摘要:

Russian Chemical Reviews 68 (12) 983 ± 1000 (1999) Micellar liquid chromatography EMBasova, VMIvanov, O A Shpigun Contents I. Introduction II. Retention models for micellar liquid chromatography III. Efficiency of micellar liquid chromatography IV. Selectivity of micellar liquid chromatography V. Retention and structure of sorbate molecules VI. Analytical application of micellar liquid chromatography VII. Comparison of the possibilities of ion-pair and micellar liquid chromatography Abstract. Background and possibilities of practical applications of micellar liquid chromatography (MLC) are considered. Various retention models in MLC, the effects of the nature and concen- tration of surfactants and organic modifiers, pH, temperature and ionic strength on the MLC efficiency and selectivity are discussed.The advantages and limitations of MLC are demonstrated. The performance of MLC is critically evaluated in relationship to the reversed-phase HPLC and ion-pair chromatography. The poten- tial of application of MLC for the analysis of pharmaceuticals including that in biological fluids and separation of inorganic anions, transition metal cations, metal chelates and heteropoly compounds is described. The bibliography includes 146 references. I. Introduction it depends on a large number of parameters: the nature of surfactants (viz., their types and charges) and concentrations, additives of various organic modifiers, ionic strength (salt addi- tives) and pH of the mobile phase and the nature of the stationary phase.Fourth, biological fluids can be injected directly into a chromatographic column without preliminary precipitation of proteins, because both anionic and non-ionic surfactants are capable of solubilising proteins. This simplifies preparation of samples and accelerates analysis of various medicinal prepara- tions. Fifth, the sensitivity of determination is increased since the intensity of luminescence, low-temperature phosphorescence and absorption are enhanced for many substances in the presence of surfactants. Sixth, the use of micellar gradients makes it possible to carry out fast separations because there is no need for column conditioning upon replacement of the mobile phase as the concentration of the free surfactant in solution is virtually constant after equilibration of the column with the surfactant at concentrations above CMC.Seventh, many authors have demon- strated the existence of a correlation between the retention and partition coefficients in the octanol ± water system and biological activity. In this context, MLC may be useful for estimating hydrophobicity of sorbates and predicting biological activity. A substantial drawback of this method is lesser efficiency of separation compared to that in classical RP HPLC. Two ways have been proposed to overcome this drawback: to use the so- called hybrid eluents, i.e., mobile phases with addition of small amounts of organic modifiers (usually, short-chain alcohols), or to raise the working temperature. The addition of alcohols to the micellar mobile phases leads not only to enhancement of the efficiency of separation, but also to changes in selectivity and to a faster analysis.Therefore, special attention has been paid in recent years to the MLC techniques using hybrid eluents. Studies of secondary chemical equilibria (acid ± base, complex- ation, formation of ion pairs and solubilisation) in reversed-phase high-performance liquid chromatography (RP HPLC) have been gaining momentum in recent years since their consideration leads to a higher selectivity of separation. The application of surfactant solutions at concentrations above the critical micellar concentra- tion (CMC) as mobile phases in RP HPLC has given birth to micellar liquid chromatography (MLC), which is an alternative to the classical RP HPLC with aqueous-organic eluents.1±4 Micellar liquid chromatography offers a number of advan- tages compared to other chromatographic methods.First, owing to the dual nature of surfactants possessing both hydrophobic and hydrophilic properties, it is possible to carry out simultaneous separation of a mixture of ionic and non-ionic sorbates. Second, owing to the ability of micelles to increase solubility of uncharged sorbates in water, they can act as organic modifiers; it is note- worthy that micellar eluents are inexpensive and non-toxic. Third, it is possible to influence effectively the selectivity ofMLCbecause EMBasova, VMIvanov, O A Shpigun Department of Chemistry, Moscow State University, Leninskie Gory, Moscow 119899.Fax (7-095) 932 88 46. Tel. (095) 939 22 77 (E MBasova, VMIvanov). Tel. (095) 939 13 82 (O A Shpigun) Despite the fact that this method has been developing for only two decades (the first paper appeared in 1980 5), several reviews 2, 6 ± 9 devoted to the discussion of basic theoretical prob- lems of MLC and to the most recent achievements 7 have been published. However, in the Russian-language literature none of the numerous monographs or reviews devoted to HPLC make any mention of this method (except for Ref. 1 where only the notion of MLC is given). Received 17 May 1999 Uspekhi Khimii 68 (12) 1083 ± 1101 (1999); translated by V D Gorokhov The aim of this review is to fill this gap.The review generalises the retention models in MLC described in the relevant literature, discusses causes of low efficiency of separation and ways of its increase and the possibilities of modifying the selectivity of separation and gives examples of the use of MLC for the analysis #1999 Russian Academy of Sciences and Turpion Ltd UDC 543.544 983 984 987 990 994 996 998984 and determination of inorganic, medicinal and biologically active compounds. II. Retention models for micellar liquid chromatography Micellar liquid chromatography is a variant of partition chroma- tography where the sorbate molecules are partitioned between three phases: stationary phase, micellar and non-micellar (aque- ous) mobile phases.In MLC, three equilibria are taken into consideration: sorbate partition between the micelle and the aqueous phase characterised by the partition coefficient PMW, sorbate partition between the micelle and the stationary phase with the partition coefficient PSM and sorbate partition between the stationary phase and the mobile aqueous phase with the partition coefficient PSW. Several equations have been proposed to relate the retention of sorbates in MLC to the concentration of micelles.10 ¡¾ 12 Armstrong and Nome 10 express this dependence as , (1) a VOPMW ¢§ 1U aMa a 1 PSW PSW Vs Ve ¢§ V0 where Vs, Ve and V0 are the volume of stationary phase, the volume of sorbate elution and the void volume of the column, respectively; V is the molar volume of the surfactant, [M] is the concentration of the surfactant in form of micelles in solution, which is calculated as [M]=c7ccr, where c is the total surfactant concentration in solution and ccr is the critical micelle concen- tration.The dependence Vs/(Ve 7 V0) on [M] should be linear, whereas the value of V(PMW 7 1) is equal to the constant of sorbate association with the micelles 9 and may be obtained experimentally from the slope of the straight line and the magnitude of the intercept of this line on the ordinate. Arunyanart and Cline Love 11 have proposed the equation relating the basic chromatographic parameter �¢ capacity factor (k0)�¢to the concentration of surfactant micelles. (2) , saK1 aMa a 1 faLsaK1 1 k0 a faLK2 where K2 is the constant of sorbate association with the micelle, f is the phase ratio, [Ls] is the ligand concentration in the stationary phase, and K1 is the binding (adsorption) constant of sorbate from the mobile aqueous phase to the stationary phase.In this case, the dependence of 1/k0 on [M] should be linear, while the constants of sorbate ¡¾ micelle association can also be easily derived from experimental data. Equations (1) and (2) are therefore equivalent.13 1 (3) k0 S , SM a k0 a Proceeding from the general model of secondary equilibria in chromatography,14, 15 Foley 12 has deduced the equation K 1 a KSMaMaMa 1 a KSMaMa k0 where KSM is the constant of sorbate ¡¾ micelle association, k0S and k0SM are the capacity factors of the sorbate and the sorbate ¡¾ micelle complex.Transformation of Eqn (2) with consideration of the fact that the sorbate ¡¾ micelle complex is not retained, since the micelle is a component of the mobile phase, gives the equations (4) k0S k0 a 1 1 a KSMaMa or (5) k10 a KkSM0S aMa a k10S . Equation (5) is equivalent to Eqns (1) and (2). EMBasova, VMIvanov, O A Shpigun It should be noted that the constant of sorbate ¡¾ micelle association derived from any equation �¢ (1), (2) or (5) �¢ is called the constant of association per monomer. As the concen- tration of micelles is expressed by the equation [M]=(c7ccr)/n, (6) where n is the aggregation number, the multiplication of the constant K2/KSM or V(PMW 7 1) by n gives the constant of association per micelle.In the studies of the retention mechanism of hydrophobic, poorly water-soluble alkylbenzenes in MLC with non-ionogenic surfactants polyoxyethylene [10]- or -[23]-lauryl ethers (Brij-22 or Brij-35, respectively), Borgerding 16 has modernised Eqn (1) and proposed a model of limited solubility. Assuming that PMW71&PMW and denoting the coefficient of sorbate parti- tion between the stationary and mobile phase as PSM= PSW/PMW, the equation , (7) k0 a 1f 1 a PSM 1 fPSW VaMa was obtained, which in the extreme case (insolubility in water) is transformed into the equation . (8) SM k0 a P f VaMa Analysis of Eqns (1) and (2) shows that as the concentration of micelles in the mobile phase increases the retention decreases (1/k0 increases), i.e., the eluting strength increases with the concentration of micelles.Equations (1) and (2) were both used for the determination of the constants of sorbate association with different surfactants in purely micellar as well as in hybrid eluents.10, 11, 17 ¡¾ 20 In both cases, straight lines with good correlation coefficients were obtained;18 the values of association constants calculated by the two methods are also in good agreement. It was noted that in the presence of the cationic surfactant cetyltrimethylammonium bromide the constants of association of a number of benzene and naphthalene derivatives with micelles are larger than those in the presence of the anionic surfactant sodium dodecyl sulfate,18 which can be explained by the preferred electrostatic interactions between the positively charged head groups of cetyltrimethylammonium bromide micelles and delo- calised electrons of the aromatic ring of sorbates.However, still higher values of the constant of sorbate ¡¾ micelle association were obtained in the case of the non-ionogenic surfactant Brij-35, which is difficult to explain if one considers solely the hydrophobic interactions. As it follows from the basic equilibria taken into consideration in the deduction of Eqns (1) and (2), the values of PMW and K2 depend only on the nature of the sorbate and the micellar system used and are independent of the nature of the stationary phase.17 It should be noted that the error in the determination of dissociation constants also depends on the nature of the sorbate and is increased with its higher hydrophobicity [the PSW values are large and the 1/PSW value in Eqn (1) is small].The addition of organic modifiers to a micellar system leads to changes in its characteristics (CMC and the aggregation num- ber),21, 22 which should in turn result in the modification of sorbate ¡¾ micelle interactions and, consequently, in changes in the retention.21, 22 The sorbate ¡¾ micelle interactions are as a rule decreased in a medium modified by alcohols.9 Indeed, the constants of sorbate ¡¾ micelle association calculated for a number of benzene and naphthalene derivatives are larger for sodium dodecyl sulfate and cetyltrimethylammonium bromide in purely aqueous micellar solutions than in solutions containing 5 vol.% or 10 vol.% of n-butanol as the modifier, which is related, according to Garcia et al.,20 to the competition between the sorbate and modifier molecules for the interaction with theMicellar liquid chromatography micelles.Addition of salts (e.g., NaCl) to the system enhances the interaction between the sorbate and the sodium dodecyl sulfate micelle; in this case, the values of association constants are even higher than those in the purely micellar mobile phases.20 Thus, the ability of micelles for selective solubilisation of the sorbate or interaction with it is the basis of separation in MLC using grafted alkyl stationary phases.It is known that the molecules of monomeric surfactants are easily adsorbed on such stationary phases thereby modifying their properties and making a definite contribution to the retention and selectivity in MLC. However, the equilibrium adsorption of surfactants on the sta- tionary phase was not taken into account in the deduction of Eqns (1), (2) and (5). Hu et al.23, 24 have given an explanation of the mechanism of chiral recognition in the separation of enantiomers of 2,20- dihydroxy-1,10-dinaphthyl- and -1,10-dinaphthyl-2,20-diyl hydro- phosphates by MLC with the use of different salts of bile acids (e.g., sodium cholate) in the mobile phases. A three-phase system proved to be ineffective for this purpose.A new model of the heterogeneous stationary phase has been proposed: the ODS (octadecyl sulfate) stationary phase adsorbs micelles, and it is into these micelles that the analytes are distributed. Based on the existence of the following equilibria in the system K1 ALs , Am+Ls K2 AMm , Am+Mm K3 (9) ALs+Mm , AMm+Ls K4 AMs , Am+Ms K5 AMs+Ls , ALs+Ms K6 AMs+Mm , AMm+Ms where A, M and L are the sorbate, the micelle and the solvated stationary phase, respectively, whereas the subscripts `m' and `s' refer to the mobile and stationary phases, respectively, the following equation was deduced k0 a K 1 3K4VLaLsa a K4VsMaMsaa Vm (10) VM aK m aMma 3VLaLsa a VsMaMsa , where VL, VMs , Vm and VMm are the volumes of solvated stationary phase, micelles adsorbed on the stationary phase, micelles of solvent and micelles in the mobile phase, repectively; [M]=c7ccr.The retention mechanism depends on the concentration of a co-solvent (acetonitrile) in the mobile phase: at low concentra- tions of acetonitrile, its partition between the micelles in the mobile and stationary phases is dominant. At higher acetonitrile concentrations, the mechanism of chiral recognition is determined by the partition between the bulk of mobile phase solution and the micelles adsorbed on the stationary phase. However, consider- ation of the equilibrium of micelle adsorption on the stationary phase did not result in the modification of the general relationship between the retention and the concentration of micellised surfac- tant: as in the case of Eqns (1) and (2), the dependence of 1/k0 on [M] is linear.In the deduction of equations, neutral organic compounds, mainly aromatic hydrocarbons, were used as model sorbates. The distribution of a sorbate in a micelle is determined by the nature of the micelle and the balance of hydrophilic and hydrophobic properties of the sorbate. The micelle has a lamellar structure. In 985 aqueous solution, it consists of a hydrocarbon nucleus, a water ¡¾ hydrocarbon layer the thickness of which is usually 2 or 3 CH2 groups and corresponds to the depth of penetration of water molecules inside the hydrocarbon part of the micelle or, con- versely, of the hydrocarbon chains inside the aqueous phase (here, the hydrophilic ¡¾ hydrophobic balance is established), a layer of hydrated polar groups and a heterogeneous layer of solution (the diffusive part of the double electrical layer in the case of ionic micelles) adjoining the micelle.22 Since the micelle is regarded as the macroscopically