aminco - Bowman spectrophotofluorimeter equipped with a 1P21 PM tube. Fluorescence lifetime measurements were performed at the Center for Fast Kinetics Research at the University of Texas in Austin using a Quantel neodymium:YAG laser system.This system has been described elsewhere.24 All chemilumines-cence measurements were made on a Turner Designs 20-000 luminometer equipped with an end-on PM tube (S-11 response) range extender Cavro automatic syringe injector and printer system. A Fisher Series 5000 Recordall was used to record the CL intensity - time profiles and a Fisher Model 300 sonic dismem brator used to sonicate appropriate surfac-tant solutions in the preparation of synthetic vesicles. A Fisher Accumet Model 320 expanded-scale pH meter equipped with a Fisher combination glass - calomel electrode was used for all pH measurements. Reagents 10,10’-Dimethyl-9,9’-biacridinium dinitrate (lucigenin) (Sigma) lO-methyl-9( 10H)-acridone (N-methylacridone) (Aldrich) hexadecyltrimethylammonium chloride (CTAC) (Kodak) polyoxyethylene(23)dodecanol (23-lauryl ether; Brij-35) (Sigma) research-grade 3-(N-dodecyl-N,N-dimethy1ammonio)propane-1-sulphonate (sulphobetaine; SB-12) (Serva Fine Biochemicals) electrophoresis-grade sodium dodecylsulphate (NaLS) (Bio-Rad) the sodium salt of deoxycholic acid (NaDC) or cholic acid (NaC) (Calbiochem), dioctadecyldimethylammonium bromide (DODAB 932 ANALYST AUGUST 1986 VOL.111 (Kodak) water-soluble (3-cyclodextrin polymer (Chinoin Bio-chemical Research Laboratory Budapest) a- P- and y-cyclodextrin (Sigma or Advanced Separation Technology), urea (Calbiochem) and 2,3-dimethylnaphthalene (9870, Aldrich) were used with no further purification.The solvents, spectranalysed dimethyl sulphoxide (DMSO) spectranalysed 1,4-dioxane and HPLC-grade water all from Fisher Scientific, were used as obtained. All other chemicals and solvents were of the best grade available and were used without further purification. Concentrated stock solutions of lucigenin (1 .O X 10-3-1.5 x 10-2 M) CTAC (0.1M.40 M) NaLS (0.05-0.20 M) SB-12 (0.25 M) NaDC (0.50 M) a- and y-cyclodextrin (0.10 M), P-cyclodextrin (0.014.10 M) urea (4.0 M) and water-soluble P-cyclodextrin polymer (12.0% m/v) were all prepared by accurately weighing the solid material and diluting to the final volume with water. A concentrated aqueous 30% (0.25 M) Brij-35 solution served as the source of Brij-35. Basic solutions were prepared from certified standard 1.00 2.50 or 5.00 M solutions of sodium hydroxide (Fisher).Stock solutions of N-methylacridone (1.0 x 10-3 M) were prepared in DMSO or in 1,4-dioxane. Hydrogen peroxide solutions were prepared by dilution of a 30% hydrogen peroxide stock solution (Sigma). The pH of the peroxide solutions was adjusted to 4.5 in order to minimise their decomposition.23 The hydrogen peroxide was standardised against standard potassium per-manganate and the diluted solutions were used within 30 min of preparation. Lower concentrations of all of these solutions were achieved by serial dilution with water or the aqueous solution of the membrane mimetic agents being studied. Procedure Sodium 1,3-(dodecylcarboxypropane)-2-sulphonate (NaDDS) was synthesised as described by Kunitake and Okahata,25 and the synthetic surfactant vesicle solutions were prepared using modified literature methods.25.26 The typical vesicle prepara-tion consisted in the sonic dispersion of DODAB or NaDDS (15.0-45.0 mg of surfactant in 2.0-6.0 ml of HPLC-grade water) at 68-78 "C using the standard 19-mm probe of the Fisher sonic dismembrator set at 35% power.Using this procedure,optically transparent solutions of vesicles were obtained within about 35-75 min sonication time. The stock vesicle solutions were filtered (Whatman No. 42 filter-paper) prior to use. After cooling this stock solution below the phase- transition temperqture of the surfactant the desired surfactant vesicle concentration could be achieved by appropriate dilutions.Attempts to prepare DODAB aggreg-ates via a different procedure,l5 which involved stirring a 0.01 M DODAB suspension for 72 h did not yield clear solutions, and we were unable to prepare DODAB vesicles using this procedure. Any solution containing hydrogen peroxide should not be sonicated as this speeds up its decomposition. A vesicle solution of hydrogen peroxide is prepared by adding a small volume of the concentrated vesicle stock solution to the peroxide-containing sample solution. The spectroscopic characterisation and solubility measure-ments of lucigenin and N-methylacridone in the membrane mimetic agents were carried out using procedures described el~ewhere.24~28 Binding constants for the interactions of lucigenin and N-methylacridone with the different membrane mimetic agents were evaluated using either the solubility or chromatographic m e t h ~ d .~ ? ~ ~ An established fluorescence quenching method was used to determine the binding constant for the interaction of hydrogen peroxide with the different surfactant membrane mimetic systems,27 with 2,3-dimethylnaphthalene as the fluorescent probe. The pK for hydrogen peroxide in different media was determined from potentiometric measurements. The general CL experimental procedure consisted in pipetting (using Hamilton microsyringes) all but one of the required reagents into a polypropylene cuvette (8 x 50 mm, 1.6 ml volume Evergreen Scientific) which was agitated and placed into the cuvette holder of the luminometer. After an incubation period of 20-60 s the last reagent was injected into the cuvette using the automatic dispensing syringe so that the final volume was always 300 pl.The injection process simultaneously initiated the CL reaction and activated the data acquisition system of the luminometer. The CL signal was collected over the entire 300-600 nm region and electronically integrated for a specified time period hence giving the integrated CL intensity. Typically a delay time (the time interval between the initiation of the CL reaction and start of data collection) of 10 or 15 s and a run time (the actual data collection period) of 10 or 15 s were used. In some experiments the continuous run mode was used in order to obtain the entire CL intensity - time curve. The specific types of CL experiments differed only with respect to the order of addition to and the volume of each of the reagents in the cuvette and the incubation time used.The experiments designed to determine the effect of surfactant micelle - vesicle concentration on CL intensity were carried out using the following conditions 100 p1 of lucigenin 50 1.11 of surfactant solution (or water for water reference system) 100 pl of hydrogen peroxide (or water for blank) an incubation period of 60 s an autoinject of 50 p1 of sodium hydroxide and a run time(= delay time) of 10 s. The CL intensity - time profiles were determined using the same procedure with the instrument in the continuous run mode. The work with cyclodextrins was conducted using the following conditions: 200 pl of the appropriate cyclodextrin solution which also contained the desired concentration of sodium hydroxide (or 200 pl of aqueous sodium hydroxide for the water reference system) 50 pl of lucigenin an incubation period of 30 s an autoinject of 50 pl of hydrogen peroxide (or water for the blank) a delay time of 1-5 s and a run time of 60 or 120 s.The CL intensity - time profiles were determined using a similar procedure with the instrument in the continuous run mode. Unless stated otherwise all integrated and peak CL intensities were corrected for the intensity of the blank which contained no peroxide. These background corrected CL values are referred to as the net CL signals. All reagent concentrations reported are the final solution concentrations, accounting for any dilutions.All pH values given were those obtained from actual measurement of the final mixture. All measurements were carried out at 24.0 k 2.0 "C in air-saturated solutions. Results and Discussion Description of Membrane Mimetic Agents The effects of three general types of membrane mimetic agents i.e. surfactant micelles (both normal and bile salt), synthetic surfactant vesicles and cyclodextrins as media for the lucigenin - hydrogen peroxide chemiluminescent reaction were examined. A comparison of the general features and properties of these different types of membrane mimetic agents is presented in Table 1. The structure charge type and critical parameters of the specific surfactants used to form micelles and vesicles are compiled in Table 2. It is important to note that under the conditions required for the lucigenin -hydrogen peroxide CL reaction it is usually possible to form micelles or vesicles at much lower surfactant concentrations compared with those necessary in water alone (Table 2).',24328 In addition to these surfactant micelles and vesicles that associate in water to form aggregated structures the cyclodex-trins make up the third type of membrane mimetic agents employed in this investigation.They possess a fixed rigid molecular structure consisting of macrocyclic glucose poly-mers that contain six or more 6-D( +)-glucopyranose groups, attached via a-( 1,4) linkages. The cyclodextrins employed in this work contained 6 7 or 8 such units and are referred to a ANALYST AUGUST 1986 VOL. 111 933 Table 1.Comparison of general features and properties of different types of membrane mimetic Micellar systems ~~~ Bile salt Synthetic Characteristic feature or property Normal micelles micelles vesicles Constituents . . . . . . . , Preparation . . . . . . . Shape/structure . . . . . . Diameter (and weight-averaged relative molecular mass) Stability . . . . . . . . . . Effect of dilution with water . . . . . . . Alkyl-substituted Salts of surfactants bile acids . . . Dissolution of the appropriate surfactant or bile salt at a concentration above its CMC in water Roughly spherical Small aggregate aggregates in which the bile molecules orient in a back-to-back manner . . . . . 0.3-0.6nm 0.20-0.26 nm . . . . . . Indefinite Indefinite .. . . . . Micellar aggregated structure is (2000-8000) destroyed if the surfactant concentration is diminished below its CMC Dialkyl-substituted surfactants Sonication of an aqueous suspension of the surfactant Ellipsoidal multi-compartment closed bi-layer Cyclodextrins C ycloam yloses Dissolution in water Truncated cone doughnut shaped 30-90 nm 0.45-1 .O nm (>106) (970-1 300) Days Indefinite Unaltered Unaltered a Adapted from reference 8. b More detailed descriptions of these membrane mimetic systems are provided in references 1-3 6-10 25 26 and 29. a-cyclodextrin (cyclohexaamylose) b-cyclodextrin (cyclohep-taamylose) and y-cyclodextrin (cyclooctaamylose) respec-tively. Also included for study was one p-cyclodextrin polymer (P-CDP).29 This material consists of P-cyclodextrin molecules cross-linked to each other by epichlorohydrin.Compared with P-cyclodextrin this polymeric material exhibits a greater solubility in water. 10,29 Characterisation of Lucigenin N-Methylacridone and Hydrogen Peroxide in Membrane Mimetic Agents Before the CL studies the effects of the different types of membrane mimetic agents on some physical properties of the CL reagent (lucigenin) analyte (hydrogen peroxide) and CL primary emitter (N-methylacridone) were determined. The degree of interaction (binding) between these species and the membrane mimetic agents was also evaluated. Lucigenin It was found that lucigenin binds to the DODAB and NaDDS vesicles [binding constant Kb = (5 k 3) X 103 M - ~ ] .The magnitude of the binding constant is in the same range as that reported for lucigenin’s interaction with the CTAC NaLS and SB-12 surfactant micelles.24 Lucigenin can also interact with the cyclodextrins particularly the larger 6- and y-cy-clodextrin. This may be caused either by partial inclusion into their hydrophobic cavity or at higher pHs by electrostatic interactions (the primary and secondary hydroxyl groups on the rim of the cyclodextrin molecule have pKa values in the range 12.1-12.66). The solubility of lucigenin in the presence of surfactant vesicles and cyclodextrins was increased slightly (1.1-5.4 times) compared with that in bulk water (water solubility = 1.3 x 10-2 M). This finding is similar to that reported for lucigenin’s solubility in aqueous micellar media .24 The results obtained in this work and that published previ0usly2~ indicate that the interactions between lucigenin and these types of ordered media do not alter appreciably lucigenin’s visible absorption maximum ( ~ 3 6 8 nm) intensity (molar absorptivity 3.76 X 104 1 mol-1 cm-l) or fluorescence emission wavelength (485 k 3 nm).Although the fluorescence quantum yield and lifetime of lucigenin are essentially unaltered in the presence of micellar Brij-35 or SB-W4 and cyclodextrin media (Qf = 0.48 t = 19.1 2 0.7 ns) there is a measurable decrease in these parameters in the presence of micellar CTAC and NaLS24 and DODAB and NaDDS vesicular media. In these latter systems the fluorescence intensity and lifetime of lucigenin decrease with increasing surfactant concentration.The pKal of lucigenin determined in these types of organised media was not changed significantly compared to that in water alone (pKal = 12.43).3O Hydrogen peroxide The binding constants for the interaction of the analyte, neutral hydrogen peroxide with micellar NaLS and CTAC, and vesicular NaDDS and DODAB were determined to be 100 k 30 170 k 45 160 k 40 and 300 k 50 M-1 respectively. The values in micellar media are in good agreement with those reported previously by Encinas and Lissi.27 We did not examine the binding interaction of hydrogen peroxide with cyclodextrins but it has been reported31 that hydrogen peroxide interacts with cyclodextrins to form 1 1 inclusion complexes. The pKa of hydrogen peroxide in water alone was determined to be 11.65 2 0.15 and was not altered appreciably by the presence of any of the membrane mimetic agents.Brown and Darwent32 have noted that the acid dissociation constant for hydrogen peroxide does not change significantly in the presence of micellar CTAC. Under very basic con-ditions in which peroxide exists in its anionic HO2- form one would expect it to exhibit stronger binding to cationic vesicles or micelles than its neutral form due to the additional favourable electrostatic interactions. On the other hand the binding of the peroxide anion to either anionic micelles or vesicles (or to the de-protonated cyclodextrin molecule 934 ANALYST AUGUST 1986 VOL. 111 Table 2. Summary of aqueous surfactant systems and parameters" Surfact ant Aggregate type CH,(CHZ)I,N+ (CH,) a- .. . . . . . . . . * * Cationic micelle Hexadecyltrimethylammonium chloride (CTAC), Dioctadecyldimethylamrnonium bromide (DODAB), 3-(N-Dodecyl-N N-dimethylarnmonio)propane-1- . . . . [CH3(CH2)1712N+(CH3)2 Br- . . . . . . . . Cationic vesicle sulphonate (SB-12), CH3(CH2)1 IN+ (CH3)2(CH2)3S03- . . . . . . . Zwitterionic micelle (CH,) (CH2)11 (OCH,CH2)230H . . . . . . . . Non-ionicmicelle CH3(CH2) l 1 0 S 0 3 - Na+ . . . . . . . . . . . . Anionic micelle Polyoxyethylene(23)dodecanol (Brij-35), Sodium dodecylsulphate (NaLS), Sodium 1,3-( dodecylcarboxypropane)-2-sulphonate (NaDDS) 9 [CH3(CH2)1 *C0KHzI[CH3(CH2)11 C02l-CHS03 Na+ . . . . . . . . . . . . . . . . Anionic vesicle ~~X(CMC)VM Nc 1.3d 0.6e 0.3f (1 .3)d 78 3.3d 1.2f 55 0.lci (0.09)d 40 8.ld 2 . 7 ~ 1.5' 62 0.01- -Sodium cholate (NaC) oH I 7.6d 8 . 3 h 5 Sodium deoxycholate (NaDC), Anionic bile salt micelle HO 0.3d 2.9h 15 a Literature values taken from references 8 9 and 24. CMC determined in this study. Literature values are given in parentheses. Aggregation number refers to the number of surfactant molecules per aggregate species. Values given were taken from the literature. CMC value in water alone. CMC value in 0.05 M NaOH. CMC value in 0.10 M NaOH. g Minimum concentration of surfactant required to form vesicles during sonication step. It should be emphasised that the minimum concentration needed to form vesicles is not equivalent to the concept of CMC used for micelles.Once formed vesicles cannot be destroyed by dilution. CMC value at pH 9.0. would be expected to be meagre or non-existent due to electrostatic repulsion. In agreement Yamaguchi et al. have reported that H02- did not form an inclusion complex with P-cyclodextrin.31 N-Methylacridone The primary emitter of the lucigenin CL system is postulated to be N-methylacridone.3" Consequently it is important to characterise N-methylacridone in the different membrane mimetic agents. N-Methylacridone was found to interact with and bind to DODAB and NaDDS vesicles [Kb = (2.0-7.0) X lo4 M - ~ ] to the same extent as with surfactant micelles.24 It had been reported previously that surfactant micelles en-hanced the aqueous solubility of N-methylacridone thereby improving the performance of the lucigenin CL reaction system.24.28 Solubility measurements revealed that the presence of surfactant vesicles and cyclodextrins could also increase N-methylacridone's solubility in water.For instance, in the presence of 0.01 M cyclodextrins the solubility of N-methylacridone was increased by factors of 1.6,3.1 and 4.1 in a- p- and y-cyclodextrin respectively compared with that in water alone (N-methylacridone water solubility = 3.5 X M). These solubility enhancements were approximately the same as those observed in 0.01 M surfactant micellar solutions.24 The presence of cyclodextrins or vesicles did not appre-ciably alter N-methylacridone's ultraviolet absorption wavelength (258 k 2 nm) or intensity [molar absorptivity (6.1 k 0.3) x 104 1 mol-1 cm-11.However there were subtle changes in the two visible absorption bands for N-methyl-acridone in the 370-420 nm wavelength region. By compari-son with spectra that were obtained for N-methylacridone in pure solvents of different polarity,Z4 it was found that the surfactant vesicles provided N-methylacridone with a micro-environment that was similar to that observed in Iow molecular weight alcohols (i.e. ethanol or propanol). Results in the bile salt micelles (NaDC NaD) were similar. The same conclusions had been reached about the effective environment of N-methylacridone in other micellar systems. 14.24 By com-parison in the presence of cyclodextrins the visible absorp-tion bands did not differ appreciably from those observed in water alone and they closely resembled those obtained in aqueous 5.0-10.0 V/V% dioxane solutions.Kondo et al. had estimated that the effective dielectric constant inside th ANALYST AUGUST 1986 VOL. 111 935 cyclodextrin cavity was in the range 47-70 corresponding to that of different water - 1,4-dioxane mixtures.33 Conse-quently it appears that the effective micro-environment experienced by N-methylacridone in cyclodextrins is slightly more polar than that in surfactant micelles or vesicles. We have previously described the fluorescence parameters of N-methylacridone in different homogeneous solvents and in micellar CTAC NaLS Brij-35 and SB-12 and in fl-cy~lodextrin.~~ In general the position and intensity of emission from N-methylacridone changed little in these ordered media compared with values observed in bulk water alone (Af = 431 _+ 2 nm Qf = 0.95 2 0.05 t = 15.3 k 0.5 ns).The results obtained in this work suggest that similar conclusions can be reached about the effects of the bile salt micelles (NaDC NaC) a- and y-cyclodextrin and the NaDDS vesicle systems on N-methylacridone’s fluorescence proper-ties. As was observed in some of the micelle systems,24 very high concentrations of some of these membrane mimetic media slightly diminished the fluorescence intensity of N-methylacridone. This is thought to be due to the presence of quenching impurities in a few of the surfactant and cyclodextrin preparations. In the presence of DODAB vesicles however a substantial reduction in the fluorescence lifetime and intensity of N-methylacridone was observed even at low surfactant concentrations.For example at 7.5 x 10-3 M DODAB the fluorescence intensity of NMA was only 65% of that observed in bulk water alone. The intensity progressively decreased as the DODAB concentration increased. This was due as suggested by Paleos et al. ,I4 to the quenching of N-methylacri-done by the bromide counter ion of the surfactant. Although not strictly constant the Stern - Volmer “constant” based on the initial slope of the plot of [(FIo/FI) - 11 vs. [Br-1 at bromide ion concentrations less than or equal to 0.10 M for the quenching of N-methylacridone by bromide in water was found to be 4 IfI 2 M - 1 . In order to ensure maximum luminescence from the NMA primary emitter purified membrane mimetic agents should be employed in any analytical application.Also for cationic surfactant systems chloride or nitrate rather than bromide or iodide should be employed as the counter ion because the former counter ions do not appreciably quench the emission of the primary emitter in this CL system. Summary The data presented in this section show that there are appreciable differences in the degree of interaction and binding possible between the CL reagents (lucigenin and hydrogen peroxide in addition to the primary emitter, N-methylacridone) and the various types of membrane mimetic agents examined. The presence of such agents affects the distribution of the reactants and alters their effective concentration at the reaction site(s) which in turn will change the rate of the CL reaction compared with that in water alone.As the net CL intensity integrated over a specified time interval depends on both the over-all CL quantum efficiency, @cL (where QcL = i.e. the product of the excitation efficiency for generation of the excited state and the lumines-cence yield of the primary emitter) and the number of molecules reacting per unit time dCldt according to equation (1),24 the observed CL intensity will also be altered owing to the presence of membrane mimetic agents. The micro-environment experienced by lucigenin , hydrogen peroxide and N-methylacridone (and also the reaction intermediates) can also be altered relative to bulk water. This can affect the over-all CL quantum efficiency and 700 ~ 0 10 20 3 Tirne/rn in Fig.1. Relative CL intensity vs. time profiles for the lucigenin -hydrogen peroxide reaction in different media A water; B 2.0 X 10-3 M Brij-35 micelles; and C 1.00 X 10-3 M CTAC micelles. Conditions 3.3 x 10-4 M lucigenin; 0.30 M sodium hydroxide; and 3.3 X M hydrogen peroxide hence the observed CL intensity (CLI). Consequently the presence of such agents could influence any CL assays based on this CL reaction system. Depending on the specific membrane mimetic agent used there should be a range of possible effects observed. Effect of Different Types of Membrane Mimetic Agents on the Lucigenin - Hydrogen Peroxide CL Reaction System In order to investigate whether a medium of membrane mimetic agents can function effectively for the lucigenin -hydrogen peroxide CL reaction system different charge-type surfactant micelles and vesicles and different size-type cyclodextrins were added to the reaction mixture and the time course of the reactions monitored.A typical representative CLI vs. time profile is shown in Fig. 1 for the lucigenin -hydrogen peroxide reaction at pH 13.45 in water alone and in the presence of CTAC and Brij-35 micelles. A change in the shape of the profiles is seen in the presence of these two micelle systems when compared with that in water. For example the maximum CL emission is greater by a factor of approximately 3.2 in the micellar media. There is also a 1.8-2.4-fold reduction in the CL half-life (ti the time required for the maximum CLI to decay to half of its original value).Other membrane mimetic agents were tested in a similar fashion and the results are compiled in Table 3 as maximum CLI and ti values. Table 3 (and similar data from other runs) shows that the presence of all size-type cyclodextrins (a- fl- and y-cyclodex-trin) cationic surfactant CTAC micelles and DODAB ves-icles and zwitterionic SB-12 and non-ionic Brij-35 micellar systems increased the maximum CL signal and reduced the CL half-life relative to that observed in water alone under the same experimental conditions. The data indicate the catalysis of this CL reaction by these membrane mimetic agents. Whereas the magnitude of the net CL light yield observed in this work for CTAC micelles is in good agreement with that previously reported for cationic CTAB micelles,14 the shape of the CLI - time profile is not.Paleos et al. 14 stated that the higher CL quantum yield observed in micelles was due to a longer emission duration and not to a higher CL intensity. As is shown in Fig. 1 our results for CTAC are in contrast to this 936 ANALYST AUGUST 1986 VOL. 111 Table 3. Effect of different organised media on the maximum net chemiluminescence intensity and half-life of the lucigenin - hydrogen peroxide reaction Experimental conditions [ L ~ c ] = 3.3 x 10-4111, [H,O,I = 3.3 x 10-4 M, [NaOH] = 0 . 3 3 ~ . . . . [LUC] = 1 .o X lop3 M , [H202] = 1.7 x M, (NaOH] = 0 . 0 1 ~ . . . . [ L ~ C ] = 2.75 x 10-4 M, [H2O2] = 2.77 x 10-3 M, [NaOH] = 0.167 M . . [Luc] = 1.8 X 10-'M . . [H202] = 8.0 X M, [NaOH] = 0 .2 8 ~ . . . . . . . . . . . . . . Medium Net CLI,,,,~i Water 200 CTAC (10-3 M) 665 Brij-35 (2 X lop3 M) 630 SB-12 (6 x M) 490 NaLS (8 x lop3 M) 36 NaDC (3 x 10 -2 M) 96 Water CTAC ( M) 105 212 Water 440 DODAB (6.6 x M) 565 95 NaDDS (5.0 x 10-3 M) Water 100 (w-CDe (5.0 x 10-3 M) 230 P-CDe (4.5 X 10-3 M) 418= 1400d y-CDe(5.0 x 1 0 - 2 ~ ) 1260 t+blmin 20.1 11.3 8.8 S G G -53 32 24.1 13.5 G -2.7 -1.9 -1.5 1.1 1.2 a Net maximum chemiluminescence intensity observed for the reaction. This value has been corrected for the background (blank) Refers to the time required for the chemilurninescence intensity to decay to half of its maximum intensity. (G = f,,2 greater compared Values obtained using a freshly prepared P-cyclodextrin solution.Values obtained using an aged P-cyclodextrin solution. chemiluminescence observed for the lucigenin - base reaction in the absence of the analyte (i.e. no peroxide). with that in water; S = t1,2 shorter compared with that in water). e CD refers to cyclodextrin. A Q '' 400 1 Q I A I - . 