homogeneous medium, it cannot be indicated unequivocally where the sorbate is localised.As the existence of a correlation between the capacity factors and partition coefficients in the octanol ¡¾ water system 8, 9 been demonstrated for many neutral organic compounds, it is believed that the basic interac- tions are hydrophobic; in this case, the sorbate is either in the hydrophobic hydrocarbon nucleus or in the layer nearest to it.In the course of partition of charged organic sorbates, in addition to hydrophobic interactions, electrostatic interactions should also be observed, including those with the stationary phase surface modified due to the adsorption of monomeric surfactants. The retention of monoprotonic,25 ¡¾ 28 diprotonic and zwitter- ionic (amino acids and peptides) 26 ¡¾ 28 compounds in MLC have been systematically studied. The retention of monoprotonic compounds is described 26 by the equation (11) k0 a 1 a KHA fLsOKHA s a KAs ¢§ Ka=aHaaU m aMa a O1 a Km A¢§ aMaKa=aHaaU , s where KHA, KAs , KHA m and KAm are the binding constants of a weak acid HA and the conjugate base A with the stationary phase ligands Ls and the micelle M, Ka is the constant ofHAdissociation in aqueous solution, [M] is the concentration of micelles calculated as the stoichiometric concentration of surfactant minus CMC.Equation (11) allows prediction of the retention at any concen- tration of micelles and protons. For the diprotonic sorbates in the mobile phase, the following equilibrium is established: Ka 2 Ka1 2H++7AB . HABH+ H++7ABH+ k0 a k0cO1 a KmcaMaU a k0zO1 a KmzaMaUKa 1 a Each particle interacts with a micelle with the equilibrium constants Kmc, Kmz and Kma, respectively, and with the stationary phase with the equilibrium constantsKsc,Ksz andKsa, respectively. The following equation was proposed for calculating the capacity factor" aHaa (12) #, ak0aO1 a KmaaMaUKa1Ka 2 aHaa2 , # ," aHaa 1 a KmcaMa a O1 a KmzaMaUKa 1 a O1 a KmaaMaUKa 1Ka 2 aHaa2 where k0c, k0z and k0a are the capacity factors of the ions HABH+, 7ABH+ and 7AB, respectively; Kma is the apparent dissociation constant in a micellar solution.Equations (11) and (12) are useful for studying the dependence of selectivity on pH of the mobile phase. They were deduced assuming the occurrence of only hydrophobic interactions of various particles (charged, uncharged, zwitterionic) with the micelle and the mobile phase with the consideration of secondary acid ¡¾ base equilibria in the mobile phase.26 Sorbate ¡¾ micelle interactions in micellar chroma- tography of zwitterions were studied by Zibas.27 Carmelo 28 used the example of cephalosporin to study the retention mechanism in the micellar liquid chromatography of zwitterions.986 Since the hydrophobicity of inorganic anions is usually low and the main interactions in such systems are electrostatic, an ion- exchange model 29 was proposed for explaining the mechanism of separation of transition metal cations in MLC.In this model, ionic micelles (and the stationary phase coated with sodium dodecyl sulfate) are regarded as ion exchangers; retention is determined by the concentrations of micelles and counter-ions. For the ion exchange equilibrium Kie 2Naas +M2ma 2Naam+M2s a (Na+ ions are counter-ions in the sodium dodecyl sulfate micelle, whereas the subscripts `m' and `s' refer to the phases of micelles and solution, respectively; Kie is constant), the following equation was obtained (13) logKapp MW=log ([Naama2Kie)72 log [Naas a , where Kapp MW is the apparent coefficient of sorbate partition between the aqueous and micellar phases. The value of KMW was calculated using Eqn (11) proposed earlier by Okada 30, 31 for the retention in the exclusion MLC and simplified assuming that the inner volume of the ODS stationary phase is negligibly small (14) V a OKMW ¢§ 1Uncm a 1 , KSWVs 1 r ¢§ V0 MW where Vr, V0 and Vs are the volume of sorbate retention (elution), the void volume of the column and the stationary phase volume, respectively; n is the partial molar volume of micelle; cm is the concentration of micelles equal, as in all the cases discussed above, to the surfactant concentration minus CMC; KMW and KSW are the coefficients of sorbate partition from the aqueous phase into the micelle and the stationary phase, respectively.Equation (14) is analogous to Eqn (1). In order to minimise partition of sorbates into the stationary phase, it is recommended to supplement the eluent with salts and also with tartaric acid, which is a complexation agent.32 If the retention fits the ion-exchange model, the dependence of log Kapp on log [Nas+] should be linear, and this was confirmed for all metals, the slope of the straight line being equal to 2. If complexation of metal ions with tartrate ions in the mobile phase (the complexes ML with the stability constant bL and MHL+ with the stability constant bHL) is taken into consider- ation, Eqn (13) is transformed into (15) (K2 is the constant of tartaric acid dissociation).1 Kapp MW HLaHL¢§a a bLaL2¢§a ! K s a2 1 a b a aNaa ieaNam aa2 (15) 1 a ObHLaHaa=K2 a bLUaL2¢§a a aNasaa2 K a # " ieaNam aa2 In this case too, a linear dependence of logKapp MW on log [Nas+] was obtained for all metal ions, which implies that complexation does not influence the partition of sorbate into the micellar phase. This is understandable since in the presence of micelles the ionisation constants of weak acids (pKa) increase,21 which leads to a lower concentration of the reactive ligand L27 in the system. Thus for tartaric acid at pH 4.35, 50% of L27 is present in the aqueous solution, whereas in the presence of sodium dodecyl sulfate its content is as low as 4%.29 Amechanism based on the stoichiometric ion-exchange model for anions in the system with a cationic surfactant was proposed by Okada and Shimizu.33 Based on the existence of equilibria of the analyte A and the counter-ion C Kies A¢§s +C¢§aq , A¢§aq +C¢§s Kiem A¢§m +C¢§aq , A¢§aq +C¢§m EMBasova, VMIvanov, O A Shpigun Kd Surf Cm Surf am +C¢§aq , (16) , aqa k0 a f where the subscripts `aq', `s' and `m' refer to the aqueous solution, the stationary phase and the micellar phase, respectively; Kd is the equilibrium constant of the counter-ion transition from the micelle to the aqueous phase; Kies and Kiem are the constants of ion-exchange equilibrium between the solution ¡¾ resin and solu- tion ¡¾ micelle systems, the following equations were deduced ! 1 a Kiemcm aC¢§ aC¢§ KiesaC¢§s a ncs(17) aqa a Kd ! OVr ¢§ V0UaCaq ¢§ a a K1ies 1 a Kiemcm , aC¢§aqa where ncs is the number of counter-ion moles in the stationary phase, which is equal to the number of moles of hexadecyltrime- thylammonium bromide adsorbed on the stationary phase; cm is the concentration of micelles; Vr and V0 are the volume of analyte retention and the void volume of the column, respectively. The concentration of the counter-ion in the aqueous phase [C¢§aq] is determined as the sum of concentrations of the added counter-ion C7 (due to dissociation of micelles) and of the counter-ions of monomeric surfactants (c 7 ccr).The retention of ions in ionic micellar chromatography can be described by the stoichiometric ion-exchange model, and it formally fits the pseudophase model where the concentration of the added counter-ion is sufficiently high; in this case, the concentration of the counter-ion formed upon dissociation of micelles may be neglected. The model of micellar exclusion chromatography 30, 31 is based on Eqn (1) and takes into account that the micelles of cetyltrime- thylammonium and dodecyltrimethylammonium chlorides 30 or sodium dodecyl sulfate can penetrate partly into the stationary phase (external solution). Micelles do not penetrate into the inner part of the stationary phase, but the latter contains surfactant monomers (internal solution) and ions are retained there owing to ionic interactions.In this case, the retention is described by the equation (18) V a OKMW ¢§ 1UVcm a 1 , V 1 r ¢§ Ve i a VsKSW where Vr, Ve, Vi and Vs are the volumes of sorbate retention, external solution, inner solution and stationary phase, respec- tively; KMW is the coefficient of sorbate partition (ratio of concentrations) between the micellar phase and the internal solution phase; KSW is the coefficient of sorbate partition (ratio of concentrations) between the stationary phase and the internal solution phase; V is the molar volume of surfactant; cm is the concentration of micelles in the eluent. The dependence of 1/(Vr7 Ve) on cm is linear for both anions (IO¢§3 , I7, Br7, NO¢§3 , NO¢§2 ) 30 and cations (Pb2+, Zn2+, Co2+).31 Thus, there are two basic models for describing MLC, viz., Eqns (1) and (2); they describe adequately the dependence of sorbate retention on the concentration of micellised surfactant and are based on three equilibria: sorbate partition between the micellar mobile phase and the aqueous mobile phase, between the micellar mobile phase and the stationary phase and between the stationary phase and the aqueous mobile phase.The major types of interactions are hydrophobic in the case of organic uncharged and charged sorbates and electrostatic for transition metal cations. The stationary phase is always modified due to the adsorption of monomeric surfactant molecules; this effect, how- ever, was taken into account only qualitatively.In many cases, irreversible surfactant adsorption on various commercial silica gels with grafted C18 groups was observed.34 ¡¾ 36 As shown in the example of cetyltrimethylammonium bromide, the amount of surfactant sorbed depends irreversibly on the sorbent character-Micellar liquid chromatography istics, the type of eluent and temperature.36 In the case of sodium dodecyl sulfate, 6% of this compound initially adsorbed on the C14 stationary phase is sorbed irreversibly.35 On the whole, irreversible adsorption is more probable for monomeric station- ary phases with a high density of grafted ligands and surfactants with a small polar head group (when both the surfactant and the grafted ligand have the alkyl chain length of more than 8 carbon atoms).35 In the early stages of MLC studies it was believed that the surfactant adsorption is arrested after reaching CMC.In 1989, Borgerding et al.,37 showed that more than 58% of the total amount of sodium dodecyl sulfate is adsorbed on Resolve C18 when its concentration in the mobile phase varies in the range of 2 to 70 CMC. For Brij-22 and Brij-35, more than 50% of the total amount of adsorbed surfactant is sorbed on Resolve C18 from solutions with the initial concentration >50 CMC. Thus, the adsorption of both non-ionogenic surfactant and anionic surfac- tant continues after attaining CMC. In their recent study, Hu and Haddad 38 tried to take into account the surfactant adsorption on the stationary phase surface.They also used a three-phase model. The stationary phase consisted of a surfactant adsorbed on an ODS sorbent. The sorbate localised in the extramicellar mobile phase Am can bind to the surfactant on the stationary phase Ss in the form of an ASs complex in conformity with the equation K1 ASs , Am+Ss and to the surfactant in the mobile phase Sm in the form of a complex ASm in accordance with the equation K2 ASsm . Am+Sm Direct transfer to the stationary phase of the sorbate bound to the surfactant in the mobile phase ASm can also take place by the equation K3 ASs+Sm. ASm+Ss The capacity factor can be expressed by the equation (19) K k0 à Vm á K2VSmâSmä , 1VSs âSsä where VSs , Vm and VSm are the volumes of surfactant adsorbed on the stationary phase bulk, of the solvent and of surfactant in the mobile phase, respectively.It follows from Eqn (19) that (20) k10 à KV1W00 s á K2KV1SmW âSsmä , where Ws=VSs [Ss] is the mass of surfactant adsorbed on the stationary phase; Vm=VSm 7V0. The dependence of 1/k0 on [Sm] (determined as the total surfactant concentration in the mobile phase) should be linear. The content of sodium dodecyl sulfate on the stationary phase was determined (16.5%) and found to be almost equal to the carbon : silicon ratio in the initial sorbent (17.1%). In other words, the octadecyl groups of the sorbent are covered to a large extent with the adsorbed surfactant, and the stationary phase may be regarded as the adsorbed surfactant.The dependence of 1/k0 on [Ss] is linear for the model sorbents studied, viz., benzyltriethylammonium chloride and phenylalanine. It should be noted that Eqn (20) is analogous to Eqn (2), the only difference consisting in that Eqn (2) includes the surfactant concentration in the form of micelles in solution (minus CMC), while Eqn (20) contains the overall surfactant concentration in solution. Consideration of some other equilibria (acid ± base equilibria in the mobile phase or adsorption of micelles on the stationary phase) does not modify the dependence of retention on the concentration of micelles. Electrostatic interactions (ion 987 exchange between metal cations and the cations of the diffusive part of the double electrical layer) are taken into consideration in the MLC model in which retention is dependent on both the concentration of micelles and the concentration of counter-ions.III. Efficiency of micellar liquid chromatography One of the major drawbacks of the MLC method is its low efficiency compared to that of RP HPLC. There are several explanations of this phenomenon in the relevant literature. One of these 39 suggests that the main factor of the band blurring is the resistance to the sorbate mass transfer because of the poor wetting of the stationary phase. Yarmchuk et al.40 presume that the resistance to the sorbate mass transfer exists due to both low rates of the sorbate efflux from the micelle and its slow desorption from the stationary phase. When constructing their model, these authors 40 took into account kinetics of sorbate adsorption onto, and desorption from, the stationary phase, on the one hand, and kinetics of the sorbate influx to, and efflux from, the micelle into the bulk of the non-micellar mobile phase, on the other.Having calculated the rates of all processes, they have shown 40 that the rate constants of the influx into the micelle are appreciably higher than the rate constant of the adsorption on the stationary phase (silica gel with the C1-grafted radicals Supelcosil LC 1) because of the differences in the distance to be covered by the sorbate. Thus 0.10 M sodium dodecyl sulfate solution contains 9.0361017 micelles cm73.Consequently, in order to bring the sorbate from the bulk of solution to the surface of the stationary phase, it is necessary to encounter about 180 micelles, whereas an appreciably shorter distance must be covered for a repeated collision with the micelle. In order to pass from the micelle to the stationary phase, the sorbate should have a definite orientation relative to the stationary phase. No definite orientation is required for the reverse transition; the sorbate can enter the micelle at any point. The rate constants of these four processes depend on the nature of the sorbates. Relatively small and well water-soluble molecules, having large rate constants of the influx into, and the efflux from, the micelle can move quickly in the bulk of solution, enter and leave micelles multiply before they reach the stationary phase surface.For such hydrophilic sorbates, the efficiency (mass transfer) is not limited by micellar equilibria. For the more hydrophobic sorbates the affinity to the micelle is higher and the rate constants of their efflux from the micelle are much lower (e.g., for pyrene the efflux rate constant is 1000 times as low as that for benzene). Con- sequently, the hydrophobic sorbates remain in the micelle for an appreciably longer time, decreasing the mass transfer between the micelle and the stationary phase. Since there is an antiparallel correlation between the rate constants of the sorbate efflux from the micelle and the constants of desorption from the stationary phase, the hydrophilic sorbates can return more often to the bulk of aqueous mobile phase and be thereby close to equilibrium.The consequence of the differences in the rates of these processes for the hydrophilic and hydrophobic sorbates is fairly good efficiency for the former and smeared peaks for the latter. In order to increase the efficiency and to improve the shape of peaks, it was proposed to add small amounts of organic modifiers to the mobile phase and to operate at higher temperatures.40 Yarmchuk et al.40 explain enhanced efficiency in the presence of 2 vol.