0 0.002 0.004 0.006 0.008 0.01 0.05 0.10 SurfactanVmol I-' Fig. 2. Variation in the net integrated CL emission intensity with surfactant concentration for the lucigenin - hydrogen peroxide reaction system A NaLS ( H 12.5); B SB-12 (pH 12.4); C CTAC (pH 12.8); and D Brij-35 &H 12.5). Line E indicates the emission intensity observed for the CL reaction in the absence of surfactant at pH 12.5. Conditions 3.30 X l o - 4 ~ lucigenin; 2.90 X 10-4111 hydrogen peroxide; and instrument delay and run times 10 s Although the earlier work had bromide (in CTAB) rather than chloride (in CTAC) as the micellar counter ion it is more likely that the difference in the results is caused by the manner in which the blank was determined and/or the fact that the earlier work used a large excess of hydrogen peroxide in the CL reaction.It has been previously noted that changes in the ratio of the concentration of reactant(s) to surfactant micelle can alter the appearance of the CL intensity vs. time c ~ r v e s . * ~ ~ ~ 8 Also in our work the time profiles have been corrected for the background CL - time curve obtained under the same conditions in the absence of hydrogen peroxide. In the previous work with CTAB,l4 no background corrections in the absence of hydrogen peroxide were made.The profiles for the reaction mixture in the micelle were instead compared directly with that in water with all reactants present. Our value for the CL enhancement due to DODAB vesicles is substantially lower than published values. 13.15 This could be explained by some of the same reasons mentioned for the CTAB example. Alternatively this discrepancy could be due to differences in the DODAB vesicle preparation. Vesicles used in this study were prepared by the sonication method,25?26 whereas sonication was not used in the prepara-tion of vesicles used in the earlier ~ o r k . 1 ~ Our enhancement factors and CL half-life data for the lucigenin - hydrogen peroxide reaction in the presence of cyclodextrins are in qualitative agreement with the recently published results of Grayeski and W00lf.19 The age of the P-cyclodextrin solutions seemed to have an effect on the results obtained.Aged cyclodextrin solutions yielded greater CL enhancements compared with results obtained using freshly prepared solutions (Table 3). This could be due to a solubility effect. Although both the aged and freshly prepared cyclodextrin solutions appeared to be homogeneous it is possible that the cyclodextrin was not completely dissolved in the freshly prepared solutions. In agreement recent dissol-ution studies using P-cyclodextrin hydrate indicate that at least 2 h are required before an equilibrium solubility concentration value is obtained in water at 40 OC.34 Alternatively this could be due to the presence of mould or other biological contami-nation in the aged solutions although none was obvious to the naked eye.A more sensitive test needs to be used to determine the presence or absence of such possible contami-nants. In contrast to the behaviour (i.e. increased CLI reduced ti) shown by the systems above all anionic membrane mimetic agents tested (e.g. NaLS NaDC NaD micelles and NaDDS vesicles) caused a reduction in the maximum CL intensit ANALYST AUGUST 1986 VOL. 111 937 (Table 3) and a considerable increase in the CL half-life compared with values seen in water. Such behaviour indicates an inhibition of the lucigenin - hydrogen peroxide CL reaction by these negatively charged membrane mimetic agents i.e., the rate term in equation (1) is diminished in these media.With such anionic surfactant systems the electrostatic repul-sion of the hydrogen peroxide anion (-OH2) from the negatively charged surface of the micelle or vesicle to which the positively charged CL reagent lucigenin is bound seems to be responsible for the inhibition of the CL reaction. It is the rate of emission and not necessarily the total light yield that is lowered in the presence of these anionic membrane mimetic systems. Effect of Membrane Mimetic Agent Concentration on the Lucigenin - Hydrogen Peroxide CL Reaction The effect of the membrane mimetic agent concentration on the net integrated CL intensity and half-life of the lucigenin -hydrogen peroxide CL reaction was also investigated.As shown in Fig. 2 the net integrated CLI - surfactant concen-tration profiles obtained in the presence of micellar Brij-35, CTAC SB-12 and NaLS for the reaction of 3.3 X 10-4 M lucigenin with 2.9 X 10-4 M hydrogen peroxide at pH ca. 12.5 can depend strongly on the surfactant concentration. At surfactant concentrations below the critical micellar concen-tration (CMC) (not all shown) the net CL intensities are about the same as in water regardless of the charge type of the surfactant employed. Above the CMC two general types of behaviour are observed. In the first as the concentration of an anionic surfactant such as NaLS (Fig. 2) and NaDC or NaC (not shown) increases the net CLI decreases until it reaches a minimum after which it increases slightly.At the minima the net integrated CLIs are diminished by factors ranging from 7 to 30 compared with water depending on the specific anionic surfactant system. Also it was found that the CL half-life increased with surfactant concentration until it reached a maximum after which it decreased slightly at higher surfac-tant concentrations. Similar CLI and half-life behaviour with surfactant concentration was observed for the NaDDS vesicle system. The shapes of these net integrated CLI - surfactant concentration profiles are suggestive of micelle or vesicle inhibition of the CL reaction. Such profiles have been observed for numerous other reaction systems,8*9 including some other types of CL reactions in the presence of surfactant assemblies .24.2* In the second type of behaviour which is observed for the other surfactant charge-types examined the net CL intensities increase to a maximum value after which they either level off (Brij-35 Fig.2 ) or decrease slowly with increasing surfactant concentration (CTAC SB-12 Fig. 2). The behaviour ob-served in the presence of DODAB vesicles (not shown) was similar to that exhibited by CTAC micelles. The CL half-life behaviour paralleled the inverse of the CLI profiles i. e. the half-life decreased as the surfactant concentration increased, reached a minimum and then slowly increased with increasing surfactant concentration (or remained constant in the case of Brij-35). The shapes of the CLI - surfactant concentration profiles and diminution of the CL half-lives are indicative of micellar - vesicular catalysis of the lucigenin - hydrogen peroxide CL reaction by these surfactant systems.Similar observations have been reported for micelle - vesicle catalyses of other reactions,1,8,9 including several CL reaction^.^^,^^.^^ The maximum CLI enhancement occurs at surfactant concentrations of 1.0 X 7.5 X and 8.5 X M for CTAC SB-12 and DODAB respectively. For the non-ionic micellar Brij-35 system the net CLI was at a maximum and independent of Brij-35 concentration in the range 2.0 X 10-3-2.5 x 10-2 M (which was the upper limit examined). Although not studied in detail the surfactant concentrations at which these CLI maxima occur appear to be independent of the pH in the pH range 10-13 and of the lucigenin concentration in the concentration range 1.0 x 10-6-4.0 x 10-4 M.For these surfactant systems the net integrated CL intensity is 1.9-4.2 times greater than that in their absence. It should be noted that the absolute magnitude of the enhancement factor in these media does depend on the pH and lucigenin concen-tration with the factor being diminished at low pH or high lucigenin concentration 2 3 Grayeski and W00lf19 had reported previously that the integrated CLI observed for the lucigenin - hydrogen peroxide reaction increased as the cyclodextrin concentration increased. Our results confirmed this for all three of the size-type cyclodextrins examined. For example with 8.0 x 10-3 M cyclodextrin the net integrated CLIs were 1.9 17.6 and 3.6 times greater in the presence of a- (3- and y-cyclodex-trin respectively relative to that in water.The greatest enhancement was observed in the presence of P-cyclodextrin. However its solubility in water is limited (ca. 1.6 X 10-2 In an attempt to use higher (3-cyclodextrin concen-trations in order to observe even greater CL enhancements, we examined several other P-cyclodextrin-containing systems. First we solubilised 0.10 M (3-cyclodextrin in aqueous 4.0 M urea. The CLI-time experiments revealed that there was essentially no enhancement observed in this system relative to water alone. Next we employed a water-soluble (3-cyclodex-trin polymer.29 In the 1.5-10.0% m/V P-cyclodextrin polymer concentration range only a small enhancement in the CLI was observed. Consequently we could not overcome the solubility constraints and all subsequent work was conducted using 0.016 M (3-cyclodextrin stock solutions.The magnitude of the CLI enhancements observed in the presence of cycIodextrins is dependent on both the solution pH and lucigenin concen-tration. For instance in the pH range 9-12 examined the CL enhancement in the presence of 4.5 X 10-3 M (3-cyclodextrin varied from 1.4 to 2.8 compared with water. Even more dramatic is the dependence on the lucigenin concentration. At pH 12.3 in the presence of 0.005% hydrogen peroxide the CLI enhancement factor was 1.0 1.2 and 3.4 at [(3-cyclodextrin]/[lucigenin] ratios of 9.4 20 and 950 respec-tively. Consequently for any analytical application relatively high [cyclodextrin]/[lucigenin] concentration ratios must be used.Possible Reasons for the Enhanced Chemiluminescence in Membrane Mimetic Agents In the absence of any other changes in the experimental conditions the observed CL enhancements in the indicated membrane mimetic agents must stem from an increase in the CL reaction rate excitation efficiency and/or luminescence efficiency of the primary emitter [equation (l)]. Enhancement due to increased efficiency of the primary emitter(s), N-methylacridone and/or lucigenin can probably be ruled out because as mentioned earlier their fluorescence quantum yields are not enhanced in the presence of the membrane mimetic agents. According to published mechanistic studies ,30736 the rate step in the lucigenin - hydrogen peroxide CL reaction pathway is postulated to involve conversion of a hydroperoxide (I) to 938 ANALYST AUGUST 1986 VOL.111 1,2-dioxetane intermediate which decomposes rapidly to form excited singlet state N-methylacridone. In competition with this CL pathway are non-CL pathways involving hydroly-sis of the hydroperoxide with resultant formation of ground-state N-methylacridone.30 Consequently there are a number of. possible explanations involving rate and/or excitation efficiency effects that could account for the CL enhancements observed in the presence of micellar CTAC SB-12 and Brij-35 DODAB vesicles and the cyclodextrins. First the presence of these membrane mimetic agents could enhance the rate of oxidation of the hydroperoxide intermediate, thereby decreasing the formation of ground-state N-methylac-ridone via the competing non-CL hydrolytic pathways.The enhanced CL observed for an acridan ester in micellar media has been attributed to this type of effect.16 Alternatively and equivalent in terms of the net effect produced the enhancements could be due to stabilisation of the hydroperoxide in the protective micro-environment of the membrane mimetic agent so that a larger fraction of the hydroperoxide reacts via the desired CL pathway. For example on formation in surfactant micelles or vesicles the hydroperoxide I could orient itself so that the remaining quaternary ammonium moiety (acridinium ring) probably resides near the surfactant surface whereas the other un-charged more lipophilic N-methylamine portion extends into the hydrocarbon core.With this orientation for I the para position of the acridinium ring is sterically shielded from subsequent attachment by the hydroxide ion or water which would lead to hydrolytic products (the non-CL pathway). Consequently the presence of surfactant membrane mimetic agents allows for a more favourable competition by the intramolecular step leading to formation of the dioxetane required for CL. Such shielding - orientation effects in micelles have been previously shown to lead to the alteration of reaction paths and selective product f0rmation.gJ~~~8 In agreement it has also been reported that hydroperoxides bind to both micelles and cyclodextrins and that they exhibit a high degree of stability in such micro-environments.*7~3~ Grayeski and Woolflg have proposed this as a possible explanation for the enhanced CL that they observed in the presence of cyclodextrins.It should also be noted that the rates of several reactions involving hydroperoxides and peroxides have been shown to be significantly altered by the presence of surfactant assemblies.32>40-42 Lastly as has been proposed by Shinkai et aZ.17 and 0thers,13~14J9 the CL enhancements may be attributed to a microscopic solvent effect (i. e. altered micro-environment). The formation of excited state N-methylacridone from the 1,Zdioxetane intermediate is thought to be facilitated by the less polar micro-environment of the membrane mimetic agents. This results in a higher excitation efficiency and increased CL.14,17 A similar explanation has been suggested recently for the CL enhancement observed for the hydrogen peroxide - Cu(I1) - 1 ,lo-phenanthroline reaction in micelles.35 Obviously more mechanistic studies are required before any conclusive and quantitative description of these membrane mimetic agent effects on the CL intensity observed from the lucigenin - hydrogen peroxide reaction is possible.Analytical Implications The results of this study clearly indicate that the development of CL procedures in appropriate membrane mimetic agents is potentially very useful in several respects. Firstly their presence can affect the sensitivity as both the maximum CL light output and duration of emission observed from the lucigenin - hydrogen peroxide reaction can be altered not only by variation of the type of membrane mimetic agent used, but also by its concentration.The ability to use these agents to manipulate the intensity or duration of emission should prove convenient and useful in a variety of analytical applications based on the lucigenin - hydrogen peroxide CL reaction system. In terms of their ability to enhance the maximum CL intensity observed from the lucigenin - hydrogen peroxide reaction the relative effectiveness of the different membrane mimetic agents follows the order P-cyclodextrin > y-cyclodex-trin > CTAC = Brij-35 > SB-12 3 cx-cyclodextrin 3 DODAB > bulk water > NaLS = NaDDS 3 NaDC. Consequently if amplified CL and hence greater sensitivity are desired then use of either the cyclodextrins or micellar CTAC Brij-35 or SB-12 should be considered.Of these cyclodextrins give the best enhancement of CL. However they are more expensive and the amplification factor is very dependent on the exact experimental conditions used. The use of CTAC or Brij-35 is also attractive. With micellar CTAC an enhancement factor of up to 3.9 can be obtained and there is the possibility of observing CL at a lower pH due to CTAC's ability to concentrate the hydroxide ion at its cationic surface. The use of Brij-35 micelles offers a four-fold CL enhancement that is independent of the Brij-35 concentration above its CMC. On the other hand the use of anionic membrane mimetic agents (NaLS NaC and NaDC surfactant micelles and NaDDS surfactant vesicles) diminished the maximum CL emission intensity (and hence sensitivity) which can give an extension of the linear dynamic range at the upper concen-tration limits.Also the duration of the lucigenin - hydrogen peroxide CL reaction in these anionic surfactants is prolonged relative to that in water. From a practical standpoint such increased duration of the light emission could allow for greater flexibility and convenience with respect to the timing of the analytical measurements. Secondly the use of membrane mimetic agents is very attractive from the analytical viewpoint as they offer the possibility of better selectivity. For example metal ions typically interfere in the lucigenin assays for hydrogen peroxide.23 However if the assay is conducted in the presence of cationic micellar or vesicular media the metal ions should be electrostatically repelled from the cationic surfactant surface with a resultant improvement in specificity.The use of appropriate micelles in this fashion has led to better selectivity in some fluorimetric assays.43 Additionally the membrane mimetic agents described in this work are compatible with high-performance liquid chromatographic separation schemes. Aqueous solutions of micelles and cyclodextrins have been shown to function effectively as mobile phases in HPLC.2,6>4"47 Consequently selectivity can be achieved by prior HPLC separation using these agents as the chromato-graphic mobile phase. The enhanced CL detection could then be achieved by post-column addition of lucigenin and NaOH as the membrane mimetic enhancement agents would already be present in the chromatographic mobile phase.This work was supported by the National Science Foundation (CHE-8215508). The fluorescence lifetime experiments and analyses of the data produced were performed at the Center for Fast Kinetics Research (CFKR) at the University of Texas at Austin. The CFKR is supported jointly by the Biotechnol-ogy Branch of the Division of Research Resources of NIH (RR00886) and the University of Texas. The travel to the CFKR was supported by a grant from the Southern Regional Education Board and Wake Forest University. Preparation and reprint costs were supported by a Wake Forest University Research and Publication Fund Grant. The authors express their thanks to Dr. Steve Atherton (University of Texas, Austin Texas) for making the fluorescence lifetime measure-ments Mr.Yoshimi Baba (Nagasaki University Nagasaki, Japan) for making preliminary measurements on the cyclodextrin systems and Michael Borgerding (Wake Forest University) for helpful suggestions. This work was presented in part at the 35th Southeastern Regional Meeting of th 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. References Hinze W. L. in Mittal K. L. Editor “Solution Chemistry of Surfactants,” Volume I Plenum Press New York 1979 p. 79, and references cited therein. Armstrong D. W. Sep. PuriJ Methods 1985 14 213. Pelizzetti E. and Pramauro E. Anal. Chim. Acta 1985,169, 1. Cline Love L. J. Grayeski M. L. Noroski J. and Wein-berger R.Anal. Chim. Acta 1985 169 355. Kornahrens H. Cook K. D . and Armstrong D. W. Anal. Chem. 1982 54 1325. Hinze W. L. Sep. Purif. Methods 1981 10 159. Cline Love L. J. Habarta J. G . and Dorsey J. G. Anal. Chem. 1984 56 1132A. Fendler J. H . “Membrane Mimetic Chemistry,” Wiley New York 1982. Attwood D. and Florence A. T. “Surfactant Systems,” Chapman and Hall New York 1983. Szej tli J. “Cyclodextrins and their Inclusion Complexes,” Akademiai Kiado Budapest 1982. Hinze W. L. Cienti. Technol. 1985 December 20. Hinze W. L. Baba Y . and Singh H. N. Abstracts of Papers, 34th Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy Atlantic City NJ 1983 Abstr. No. 296. Nikokavouras J. Vassilopoulos G. and Paleos C. M. J. Chem.SOC. Chem. Commun. 1981 1082. Paleos C. M. Vassilopoulos G. and Nikokavouras J. J. Photochem. 1982 18 327. Nikokavouras J. and Vassilopoulos G . Z . Phys. Chem. (Leipzig) 1984 265 618. McCapra F. Acc. Chem. Res. 1976 9 206. Shinkai S . Ishikawa Y. Manabe O. and Kunitake T., Chem. Lett. (Jpn.) 1981 1523. McCapra F. Roth M. Hysert D. and Zaklika K. A. in Cormier J. Editor “Chemiluminescence and Biolumines-cence Papers International Conference,” Plenum Press New York 1973 p. 319. Grayeski M. L. and Woolf E. J. J . Lumin. 1985 33 115. Tovmasyan A . P. Galstyan G. G. and Uloyah S. M. Gig. Tr. Prof. Zabol. 1979 No. 3 58; Chem. Abstr. 1979 90, 191834r. Montano L. A . and Ingle J. D. Anal. Chem. 1979,51,919, 926. Carter T. J. N. and Kricka L. J. in Kricka L.J. and Carter, T. J. N. Editors “Clinical and Biochemical Luminescence,” Marcel Dekker New York 1982 pp. 135 and 144. ANALYST AUGUST 1986 VOL. 111 American Chemical Society Charlotte NC November 9th, 1983 (Abstract No. 160). 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37, 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 939 Malehorn C. L. Riehl T. E . and Hinze W. L. Analyst, 1986 111 941 and unpublished results. Hinze W. L. Riehl T. E. Singh H. N. and Baba Y . Anal. Chem. 1984 56 2180. Kunitake T. and Okahata Y. Bull. Chem. SOC. Jpn. 1978, 51 1877. Kunitake T. and Okahata Y. J. Am. Chem. SOC. 1977 99, 3860. Encinas M. V. and Lissi E. A . Photochem. Photobiol. 1983, 37 251. Klopf L. L. and Nieman T. A . Anal. Chem. 1984,56,1539. Fenyvesi E. Szilasi M. Zsadon B. Szejtli J . and Tudos, F. in Szejtli J. Editor “Proceedings of the First International Symposium on Cyclodextrins,” Akademiai Kiado Budapest, 1982 p. 345. Maskiewicz R. Sogah D. and Bruice T. C. J . Am. Chem. SOC. 1979 101 5347 5355. Yamaguchi S. Miyagi C. Yamakawa Y. and Tsukamoto, T. Nippon Kagaku Kaishi 1975 3 562. Brown J. M. and Darwent J. R. J . Chem. SOC. Chern. Commun. 1979 169 and 171. Kondo H. Nakatani H. and Hiromi K. J. Biochem. (Tokyo) 1976 79 404. Jozwiakowski M. J. Connors K. A . Carbohydr. Res. 1985, 143,51. Yamada M. and Suzuki S . Anal. Lett. 1984 17 257. Janzen E. G. Pickett J. B. Happ J. W. and DeAngelis W., J. Org. Chem. 1970 35 94. Jaeger D. A. and Robertson R. W. J. Org. Chern. 1977,42, 3298. Sutter J. K. and. Sukenik C. N. J. Org. Chem. 1982 47, 4174. Matsui Y . Naruse H. Mochida K. and Date Y. Bull. Chern. SOC. Jpn. 1970 43 1909; 1910. Rhodes C. T. Can. J. Pharm. Sci. 1967,3 16. Eremin A. N. and Metelitsa D. I. React. Kinet. Catal. Lett., 1985 27 47. Purnanand D. R. K . J. Surf. Sci. Technol. 1985 1 69. Hinze W. L. Singh H. N. Baba Y. and Harvey N. G., Trends Anal. Chern. 1984 3 193. Armstrong D. W. and Henry S. J. J . Liq. Chromatogr., 1980 3 657. Armstrong D. W. and Stine G. Y. Anal. Chem. 1983 55, 2317. Armstrong D. W. and Nome F. Anal. Chem. 1981 53, 1662. Borgerding M. F. and Hinze W. L. Anal. Chem. 1985 57, 2183. Paper A51435 Received November l l t h 1985 Accepted March 3rd 198

ISSN:0003-2654    

DOI:10.1039/AN9861100931

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST AUGUST 1986 VOL. 111 941 Improved Determi Measurement of L Assemblies nation of Hydrogen Peroxide or Lucigenin by lucigenin Chemiluminescence in Organised Cheryl L. Malehorn,* Terrence E. Riehlt and Willie L. Hinze Department of Chemistry Analytical Micellar Institute Wake Forest University P. 0. Box 7486, Winston-Salem NC 27109 USA Chemiluminescence (CL) from the reaction of 10,1O'-dimethyl-9,9'-biacridinium dinitrate (lucigenin) with hydrogen peroxide in the presence of different organised assemblies has been evaluated as a means of improving the CL assay for the determination of hydrogen peroxide. If the assay is conducted in the presence of 34 N-d od ec y I - N N-d i met h y I a m m o n i 0) p ro pa n e- 1 -s u I p h o n ate h exa d ecy I t r i m et h y I a m m o n i u m c h I o ride o r polyoxyethylene(23)dodecanol (Brij-35) micelles or (J-cyclodextrin improvements in the sensitivity by factors of 1.9-7.6 were observed relative to that in water.The best sensitivity is achieved in micellar Brij-35 in which the CL intensity is linearly proportional to hydrogen peroxide from 3.0 x 10-7 up to 1.0 x 10-3 M (the highest concentration examined). The precision in Brij-35 is good with the relative deviation at the 95% confidence level being 3.2 f 2.2%. The selectivity of the lucigenin CL assay for hydrogen peroxide is also improved in the Brij-35 medium. Additionally the usable pH range in which analytically useful CL can be observed is extended in the Brij-35 or CTAC micellar systems relative to water. Preliminary data suggest that these advantages are also possible if one uses the reaction to quantitate lucigenin rather than hydrogen peroxide.Keywords Hydrogen peroxide determination; lucigenin chemiluminescence; Brij-35; cyclodextrins; lucigenin determination Hydrogen peroxide is often the focus of clinical environ-mental and biological studies and it is used in many industrial and related processes as an oxidising bleaching and sterilising agent. For these reasons it is important that new and improved analytical methods be developed for the determination of trace amounts of hydrogen peroxide.' In addition such methods are also potentially useful for the monitoring of processes or for the indirect quantitative determination of other substances. For example one can monitor the activities of enzymes that specifically catalyse the oxidation of biological materials in the presence of oxygen with the formation of hydrogen peroxide.The substrates or the enzymes in these reactions can be indirectly determined by measurement of hydrogen peroxide.2 Finally methods for hydrogen peroxide determination can also serve as chemical dosimeters for measuring radiation because it is formed from dissolved oxygen in water after exposure to various forms of radia-ti0n.3~4 Many methods for the quantitative determination of hydrogen peroxide have been reported,lJ+'" and chemi-luminescence (CL) procedures appear to be among the most popular and sensitive of these.2.11 Lucigenin (10,lO'-dimethyl-9,9'-biacridinium dinitrate) has been employed previously as a reagent for the CL determination of hydrogen peroxide in water polymer resin and air samples and in biological media.2.4.7-15 The CL assay is based on the fact that lucigenin is oxidised by hydrogen peroxide in basic solution to form the excited state of N-methylacridone.The light emission ob-served from N-methylacridone is proportional to the peroxide concentration.2JlJ4 The primary purpose of this study was to determine the effectiveness of organised assemblies (also termed membrane mimetic agents or ordered media) such as surfactant micelles and cyclodextrins in'increasing the sensi-tivity and selectivity of the lucigenin CL assay for the determination of hydrogen peroxide. In a physico-chemical * Present address Department of Chemistry Indiana University, t Present address Department of Biology Washington University, Bloomington IN 47405 USA.St. Louis MO 63130 USA. investigation we had previously shown that both the maxi-mum CL intensity and duration observed for the lucigenin -hydrogen peroxide reaction system could be dramatically altered by the presence of surfactant micelles and cyclodextrin media.16.17 This paper demonstrates the use of different types of organised assemblies for improving both the sensitivity and selectivity of the lucigenin CL assay for the determination of hydrogen peroxide or lucigenin. In addition it is shown that with the use of an appropriate organised assembly solvent system such as hexadecyltrimethylammonium