% ± 5 vol.% of alcohols by the better mass transfer in the mobile phase owing to the increase in the rate of sorbate efflux from the micelle since the non-micellar part of the mobile phase becomes less polar.These authors relate the higher efficiency after an increase in temperature [for small sorbates of the benzene and phenol type the number of theoretical plates (N) is increased by 50%± 100% at 50 ± 60 8C] to both upgraded micellar kinetics and lower viscosity of the mobile phase. Addition of 3% of propanol to the mobile phase and an increase in temperature to 40 8C make it possible to attain efficiencies close to that of RP HPLC with aqueous ± organic eluents.39, 41 Enhanced efficiency is promoted988 by the decrease in the surfactant concentration.40 Indeed, Bor- gerding et al.42 have shown that the use of elevated surfactant concentrations can result in the total loss of efficiency.Another factor influencing the height equivalent to a theoretical plate (HETP) is the linear rate.40 For small molecules (benzene), the derived dependence was almost analogous to that of usual RP HPLC, whereas for large molecules (naphthalene, anthracene) the efficiency drop with the increasing rate was appreciably larger. Yarmchuk et al.40 recommend to operate at low flow rates. The low efficiency of MLC is related 36, 43 ± 45 to the surfactant adsorption on the stationary phase surface, which results in poor mass transfer in the stationary phase.43, 44 Direct relationship between the amount of surfactant in the stationary phase and the decrease in efficiency was observed.43 Based on the determination of diffusion coefficients for a series of sorbates (2-naphthol, p-nitroaniline, p-nitrophenol, naphthalenedisulfonic and naphthalenetrisulfonic acids) in the absence and the presence of sodium dodecyl sulfate and meth- anesulfonate in the mobile phase, Borgerding and Hinze 43 attempted to elucidate which of the mass transfer processes (in the mobile or the stationary phase) is causative of poor efficiency in MLC.All sorbates may be divided into two groups: the sorbates which interact actively with the micelles (2-naphthol, p-nitroani- line and p-nitrophenol) and those which do not interact and are repelled or excluded from the micelles (sodium 2,6-naphthalene- disulfonate, sodium 4,5-dihydroxynaphthalenedisulfonate and sodium 1,3,6-naphthalenetrisulfonate).The sorbates of the first group have noticeably lesser diffusion coefficients in the micellar mobile phases. Thus the diffusion coefficients of 2-naphthol and sodium 1,6-naphthalenedisulfonate are 6.061076 and 6.7161076 cm2 s71 in water, 5.6761076 and 4.7361076 cm2 s71 in 0.1 M MeSO3Na solution, and 0.9061076 and 4.061076 cm2 s71 in 0.1 M sodium dodecyl sulfate solution. If the band blurring determined the slow mass transfer in the micellar phase, one would observe different values of efficiency for the two groups of sorbates. Besides, no appreciable variation of N was noticed at different surfactant concentrations in the mobile phase. All this indicates that the most probable factor eliciting the deterioration of the efficiency is the resistance to mass transfer in the stationary phase. The MLC efficiency is strongly dependent on the flow rate.For example, in the case of p-nitroaniline the value of N for a 30064 mm column with the cyanopropyl stationary phase increases from 400 to 1300 upon the decrease in the mobile phase flow rate from 1.5 to 0.25 ml min71. However, this cannot be regarded as the proof in favour of only a decelerated process of mass transfer in the mobile phase because, according to the van Deemter equation, both members contribu- ting to HETP owing to the resistance to the mass transfer in the mobile and the stationary phases are proportional to the linear flow rate. Borgerding et al.42 believe that the enhancement of efficiency in MLC with the micelles of ionic surfactants occurs owing to changes in the mobile phase: the repulsion barrier decreases because of the lower density of the electrical charge on the surface of ionic micelles.Indeed, the addition of alkanes has no effect on the surface density of charge and does not lead to higher efficiency for strongly hydrophobic sorbates. Addition of alcohols does not result in higher efficiency ofMLCwith non-ionogenic surfactants, e.g., Brij-35. In order to decrease the resistance to the mass transfer and to improve the MLC efficiency, Berthod and Roussel 46 proposed to use hybrid mobile phases. The addition of short- and medium- chain alcohols induces desorption of surfactants from the sta- tionary phase, which leads to the improvement of mass transfer into the stationary phase (and from the stationary phase) and also ensures better wetting.37, 39, 41, 42 This effect is enhanced with increase in the modifier concentration and hydrophobic- ity.19, 37, 47 A study 41 carried out on a broad set of sorbates showed that the higher efficiency of the column is observed not for all sorbates; nonetheless, it was concluded that the cause of EMBasova, VMIvanov, O A Shpigun higher efficiency is the decrease in the amount of surfactant sorbed on the stationary phase.Another mechanism of the effect of alcohols on the efficiency was proposed by Armstrong et al.44 In their opinion, the use of hybrid eluents makes possible changes in the structure of the stationary phase itself from the block-like to the brush-like one, which can also improve the mass transfer into the stationary phase (and from the stationary phase).It is noted therewith that the use of hybrid eluents increases appreciably the efficiency in the case of hydrophobic rather than hydrophilic sorbates.35, 43 ± 45 As there is no consensus about the cause of poor efficiency in MLC, investigators come continuously back to this prob- lem.35, 45, 48 Thus investigation of the effect of the nature of the organic modifier (ethyl acetate, n-propanol) on the efficiency of a column packed with different sorbents with the graftedC1, C8, C18 or cyanopropyl groups in the elution by mobile phases with the surfactant concentrations above or below CMC made it possible to reveal a correlation between the column efficiency and the extent of wetting of the surface of grafted sorbents with an organic solvent.Bailey and Cassidy 45 studied the effect of sodium dodecyl sulfate concentration (0 ± 0.35 M) at the ionic strength of 0 ± 0.3 M (NaClO4) and of n-propanol (0 ± 0.8 M) on the column efficiency. In the absence of micelles in the mobile phase, the efficiency of the column varies insignificantly. This suggests that the decrease in the amount of sodium dodecyl sulfate adsorbed on the stationary phase is not the cause of the efficiency increase in the presence of alcohols, contrary to what was thought by some investiga- tors.35, 39, 41, 42 In the absence of micelles, at low alcohol concen- trations (not exceeding 0.13 M), the column efficiency for benzaldehyde and acetophenone reaches its maximum, and this may be attributed to modification of the structure of the sta- tionary phase surface.However, at high alcohol concentrations, which are commonly used for achieving better efficiency in MLC, the column efficiency remains approximately constant, which may be indicative of the absence of modification of the stationary phase surface in the presence of n-propanol. Studies of the factors that influence the mass transfer in the mobile phase in the range of sodium dodecyl concentrations from CMC to 0.3 M showed that at concentrations of 0.005 ± 0.012 M this dependence passes through a maximum, the position of which depends on the sorbate hydrophobicity.At higher concentrations the efficiency decreases, which is consistent with the data reported by Yarmchuk et al.40 In the presence of n-propanol, the maximum is broader and less clearly pronounced, whereas the column efficiency increases from 200 ± 2200 to 700 ± 3700 theoretical plates. These results led the authors to conclude that mass transfer in the stationary phase is not the main cause of low efficiency in MLC; this is apparently related to the processes of micellisation. In a micellar system, two oppositely directed processes take place continuously: desintegra- tion of micelles and their formation. At the same time, two basic processes occur in the micellar system: rapid exchange of the surfactant monomer with the micelle [the lifetime of the surfactant molecule in the micelle is of the order of 1077 s (see Ref.22)] and the process of formation and disintegration of micelles [the half- life of the micelle is 1073 ± 1 s (see Ref. 22)]. The latter can in turn be split into two processes: (1) a series of stepwise additions of single molecules to the surfactant aggregates with formation of micelles and (2) association of premicellar aggregates with formation of micelles. Either of these processes can influence the sorbate mass transfer between the eluent and the stationary phase. The addition of alcohols can influence these processes and enhance the kinetics of mass transfer in the mobile phase. Villanueva Camanas et al.49 studied the decrease in the efficiency due to the addition of alcohols to the mobile phase containing sodium dodecyl sulfate micelles in the elution of L-DOPA, 2-methyl-DOPA, norepinephrine, adrenalone, dopa- mine, isoproterenol on Spherisorb ODS-2 C18.Analogous regu- larities were revealed earlier in the elution of amines.50, 51 It is presumed that the mass transfer is brought about through theMicellar liquid chromatography following mechanisms: (1) direct transfer from the micelles to the stationary phase, which is opposed by the electrical repulsion between them and (2) transfer through a continuous aqueous pseudophase. In the presence of alcohol, the polarity of the continuous aqueous pseudophase is decreased. On the one hand, this increases the rate of transfer of hydrophobic sorbates through this pseudophase, while on the other hand, the solubility and the rate of transfer of highly hydrophilic catecholamines is decreased, which accounts for the decrease in the efficiency.Analysis of the band blurring in MLC was performed 35, 52 using the Knox equation (21) h=An1/3+B/n+Cn , where h=H/dp is the reduced height equivalent to a theoretical plate,H is the height equivalent to a theoretical plate, dp is the size of stationary phase particles, n=udp/Dm is the reduced linear flow rate, u is the linear flow rate of the mobile phase, Dm is the coefficient of sorbate diffusion in the mobile phase, A is the coefficient which characterises the quality of the column packing and the decelerated mass exchange in the mobile phase outside the particles, B is the coefficient accounting for the contribution to the band blurring from the longitudinal diffusion and C is the coefficient describing the decelerated mass exchange in the sta- tionary phase.The dependences of h on n were determined 33 on one column using first an aqueous ± organic mobile phase, then the micellar mobile phase and then again the same aqueous-organic phase. Two types of alkylated silica gels were used, viz., a monomeric C18 sorbent and a C14 sorbent with high density of grafted alkyl groups (3.5 mmol m72), and two types of surfactants, viz., sodium dodecyl sulfate (anionic surfactant) and Brij-35 (non-ionogenic surfactant). An increase in h was shown to occur mainly due to the higher values of coefficient A: for benzene with the aqueous methanol as a mobile phase, h=4.9 at n=10, whereas in the presence of sodium dodecyl sulfate at n=10.9 the contribution from the coefficient A is 2.36 (48% methanol) and 6.9 (63% methanol), respectively.In micellar eluents, the contribution from the coefficients B and C also increases resulting in a stronger band blurring and lower efficiency. The decrease in the diffusion coefficient of sorbates upon their inclusion into the micelles induces the increase in B. The coefficient of sorbate diffusion in the stationary phase depends on the nature of the sorbate and decreases as the hydrophobicity increases; highly hydrophobic sorbates are transferred from the micelle directly into the sta- tionary phase by-passing the extramicellar aqueous phase in which they are poorly soluble.The value of coefficient C increases in a micellar system (e.g., up to 700% for propylbenzene). The decrease in the mass transfer coefficients may also be associated with poor wetting.2, 39 The most important reason, however, is believed to be the surfactant adsorption on the stationary phase.16, 17, 43, 44, 50 In the first place, adsorption of a surfactant on the stationary phase surface [a C18 sorbent adsorbs 70 mg g71 of Brij-35, and a C14 sorbent adsorbs 140 mg g71 (or 2.3 mM m72) of sodium dodecyl sulfate] leads to a decrease in the pore volume, obstructs the smallest pores with diameter <7 nm and reduces appreciably the surface area of the stationary phase and its permeability.35 The consequence of this is the increase in the coefficient A values.Secondly, adsorption of a surfactant increases thickness of the organic layer of the stationary phase and decreases the mass transfer rate and the diffusion coefficients in the stationary phase and the coefficients of sorbate diffusion in the mobile phase.16, 42, 43, 46 This results in the increased values of the coefficients B and C. The surfactant adsorbed on the stationary phase can be eluted with a pure solvent; however, the adsorption may prove to be partially irreversible.32 ± 34 Following operation with the use of micellar phases, the efficiency of a column drops drastically as is evident from the values of the coefficients A, B and C in the Knox equation for benzene.This is also the consequence 989 of changes in the porosity and permeability of the stationary phase. Mobile phase A B C 3.1 0.22 0.2 70.7 1.1 3.2 5.7 204 Methanol ± water (30 : 70) 1.4% Sodium dodecyl sulfate in water Methanol ± water (30 : 70) after operation with sodium dodecyl sulfate Lavine and Hendayana 52 studied the effect of temperature in the range of 25 ± 45 8C on the dependence of h on n for acetophenone, nitrobenzene, benzene, methyl benzoate and tol- uene in MLC with an aqueous solution of sodium dodecyl sulfate. The values of the coefficients A, B and C for p-nitrophenol in 0.02 M sodium dodecyl sulfate solution were found to be 2.340.14, 7.430.69 and 0.0740.013 at 25 8C and 1.060.13, 11.40.65 and 0.0260.010 at 45 8C, respectively. The decrease in the coefficient B at 35 8C (5.160.66) is anomalous and inexpli- cable.52 Noticeable enhancement of chromatographic efficiency with the increase in temperature occurs owing to the decrease in values of both the coefficients A (flow anisotropy) and C (mass transfer in the stationary phase).The decrease in A values with the increase in temperature may be explained by the decrease in the constant of sorbate partition between the micelle and the bulk of solution [K2, Eqn (2)] and, consequently, by the decrease in the proportion of sorbate bound to the sodium dodecyl sulfate micelles in the mobile phase. A decrease in the eddy diffusion is eventually attained. The decrease in the values of the coefficient C with the increase in temperature may be explained by the decrease in the adsorption of a surfactant on