chloride (CTAC) the pH requirements for efficient CL are modestly reduced.Compared with a homogeneous aqueous medium, the presence of micellar polyoxyethylene(23)dodecanol (Brij-35) improves the sensitivity precision and specificity of the CL assay. Experimental bpparatus Absorption spectra and measurements were made with a Varian Cary 219 UV - visible spectrophotometer. The chemi-luminescence measurements were made using a Turner Designs Model 20-000 luminometer equipped with an end-on photomultiplier tube (S-11 response) a Model 2-069 range extender (neutral density 2 filter 1% light transmission) a Cavro automatic pipette injector and a printer system. A Fisher Series 5000 Recordall was used to record the CL intensity - time profiles. The pH measurements were made using a Fisher Accumet Model 320 expanded-scale research pH meter equipped with a Fisher combination glass - calomel electrode.Reagents 10,10'-Dimethyl-9,9'-biacridinium dinitrate (lucigenin) (Sigma) hexadecyltrimethylammonium chloride (CTAC) (Kodak) polyoxyethylene(23)dodecanol (23-lauryl ether: Brij-35) (Sigma) research-grade 3-(N-dodecyl-N N-dimethy1ammonio)propane-1-sulphonate (sulphobetaine; SB-12) (Serva Fine Biochemicals) electrophoresis-grad 942 ANALYST AUGUST 1986 VOL. 111 dodecylsulphate (NaLS) (Bio-Rad) the sodium salt of deoxy-cholic acid (NaDC) (Calbiochem) dioctadecyldimethyl-ammonium bromide (DODAB) (Kodak) and P-cyclodextrin (Sigma or Advanced Separation Technology) were used with no further purification. Fisher HPLC-grade water was used in the preparation of all aqueous solutions.All other chemicals and solvents were of the best grade available and were used without further purification. Concentrated stock solutions of lucigenin surfactant micelles and cyclodextrins were prepared by accurately weighing amounts of the solid materials and diluting to final volume with HPLC-grade water. Basic solutions were pre-pared from Fisher certified standard 1.00 2.50 or 5.00 M solutions of sodium hydroxide. Lower concentrations of all of these solutions were achieved by serial volumetric dilution with water or the appropriate aqueous solution of the organised assembly being studied. A concentrated aqueous 0.25 M Brij-35 solution served as the source of Brij-35. It should be cautioned here that, depending on the supplier Brij-35 and other non-ionic surfactants usually contain variable amounts of oxidising impurities which can range from 0.04-0.22% hydrogen peroxide equivalents.18J9 Consequently it is imperative to determine appropriate blanks in order to compensate for the background CL caused by the peroxide present in the non-ionic surfactant preparations.Several suppliers now sell highly purified (i.e. very low peroxide) grade non-ionic surfactants but these were not used in this study. Hydrogen peroxide standards were prepared by diluting an aqueous 30% hydrogen peroxide stock solution (Sigma) that was standardised against standard potassium permanganate (sodium oxalate was the primary standard).20 The concen-tration and stability of the peroxide solutions were also determined by monitoring the UV absorbance of hydrogen peroxide at 200 240 or 304 nm.6.21 The diluted hydrogen peroxide solutions were maintained at pH 4.5 (adjusted with HC1) and direct exposure to light was avoided in order to minimise peroxide decomposition.1 ~ 2 The dilute hydrogen peroxide solutions were used within 30 min of preparation. All glass and plastic ware was cleaned by soaking in 50% nitric acid for at least 8 h and then rinsing with distilled water. All glassware was dried in an oven at 150 "C. Procedure The DODAB synthetic surfactant vesicles were prepared as described previously.17 One should not sonicate any peroxide analyte-containing solutions in order to prevent possible alteration of the hydrogen peroxide concentration. The general CL experimental procedure consisted of pipetting (using Hamilton microsyringes) all but one of the required reagents into a polypropylene cuvette (9 X 50 mm, 1.6 ml volume Evergreen Scientific) which was agitated and placed in the cuvette holder of the Turner Designs lumin-ometer.After an appropriate incubation period (20 or 30 s) the last reagent was injected into the cuvette using the automatic dispensing syringe so that the final volume was always 300 p1. The injection process simultaneously initiated the CL reaction and activated the data acquisition system of the luminometer. The CL signal was collected over the 300-600 nm region and electronically integrated for a specified time period hence providing the integrated CL intensity. Typically a delay time (time interval between the initiation of the CL reaction and start of data collection) of 10 s and a run time (actual data collection period) of 10 s were utilised.The experiments designed to optimise the solution pH were conducted using the following conditions 100 1.11 of lucigenin, 50 pl of organised assembly media (or water for water reference system) 100 pl of hydrogen peroxide (or water for blank) an incubation period of 60 s an autoinject of 50 yl of the appropriate concentration of sodium hydroxide and a run time = delay time of 10 s. The conditions employed for the determination of the hydrogen peroxide calibration data were 100 y1 of sodium hydroxide (in water or appropriate organised assembly) 100 y1 of lucigenin an incubation period of 20 or 30 s an autoinject of 100 y1 of hydrogen peroxide (or water for blank) a delay time of 10 or 15 s and a run time of 10 or 15 s.The work with P-cyclodextrin was typically conducted using the following conditions 200 yl of appropriate (J-cyclodextrin concentrated solution that contained the desired concen-tration of sodium hydroxide (or 200 1.11 of NaOH in water for the water reference system) 50 yl of lucigenin an incubation period of 30 s an autoinject of 50 pl of hydrogen peroxide (or water for blank) a delay time of 1-5 s and a run time of 60 or 120 s. The recommended optimised procedure for the deter-mination of hydrogen peroxide is as follows 50 1.11 of 1.90 X 10-2 M (or other desired concentration) sodium hydroxide in 7.5 X M CTAC) 50 p1 of 2.00 X 10-3 M lucigenin an incubation period of 30 s an autoinject of 200 p1 of the hydrogen peroxide sample a delay time of 10 s and a run time of 10 s.All integrated or peak CL intensities were corrected for the intensity of the blank which contained no peroxide. These background corrected CL values are referred to as net CL intensities. All reagent concentrations reported in the follow-ing text are the final solution concentrations accounting for any dilutions. Unless otherwise stated the analytical hydrogen peroxide calibration graphs were obtained using 6-8 different hydrogen peroxide concentrations with 4-6 data points being taken at each hydrogen peroxide level. The detection limit for hydrogen peroxide or lucigenin was defined as the concentration yielding an analytical signal equal to three times the standard deviation of the blank (signal to noise ratio = 3).All measurements were carried out at 25 k 2 "C in air-saturated solutions. M Brij-35 (or in 6.0 X Results and Discussion Hydrogen Peroxide Stability A brief stability study indicated that acidified (pH 4.5) hydrogen peroxide stock solutions (in the 2.00-10.0% mlV hydrogen peroxide concentration range) appeared to be more stable in the presence of non-ionic micellar Brij-35 than in acidified water alone. Previously the use of the surfactant nonylphenylpolyoxyethylene phosphate (NPPEP) was shown to stabilise hydrogen peroxide against oxidative deterioration during storage.23 Presumably solutions of both the Brij-35 and NPPEP retard the decomposition of hydrogen peroxide by complexing trace amounts of metal ions in their poly-oxyethylene shells.Optimisation of Experimental Parameters Concentration and type of organised assembly employed The results of our previous physico-chemical study indicate that the most promising organised assemblies with respect to their ability to amplify CL from the lucigenin - hydrogen peroxide reaction are micellar CTAC Brij-35 and SB-12, vesicular dioctadecyldimethylammonium bromide (DODAB) and P-cyclodextrin.17 The CL intensity observed from the lucigenin - hydrogen peroxide reaction as a function of the concentration of these organised assemblies has also been reported previously. 17-25 The data indicate that the optimum surfactant concentrations are 1.00 x lO-3,6.0 X 10-3 and 1.0 X 10-3 M for CTAC SB-12 and DODAB respectively.With non-ionic micellar Brij-35 maximum CL is observed at Brij-35 concentrations 3 2.0 x 10-3 ~ . 1 7 For P-cyclodextrin, the CL intensity increased with increasing (J-cyclodextrin concentration up to its solubility limit in water (0.016 M). 1725 In the work reported in this study the final @-cyclodextrin concentration range was usually 4.0 x 10-3-8.0 x 10-3 M ANALYST AUGUST 1986 VOL. 111 943 Effect of variation of p H and reagent concentrations In order to optimise the conditions for the lucigenin - hydrogen peroxide reaction for analytical purposes in the four most promising systems (i.e. CTAC SB-12 Brij-35 and P-cyclodextrin) a number of parameters such as pH lucigenin concentration and order of reagent addition were investi-gated.pH. The dependence of the net relative CL signal on the pH of the reaction solution is shown in Fig. 1 indicating that the optimum pH is ca. 12.0 ? 0.1 in both bulk water and the micellar media. The profile in the presence of 0.0045 M P-cyclodextrin (not shown) was similar in shape but of greater intensity (1.4-2.8 times) than that shown for bulk water in the pH region examined (pH 9.0-12.0). It should be noted that P-cyclodextrin functions as an effective buffer in the pH range 11.1-13.6 (8-cyclodextrin possesses 14 secondary and seven PH Fig. 1. Net integrated chemiluminescence intensity vs. final reaction pH in A 2.0 x 10-3 M Brij-35; B 1 .O x 10-3 M CTAC; and C water media. Conditions 3.38 X 10-4 M lucigenin; 3.30 x 10-4 M hydrogen peroxide; 25.0 "C; instrument delay and run times 10 s 104 t 1 HpOp/mol I-' Fig.2. Calibration graphs of log (net integrated CL intensity) vs. log (hydrogen peroxide molar concentration) for the lucigenin - hydrogen peroxide CL assay conducted in A 1.0 x 10-3 M CTAC pH 11.6; B, aqueous solution pH 12.6; C 1.0 x 10-3 M CTAC pH 10.7; and D, aqueous solution pH 10.9. Conditions lucigenin = 1.0 x 10-3 M ; 25.1 "C; delay time = run time = 10 s primary hydroxy groups pKa ca. 12.1 12.617). The optimum pH determined in this work is about one pH unit lower than that reported for the cobalt(I1)-catalysed lucigenin - hydrogen peroxide reaction in bulk water.7 The strengthening of the net CL signal with increasing pH is presumably associated with the increased dissociation of hydrogen peroxide (pK = 11.65) to form the reactive H02-anion in the pH range 10-12.17J6 The slight decrease in CL signal with increasing pH in the pH range 12.1-13.0 could be due to the conversion of lucigenin to its pseudobasic form One criticism of hydrogen peroxide analyses based on the lucigenin CL reaction is that the assay has to be conducted under very basic conditions forefficient CL.2 However the use of micellar CTAC or Brij-35 results in a significant reduction of the pH requirements necessary for attainment of analytically useful CL for this reaction system.For example, inspection of Fig. 1 reveals that the net CL intensity observed in Brij-35 at pH 10.6 can only be achieved in bulk water alone at a pH of 12.1! Consequently in the presence of these organised assemblies it is possible to conduct hydrogen peroxide CL assays at lower solution pH without any diminution in sensitivity (refer also to the hydrogen peroxide calibration graphs in Fig.2). This micellar feature could be very useful in certain hydrogen peroxide assays such as those involving coupled enzymatic systems. Lucigenin. The parameter next optimised was the lucigenin reagent concentration. Under all reaction conditions in this study a linear increase was observed in both the peak height and net integrated CL intensity with increasing lucigenin concentration in the range 5.0 X 10-7-5.0 X M-(the slopes of the log CL intensity vs. log [lucigenin] graphs were 0.90 2 0.10). However the signal to noise ratio did not increase correspondingly with the lucigenin concentration.Conse-quently a final lucigenin concentration in the range 3.0 x 10-4-10.0 x 10-4 M was chosen as the optimum concentration for the subsequent analytical hydrogen peroxide assays in the different surfactant organised assemblies. For the P-cyclodextrin system the CL enhancement factor was strongly dependent on the cyclodextrin to lucigenin concentration ratio [(3-~yclodextrin]/[lucigenin] employed. 17 At pH 12.3 in the presence of 0.005% hydrogen peroxide the CL amplification factor was 1.2 and 3.4 at [P-cyclodextrin]/ [lucigenin] ratios of 20 and 950 respectively. Hence owing to the limited solubility of (3-cyclodextrin in water and the need for a high [ P-cyclodextrin]/[lucigenin] ratio a final lucigenin concentration of 1.8 x 10-6 M was chosen for the subsequent analytical work using (3-cyclodextrin.(PKa 12.4).26 Order and timing of addition of reagents The last parameters to be investigated concerned the order and timing of the addition of the different reagents. The procedures typically involved the addition of two of the required reagent combinations (e.g. H 2 0 - NaOH or organ-ised assembly media - NaOH together with lucigenin and hydrogen peroxide) into the cuvette and waiting for a specified incubation time period before automatic injection of the third reagent. It was found that there was no significant difference in the quantitative results as a function of waiting time (incubation period) as long as the period was 3 20 s. Also the order of addition of the reagents was found to have no effect on the net magnitude of the CL signal.However the order did have an appreciable effect on the signal to noise ratio. The optimum signal to noise ratio was consistently achieved in all media in which the hydrogen peroxide was the last reagent to be added. This may be explained as follows. The background CL signal in the absence of hydrogen peroxide is presumably due to the reaction of lucigenin with molecular oxygen in the basic medium. This reaction has been reported to be CL even in the absence of hydrogen peroxide.26 Consequently if the base and lucigenin are added first and allowed to react durin 944 ANALYST AUGUST 1986 VOL. 111 the 220 s incubation period before introduction of the blank solution (or hydrogen peroxide for the reaction mixture) the background signal is more reproducible.Analytical Parameters and Calibration Data Assay for hydrogen peroxide Fig. 2 shows the log - log calibration graphs for hydrogen peroxide in the absence and presence of CTAC micelles at several different pH values. Such graphs were typically linear (correlation coefficients 20.99) over at least three orders of magnitude and had a slope of unity. Table 1 summarises the data obtained in the different organised assemblies and compares the data with that obtained in water alone. The detection limits for a signal to noise ratio of three differed depending on the exact experimental conditions and instru-ment sensitivity setting(s) employed. However in almost all instances the detection limits were lowered to approximately the same degree as one would have predicted based on the CL enhancement factors that have been reported for the lucigenin - hydrogen peroxide reaction in these organised assem-blies.13J7,*5 In some instances the improvement in detection limits was greater than that expected from the enhancement factor alone (see last Brij-35 entry Table 1).This is due to a combination of the CL amplification factor and the better signal to noise ratio achieved in the micellar Brij-35 medium. The reverse situation was observed for the assay in p-cyclodex-trin i.e. a less favourable signal to noise ratio off-set the substantial CL amplification factor in this reaction medium (refer to last entry Table 1). In general the detection limits achieved in both the organised assembly media and in water alone under optimum conditions were better by at least factors of 10-50 compared with previously published values for the determination of hydrogen peroxide using the lucigenin CL reaction systern.7>gJ4 The lowest detection limit achieved in this study was 3.3 x 10-7 M hydrogen peroxide in the presence of Brij-35 micelles.If one accounts for the difference in the definition of the signal to noise ratio (3 in this work 2 in published reports), then this value is comparable to the detection limits reported for the peroxide assay using the 2,4,6-trichlorophenyl oxalate (TCPO) - perylene CL system.27 It may be possible to improve the detection limits further (Table 1) by at least another factor of 5-10 by (i) alteration of the volume(s) of reagents to analyte employed in order to reduce the sample analyte dilution factor (the analyte hydrogen peroxide was diluted by a factor of 3 in this work); (ii) by switching from the batch to the continuous flow mode which has been reported to improve the signal to Table 1.Summary of parameters for working graphs [log (integrated CL intensity) vs. log (hydrogen peroxide concentration)] Enhancement factorb -2.4 0.5 --1.3 -1.9 -3.1 -1.9 -2.9 3.3 --14.4 Slope 1.04 0.99 1 .00 1 .00 1.02 0.99 1.01 0.98 1.03 1.08 1.01 1.03 1.05 0.99 1.02 1 .oo 1.31 0.99 Intercept 6.00 6.18 6.62 6.32 4.99 4.98 5.49 5.66 5.2 5.8 6.4 6.7 5.60 5.86 6.33 6.75 6.63 6.62 Hydrogen peroxide concentration range examinedh 1 x 10-6-1 x 10-3 1 x 10-6-1 x 10-3 3 x 10-6-6 x 10-4 2 x 10-4-2 x 10-3 2 x 10-4-2 x 10-3 3 x 10-6-2 x 10-3 3 x 10-6-2 x 10-3 3 x 10-5-3 x 10-3 3 x 10-5-3 x 10-3 3 x 10-6-6 x 10-4 3 x 10-6-6 x 10-4 2 x 10-5-2 x 10-3 2 x 10-6-2 x 10-3 3 x 10-6-6 x 10-4 3 x 10-5-6 x 10-4 2 x 10-6-6 x 10-4 3 x 10-6-6 X 10-4 3 x 10-6-6 x 10-4 Correlation coefficient 0.9981 0.9998 0.9998 0.9992 0.9943 0.9995 0.9998 0.9999 0.9976 0.9979 0.9999 0.9998 0.9992 0.9999 0.9997 0.9999 0.9994 0.9989 Limit of detectionh (SNR' = 3) 2.6 X 10-6 7.5 x 10-7 8.3 x 10-7 1.6 x 10-6 I 3.7 x 10-6 9.5 x 10-7 8.5 x 10-7 4.4 x 10-7 7.4 x 10-7 3.3 x 10-7 2.1 x 10-6 2.5 x 10-6 2.6 X 10-5 8.0 x a The final reagent concentrations are given in the footnotes below for each system studied.Typically the calibration graphs were prepared using 6-8 different analyte concentrations each of which was determined in duplicate or triplicate. Luc and HP are the abbreviations for lucigenin and hydrogen peroxide respectively. Enhancement factor refers to the net integrated CLI in the ordered media divided by the net integrated CLI in water. Conditions [Luc] = 5.0 x 10-4 M pH = 12.05 HP autoinjected delay time = run time = 10 s. [SB-121 = 6.0 x 10-3 M. Conditions [Luc] = 3.3 x 10-4 M pH = 11.65 HP autoinjected delay time = run time = 15 s. [NaDC] = 0.030 M. [DODAB] = 1.0 x 10-3 M. Not determined.j [CTAC] = 1.0 x 10-3 M. Conditions pH = 13.5 [Luc] = 3.3 x 10-4 M HP autoinjected delay time = run time = 10 s. I Conditions [Luc] = 1.0 x 10-3 M pH = 12.85 HP autoinjected delay time = run time = 10 s. [Brij-35] = 1.0 x 10-3 M. Conditions [Luc] = 3.3 x 10-4 M pH = 11.5 HP autoinjected delay time = run time = 10 s. O [Brij-35] = 2.0 x 10-3 M. P Conditions [Luc] = 1.8 x 10-6 M pH = 13.40 hydroxide autoinjected delay time = 1 s run time = 120 s. q [p-CD] = 5.0 x 10-3 M. g Conditions pH = 12.0 [Luc] = 1.0 x 10-3 M HP autoinjected delay time = run time = 10 s. SNR = Signal t o noise ratio ANALYST AUGUST 1986 VOL. 111 945 noise ratio for some analyses conducted in aqueous2 and micellar media13.36; and/or (iii) by elimination of the lumin-ometer’s range extender (a neutral density No.2 filter placed between the sample cuvette and the detector allows 1.0% light transmission at 450 nm). If any improvement in detection by these means is possible then the Brij-35 micellar modified lucigenin procedure may prove to be competitive with the solid-state CL reactor recently developed for the detection of hydrogen peroxide based on the TCPO - perylene CL system (detection limit 1.5 X 10-8 M) which is conducted in a mixed hydro - organic medium.28 It is also important to comment on the effects of pH on the sensitivity of this hydrogen peroxide CL assay in organised assemblies and in bulk water. The data shown in Fig. 2 illustrate the potential use of micellar media to relax the usually very stringent (i.e. very basic) requirements for the lucigenin CL assay for hydrogen peroxide .2J2 Comparison of the upper two lines reveals that a two-fold improvement in sensitivity is obtainable if the assay is conducted in 1.0 x 10-3 M CTAC micelles rather than in bulk water even though the hydroxide ion concentration is an order of magnitude less (line A CTAC pH 11.6; line B water pH 12.6).Even more dramatic is the almost six-fold increase in sensitivity observed in CTAC at still lower pH but of comparable basicity to that in water (refer to lines C and D CTAC at pH 10.7 bulk water at pH 10.9). Comparison of the two middle calibration lines indicates that even if the hydroxide ion concentration is about 100 times less the sensitivity observed in micellar CTAC medium is diminished by a factor of only 1.4 compared with that in water (water system at pH 12.6; CTAC system at pH 10.7).In aqueous solution buffered at pH 10.4 a lower limit of 1.0 X 10-5 M hydrogen peroxide has been reported using the lucigenin CL assay.l4 At the same pH in the presence of 1.0 x 10-3 M CTAC the lower limit is estimated to be S3.5 x 10-6 M hydrogen peroxide. Similar data (not shown) were also obtained for the lucigenin CL assay conducted in micellar Brij-35. Hence the use of a micellar Brij-35 or CTAC medium for the lucigenin hydrogen peroxide assay has the advantage of a greater pH range in which analytically useful CL data can be obtained. This feature should prove to be desirable in a number of applications based on the lucigenin - hydrogen peroxide CL system particularly in the development of coupled enzymatic reaction schemes for the CL quantitation of biological substances *,11,13 In order to compare the reproducibility of the hydrogen peroxide assay in micellar Brij-35 with that in bulk water net integrated CL values for sets of six samples of hydrogen peroxide at each of seven concentrations (0-5.8 x M) were determined (Table 2).At the 95% confidence level the relative deviation for each determination is relatively small in the micellar medium ranging from 0.9 to 6.4%; the mean relative deviation is 3.2 & 2.2%. This suggests a high level of precision for the Brij-35 micellar modified lucigenin proce-dure. By contrast the relative deviation at the 95% con-fidence level in water alone is much worse varying from 1.0 to 56.7% with a mean relative deviation of 19.3 k 23.0%.A relative deviation of 11% has been reported for the determi-nation of hydrogen peroxide in water using the cobalt(I1)-catalysed lucigenin CL reaction.8 Although not extensively examined analytical data were also obtained in several anionic micellar systems i.e. sodium dodecylsulphate and sodium deoxycholate (NaDC). Use of these anionic micellar organised assembly systems had previ-ously been shown to reduce the maximum CL intensity and prolong the duration of emission observed from the lucigenin -hydrogen peroxide reaction compared with that in water.” Analytical data for NaDC (Table 1) reveals that the relative integrated CL intensity is only half that observed in water, which results in a hydrogen peroxide detection limit that is twice that in water.Hence one advantage to using these anionic media is that they allow for an extension of the linear dynamic range at the upper concentration limit. In situations where maximum sensitivity is not required their use may allow for greater flexibility in the timing of the analytical measurements as their presence prolongs the duration of the lucigenin - hydrogen peroxide CL reaction. Assay for lucigenin A brief investigation was conducted in order to evaluate the feasibility of determining lucigenin using the lucigenin -hydrogen peroxide CL reaction system in the presence of organised assemblies. The preliminary results (i. e. analytical calibration parameters) are presented in Table 3.The data Table 2. Statistical analysis of calibration data hydrogen peroxide concentration vs. integrated chemiluminescence intensity in the absence and presence of Brij-35 micellesa Water system Brij-35 micellar system Peroxide concentrationb 0 0.29 0.59 2.94 5.87 29.4 58.7 Mean (n = 6) CL intensityc-d 2.4 * 1.3 6.7 k 3.8 1 2 2 1 53 & 1.8 107 * 1.1 517 k 12 1018 f 92 Relative deviation % o ~ . ~ 54.2 56.7 8.3 3.4 1 .0 2.3 9.0 Mean (n = 6) CL intensitycJ 9.6 k 0.6 26.7 f 1.7 42.8 k 0.6 174 f 3.8 340 f 13 1659 f 15 3437 k 46 Relative deviation % o ~ ~ 6.3 6.4 1.4 2.2 3.8 0.9 1.3 Mean (n = 7) 19.3 f 23.0 Mean (n = 7) 3.2 k 2.2 a Conditions [Luc] = 3.3 x 10-4 M [NaOH] = 3.16 x 10-3 M delay time = 10 s run time = 10 s T = 25 “C.For the micelle system, [Brij-351 = 2.0 x 10-3 M. b (In moll-’) X 105. Integrated chemiluminescence intensity. 95% confidence level. Calculated as 100[mean integrated CL intensity - lower integrated CL intensity limit]/mean integrated CL intensity. Equations for calibration line: log (net integrated CL intensity) = 6.33 + 1.02 log [H202] (in water); log (net integrated CL intensity) = 6.75 + 1.00 log [H20,] (in Brij-35). Correlation coefficient 0.9997 (water); 0.9999 (Brij-35) 946 ANALYST AUGUST 1986 VOL. 111 Table 3. Summary of parameters for working graphs [log (integrated CL intensity) vs. log (lucigenin concentration)] Lucigenin Limit of Enhancement concentration Correlation detection/M Mediuma factorb Slope Intercept range examinedh coefficient (SNR = 3) Waterc .. . - 0.80 5.11 5 X 10-7-5 x 10-3 0.9996 e 0.9971 8.5 x 10-7 Water' . . . . - 0.98 6.28 5 x 10-6-5 x 10-4 Brij-35f.g . . 4.0 0.92 6.58 5 X 10-7-5 X 0.9984 2.9 x 10-7 Waterh . . . . - 0.97 6.75 5 X 10-7-3 X 10-4 0.9972 1.7 x 10-7 0.9984 1.2 x 10-7 CTACc*d . . 2.6 0.71 5.14 5 X 10-7-5 x 10-3 0.9998 P-CDh,' . . . . 