the stationary phase surface and the higher rate of sorbate mass transfer between the mobile and stationary phases.In addition, the fluidities of the grafted alkyl chains and the adsorbed surfactant are increased at higher temperature, which also decreases the resistance to the mass transfer in the stationary phase. The values of the coefficient B increase considerably with temperature due to both the lower viscosity of bulk of the solvent and transition of part of the sorbate from the micelle to the solution. In the example of separation of 12 polycyclic aromatic hydro- carbons on Nova-Pak C18, the influence of the nature and the content of the alcohol and of the column temperature on the efficiency of separation in MLC with the use of sodium dodecyl sulfate was studied.53 The number of theoretical plates increases with the increase in the surfactant concentration in the mobile phase up to 0.13 M.Upon transition from methanol to 2- propanol, the peak becomes more symmetrical, and one observes an increase in the efficiency for less hydrophobic sorbates (fluorene and chrysene). For more hydrophobic dibenzo[ac]anthracene, more symmetrical peaks were obtained with n-butanol, although in this case the changes in the efficiency were insignificant. It should be noted that in contrast to conven- tional HPLC, in MLC the number of theoretical plates is commonly calculated by the following formula 54 (22) N à â41:7ÖtR=w0:1Ü2ä , ÖB=A á 1:25Ü where tR is the retention time, w0.1 is the peak width at its 10% height, B/A is the factor of peak asymmetry.An increase in temperature leads to insignificant enhancement of the efficiency and appreciable improvement of the shape of peaks. In this case, the effect of the flow rate on N at different temperatures depends on the sorbate hydrophobicity. Studies of the effect of alcohols as modifiers of the mobile phase showed that the addition of small amounts of propanol or pentanol led to higher efficiency because of considerable decrease in the values of the coefficients A and C in Eqn (21).52 An increase in the coefficient B upon addition of an alcohol seems to occur due990 to the lower viscosity of the mobile phase and a shift of sorbate equilibrium in the micelle ± solvent volume system.Consequently, the mechanisms of the influence of the increase in temperature and addition of alcohols on the separation efficiency in MLC are similar in many respects. Thus, Okada 32 as well as Lavine and Hendayana 52 related the deterioration of MLC efficiency to an appreciable increase in the coefficients A and C in the Knox equation. Flow anisotropy, resistance to mass transfer in the stationary phase and molecular diffusion contribute to the band blurring. To enhance the efficiency in MLC, it is recommended to use hybrid eluents.IV. Selectivity of micellar liquid chromatography The selectivity of separation in MLC is influenced by the type and concentration of the surfactant, the nature and concentration of the organic modifier, pH, ionic strength, temperature and the nature of the stationary phase. Of these factors, the effects of surfactant and modifier concentrations have been studied most extensively. In MLC, the eluting strength of purely micellar eluents is small, it increases with increase in the micelle concen- tration in the mobile phase.47 1. Effect of surfactant concentration Foley 12 deduced Eqn (23) for calculating the optimum concen- tration of micelles [rather than of the micellised surfactant as in Eqns (1) and (2)] (23) pMopt&log KSM721 log k0s , k01 where pMopt=7log [M], KSM is the constant of sorbate association with the micelle calculated per micelle rather than per monomer, k0s is the factor of sorbate capacity (the extreme value), k01 is the sorbate capacity factor at a surfactant concen- tration of 1 mol litre71. The optimum pM values for a broad spectrum of sorbates and three surfactants (sodium dodecyl sulfate, cetyltrimethylammo- nium bromide and Brij-35) were calculated using Eqn (23).The concentrations of surfactants calculated from Eqn (23) fall into a wide range and are the optima as regards selectivity rather than retention. For dihydropyridines on a C18 stationary phase, an increase in the surfactant concentration induces a trend towards changes in the retention mechanism from three partition equilibria to direct transfer from the micelles to the stationary phase.55, 56 Madamba-Tan et al.57 used isocratic data to deduce the equation for predicting the retention times in the gradient regime of elution with a gradient of micelle concentration tR=(t0/b){71/k 00 + [(1/k 00)2+2b(17f)]1/2} +t0+tD (24) where t0 is the dead time of the column; b is the parameter of the gradient steepness; k00 is the capacity factor in the initial mobile phase; tD is the delay time, i.e., the time interval before the gradient does reach the upper part of the column; f is the part of the column through which the sorbate band has already passed and which is calculated from the equation (25) f à tD 0=tR à âtDÖ1 á 1=k00Üä .tR 2. Effect of organic modifier concentration The addition of an organic solvent to micellar eluents leads to simultaneous increase in the eluting strength and selectivity for some ionic and non-ionic sorbates.47, 58, 59 The increase in selec- tivity is related to the existence of competitive partition in MLC and to the unique capacity of micelles for solubilising separately molecules of sorbates and organic solvents. Subsequent studies of the role of organic modifiers and micelles in the control of the eluting strength and selectivity in MLC in the example of small peptides, phenols, substituted benzenes and benzoic acids in a sodium dodecyl sulfate ± water ± EMBasova, VMIvanov, O A Shpigun 2-propanol system confirmed simultaneous increase in the eluting strength and selectivity;60 in this case, the selectivity varies systematically and monotonically as the concentration of 2-propanol in the mobile phase increases.The concentration of micelles has the opposite effect on selectivity compared to that of 2-propanol. For those ion pairs the selectivity of which decreases as the concentration of 2-propanol increases, one observes enhancement of selectivity with the increase in concentration of micelles and vice versa. Thus, although the eluting strength of the mobile phase is increased with the increase in the concentration of micelles and organic solvent, the effects of these parameters may be quite different and even opposite. Micelles and 2-propanol compete for the interaction with sorbates, which influences the selectivity and retention.The dependence of retention on the volume proportion (j) of organic modifier is described by the equation 47 (26) ln k 0=7Shybj + ln k 00, where Shyb is the parameter of solvent strength, ln k00 is the apparent (extrapolated) coefficient of sorbate capacity in a purely micellar mobile phase. Equation (26) is analogous to Eqn (1) used for describing changes in the retention with the variation of the volume proportion of the organic modifier in RP HPLC. Combination of Eqns (26), (1) and (2) gives relations (27) and (28). (27) ln(PSWF)= 7Ss+ln(P0SWF) , (28) ln(KMW[M] + 1)=7Sm+ln(K0MW[M]+1), where Ss and Sm are the parameters which characterise the sensitivity to changes in the sorbate partition from the bulk of solution into the stationary phase and the micelles with variation of j; K0MW is the constant of sorbate binding to a micelle, P0SW is the coefficient of sorbate partition between the mobile and stationary phases for aqueous micellar eluents, i.e., in the absence of an organic modifier.The parameter Shyb depends on Ss and Sm (29) Shyb=Ss7Sm. Several conclusions can be drawn from Eqns (27) ± (29). First, the magnitude of the solvent strength parameter in hybrid eluents is smaller than that in aqueous-organic eluents. Second, Sm and, consequently, Shyb depend on the micelle concentration. For two sorbates (a and b) and at two concentrations of organic solvent (j1 and j2) in the mobile phase, the following equation is valid (30) ln a27ln a1=7(S bhyb7S ahyb) (j27j1), while combination of Eqns (26), (29) and (30) gives Eqn (31).(31) ln a=(dSm7dSs)j + lna0, where a=k0b=k0a is the selectivity coefficient (k0b > k0a), a0 is the selectivity coefficient in a purely aqueous micellar eluent, dSm and dSs are the differences in Sm and Ss for compounds a and b.It follows from Eqn (31) that the selectivity (ln a) depends linearly on the volume proportion of the organic modifier in the mobile phase and decreases or increases monotonically depending on the dSm/dSs ratio for two sorbates. Thus for the majority of substituted benzenes, amino acids and peptides, the reverse dependence between retention and Shyb is observed: the more strongly compounds are retained, the smaller is the Shyb value.In this case, for most aromatic compounds, amino acids and proteins a direct dependence between retention and Sm and a reverse dependence between retention and Ss are observed. The influence of the sorbate structure on the value of Shyb is reflected mainly on changes in Ss. The influence of the concentrations of an organic solvent and micelles on different equilibria occurring in the system can clearly be seen if Eqn (1) or (2) is transformed into (32).Micellar liquid chromatography SW (32) KaMWâMä á 1 Pb Pa a à à aSW , aMW SW!, ! KbMWâMä á 1 where aMW is the parameter characterising the selectivity of binding of sorbates to the micelles, aSW is the parameter character- ising the selectivity of partition into the stationary phase.The values of PSW and KMW decrease as the concentration of 2-propanol increases; however, the extent of decrease in these parameters is different for different sorbates, which leads to changes in selectivity. The concentration of micelles, in turn, influences the selectivity at a fixed concentration of an organic modifier mostly through the parameter aMW: as the concentration of micelles increases the parameter aMW decreases faster than the parameter aSW, which is independent of [M]. The selectivity is increased because of this effect. 3. Effect of the nature of the organic modifier The nature of the organic modifier of the mobile phase influences the selectivity of separation in MLC; however, this effect is not the same as that in RP HPLC.Usually lower alcohols from methanol to pentanol 9, 35, 39, 41, 42, 47, 55, 57, 58, 60, 61 ± 70 and also a pentanol ± heptanol mixture (3 : 1) 71 are used as modifiers in MLC. Diols,61 dipolar aprotic solvents,61 acetonitrile and tetrahydrofuran 67 were also used for this purpose. The principles of the choice of solvent selectivity in RP HPLC are based on the Snider triangle of selectivity, but this approach proved to be inapplicable to MLC.67 The selectivities of separation attainable with 2-propanol, aceto- nitrile and tetrahydrofuran belonging to different types of solvents are close,67 whereas those with 2-propanol and n-butanol belonging to the same type exhibited strong differ- ences.9, 65, 67 For example, in a model mixture of 15 polyaromatic hydrocarbons (PAH) in the absence of the modifier (0.15 M sodium dodecyl sulfate) in the mobile phase, many compounds are co-eluted, particularly the most hydrophobic ones.65 In the presence of 15 vol.% methanol, the peaks of all compounds are well resolved. The quality of separation with 2-propanol instead of methanol is similar; in this case, the time of analysis is reduced from 45 to 25 min.However, the addition of only 7 vol.% of n-butanol degrades separation. On the one hand, the addition of alcohols to the micellar mobile phase modifies appreciably the partition of sorbates between the micelles and the bulk of aqueous phase which becomes less polar.64 On the other hand, alcohols decrease the amount of surfactant sorbed on the stationary phase and solvate the grafted hydrocarbon phase.Considering that n-butanol is endowed with a higher solvation capacity than methanol or 2-propanol, this should exert stronger effect on the partition of sorbates over three phases. This effect is particularly noticeable for more hydrophobic compounds. The parameter Shyb reflects the extent of solvation of sorbates by organic solvents, localisation of the molecules of sorbates and organic solvents in the micelles 9 and hence, should depend on the nature of both the sorbate and the surfactant. Khaledi et al.47 have found that for 14 aromatic compounds belonging to different classes the values of Shyb in the cetyltrimethylammonium bromide micellar phase, modified by methanol, 2-propanol and n-butanol, vary in the order Sn-butanol > S2-propanol > Smethanol, which coincides with the increase in the eluting strength of these modifiers in RP HPLC.An analogous sequence was obtained by Rodriguez Delgado et al.65 for 15 PAH in the mobile micellar phase of sodium dodecyl sulfate modified with the same alco- hols.65 The higher values of the parameter Shyb for 2-butanol indicate that it interacts with the micelles more actively, solvates the sorbates more effectively and can compete better with the micelles for the interaction with sorbates than does methanol. Variation of Shyb as a function of surfactant concentration is also different for different modifiers: as the concentration of sodium dodecyl sulfate in the mobile phase increases from 0.06 to 0.14 M, the DSm varies by *0.2 for methanol, *0.4 for 2-propanol and*2.0 for n-butanol.65 991 Less hydrophobic sorbates interact strongly with the micelles (e.g., anthracene with cetyltrimethylammonium bromide 47 or naphthalene, acenaphthylene, fluorene and anthracene with sodium dodecyl sulfate 65).They are less accessible to such polar solvents as methanol. Consequently, the addition of modifying alcohols will have a weaker effect on their retention compared to that in RP HPLC. The interaction of sorbates with the micelles decreases as the hydrophobicity of alcohols increases. This effect is particularly well manifested for hydrophobic sorbates (having large partition coefficients in a water ± micelle system) because of the competition between this alcohol and the micelles for the interaction with the sorbate.The order of the variation of Shyb as a function of the nature of the alcohol used proved to be the same in the systems with both cetyltrimethylammonium bromide 47 and sodium dodecyl sul- fate.65 It should be noted that the proportion of an organic modifier added to hybrid eluents does not usually exceed 10 vol.%. At larger contents of organic solvents, the retention mechanism can be changed to reach eventually that typical of RP HPLC; in this case, the system will lose all its advantages in selectivity. It was shown 58, 59 that up to 20 vol.% of 1-propanol could be added to the micellar eluent without formation of the usual aqueous- organic system.Based on the equations deduced for RP HPLC, Madamba- Tan et al.71 have developed a new approach to the gradient elution with organic modifiers in MLC. When such gradients are used, equilibrium is established rapidly owing to a narrow interval of organic modifier concentrations in the gradient. This results in the reduction of the analysis time. Practical applications of propanol and acetonitrile gradients have been described.72 For example, separation of dansylated and bisdansylated amino acids in the isocratic regime requires 67 min. A gradient of 2-propanol (from 3 vol.% to 15 vol.%) in 5 min with subsequent isocratic elution for another 20 min allows separation of these compounds (the total separation time is 32.5 min, including the time necessary for the equilibrium onset).According to Eqn (26), the retention of sorbates in MLC decreases as the concentration of organic modifier increases. This holds for the most systems studied in a narrow range of modifier concentrations (3 vol.% ± 15 vol.