2.3 0.82 6.33 5 X 10-7-3 x a The final reagent concentrations are given in the footnotes below for each system examined. Enhancement factor refers to the ratio of the net integrated CLI in the ordered media to that obtained in water alone. Conditions [HP] = 2.9 x 10-4 M [NaOH] = 1.0 x 10-2 M hydroxide was autoinjected delay time = run time = 10 s. [CTAC] = 1.0 X 10-3 M. Not determined. Conditions [HP] = 3.3 x lop4 M pH = 12.25 lucigenin was autoinjected delay time = 15 s run time = 10 s.g [Brij-351 = 2.0 x 10-3 M. Conditions [HP] = 0.005 Yo V/V [NaOH] = 5.0 x 10-2 M lucigenin was autoinjected delay time = 2 s run time = 60 s. [p-CD] = 4.0 X M. indicate that the presence of P-cyclodextrin and Brij-35 ordered media can result in improvements in the sensitivity. It is thought that further optimisation of the instrumental and experimental conditions will result in even greater sensitivity enhancements due to the presence of organised assemblies. It should be noted that the same trends with regard to the pH requirements for the lucigenin assay in the presence of micellar CTAC and Brij-35 were observed as have previously been mentioned for hydrogen peroxide.The primary purpose of the development of improved assays for lucigenin is not due to the need to quantitate lucigenin itself but to determine lucigenin-labelled compounds. The covalent coupling of synthetic CL reagents such as luminol to antibodies and antigens has led to the development of some sensitive CL immunoassays.11J9 The use of lucigenin in this regard has been extremely limited but some applications are mentioned in patents.30331 The use of organised assemblies appears promising in terms of helping to further enhance the sensitivity observed for such immunoassays that utilise lucigenin or other acridinium esters as the labelling reagents. 13 Interference Study A lack of specificity is perhaps the most pressing and prevalent problem encountered in most CL assays.2 The lucigenin -hydrogen peroxide reaction system also permits the deter-mination of over 20 other species.32-34 Different organic reducing agents can also react with lucigenin to produce CL.353 Therefore the effects of certain foreign substances on the lucigenin - hydrogen peroxide CL reaction were examined in both water and the micellar Brij-35 media and the results obtained are compared in Table 4.It is evident from the data that under the experimental conditions employed Ag(I), Cu(I1) and Co(1I) ions creatinine and ascorbic acid pose interference problems in the aqueous medium. In contrast, the presence of Brij-35 micelles is seen to be effective in eliminating the m-eta1 ion interference presumably because the metal ions present in solution which can catalyse or inhibit the lucigenin CL reaction are complexed by the poly-oxyethylene moiety of the Brij-35 surfactant.The use of CTAC should provide similar results due to electrostatic repulsion of metal ions from the cationic micelle surface where the CL reaction is presumably occurring,17 but this was not examined. Also the slight interference due to creatinine, observed in an aqueous medium is eliminated in a Brij-35 medium. It has been reported that the presence of micellar media can alter the reactivity and selectivity of organic reductants towards lucigenin.36 The negative interference due to ascorbic acid is concen-tration dependent. This interference is probably due to the Table 4. Effect of other substances on the peroxide CLH YO of peroxide CL signalc In 1.25 X 1 0 - 2 ~ Species added (concentration)b In water Brij-35 Ni(I1) (1.67p.p.m.) .. . . . 109 106 Ag(1) (1.67p.p.m.) . . . . . . 353 104 Cd(I1) (1.67p.p.m.) . . . . . . 95 103 Cu(I1) (1.67p.p.m.) . . . . . . 46 98 Fe(II1) (1.67 p.p.m.) . . . . 102 105 Co(II)(1.67p.p.m.) . . . . . 673(613)d 85 ( 101)d Glucose (0.50 g 1-1) . . . . . . 105 (102)~ 103 (103)e Cr(V1) (1.67p.p.m.) . . . . 106 101 Pb(I1) (1.67p.p.m.) . . . . . . 99 102 Fructose (0.03 g 1-1) . . . . . . - ( 103)f - (97)f Glutathione (0.05 g 1-1) . . . . - (99)f - (101)' Ascorbic acid (10 mg 1- 1) . . . . 0 (99)g 0 (99)g Creatinine(0.30gl-*) . . . . 112 103 a Conditions [Luc] = 3.3 x 10-4 M [HP] = 5.8 x M [NaOH] = 0.167 M delay time = 10 s run time = 10 s T = 25 "C.Concentration of added species in the final solution. Refers to the net integrated CL signal for the Luc - HP reaction in the presence of the added species divided by that for the reaction in the absence of the species multiplied by 100. A value of 100% indicates no change in the peroxide signal; a value of 0% indicates the background CL level. Value for 1.00 p.p.m. added cobalt(II) [NaOH] = 0.0316 M. Value for 0.125 g 1-1 added glucose [NaOH] = 0.0316 M. Value obtained with [NaOH] = 0.0316 M. g Value for 3.3 mg 1-1 added ascorbic acid [NaOH] = 0.0316 M. reaction of the analyte hydrogen peroxide with the ascorbic a~id.37~38 It has been demonstrated that the intensity of CL observed from the lucigenin - hydrogen peroxide system is depressed by ascorbic acid in a concentration-dependent manner .39 The selectivity of the lucigenin CL assay for the determi-nation of hydrogen peroxide is seen to be much improved when the assay is conducted in micellar Brij-35.This improved specificity is potentially the most significant advantage of utilising organised assemblies in CL procedures. Conclusions The results presented in this work demonstrate that the use of appropriate organised assembly reaction media can substan-tially improve the analytical performance of the lucigenin CL assay for hydrogen peroxide. Advantages include improved sensitivity better specificity and an extended usable pH range for the assay. With these improvements the lucigenin assa ANALYST AUGUST 1986 VOL. 111 947 has now become competitive with other popular CL assays for the determination of hydrogen peroxide.It had been reported that solubility difficulties encountered with the lucigenin CL reaction system could be eliminated via the use of ordered rnedia.l7936>4O Consequently in view of the results of this study and work summarised elsewhere,13 it is clear that the use of different organised assemblies should be beneficial in many other CL assays and lead to improvements in the analytical performance exhibited by them. The support of this research by the National Science Founda-tion (CHE-8215508) is gratefully acknowledged. Some prep-aration and reprint costs were supported by a Wake Forest University Research and Publication Fund Grant. This work was taken in part from the MS Thesis of Cheryl L.Malehorn, Wake Forest University August 1985. The authors thank Professor Purnendu K. Dasgupta (Texas Tech University) and a referee for their helpful comments with regard to this work. Parts of this work were presented at the 35th Southeastern Regional Meeting of the American Chemical Society Char-lotte NC November 9th 1983 (Abstract No. 160). 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. References Schumb W. C. Satterfield C. N. and Wentworth R. L., “Hydrogen Peroxide,” Reinhold New York 1975. Seitz W. R. CRC Crit. Rev. Anal. Chem. 1981 1 and references cited therein. Armstrong W. A. and Humphreys W. G. Can. J. Chem., 1965 43 2576. Sarylov V. N. Dubovenko L. I . Guta A. M. Komash-chenko V. N.and Petrov V. N. Khim. Tekhnol. Vody 1982, 4,45. Guilbault G. G. “Practical Fluorescence,” Marcel Dekker , New York 1973. Snell F. D. “Photometric and Fluorometric Methods of Analysis (Nonmetals),” Wiley New York 1981 p. 690. Tovmasyan A. P. Galstyan G. G. and Uloyan S. M. Gig. Sanit. 1978 5 75; Chem. Abstr. 1978 89 8666011. Tovmasyan A. P. Galstyan G. G. and Uloyan S. M. Gig. Tr. Prof. Zabol. 1979 No. 3 58; Chem. Abstr. 1979 90, 191834r. Ehrlich S. H. and Capone S. M. Res. Disclos. 1982,215,64. Totter J. R. in Johnson F. H. and Haneda Y. Editors, ‘‘Bioluminescence in Progress,” Princeton University Press, Princeton 1966 p. 23. Campbell A. K. and Simpson J. S. A. Tech. Metab. Res., 1979 B213 10. Carter T. J. N . and Kricka L. J. in Kricka L. J. and Carter, T.J. N. Editors “Clinical and Biochemical Luminescence,” Marcel Dekker New York 1982 pp. 136 and 140. Malehorn C. L. Riehl T. E . Hoshino H. and Hinze W. L., Am. Lab. 1986 in the press. “Methods of Bioluminescence Analysis,” Boehringer Mann-heim Biochemicals Indianapolis IN 1985 Section 3.1. 15. 16. Babko A. K. Dubovenko L. I. andTerletskaya A. V. Ukr. Khim. Zh. 1966 32 1326. Hinze W. L. Baba Y. and Singh H. N . Abstracts of Papers, 34th Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy Atlantic City 1983 Abstract No. 296. 17. Riehl T. E. Malehorn C. L. and Hinze W. L. Analyst, 1986 111 931. 18. Ashani Y. and Catravas G. N. Anal. Biochem. 1980 109, 55. 19. Magill A, and Becker A. R. J . Pharm. Sci. 1984,73 1663. 20.Ayres G. H. “Quantitative Chemical Analysis,” Harper and Row New York 1958 pp. 440-442 and 655-658. 21. Hwang H. and Dasgupta P. K. Environ. Sci. Technol. 1985, 19 256. 22. Bostick D. T. and Hercules D. M. Anal. Chem. 1975 47, 447. 23. Grifo R. A. US Pat. 3 394 993 1966; Chem. Abstr. 1968,69, P68677f. 24. Fendler J. H. “Membrane Mimetic Chemistry,” Wiley New York 1982. 25. Grayeski M. L. and Woolf E. J. J. Lumin. 1985 33 115. 26. Maskiewicz R. Sogah D. and Bruice T. C. J. Am. Chem. SOC. 1979 101 5347 5355. 27. Williams D. C. Huff G. F. and Seitz W. R. Anal. Chem., 1976 48 1003. 28. Zoonen P. V. Kamminga D. A. Gooijer C. Velthorst, N. H. and Frei R. W. Anal. Chim. Acta 1985 167 249. 29. Weeks I. Woodhead J. S. Campbell A. K. and McCapra, F. in Hunter W. M. and Corrie J. E. T. Editors, “Immunoassay and Clinical Chemistry,” Churchill Living-stone Edinburgh 1983 pp. 524-531. Maier C. L. US Pat. 4 104 029,1978; Chem. Abstr. 1979,90, 3605 1 q . Maier C. L. Br. Pat. 1578 275,1980; Chem. Abstr. 1981,95, 3030g. MacDonald A. Chan K. W. and Nieman T. A. Anal. Chem. 1979 51 2077. Montano L. A. and Ingle J. D. Anal. Chem. 1979,51,919, 926. Fernandez-Gutierrez A. and de La Pena A. M. in Schul-man S. G. Editor “Molecular Luminescence Spectroscopy, Part I,” Volume 77 Wiley New York 1985 Chapter 4 pp. Veazey R. L. and Nieman T. A. Anal. Chem. 1979 51, 2092. Hinze W. L. Riehl T. E. Singh H. N. and Baba Y. Anal. Chem. 1984 56 2180. Barakat M. Z . Bassiouni M. and El-Wakil M. J. Sci. Food Agric. 1972 23 1141. Pokorny J. Poskocilova H. and Davidek J. Nahrung 1981, 25 K29. Nikokavouras J. and Vassilopoulas G. Monatsh. Chem., 1983 114 255. Klopf L. L. and Nieman T. A. Anal. Chem. 1984,56,1539. 30. 31. 32. 33. 34. 463-475. 35. 36. 37. 38. 39. 40. Paper A51436 Received November 25th 1985 Accepted March 3rd 198

ISSN:0003-2654    

DOI:10.1039/AN9861100941

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST AUGUST 1986 VOL. 111 949 Titrations in Non-aqueous Media Part 1. Determination of Factors Influencing the Basicity of Schiff Bases in Nitrobenzene Solvent Esma Kill9 and Turgut Gunduz” Department of Chemistry Faculty of Science University of Ankara Ankara Turkey Potentiometric titrations of Schiff bases with perchloric acid in nitrobenzene solvent were carried out in order to obtain information about the factors that influence the basicities of the Schiff bases. The Schiff bases were synthesised from aromatic aldehydes and amines bearing an aromatic ring. Some contained a methylene group between the aromatic ring and the imino group and others contained a hydroxy group ortho to the imino group. For each Schiff base an S-shaped potentiometric titration curve was obtained.From these curves half-neutralisation potentials were determined and from the latter the pK,’ values. In these determinations the half-neutralisation potential of 0.001 M hexa(p-to1uidino)cyclotriphosphazatriene [N3P3( NH-p-t~l)~] in nitrobenzene with perchloric acid was taken as the standard. Keywords Non-aqueous titration; potentiometric titration; Schiff bases Potentiometric titrations in non-aqueous media yield valuable information about the basicity or acidity of compounds.’-12 In this work a number of Schiff bases (compounds containing the -CH=N- group) were titrated potentiometrically with perchloric acid in nitrobenzene solvent. These Schiff bases contained aromatic aldehydes (benzaldehyde l-naphthal-dehyde salicylaldehyde and 2-hydroxy-1-naphthaldehyde) as the aldehyde component and contained amine components bearing an aromatic ring.Some contained a methylene group between the aromatic ring and the imino group and others contained a hydroxy group ortho to the imino group. The Schiff bases synthesised from these aldehydes and amines are listed in Table 1. Schiff bases are very weakly basic compounds,13 so their titration requires great care. As nitrobenzene is a good non-aqueous solvent with a dielectric constant of E = 36 good S-shaped titration curves (mV versus ml or mV versus equivalent of acid/equivalent of base) were obtained. As an example the titration curve of salicylidene-2-hydroxy-l-naphthylamine is shown in Fig. 1. By careful examination of the curves half-neutralisation potentials (HNP) of the Schiff bases were determined with an accuracy of 2-3 mV and from these the corresponding pK,’ values were calculated.In determining half-neutralisation potentials the half-neutralisation potential of 0.001 M hexab-to1uidino)-cyclotriphosphazatriene [N3P3(NH-p-tol)6] base with per-chloric acid in nitrobenzene solvent was taken as the stan-dard.6 The half-neutralisation potential and corresponding pK,’ value of this compound are 213 mV and 3 respectively. In calculations of pK,’ values 59 mV was taken as correspond-ing to one pK,’ unit. The half-neutralisation potentials and corresponding pKa’ values of the Schiff bases are given in Table 1. In order to minimise homoconjugations within the titra-tions 0.001 M solutions ’of Schiff bases in nitrobenzene were carefully prepared.At this concentration very smooth titra-tion curves were observed. A 0.034 M solution of perchloric acid in nitrobenzene was prepared to minimise errors that would result from dilution of the solutions. Close investigation of the pK,’ values showed that the basicities of Schiff bases are influenced by several factors (1) the aldehyde component; (2) the amine component; (3) * To whom correspondence should be addressed. methylene groups; (4) one hydroxy group; ( 5 ) two hydroxy groups; and (6) a- and @-amino substitutions. All these are discussed in detail under Results and Discussion. Experimental Apparatus An Orion Model 801A digital pH meter equipped with a glass and a calomel electrode was used throughout. The saturated KCl solution of the calomel electrode was emptied and the electrode washed several times with anhydrous methanol.After drying it was filled with saturated KCl solution in methanol and a pressure of 2 cmHg was applied to the solution in the calomel electrode in order to prevent diffusion of the solution into the electrode. After each titration the electrode was first washed with anhydrous methanol to remove nitro-benzene solution from the surface then with anhydrous methanol. After each use the electrodes were first dipped in pure nitrobenzene to remove trace amounts of methanol. A magnetic stirrer was used in the titrations. Titrations were carried out in a 50-ml well-shaped beaker wound with copper wire. The copper wire and all other electrical components were properly earthed.For titrations a semimicro-burette reading to 0.01 ml was used. Chemicals Nitrobenzene. Nitrobenzene (Merck) was used after purifi-cation as follows. A 10-g amount of P205 was added to 1 1 of nitrobenzene. After thorough shaking the mixture was left overnight and distilled twice under reduced pressure with the aid of a suction oil pump equipped with a liquid air-cooled trap. Nitrobenzene purified in this way was straw yellow and its refractive index was 1.5513. Perchloric acid solution in nitrobenzene. Prepared from pure 70% perchloric acid (Merck). In all titrations anhydrous 0.034 M perchloric acid solution prepared as follows was used. With a micro-pipette 0.72 ml of 70% perchloric acid was added dropwise to 5 ml of ice-cooled pure acetic anhydride (otherwise a vigorous reaction took place and a dark brown solution was obtained instead a light yellow one).The light yellow solution was left for 5-6 h at room temperature then a 1.00-ml volume was introduced into a 50.0-ml calibrated flask and diluted to volume with nitrobenzene. The concentration of this solution was determined against the primary standar 950 ANALYST AUGUST 1986 VOL. 111 ~ ~ ~~ ~~ Table 1. Schiff bases titrated with perchloric acid in nitrobenzene and their pK,' values and half-neutralisation potentials (HNP) Schiff base PK,' -2.60 HNP/mV 535 HNP/mV -2.75 545 -1.75 485 -1.90 -1.65 -0.95 -0.17 -0.20 500 -1.40 -1.00 0.40 0.50 474 111 9 HO CH= N OH 450 440 IV =NcH23 362 399 V acH=NcH- OH at=:cH24 HO VII 355 317 1.26 0.50 365 a;= NcH2-p 0 1.40 315 0.85 350 Vlll acH=Nm OH 505 -2.12 -2.06 -1.10 -2.50 505 IX aCH=" X -0.95 440 455 aCH=" -2.50 525 530 X ANALYST AUGUST 1986 VOL.111 951 I _ _ _ 2oo t CH=N \ XIV 800 I 1 .o rnequiv. acidslmequiv. base Fig. 1. Titration curve of salicylidene-2-hydroxy-1-naphthylamine (XIV) in nitrobenzene Table 2. Comparison of the pKa’ values of congener pairs of Schiff bases when the amine component is aromatic Pair PKa ’ Pair PKa’ xx . . . . -2.06 XII . . . . -2.75 IX . . . . . . -2.12 I . . . . . . -2.60 xv . . . . -0.95 XI11 . . . . -1.90 IV . . . . . . -1.00 I1 . . . . . . -1.75 XIV . . . . -1.65 XXII . . . . -2.50 I11 . . . . . . -1.40 XI . . . . . . -2.50 Table 3.Influence of the aromatic amine component on the basicity of Schiff bases Pair PKa ’ Pair PKa’ I . . . . . . -2.60 I11 . . . . . . - 1.40 I1 . . . . . . -1.75 IV . . . . . . -1.00 XII . . . . -2.75 XIV . . . . - 1.65 XI11 . . . . -1.90 X V . . . . . . -0.95 Table 4. Influence of the aldehyde component on the pKa’ values when amine component is aliphatic Pair PKa’ Pair PKa’ XIX . . . . 0.85 XVIII . . . . 0.50 VIII . . . . 1.40 VII . . . . . . 1.26 XVII . . . . -0.20 XVI . . . . -0.17 VI . . . . 0.50 V . . . . . . 0.40 diphenylguanidine and was found to be 0.034 M. This solution was stable for 3-4 months if it was kept in a dark flask in a refrigerator, Schiff bases. The Schiff bases titrated in this work were all synthesised and dried properly.14J5 Results and Discussion In general the factors that influence the basicity of com-pounds are (1) inductive effects (substituent alkyl and aryl effects) (2) steric effects (F-strains and B-strains) (3) solvent effects (dipole moment dielectric constant acidity and basicity) (4) hydrogen bonding and (5) resonance effects.In this work we observed that with regard to the basicity of Schiff bases in nitrobenzene inductive and resonance effects are strong and hydrogen bonding effects are weak. These are discussed below in detail. Table 5. Influence of a methylene group next to the imino group on the basicity of Schiff bases Pair PKa ’ Pair PK,’ I . . . . . . -2.60 XI1 . . . . . . -2.75 v . . . . . . 0.40 XVI . . . . -0.17 I11 . . . . . . -1.40 XIV .. . . -1.65 VII . . . . 1.26 XVIII . . . . 0.50 IV . . . . . . -1.00 xv . . . . -0.95 VIII . . . . 1.40 XIX 0.85 I1 . . . . . . -1.75 XI11 - 1.90 V I . . . . . . 0.50 XVII . . . . -0.20 . . . . . . . . 200 > E (yjp (*’ 0 0 x” W = N G I (B)m XXI 700 6ooi , 2.0 800 0 1 .o rnequiv. acidshnequiv. base Fig. 2. Influence of hydroxy groups on the basicities of congener Schiff bases The basicity of Schiff bases containing an aromatic amine ring without a hydroxy group (such as XI) is not influenced by changes in the aldehyde component. Thus whether the aldehyde component is benzaldehyde 1-naphthaldehyde, salicylaldehyde or 2-hydroxy-l-naphthaldehyde the basicity of the Schiff bases does not change. As an example six congener pairs (a congener pair means two Schiff bases that differ only on one side of the -CH=N-group) are given in Table 2.An explanation of this observation could be as follows. In the congener pairs the basicity of the imino group is directly under the influence of the amine ring but only indirectly under that of the aldehyde ring. Resonance between the imino group and the amine aromatic ring is much easier than that between the imino group and the aldehyde aromatic ring. Secondly the strength of the hydrogen bond between the hydroxy group of 2-hydroxy-1-naphthaldehyde and the imino group is almost identical with the strength of the hydrogen bond between the hydroxy group of salicylaldehyde and the imino group. The basicity of Schiff bases having an identical aldehyde component but different aromatic amine components is influenced by changes in the aromatic component.Some examples are given in Table 3. The Schiff bases containing 952 ANALYST AUGUST 1986 VOL. 111 Table 6. Influence of hydroxy groups on the basicity of Schiff bases Group of Group of bases PKa ’ bases PKa’ X I . . . . . . -2.50 XXII . . . . -2.50 I . . . . . . -2.60 XI1 . . . . . . -2.75 111 . . . . . . -1.40 XIV . . . . -1.65 x . . . . . . -0.95 XXI . . . . -1.10 1 1 . . . . . . -1.75 XI11 . . . . -1.90 IV . . . . . . -1.00 x v . . . . . . -0.95 Table 7. Comparison of basicities of Schiff bases containing one and two hydroxy groups Pair PKa’ Pair PK,’ I . . . . . . -2.60 XI1 . . . . . . -2.75 I11 . . . . . . -1.40 XIV . . . . -1.65 v .. . . . . 0.40 XVI . . . . -0.17 V I I . . . . . . 1.26 XVIII . . . . 0.50 VI . . . . . . 0.50 XVII . . . . 0.20 VIII . . . . 1.40 XIX . . . . 0.85 Table 8. Comparison of basicities of Schiff bases containing a- and &substituted naphthylamines Pair PKa’ Pair PK,’ x x . . . . . . -2.06 IX . . . . . . -2.12 X I I . . . . . . -2.75 I . . . . . . -2.60 0 r 0 1 .oo c P 0 / I (CIII) tn ;Lm -2.00 -3.00 -2.00 -1.00 0 1.00 pK,‘(Schiff bases having one OH group) Fig. 3. pKa’ values of Schiff bases with one hydroxy group versus the pKa’ values of their congener Schiff bases phenyl group are nearly 0.6 pK,’ units more basic than their congeners containing a naphthyl group. This result is in contrast with the previous observations. 16 The basicity of Schiff bases bearing an aliphatic amine component (such as VI) are influenced by changes in the aldehyde component.For instance 2-hydroxy-1-naphthylal-2-hydroxybenzylimine and salicylal-2-hydroxybenzylimine have pK,’ values of 0.85 and 1.40 respectively (Table 4). Salicylaldehyde-containing Schiff bases are more basic than the 2-hydroxy-1-naphthaldehyde-containing congener bases. The mean difference in basicity in the two groups of Schiff bases is roughly 0.6 pK,’ unit. This probably is due to the methylene group between the imino group and the aromatic ring which prevents the resonance between the latter two. As a result the imino group is under the influence of the aldehyde ring and resonance occurs between the aromatic ring of the aldehyde and the imino group.As the aldehydes are different the rates of resonance must also be different. When the imino group comes under the influence of either the aldehyde aromatic ring or the amine aromatic ring, phenyl-containing Schiff bases are nearly 0.6 pK,’ units more basic than their naphthyl-containing congener pairs. This does not mean that the resonance is of the same magnitude in both instances but simply indicates the difference in the relative pK,’ values. A Schiff base containing a phenyl group is more basic than its congener Schiff base containing a naphthyl group. Of two Schiff bases that which has a methylene group next to the imino group is approximately 2.3 pK,’ units more basic than its congener pair (Table 5). This means that the hydrogen bonding in these bases is weak otherwise the methylene group would not change the basicity of the Schiff bases so much.The Schiff bases studied are of three types with (a) no hydroxy group (b) one hydroxy group and (c) two hydroxy groups. The influence of hydroxy groups on basicity is shown in Fig. 2. Any Schiff base containing one hydroxy group is less basic than its congener containing no hydroxy group. On the other hand any Schiff base containing two hydroxy groups is more basic than its congener containing one hydroxy group (Table 6 ) . As can be seen from Table 6 any Schiff base containing two hydroxy groups is at least as basic as its congener containing no hydroxy group. This result indicates that although physically unfavourable, hydrogen bonding between hydroxy and imino groups jumps between the hydroxy and hydroxy groups and frees the imino group from hydrogen bonding.This shows that there must be a weak hydrogen bond between the hydroxy and imino groups. The first hydroxy group plays two important roles (1) it forms hydrogen bonds with the nitrogen atom of the imino group and blocks the lone pair of the nitrogen atom and decreases its basicity and (2) it furnishes electrons partly to the aromatic ring. The second effect is very obvious when a second hydroxy group is present. As stated above a second hydroxy group renders a Schiff base more basic than its congener base that has no hydroxy group. This means that two hydroxy groups free the lone pair that is entirely hydrogen bonded and in addition they furnish some electron clouds to the aromatic rings on both sides of the imino group.Moreover Schiff bases containing two hydroxy groups are roughly one pK,’ unit more basic than their congeners with one hydroxy group (Table 7). A Schiff base containing a P-substituted naphthylamine component is more basic than its a-substituted congener (Table 8). When the pK,’ values of Schiff bases having one hydroxy group are plotted against those for the congeners with two hydroxy groups a straight line is obtained (Fig. 3). By means of this straight line it is possible to find the pK,’ value of a Schiff base that is not soluble if the pK,’ value of its congener is known. We acknowledge the valuable assistance of Dr. Z . K i l i ~ and Dr. K. Giindiiz who helped in preparing the manuscript and in drawing the diagrams.References 1. 2. 3. Hall N. F. and Conant J. B. J. Am. Chem. SOC., 3047. Conant J. B. and Hall N. F. J. Am. Chem. SOC., 3062. Hall N. F. Chem. Rev. 1931 8 191. 927 49, 927 49, 4. Fritz J. S . Anal. Chem. 1950 22 578; 1952 24 674. 5. Fritz J. S. and Fulda M. D. Anal. Chem. 1953 25 1837. 6. Feakins D. Last W. A. and Shaw R. A. J . Chem. SOC., 1964,2845; 1964 4464. 7. Nabi S. N. and Shaw R. A. J. Chem. SOC. Dalton Trans., 1974 1618. 8. Shaw R. A. Pure Appl. Chem. 1980,52 1063. 9. Dhathathreyan K. S. Krishnamurthy S. S . Vasudeva Murthy A. R. Shaw R. A. and Woods M. J. Chem. Soc., Dalton Trans. 1982 1549 ANALYST AUGUST 1986 VOL. 111 953 10. Middlestone M. Nabi S. N. and Shaw R. A. J . Chem. SOC., Dalton Trans. 1975 2634. 11. Feakins D. Last W. A. and Shaw R. A. J. Chem. SOC., 1964 2804. 12. Feakins D. Nabi S. N. Shaw R. A. and Watson P . , J . Chem. SOC. A 1969,2468. 13. Gritchfield F. E. and Johnson J. B. Anal. Chem. 1957,29, 957. 14. Dikmen C. and Gunduz T. Chem. Ber. 1956 2637. 15. Gundiiz T. and Onba~ioglu I . paper presented at TUBfTAK VIIth Scientific Congress KuSadasl Aydin, Turkey 1980. Cram D. J. and Hammond G. S . “Organic Chemistry,” Second Edition McGraw-Hill New York 1964 p. 219. Paper A51442 Received December 3rd 1985 Accepted January 20th 1986 16

ISSN:0003-2654    

DOI:10.1039/AN9861100949

出版商:RSC    

年代:1986

数据来源: RSC