% of 2-propanol).67 However, the dependence of ln k0 on j is not linear for some amino acids and alkylbenzenes with mobile phases containing sodium dodecyl sulfate or cetyltrimethylammonium bromide 47 and for some benzene and naphthalene derivatives in a sodium dodecyl sulfate ± n-butanol ± water system. For some other compounds, the dependence of ln k0 on j was linear only for the mobile phases supplemented with methanol.72 For this reason, a new model was proposed 72 for describing the variation of the sorbate retention time inMLCwith changes in the organic solvent concentration.In the example of separation of a mixture of catecholamines on a column packed with Spherisorb ODS-2, the dependences of k0, 1/k0 and log k0 on [M] and j were approximated using different polynomials up to the second power inclusive using the factor planning method (the factors planned were the surfactant con- centration, the micelle concentration [M] (up to 0.15 mol litre71) and the propanol concentration j (up to 10 vol.%). The depend- ence of k0 on [M] and j was shown to be a poorly expressed asymmetric hyperbolic surface with an extremum situated near the maximum values of [M] and j.The addition of 10 vol.% of propanol and 0.035 Msodium dodecyl sulfate led to a 58% ± 78% decrease in k0, whereas the increase in surfactant concentration in a purely micellar eluent from 0.035 to 0.15 M resulted in a 81% reduction of k0. The least approximation error was obtained using the equation (33) 1/k 0=A[M]+Bj+C[M]j+D , where [M] is the micelle concentration, j is the volume proportion of the organic modifier and A, B, C and D are constants.992 Equation (33) gives a satisfactory description of the retention of amino acids, peptides, phenols and some other aromatic compounds (the coefficient of correlation between the calculated (34) and experimental k 0 values is above 0.996).The following equation was also proposed to describe reten- tion:1/k 0=A[M]+Bj+C[M]j+Dj2+E, where A, B, C, D and E are coefficients. For catecholamines, amino acids and phenols the coefficients of different members of Eqns (33) and (34) are the same, the contribution of the term Dj2 to Eqn (34) is insignificant. Equation (33) gives too low values for the most hydrophobic surfactants. Equation (34) provides a better description of the K , (35) k0 à retention.70 The following equation was proposed #," # " 1 á KAMÖ1 á KMDjÜ Ö1 á KADjÜâMä SWÖ1 á KSDjÜ 1 á KADj where KAD, KMD and KSD are constants reflecting the relative variation of the sorbate concentration in the aqueous phase, in the micelles and in the stationary phase, respectively, in the presence of a modifier compared to that in the purely micellar solution.Garcia Alvarez-Coque et al.70 examined separation of nine sorbates on a column packed with Spherisorb ODS-2 (C18) with a propanol-containing mobile phase of sodium dodecyl sulfate in order to assess the applicabilities of Eqns (33), (34) and (35) for the description of retention. The best results were obtained using Eqn (35). Thus in the description of retention of pyrene, Eqn (35) gives a substantially higher (> 50%) accuracy than does Eqn (33). Analogous mean relative deviations of the calculated k0 values from the experimental ones were obtained for other sorbates. For the purely micellar eluents, Eqn (33) is transformed into well- known Eqn (2).Thus, the simultaneous influence of both the surfactant and the organic modifier concentrations on the retention time was again demonstrated. It is these two factors that play the crucial role in the optimisation of separation conditions in MLC. Strasters et al.58 have proposed an iterative-regressive strategy for the choice of optimum conditions for the separation of model mixtures of ionogenic compounds: 13 amino acids and peptides with a sodium dodecyl sulfate ± 2-propanol ± water mobile phase and 15 phenols with a myristyltrimethylammonium bromide ± 2-propanol ± water mobile phase. Prediction of the retention of sorbates proved to be possible on the basis of only five experi- ments even when the values of parameters were far apart in the parametric space.The time of analysis was decreased 3 ± 7-fold and the selectivity was enhanced. In order to solve the problems of three-parametric optimisa- tion in MLC where the variables are the concentrations of a surfactant, an organic modifier and pH of the mobile phase, a Turbo-Pascal (version 5.5) programme has been developed.59 In the proposed model, it is necessary to derive 15 initial chromato- grams in order to plot the function of response ± criterion of separation (the resolution of a pair of components which are most difficult to separate and the total time of analysis are regarded as the separation criterion). This model is applicable for optimising the separation of a mixture of 9 amino acids and peptides using a hybrid eluent containing a phosphate buffer, sodium dodecyl sulfate and 2-propanol.Rukhadze et al.73 have established the conditions for the separation of a mixture of five barbiturates by the three-factor method using the micelles consisting of sodium dodecyl sulfate and a mixed organic modifier [pentanol ± heptanol (1 : 3)] at pH 2.5; the separation time was 40 min. Potential advantages of the application of artificial neuronal computational networks instead of classical statistical methods for simulating the dependence of the retention time on the concentrations of surfactant and organic modifier have been demonstrated using a mixture of 27 dihydropyridines 68 or a EMBasova, VMIvanov, O A Shpigun mixture of 23 benzene derivatives and PAH.69 The effects of the network architecture (the binding mode of individual neurons), the nodal activation function, the number of layers and the mode of the data processing were studied.The linear activation function was shown to be optimal. It is necessary to use networks with the recurrent bonds; the best results were obtained using the loga- rithms of capacity factors as the output data.69 In the case of dihydropyridines, 1 ± 3 neurons and recurrent networks were used.68 Different transmission functions (two sigmoidal, one logarithmic and one linear) were activated in the node of the hidden layer; of these the logarithmic function proved to be the optimum. The optimum number of nodes in the hidden layer was equal to three. The use of the magnitude reciprocal to the capacity factor enhances the performance of neuronal networks.It was shown 68 that there are appreciable differences between the possibilities of simulation of the retention of dihydropyridines with the aid of neuronal networks and by means of empirical equations. 4. Effect of the nature of the surfactant The selectivity of separation in MLC may be influenced not only by the variation of the surfactant concentration but also by changes in the nature of the surfactants (e.g., the alkyl chain length and charge). Thus a mixture of six vanillins was successfully separated using a mobile phase of 0.02 M sodium dodecyl sulfate solution (pH 3).74 However, when cetyltrimethylammonium bromide was used, the retention time of four of six vanillins proved to be longer than in the case of sodium dodecyl sulfate, although in the latter case the mobile phase contained an appreciably larger number of micelles (the CMC of cetyltrimethy- lammonium bromide at 30 8C and of sodium dodecyl sulfate at 25 8C in distilled water are 0.9 and 8 mmol litre71, respectively).Thus, the interaction of the vanillins studied with the stationary phase modified by cetyltrimethylammonium bromide is stronger than that with the stationary phase modified by sodium dodecyl sulfate. Analysis of the regression dependences carried out with the aid of Eqn (2) showed that when sodium dodecyl sulfate is used, the elution order of the sorbates correlates with the magnitude of F[Ls]K1, and hence the selectivity is determined mostly by the stationary phase ± sorbate interactions.It should be noted that for a number of compounds the retention time also correlates with K2. A decrease in selectivity with the increased concentration of sodium dodecyl sulfate, and hence with the quantity of micelles in the mobile phase, also indicates that the sorbate ± micelle interaction begins to play a more important role. For cetyltrimethylammonium bromide, the order of elution correlates with changes in both F[Ls]K1 and K2, i.e., the selectivity of separation is determined by the interactions of sorbates with both the micelles and the stationary phase. The observed differences in selectivities are determined, in the opinion of some investigators,33, 44, 75 by the nature of the surfactant interaction with the stationary phase and by the properties of the modified surface.It follows from the models of surfactant adsorption that in the adsorption of sodium dodecyl sulfate the hydrophobic alkyl radical is inserted into the stationary C18 phase, whereas the hydrophilic polar heads are directed towards the bulk of mobile aqueous phase and form a hydro- philic layer, which results in a decrease in the stationary phase polarity. In the case of cetyltrimethylammonium bromide, the head groups seem to be localised nearer to the surface owing to additional hydrophobic interactions of the N-methyl groups of ammonium salts with alkyl radicals of the stationary phase.It is possible that they are partially incorporated into the grafted phase. A part of cetyltrimethylammonium bromide can be sorbed irreversibly; the positively charged head groups of the surfactant can form strong hydrogen bonds with the residual silanol groups of the stationary phase and at the sites of defects in the grafted layer. All this leads to an increase in the stationary phase hydro- phobicity. Indeed, when aqueous-organic eluents are used before and after work with the micellar phases, the retention times ofMicellar liquid chromatography compounds where sodium dodecyl sulfate is used are decreased, which points to a decrease in the hydrophobicity of the mobile phase, but they are increased with the use of cetyltrimethylammo- nium bromide, which suggests an increase in its hydrophobicity.Lavine et al.74 presume that the irreversible adsorption of cetyltrimethylammonium bromide leads to a reduced activity of the residual silanol groups, which compensates for the increase in the stationary phase polarity associated with the presence of the charged head alkylammonium groups on the surface. On the whole, the stationary phase becomes less polar and more hydro- phobic. Since compounds under separation (vanillins) are weak acids (pKa=7.4 ± 8.9), they are characterised by specific inter- actions (hydrogen bonding) with the negatively charged layer of the head groups of sodium dodecyl sulfate, which results in higher selectivity than that in the case of cetyltrimethylammonium bromide.Lavine et al.76 have shown that the micelles of cetyltrimethyl- ammonium bromide possess better selectivity with regard to compounds containing acidic functional groups. This is explained by the existence of a secondary chemical equilibrium which includes proton transfer from the ionogenic compound to the water molecules in the Stern layer of the surfactant micelle. The decrease in Ka from 0.5 to 3.0 can lead to the incorporation of the guest molecule into the cationic micelle.77 The surfactants containing two or more hydrophobic `tails' in the monomer, e.g., didodecyldimethylammonium bromide or dihexadecyl phosphate, form vesicles. The selectivity of separa- tion with the use of vesicles in the mobile phase differs from that in MLC, which is associated with the increase in the number of sites accessible for solubilisation from five (in a micelle) to nine (in a vesicle).78 5.Effect of pH The effect of pH on selectivity is determined by the occurrence of different secondary equilibria in the mobile phase and the surfactant-induced shift of the ionisation constants of sorbates.26 The influence of pH on the selectivity of separation of amino acids (e.g., tryptophan and phenylalanine) was studied in aqueous, ion- pair and micellar eluents. The maximum selectivity was observed for the mean pKa value of a pair of sorbates. The optimum selectivity aopt may be calculated by the equation (36) aopt à 2SSm á SSm 1=2ÖSSmKi á KjÜ=ÖKiKjÜ1=2 2SSm á SSm 1=2ÖSSmKj á KiÜ=ÖKiK j Ü1=2 , where SSm is the mean self-selectivity (the ratio of capacity factors of two sorbate forms 14), Ki* and Kj* are the apparent dissociation constants of sorbates i and j in aqueous, non-aqueous and micellar media.The selectivity of separation of this pair of amino acids is a maximum in aqueous medium, ion-pair chromatography and micellar liquid chromatography at pH 2 ± 2.5, *4 and 5 ± 5.5, respectively. It should be noted that the surfactant-induced shifts of pKa values do not necessary result in enhanced selectivity. For example, for compounds having different pKa values and hydro- phobicity (isomeric chlorophenols), the micelles influence both parameters.79 6. Effect of the nature of the stationary phase The effect of the nature of the stationary phase is usually considered in MLC as a variation of the column efficiency due to different adsorption of surfactants.Alkylated silica gels C1, C8, C14 and C18 with different densities of grafted alkyl groups and cyanopropyl phases were studied. Yang and Khaledi 80 have demonstrated the advantages (higher selectivity and shorter analysis time) of fluorinated, chemically bound stationary phases compared to the alkyl ones for the separation of mixtures of amino acids, peptides, sulfonamides, phenols and alkylphenols 993 using the mobile phases containing anionic and cationic surfac- tants. Model mixtures of vanillins were separated on columns packed with Apex IC18 and Apex IC8 (5 mm) using the same mobile phases (0.02 M surfactant solution).74 Both stationary phases provide effective separation upon elution with sodium dodecyl sulfate solutions; in this case, the retention is determined by the interaction of sorbates with the stationary phase.The retention times of sorbates on the C8 phase are longer than those on the C18 phase, despite the fact that the latter is more hydro- phobic and contains 13.3% of carbon (theC8 phase contains 9.0% of carbon). This may be due to poorer sorption of sodium dodecyl sulfate on the C18 phase because of the higher density of grafted groups.81 The use of mobile phases containing the cetyltrimethylammo- nium bromide micelles also results in a stronger retention on the C8 phase than on theC18 phase; however, in this case an additional effect consists in the alteration of the order of elution of some compounds.The low adsorption level on theC8 phase could be the cause of the stronger retention than that on the C18 silica gel; this should not however result in the reversal of the elution order. The higher activity of the residual silanol groups on the C8 sorbent compared to that on the C18 sorbent can neither explain the observed phenomenon since the elution order of sorbates in the methanol ± water system is the same on both stationary phases. For this reason, Lavine et al.74 have presumed that small premicellar aggregates of cetyltrimethylammonium bromide are formed inside the stationary phase and modify the retention of sorbates.To explain the differences in selectivity observed in MLC, the interaction of sodium dodecyl sulfate and cetyltrimethyl- and dodecyltrimethylammonium bromides with Apex I C18,82 Apex I cyanopropyl and Apex I C8 83 was studied using HPLC and 13C NMR. It was shown that the surfactant is adsorbed on the stationary phase and that this provides for a selective separation of isomeric vanillins. The use of sodium dodecyl sulfate makes it possible to achieve better separation of hydrophilic compounds because it forms a hydrophilic, negatively charged layer on the stationary phase surface. During adsorption of ammonium surfactants, the head groups are incorporated, at least partly, into the grafted phase owing to hydrophobic interactions of their methyl groups with the C18 radicals of the stationary phase.82 In the separation of a mixture of hydrophilic compounds with the aid of these cationic surfactants, the best results were obtained at high concentrations of the surfactant in the mobile phase, when the sorbate ± micelle interactions are advantageous. The length of the hydrocarbon chain of the stationary phase is an important factor in the separation with the use of cetyltrimethyl- and dodecyltri- methylammonium bromide micelles: the behaviour of hydrophilic aromatic compounds on the C8 and C18 phases is different.83 It was noted 83 that when sodium dodecyl sulfate as well as cetyltrimethylammonium bromide are employed, the retention times of some ionogenic compounds are increased.This unusual phenomenon may be explained by strong interactions of the polar groups of the surfactant and the nitrile groups. 7. Effect of temperature The effect of temperature is the least studied aspect of MLC. It is obvious from general speculations that the temperature should be higher than the Kraft point, which corresponds to the lower temperature limit for the existence of micelles (23 8C for cetyl- trimethylammonium bromide 73). Researchers usually operate at higher temperatures (40 ± 60 8C, most often at 40 8C) mainly because elevated temperatures enhance the chromatographic system efficiency and reduce the time of analysis. The effect of temperature on the retention in MLC was investigated in thermodynamic terms.63, 84 The retention of anthracene and its derivatives on C18 in a sodium dodecyl sulfate ± 1-propanol ± phosphate buffer (pH 2.1) system was studied in the temperature range 303 ± 353 K.84 The results994 obtained were interpreted using two models.There is no correla- tion between the enthalpies calculated in both models and the physicochemical properties of sorbate molecules. In the model based on the Van't Hoff equation without consideration of individual equilibria, the values of standard entropies correlate with the polarity of sorbate molecules. Villanueva Camanas et al.49 did not succeed in enhancing the efficiency of separation of some 6-mercaptopurine-based medic- inal preparations by using sodium dodecyl sulfate and varying the temperature from 20 to 60 8C.They noted the influence of both organic modifiers (alcohols, acetonitrile) and temperature on the resolution owing to changes in selectivity; however, these param- eters affect little the efficiency and their variation has insignificant practical application. As the studies devoted to a systematic investigation of the effect of temperature in MLC are very scarce, it is impossible to draw any generalising conclusions. It may only be said that it is advisable to use elevated temperatures in MLC for decreasing the viscosity of mobile phases. This increases the column efficiency and stabilises the operation of pumps the working pressures of which are restricted. 8. Effect of ionic strength There are no separate studies concerning the investigation of the effect of ionic strength in MLC.Since the addition of salts leads to a lower CMC and, consequently, to an increase in the micelle concentration, the effect of ionic strength on selectivity may be described as the influence of the increasing concentration of micelles. The selectivity of separation in micellar chromatography3 differs from that in conventional ion-exchange 85 and ion-pair chromatography. In MLC, after the column equilibration with 1074 M cetyltrimethylammonium bromide solution, anions are eluted with 0.025 MNaCl solution in the following sequence: IO¡ < NO¡3 < Br7 < NO¡3 < I7 (see Ref. 30). The use of eluents containing high concentrations of an organic modifier (lower alcohols or acetonitrile) leads to degradation of micelles (in the presence of>43.5% of methanol no micelles are formed at all 86), so that retention is controlled only by the extent of surfactant adsorption on the stationary phase, and the selectivity becomes the same as that in ion-exchange or ion-pair chromatography.On the whole, it should be noted that studies of the effects of various factors on the selectivity in MLC should consider all three of the equilibria established in the system. In the first place, there is an interaction of sorbates with micelles, which is characterised by the association (binding) coefficient K2, and with the stationary phase characterised by the partition coefficient PSW; they have opposite effects on retention.V. Retention and structure of sorbate molecules In RP HPLC, three types of simulation of retention parameters (or their functions of parameters characterising specific features of the sorbate structure and influencing its chromatographic behav- iour) are used, viz., simple additive models, a retention ± property correlation and a retention ± structure correlation.1 However, in ion-pair chromatography the complexity of the processes occur- ring prevents construction of models which describe the effect of structural characteristics of sorbates on retention. Nonetheless, there are indications 1 that the parameters of retention models may be related to the contributions from different structural fragments and to the Gantsch p-constants of hydrophobicity.It would seem that in MLC where interactions are more numerous than in RP HPLC and ion-pair chromatography, construction of such models is not feasible. However, the existence of a clear-cut dependence of solubilisation on hydrophobic and hydrophilic properties of sorbates has stimulated the attempts to apply correlations of the retention ± property type, which relate the capacity factor and the partition coefficient in a 1-octanol ± water system (POW). EMBasova, VMIvanov, O A Shpigun Thus Khaledi and Breyer 87 have revealed that the values of k0 correlate well with the log POW for 35 related compounds on the phenyl phase with a cationic surfactant; however, the dependence of log k0 on log POW forms a plateau at high values of the partition coefficient, i.e., for hydrophobic compounds. For aromatic compounds on the stationary C8 and C18 phases, a better correlation was obtained for the dependence of log k0 on log POW, as in RP HPLC.The existence of a linear correlation log k0=f(logPOW) was established for monosubstituted benzenes with sodium dodecyl sulfate, cetyltrimethylammonium bromide and Brij-35 as the mobile phases,88 for a number of phenols and other monosubstituted benzenes with alcohol-modified mobile phases of sodium dodecyl sulfate and cetyltrimethylammonium bromide 76 and for a series of surfactants with mobile phases containing sodium dodecyl sulfate, cetyltrimethylammonium bromide and Brij-35.89 A correlation between the log POW and the logarithms of the constant of sorbate association with the micelle was established.76, 87, 90 A linear correlation was found to exist between the log k0 (the capacity factor in the absence of micelles) and log POW.87 The dependence of k0 and log k0 on log POW was studied by MLC for a series of benzene derivatives and surfactants on a stationary C8 phase.91 m (k0 at zero micelle concentration) and PMW Estimates of k0, k0 were derived for 26 1,4-dihydroxypyridines, and it was shown that they correspond to the hydrophobicity scale.92 The best hydro- phobicity scale is the k0 7 log POW correlation for the mobile phase of 0.031 M sodium dodecyl sulfate containing 3% of n-propanol; this phase is described by the equation k0=48=3 log POW 7 98.49. The differences between the exper- imental and calculated values of log POW for 20 of the 26 compounds studied were found to be less than 3%.A linear dependence was shown to exist between the free energy of transfer of a compound from the aqueous to the micellar phase in a water ± octanol system for benzene and naphthalene derivatives in micellar 18 and hybrid 20 eluents, as well as for substituted phenols.93 The slope of the straight lines is virtually independent of the nature of the surfactant (sodium dodecyl sulfate, cetyltrimethylammonium bromide or Brij-35) or of the presence of an organic modifier.20 This indicates that in all cases the reactions of sorbates with the micelles are based on hydro- phobic interactions. For alkyl-substituted phenols, the value of the free energy of transfer of the aliphatic hydrocarbon groups is less negative for the sodium dodecyl sulfate micelles than in the octanol ± water system, whereas for the phenols containing the OR, COR and COOR substituents the contribution to the hydro- phobicity is larger in the presence of sodium dodecyl sulfate than in the octanol ± water system.93 These differences may be due to the fact that solubilisation of different sorbates can occur in the micelle at different sites which differ in dielectric constants, local microviscosity, steric factors and the occurrence of electrostatic interactions with the ionic head groups of surfactants for charged sorbates.The second group of correlations relates retention to the structure of sorbates described with the aid of various descrip- tors,1 the simplest of which is the number of the carbon atoms in the molecule (nC).In the example of n-alkylbenzenes and alkylphenols in sodium dodecyl sulfate and cetyltrimethylammo- nium bromide-containing systems, including those with propanol added,19 it was shown that in contrast to RP HPLC where the dependence of log k0 on nC is linear, in MLC with both purely micellar and hybrid eluents only the dependence of k0 on nC is linear, whereas the dependence of log k0 on nC is better described by the quadratic function. Analogous dependences were derived in the studies of the retention of alkylbenzenes in the mobile phases containing Brij-22 or Brij-35.16 Alkylphenols in the micellar mobile phases of sodium dodecyl sulfate are an excep- tion.For this group of compounds the dependence of log k0 on nC is linear, while the dependence of k0 on nC is quadratic, as in RP HPLC.19 For phenols and a hybrid sodium dodecyl sulfate mobile phase, the differences in k0 and log k0 are too small to allow one toMicellar liquid chromatography state with certainty which of the dependences is better approxi- mated by the straight line. The cause of this phenomenon is not sufficiently clear. Khaledi et al.19 proposed two possible explan- ations. The first explanation is based on the difference in the location of various sorbates in (or on) the micelle since the heterogeneous nature of the micelles creates a unique situation where different sorbates in the same mobile phase have micro- environments differing in polarity. The second explanation relies on the model of interaction indices,94, 95 according to which the dependence of log k0 on the number of carbon atoms for aqueous- organic eluents is described by the equation (37) log k 0=(log g)n 2C +(log a)nC+log b , where log a and log b are constants and log g is expressed by the Eqn (38) (38) logk0g à DVxDIxDCxCMIM , 2:3RT where DVx and DIx are the changes in the sorbate volume and in the interaction index in a homologous series, DCx and CM are the proportionality constants for the sorbate and the mobile phase, IM is the index of interaction for the mobile phase.Usually the value of log g for aqueous ± organic eluents is so small that the first term of Eqn (37) may be neglected.It is probable that in MLC the product of DVx DIx is somewhat larger for some homologues and this causes the departure of the dependence of log k0 on nC from linearity. Nonetheless, Khaledi et al.19 have succeeded in revealing linear dependences of the logarithms of sorbate ± micelle associa- tion (binding) constants (K2) and of partition coefficients of sorbates between the extramicellar mobile phase and the sta- tionary phase (PSW) on nC. Analysis of these dependences showed that the K2 values for homologues are larger in the cetyltrimethy- lammonium bromide micelles than in the sodium dodecyl sulfate micelles, which points to a stronger interaction of sorbates with the cetyltrimethylammonium bromide micelles owing to the availability of hydrophobic environments. The slope of these straight lines is the measure of free energy calculated for the transfer of one methylene group from the bulk of solution into the micelle (dependence of K2) or into the stationary phase (PSW), and this slope is steeper for cetyltrimethylammonium bromide than for sodium dodecyl sulfate, which is indicative of the higher affinity of the methylene group of cetyltrimethylammonium bromide due to its larger hydrophobicity.(39) WS á nmDGmWS RT Vm Vs á k0 à 1 exp Hinze and Weber 96 proposed a more rigorous approach to the rationalisation of the non-linearity of the dependence of log k0 on the number of the carbon atoms in the surfactant molecule in MLC in their paper entitled `Why the relationship between the logarithm of k0 and homologue number in micellar chromato- graphy is not linear '.In RP HPLC, it is implied that each CH2 increment in the chain of a homologue makes a constant contribution to the free energy of sorbate transfer between the mobile and stationary phases. In MLC, the retention of small- sized polar sorbates that are well soluble in water and poorly soluble in micelles is basically determined by the partition between the surfactant-modified stationary phase and the non-micellar mobile phase. In a homologous series, compounds become less water-soluble as their size increases, and in the extreme case of total insolubility in water the sorbates are partitioned between the chemically similar micellar phase and the surfactant-modified stationary phase. For this process, the free energy of one methylene group transfer from the mobile to the stationary phase (DGm) is virtually equal to zero.As in MLC the DGm value depends on the homologue size, the dependence of ln k0 on the number of methylene groups in the sorbate (nm) should not be linear. DGfg 995 exp XS á nmDGmXS RT Vx Vs á DGfg where the subscripts `X', `W' and `S' refer to the micellar, aqueous and stationary phases; V stands for the corresponding volumes of phases and the superscripts `fg' and `m' refer to the functional and methylene groups, respectively. The dependence of ln k0 on nm is described by the equation ln k 0=7ln [(17Fx)exp(a+nmb) + Fxexp(c+nmd)] , (40) where Fx=Vx/Vm is the ratio of the volumes of micelles and stationary phase; a, b, c and d are constants.The parameters of a non-linear regression were calculated, the free energies of the methylene group transfer from the non-micellar part of the mobile phase (pure water or water containing 5 vol.% of alcohols) into the surfactant-modified stationary phase [parameter b, Eqn (40)] and the free energy of the methylene group transfer in the micelle ± stationary phase system [parameter d, Eqn (40)] were estimated for alkylbenzenes with the use of sodium dodecyl sulfate ± 1-pentanol ± water and Brij-35 ± water systems and also for methyl esters of fatty acids with the use of a sodium dodecyl sulfate ± 1-propanol ± water system.Recent years witnessed the appearance of more complex models which relate retention to the sorbate structure and which are based on the statistical analysis and the method of analysis of basic components.97, 98 In the example of 17 surfactants with three different micellar mobile phases (sodium dodecyl sulfate, cetyl- trimethylammonium bromide, Brij-35) it was shown that reten- tion can be adequately described using only two of the nine descriptors: L/B (ratio of the maximum molecule length to its width) and one of the topological descriptors.97 The contribution from the topological descriptors is most important in the case of cationic and anionic surfactants, whereas the electrostatic inter- actions are particularly substantial in the case of sodium dodecyl sulfate.Correlation equations of the types log k0=f (F, L/B, c) and log k0=f (n, L/B, c), where F is the correlation factor, c is the surfactant concentration in the micellar mobile phase, n is the number of p-electrons in the molecule, were derived. These equations are easier to use because the values of the corresponding descriptors can be calculated directly from the structural formulae of compounds. Analogous correlations can also be established for the partition coefficients PSM, PSW or PMW since they can be calculated from experimental chromatographic data using Eqn (1) or (2).76 For sodium dodecyl sulfate, correlations of the type PSM (PSW or PMW)=f(L/B, F) were obtained; however, the regression is deteriorated in the sequence PSM>PMW> PSW because of the increased uncertainty of calculations.97 In the hybrid mobile phases which contain sodium dodecyl sulfate with additions of methanol, 2-propanol or n-butanol the retention of unsubstituted PAH is described by the equation considering the size and shape of molecules.(41) log k 0=aF+bL/B+c, where a, b and c are constants.98 This model does not however describe methyl-substituted PAH. Therefore, new models were proposed which include electronic parameters allowing consideration of the distribution of electronic density in the molecule, viz., dipole moments (DPMON) and the share of the non-polar unsubstituted surface area (NUSA/TSA) [Eqn (42)].log k 0=aF+bL/B+cDPMON+dNUSA/TSA+e, (42) where a, b, c, d and e are constants. Equation (42) gives a satisfactory description of the retention of both unsubstituted and substituted surfactants upon addition of <20 vol.% of methanol or 2-propanol or<6 vol.% of n-butanol. The influence of molecular structure on the constants of the sorbate ± micelle and sorbate ± stationary phase binding has been studied.99, 100 The existence of a clear-cut linear dependence between the constants of the sorbate ± micelle and sorbate ±996 stationary phase binding was demonstrated.100 Fairly good quantitative dependences of the constants of sorbate ± micelle and sorbate ± stationary phase binding on solvatochromic proper- ties were obtained.99, 100 The correlation between k 0 and solvato- chromic parameters is better in the presence of tetradecyltrimethylammonium bromide micelles, whereas the correlation between log k 0 and solvatochromic properties is better in the presence of sodium dodecyl sulfate micelles; the addition of 7% of 2-propanol or 5% of butanol does not modify the strong correlations revealed for tetradecyltrimethylammo- nium bromide.99 An important role in the retention of neutral sorbates is played by both the process of cavity formation and the hydrogen bonding.The examples considered above indicate that such models not only allow prediction of the retention of various sorbates but are also helpful for revealing interactions of sorbates with stationary and mobile phases.Guermouche et al.101 studied theoretical aspects of the retention of 15 aromatic compounds in MLC using the zwitterion surfactant n-dodecyl 3-(N,N-dimethylamino)propane-1-sulfonate (DDDMAPS) and calculated the constants of equilibrium parti- tion of sorbates using Eqn (1). Compared to the cationic and anionic surfactants in the DDDMAPS-containing mobile phases, the lowest values of KSW and KMW were obtained, which leads to larger capacity factors at a constant surfactant concentration. Sorbates are retained owing to hydrophobic interactions with the stationary phase and electrostatic interactions with the surfactant in the micelles as well as in the modified stationary phase. Molecular interactions were studied using the method of linear dependences of solvation energy.101 The sorbate size and basicity were shown to be the dominant factors affecting the retention.The dipolar environment in the zwitterionic DDDMAPS is larger and this influences the partition of sorbates: acidic sorbates are more easily bound to this surfactant. A correlation between the retention and biological activity was established using models of the retention ± property type. Breyer et al.102 revealed a clear-cut correlation between k 0 and biological activity, measured as log 1/c, where c is the concen- tration required for a 50% inhibition of growth, for 26 para- substituted phenols. The advantage of this approach consists in that it uses a single parameter, viz., retention in MLC, whereas normally three descriptors are used for estimating biological activity: logPOW, acid dissociation constant and resonance parameter.Owing to the unique properties of surfactant micelles in MLC, both hydrophobic and electronic interactions are combined in one parameter, viz., k0. The nature and acidity of the stationary phase influence the correlation between k0 and log 1/c; the optimum system is cetyltrimethylammonium bromide supplemented with 10 vol.% of 2-propanol. Escuder-Gilabert et al.103 studied qualitative and quantitative correlations between the retention and the anesthetic strength, equianesthetic concentration, duration of action, toxicity and the half-life of anesthetics. Clear-cut correlations between k0 and the anesthetic strength of compounds were established.Deriving correct correlations requires reliable determination of the dead time (t0) which is problematic in MLC. In MLC, the magnitude of t0 is measured with respect to the introduction of water,93, 104 salt solutions (NaNO3,25, 37 NaI and KI 11, 19) and acetonitrile.66 Torres Lapasio et al.105 proposed four different criteria for determining the characteristic time at the beginning of chroma- tography (the time of the exit of the first peak, first perturbation) that may be called reference time. 1. Search for the first basic maximum or minimum. 2. Formation of groups reducing noises to a definite scale. 3. Formation of groups for scaling the slopes of the basal line noises.4. First-derivative method consisting in the detection of an important change in the basal line slope upon the occurrence of a considerable perturbation. The best results are given by the noise- scaling criterion; in this case it is advisable to measure the EMBasova, VMIvanov, O A Shpigun reference times upon addition of standard solutions or micellar solutions of sorbates. Determination of the reference time using a large number of different mobile phases has shown that it does not change noticeably as the mobile phase composition is modified. VI. Analytical application of micellar liquid chromatography The development of the theory of MLC was based on the studies of different organic compounds: non-ionogenic (substituted benzenes, phenols, naphthalenes, anthracenes, etc.) as well as ionogenic (amino acids and peptides), whereas practical uses of this method lie in the analysis of medicinal preparations (Table 1).Such an analytical application reflects several advantages of MLC: selectivity and capacity for simultaneous separation of ionic and non-ionic components of samples, direct introduction of biological samples without preliminary precipitation of proteins or extraction of drugs, as well as the possibility of fluorescence enhancement or stabilisation. Simplification of the sample prep- aration for the analysis of biological specimens is achieved owing to solubilisation of proteins on micellar aggregates since the protein ± micelle aggregates have a large size and are not retained by the stationary phase.Low detection limits make it possible to detect specific compounds not only in pharmaceutical formula- tions but also in urine. The possibility of enhancement of sensitivity of determination of some compounds in the presence of surfactants has been demonstrated. Thus the oxidation current of acetaminophen is increased in the presence of sodium dodecyl sulfate micelles.108 An electrochemical detector with a cell of the `reflecting wall of the three-electrode configuration' type with a carbon-fibre micro- electrode was designed for detecting acetaminophen. The compat- ibility of the separation byMLCwith amperometric detection was first demonstrated by Khaledi and Dorsey;126 dopamine was determined using a glassy-carbon electrode.124 Steroids were determined in urine by the MLC method and detected using the terbium-sensitised fluorescence.106 The micellar environment promotes an efficient Forster energy transition from donors (steroids) to acceptors [terbium(III) ions].This method is highly sensitive and allows determination of as little as 100 pg of steroids in urine. This detection technique is appreciably more sensitive than the spectrophotometric detection,114 which is applicable only for the analysis of initial pharmaceutical prepara- tions and is unsuitable for the analysis of biological samples. The luminescence [sensitised by a terbium(III) ± surfactant complex] in the inverted micelles in non-aqueous solutions (ethyl acetate or its mixture with hexane) was used for the detection of theophylline;120 the presence of paraxanthin does not interfere with this determination due to the low efficiency of energy transfer to the lanthanide ion.This method minimises the fluorescence quenching induced by the use of inverted micelles in non-aqueous solvents. Examples of the use of micellar mobile phases in the inorganic analysis are not numerous: separation of inorganic anions;30, 32, 85 cations of transition metals;9, 31 chelates;127 ± 131 heteropoly- acids;132, 133 different forms of arsenic,77, 134, 135 mercury,77, 135 selenium 77 and tin;77, 136 organic and inorganic compounds containing heavy metals.137 Thus the separation of five anions on a column (25065 mm) packed with ODS (5 mm) using 1.3661071 M aqueous cetyltri- methylammonium chloride as the eluent at a flow rate of 1.5 ml min71 was effected within 10 min.85 Detection was carried out at 205 nm.The order of elution: IO¡3 <NO¡2 <Br7<NO¡3 < I7 is analogous to the selectivity sequence for a typical, strongly basic anion exchange resin. The use of micellar exclusion chromatography on a column packed with the Asahipak GC-310H sorbent, which is a polyvinyl alcohol gel, proved to be effective for the separation of both anions 30 and cations.32 In the separation of anions the elution order varies with the modification of hydrophobicity of theMicellar liquid chromatography Table 1. Determination of components of medicinal preparations by MLC using sodium dodecyl sulfate.Detector Components of medicinal preparations Organic modifier FL SP acetonitrile 7 EC SP SP SP SP n-propanol n-butanol n-propanol """n-pentanol 7 7 7 7 Steroids Amiphenazole, amiloride, amphetamine, clostebol, ephedrine, phenylpropanolamine, methanedienone, methoxyphenamine, nandrolone, spironolactone Acetaminophen Clenbuterol Diuretics Caffeine, theophylline, theobromine N-(1-Benzo[b]thien-2-yl)ethyl-N-hydroxyurea, its N-dehydroxylated metabolite Sulfonamides Amiloride, bendroflumethazide, chlorothalidone, spironolactone, triamerene SP SP SP n-propanol "n-pentanol a Diuretics Catecholamines Anabolic steroids SP n-propanol Proteinic amino acids (1.3 ± 1.5)6 61077 mol litre71 7 7 7 SP n-propanol " SP 7 7 7 SP n-pentanol Diuretics b-Blockers (atenolol, metoprolol, oxyprenolol, diuretics (amiloride), bendroflumethazide, hydralazine Theophylline 6-Mercaptopurine, 6-thioguanidine 4-Formylaminoantipyrine, 4-aminoantipyrine, 4-methylaminoantipyrine (R)- and (S)-Naproxenes 10.54 mg ml71 ± ± 17.0 ng ml71 7 7 7 EC FL Dopamine Theophylline 7ethyl acetate, hexane Note: FL, fluorescent; SP, spectrophotometric; EC, electrochemical (amperometric) detectors, respectively; a microemulsion.surfactant alkyl chain: with 0.06 or 0.01 M cetyltrimethylammo- nium chloride mobile phase, the anions are eluted in the sequence IO¡3 < I7 < Br7 < NO¡2 < NO¡3 , whereas in the presence of dodecyltrimethylammonium chloride this order is IO¡3 <Br7 < NO¡2 < I7<NO¡3 (see Ref.30). Using micellar exclusion chromatography, Okada 29 sepa- rated Pb(II), Co(II) and Cu(II) in 0.05 M sodium dodecyl sulfate containing 0.08 mol litre71 of NaCl (pH 4.0) and also a mixture of cations of rare-earth elements (Yb, Ho, Dy, Tb and Pr) and of transition metals [Cu(II), Pb(II), Zn(II), Ni(II), Co(II) and Mn(II)] in the gradient mode of elution with a mixture of 0.025 M sodium dodecyl sulfate and 0.08 M a-hydroxyisobutyric acid (pH 4.0) or with 0.025 Msodium dodecyl sulfate and 0.08 Mtartaric acid (pH 4.3).31 Separation of 8 transition metal cations [Fe(III)<Cu(II)< Zn(II) < Ni(II) < Pb(II) < Co(II) < Cd(II) < Mn(II)] was performed in 20 min on a column with Inertsil ODS-2 with an aqueous mobile phase containing 0.0945 mol litre71 of sodium dodecyl sulfate and 0.0684 mol litre71 of tartaric acid (pH 4.02).29 Miura et al.128 determined the distribution constants of 2-(2- thiazolylazo)-4-methylphenol (TAMP) and 2-(2-thiazolylazo)-5- dimethylaminophenol (TAAP) in solutions of non-ionic surfac- tants [poly(oxyethylene)-4-nonylphenyl ether (PONPE), poly(oxyethylene)oleyl ester (POOE), Triton X-100, POL-10]; these constants were found to decrease as the length of the oxyethylene chain increased. Rapid elution of TAMP and TAAP is achieved using surfactants with shorter oxyethylene chains 997 Ref.Object of analysis Detection limit 106 107 100 pg 0.07 ± 10.95 mg ml71 urine " 108 109 110 111 112 0.02 mg ml71 71.7 ± 2.9 mg ml71 0.36 ± 1.23 mg ml71 0.08 ± 0.1 mg ml71 """"" 113 114 115 48 116 0.029 ± 0.55 mg ml71 2 ± 16 mg ml71 6.4 ± 860 ng ml71 117 118 119 urine, milk pharmaceutical preparations pills "capsules, pills pills, suspension, gel, solution for injections pharmaceutical preparations the same pills, capsules 76 mg ml71 120 121 122 pills serum plasma human liver 123 microsomes 7 7 124 9 ng ml71 7 125 because the reagents have the higher affinity to the micellar phase. The retention time is reduced as the number of oxy- ethylene units in PONPE and POOE decreases.PONPE-20 was used for the separation of metal chelates with TAAP.127, 128 Of the nine chelates studied, only the V(V) chelate is eluted selectively, the peak of Fe(III) chelate overlaps with the peak of the solvent front, while the reagent itself and other complexes are strongly retained (and some of them are possibly degraded) in the column.127 Probably, the use of hybrid eluents (supplemented with organic modifiers) would yield better results. Separation of Cu(II), Zn(II), Ni(II) and Co(II) takes 45 min using a mobile phase containing 0.006 mol litre71 of sodium dodecyl sulfate, 0.002 mol litre71 of 8-hydroxyquinoline-5-sulfonic acid and pH gradients of 3.5 ? 3.9.131 Determination of aluminium with a detection limit of 1 ppb in the form of chelates in human serum with 8-hydroxyquinoline was performed on a column packed with CAPCELL PAK MF ph-1 with an eluent containing 20 mass%of acetonitrile, 0.01 Msodium dodecyl sulfate, 1074 M EDTA, 0.01 M N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid-based buffer (pH 7.0) with fluorescent detection (370 ± 504 nm).130 The contents of Ni(II), Co(II) and Cu(II) in metal alloys were determined using chelates with sodium diethyldithio- carbamate and the mobile phase containing 0.03 M cetyltrimethyl- ammonium bromide, 1074 M sodium diethyldithiocarbamate and 45% of 1-propanol (spectrophotometric detection).129 The detec-998 tion limits of Ni(II), Co(II) and Cu(II) were 68.1, 67.2 and 35.4 pg, respectively; the separation time did not exceed 6 min.Compatibility of MLC with the mass-spectrometric determi- nation with the use of induction plasma was demonstrated in the example of determination of arsenic using the cetyltrimethylam- monium micelles.135 In this case, the detection limits (90 ± 300 pg) were comparable with that in ion-pair chromatography (0.4 ± 1.2 ng cm73 at a loop volume of 250 mm3 with detection by atomic-fluorescent spectrometry with preliminary generation of hydrides 138) but they were inferior to that in ion-exchange chromatography (20 ± 91 pg 139). This may be due to the larger volumes of peaks and a complex composition of the mobile phase and, especially, to the influence of sodium ions from buffer solution which can suppress ionisation of arsenic.A series of studies 77, 135, 136 were devoted to the use of vesicles in MLC. If a micelle is a closed monolayer, then a vesicle is a closed bilayer. The cavity is filled with the substance which is contained in the environment.22 To be able to form vesicles the surfactant molecules should contain two (or more) hydrophobic hydrocarbon radicals with 16 ± 18 carbon atoms. These require- ments are met, for example, by didodecyldimethylammonium bromide 135, 136 and dihexadecyl phosphate.75, 136 These mobile phases can be used in combination with selective detection techniques, such as atomic-emission spectrometry with induction plasma (also after formation of hydrides), mass-spectrometry with induction plasma, atomic-absorption spectrometry of cold vapour and atomic-absorption spectrometry with electrothermal atomisation.39 , 40 and SiW11O839¡ within 7 min. Interesting results were obtained in MLC of heteropoly- anions:132, 133 a mixture of BW12O5¡ 40 , SiW12O440¡ and SiMo12O440¡ was separated within 10 min and a mixture of SiW11FeO5¡ BW12O5¡ However, the results of the application of MLC in inorganic analysis, even though being interesting per se, are basically inferior to those achieved in ion-pair or ion-exchange chromatogra- phy.140 ± 142 VII. Comparison of the possibilities of ion-pair and micellar liquid chromatography As the role of micelles is similar to that of organic modifiers owing to their ability to increase solubility of uncharged sorbates in water, MLC is usually compared to RP HPLC.Separation of uncharged and hydrophobic compounds, for which RP HPLC is widely used, is the field in which MLC is inferior as regards its efficiency, eluting strength and selectivity. The efficiency of columns in MLC is appreciably lower than that in RP HPLC and decreases as hydrophobicity of compounds increases. The micellar eluents are weak eluents; therefore, the retention times of hydrophobic sorbates are long. Selectivity is also low because of the similarity of processes which occur in the mobile and sta- tionary phases,19, 66 although in a number of instances the reversal of the elution order is observed.143 ± 145 All of these characteristics of the chromatographic system can be enhanced by the introduc- tion of organic modifiers into the eluent; however, MLC will not all the same be competitive with RP HPLC in the separation of hydrophobic sorbates.An advantage of the MLC method is the possibility of separating mixtures of ionic and non-ionic sorbates. Another method allowing solution of this problem is ion-pair chromatog- raphy. These methods have much in common: both use surfac- tants; electrostatic interactions play a substantial role (though if in ion-pair chromatography this role is crucial and determines retention, in MLC it is secondary), while organic modifiers regulate the eluting strength and selectivity of separation. Also, there are differences in principle between these two methods.Micellar solutions are considered in MLC as macroscopically homogeneous systems composed of two phases: the micelles and the bulk of aqueous solution. In the ion-pair chromatography, the mobile phases are homogeneous and consist solely of surfactant EMBasova, VMIvanov, O A Shpigun monomers. Furthermore, the amount of surfactant adsorbed on the stationary phase increases with the concentration of the ion- pairing agent in the mobile phase. Descriptions of MLC by different authors are contradictory: some believe that the amount of adsorbed surfactant is almost constant after it exceeds CMC, while others argue that the surfactant adsorption remains sub- stantial at concentrations several tens of times that of CMC. Thorough comparison of the possibilities of both methods was carried out in the example of separation of amino acids and peptides.66, 146 The separation was performed using 2-propanol- modified sodium dodecyl sulfate-containing mobile phase.Inves- tigation of the dependence of retention on the alcohol concen- tration showed that in the case of ion-pair chromatography the parameter of the solvent strength S correlates well with the logarithm of capacity factor in a purely aqueous phase (ln k 00). The selectivity of separation drops with the increasing content of the organic solvent in the mobile phase. Such dependences are also characteristic of RP HPLC. Totally different regularities were observed in MLC: variation of S proved to be reciprocal to changes in ln k00, whereas the separation selectivity of most sorbate pairs increased as the content of the organic solvent in the mobile phase increased.In MLC and ion-pair chromatography, the surfactant con- centration has different effects on the retention of sorbates. In ion- pair chromatography the retention time of oppositely charged sorbates (at pH 2.5, the amino acids and peptides studied have a partial positive charge) increases with the concentration of the ion-pairing agent due to the larger amount of surfactant adsorbed on the stationary phase surface and forms a plateau upon saturation of the stationary phase. In MLC, the retention time is decreased as the total surfactant concentration in the system increases. In ion-pair chromatography there is no clear-cut dependence of selectivity on the ion-pair reagent concentration; though it is usually increased with the increase in concentration, particularly when sorbates differ in their charges. In MLC, the selectivity is as a rule enhanced as the micelle concentration increases.In ion-pair chromatography, the eluting strength of the mobile phase decreases as the ion-pairing agent concentration increases, whereas in MLC it increases with the concentration of micelles. Thus, the enhancement of selectivity and eluting strength occurs in parallel in MLC and antiparallel in ion-pair chromatog- raphy. Many investigators noted better reproducibility of the reten- tion characteristics in MLC compared to that in ion-pair chroma- tography.In the latter case, one often has to operate in the range where the retention is strongly dependent on the ion-pairing agent concentration; therefore, even minor changes in the mobile phase composition lead to a fairly appreciable variation of retention. Both methods were shown to have lower column efficiency compared to that in RP HPLC. Thus a new 25-cm long column packed with a C18 sorbent had an efficiency of 14 000 theoretical plates in the elution with an acetonitrile ± water mixture (25 : 75). In the presence of sodium dodecyl sulfate, its efficiency dropped to 5500 theoretical plates in both MLC and ion-pair chromatogra- phy. Nonetheless, the ion-pair chromatography has an advantage compared to MLC as regards the efficiency owing to the variation of the hydrophobic nature of the ion-pairing agent. It should be noted in conclusion that MLC is an interesting method with the unique possibilities of selectivity variation; its theory has been well developed.The micellar liquid chromatog- raphy is an alternative to the RP HPLC method. Analytical chemists consider this method as the most useful in the analysis of biological specimens. References 1. V D Shatts, O V Sakhartova Vysokoeffektivnaya Zhidkostnaya Khromatografiya (High-Performance Liquid Chromatography) (Riga: Zinatne, 1988)Micellar liquid chromatography 2. J G Dorsey Adv. Chromatogr. 27 167 (1987) 3. M G Khaledi Biochromatography 3 20 (1988) 4. M G Khaledi Trends Anal. Chem. 7 293 (1988) 5. D W Armstrong, S J Henry J.Liquid Chromatogr. 3 657 (1980) 6. K Saitoh Kagaku 45 884 (1990); Ref. Zh. Khim. 1 B 2713 (1992) 7. J G Dorsey,W T Cooper, J F Wheeler, H G Barth, J P Foley Anal. Chem. 66 500R (1994) 8. M J Medina Hernandez, M C Garcia Alvares-Coque Analyst 117 831 ( 1992) 9. M L Marina,M A Garcia J. Liquid Chromatogr. 17 957 (1994) 10. D W Armstrong, F Nome Anal. Chem. 53 1662 (1981) 11. M Arunyanart, L J Cline Love Anal. Chem. 56 1557 (1984) 12. J P Foley Anal. Chim. Acta 231 237 (1990) 13. D W Armstrong Sep. Purif. Methods 14 213 (1985) 14. J P Foley,W E May Anal. Chem. 59 102 (1987) 15. J P Foley,W E May Anal. Chem. 59 110 (1987) 16. M F Borgerding, F H Quina,W L Hinze, J Bowermaster, H M McNair Anal. Chem. 60 2520 (1988) 17. A Berthod, I Girard, C Gonnet Anal.Chem. 58 1359 (1986) 18. ML Marina, S Vera, A R Rodriguez Chromatographia 28 379 (1989) 19. M G Khaledi, E Peuler, J Ngeh-Ngwainbi Anal. Chem. 59 2738 (1987) 20. M A Garcia, S Vera,M L Marina Chromarographia 32 148 (1991) 21. S B Savvin, R K Chernova, S N Shtykov Poverkhnostno-Aktivnye Veshchestva (Surfactants) (Moscow: Nauka, 1991) 22. A I Rusanov Mitselloobrazovanie v Rastvorakh Poverkhnostno- Aktivnykh Veshchestv (Micelle Formation in Surfactant Solutions) (St.-Petersburg: Khimiya, 1992) 23. W Hu, T Takeuchi, H Haraguchi Chromatographia 33 58 (1992) 24. W Hu, T Takeuchi, H Haraguchi Chromatographia 33 63 (1992) 25. A H Rodgers, J K Strasters, M G Khaledi J. Chromatogr. 636 203 (1993) 26. A H Rodgers, M G Khaledi Anal.Chem. 66 327 (1994) 27. S A Zibas Anal. Chim. Acta 299 17 (1994) 28. G P Carmelo Analyst 120 53 (1995) 29. T Okada Anal. Chem. 64 589 (1992) 30. T Okada Anal. Chem. 60 1551 (1988) 31. T Okada Anal. Chem. 60 2116 (1988) 32. T Okada J. Chromatogr. 538 341 (1991) 33. T Okada, H Shimizu J. Chromatogr. A 706 37 (1995) 34. J H Knox, R A Hartwick J. Chromatogr. 204 3 (1981) 35. A Berthod, M F Borgerding,W L Hinze J. Chromatogr. 556 263 (1991) 36. B Canas-Montalvo, R C Izquierdo-Hornillos J. Liquid Chromatogr. 17 1464 (1994) 37. M F Borgerding, W L Hinze, L D Stafford, G W Fulp Jr., W C Hamlin Jr. Anal. Chem. 61 1353 (1989) 38. W Hu, P R Haddad Anal. Commun. 35 191 (1998) 39. J G Dorsey,M T DeEchegaray, J S Landy Anal. Chem.55 924 (1983) 40. P Yarmchuk, R Weinberger, R F Hirsch, L J Cline Love J. Chromatogr. 283 47 (1984) 41. J S Landy, J G Dorsey Anal. Chim. Acta 178 179 (1985) 42. M F Borgerding, R L Williams Jr , W L Hinze, F H Quina J. Liquid Chromatogr. 12 1367 (1989) 43. M F Borgerding, W L Hinze Anal. Chem. 57 2183 (1985) 44. D W Armstrong, T J Ward, A Berthod Anal. Chem. 58 579 (1986) 45. R Bailey, P M Cassidy Anal. Chem. 64 2277 (1992) 46. A Berthod, A Roussel J. Chromatogr. 449 349 (1988) 47. M G Khaledi, J K Strasters, A H Rogers, E D Breyer Anal. Chem. 62 130 (1990) 48. A J I Ward, J H Han, O Donghue, B K Lavine, in Pittsburgh Conference and Expo on Analytical Chemistry and Applied Spectro- scopy (Abstracts of Reports), New York, 1990 p. 111 49.R M Villanueva Camanas, J M Sanchis Mallols, J R Torres Lapasio, G Ramis-Ramos Analyst 120 1767 (1995) 50. P Menendez Fraga, E Blanco Gonzalez, A Sanz-Medel Anal. Chim. Acta 212 181 (1988) 51. F Palmisano, A Guerrieri, P G Zambonin, T R I Cataldi Anal. Chem. 61 946 (1989) 52. B K Lavine, S Hendayana J. Liquid Chromatogr. Relat. Technol. 19 101 (1996) 53. M A Rodriguez Delgado,M J Sanchez, V Gonzalez, F Garcia Montelongo J. Liquid Chromatogr. Relat. Technol. 19 187 (1996) 999 54. J P Foley, J G Dorsey Anal. Chem. 55 730 (1983) 55. J M Saz,M L Marina J. Chromatogr. A 687 1 (1994) 56. M A Garcia,M L Marina J. Liquid Chromatogr. Relat. Technol. 19 1757 (1996) 57. L S Madamba-Tan, J K Strasters,M G Khaledi J. Chromatogr. A 683 321(1994) 58.J K Strasters, E D Breyer, A H Rodgers,M G Khaledi J. Chromatogr. 511 17 (1990) 59. J K Strasters, S T Kim,M G Khaledi J. Chromatogr. 586 221 (1991) 60. A S Kord, M G Khaledi Anal. Chem. 64 1894 (1992) 61. W L Hinze, Z Fu, F S Sadek, in Pittsburgh Conference and Expo on Analytical Chemistry and Applied Spectroscopy (Abstracts of Reports), Atlantic City, 1987 p. 578 62. M A Garcia, S Vera,M Bombin,M L Marina J. Chromatogr. 646 297 (1993) 63. F P Tomasella, J Fett, L J Cline Love Anal. Chem. 63 474 (1991) 64. M A Rodriguez Delgado,M J Sanchez, V Gonzalez, F Garcia Montelongo Chromatographia 38 342 (1994) 65. M A Rodriguez Delgado,M J Sanchez, V Gonzalez, F Garcia Montelongo Anal. Chim. Acta 298 423 (1994) 66. M G Khaledi Anal.Chem. 60 876 (1988) 67. A S Kord, M G Khaledi J. Chromatogr. 631 125 (1993) 68. O Jimenez, I Benito,M L Marina Anal. Chim. Acta 353 367 (1997) 69. O Jimenez,M A Garcia,M L Marina J. Liquid Chromatogr. Relat. Technol. 20 731 (1997) 70. M C Garcia Alvarez-Coque, J P Torres Lapasio, J J Baeza Baeza Anal. Chim. Acta 324 163 (1996) 71. L S Madamba-Tan, J K Strasters,M G Khaledi J. Chromatogr. A 683 335 (1994) 72. J R Torres Lapasio, R M Villanueva Camanas, J M Sanchis- Mallols, M J Medina Hernandez,M C Garcia-Alvarez-Coque J. Chromatogr. 639 87 (1993) 73. M D Rukhadze, G S Bezarashvili, M V Sebiskeradze, V R Meyer Anal. Chim. Acta 356 19 (1997) 74. B K Lavine, S Hendayana, J Tetreault Anal. Chem. 66 3458 (1994) 75. B K Lavine,W T Cooper, Y He, S Hendayana, J H Han, J Tetreault J.Colloid Interface Sci. 165 497 (1994) 76. B K Lavine, A J White, J H Han J. Chromatogr. 542 29 (1991) 77. A L Underwood Anal. Chim. Acta 140 89 (1982) 78. A Sanz-Medel, B Aizpun, J M Marchante, E Segovia, M L Fernandez, E Blanco J. Chromatogr. A 683 233 (1994) 79. M G Khaledi, J K Strasters, S C Smith Anal. Chem. 63 1820 (1991) 80. S Yang,M G Khaledi, in Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy (Abstracts of Reports), Chicago, 1994 p. 1224 81. A Berthod, I Girard, C Gonnet Anal. Chem. 58 1356 (1986) 82. B K Lavine, S Hendayana, W T Cooper, Y He J. Liquid Chromatogr. Relat. Technol. 20 351 (1997) 83. B K Lavine, S Hendayana, W T Cooper, Y He J. Liquid Chromatogr.Relat. Technol. 20 377 (1997) 84. P R Bedard, M L Cotton J. Liquid Chromatogr. 14 1113 (1991) 85. F G P Mullins, G F Kirkbright Analyst 109 1217 (1984) 86. A Berthod, I Girard, C Gonnet Anal. Chem. 58 1362 (1986) 87. M G Khaledi, E D Breyer Anal. Chem. 61 1040 (1989) 88. F Gago, J Alvarez-Builla, J Elguero, J C Diez-Maso Anal. Chem. 59 921 (1987) 89. V Gonzalez,M A Rodriguez Delgado,M J Sanchez, F Garcia- Montelongo Chromatographia 34 627 (1992) 90. K T Valsaraj, L J Thribodeaux Sep. Sci. Technol. 25 369 (1990) 91. M A Garcia,ML Marina J. Chromatogr. A 687 233 (1994) 92. I Benito, J M Saz,M L Marina J. Liquid Chromatogr. Relat. Technol. 21 331 (1998) 93. R Graglia, E Pelizzetti Anal. Chim. Acta 212 171 (1988) 94. P Jandera J. Chromatogr. 314 13 (1984) 95. P Jandera Chromatographia 19 101 (1984) 96. W L Hinze, S G Weber Anal. Chem. 63 1808 (1991) 97. M A Rodriguez Delgado,M J Sanchez, V Gonzalez, F Garcia- Montelongo Fr. J. Anal. Chem. 345 748 (1993) 98. M A Rodriguez Delgado,M J Sanchez, V Gonzalez, F Garcia- Montelongo J. Chromatogr. A 697 71 (1995) 99. S Yang,M G Khaledi J. Chromatogr. A 692 301 (1995) 100. H Zou, Y Zhang, P Lu Anal. Chim. Acta 310 461 (1995)EMBasova, VMIvanov, O A Shpigun 1000 101. M H Guermouche,D Habel, S Guermouche Fluid Phase Equil. 147 301 (1998) 102. E D Breyer, J K Strasters,M G Khaledi Anal. Chem. 63 828 (1991) 103. L Escuder-Gilabert, S Sagrado, R M Villanueva Camanas, M J Medina Hernandez Anal. Chem. 70 28 (1998) 104. F P Tomasella, L J C Love Anal. Chem. 62 1315 (1990) 142. O A Shpigun, Yu A Zolotov Ionnaya Khromatografiya i ee Prime- nenie v Analize Vod (Ion Chromatography and Its Application for Analysis of Waters) (Moscow: Moscow State University, 1990) 143. D W Armstrong, G Y Stine Anal. Chem. 55 2317 (1983) 144. M Arunyanart, L J Cline Love J. Chromatogr. 342 293 (1985) 145. J P Berry,W G Weber J. Chromatogr. Sci. 25 307 (1985) 146. M G Khaledi, in Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, (Abstracts of Reports), Atlanta, 1993 p. 1000 a�Russ. J. Anal. Chem. (Engl. Transl.) 105. J P Torres Lapasio, J J Baeza Baeza,M C Garcia Alvares-Coque J. Liquid Chromatogr. Relat. Technol. 19 1205 (1996) 106. M Amin, K Harrington, R von Wandruszka Anal. Chem. 65 2346 (1993) 107. I Carretero, A Maldonado, J J Laserna, E Bonet Domingo, G Ramis Ramos Anal. Chim. Acta 259 203 (1992) 108. W Peng, T Li, E Wang Anal. Chim. Acta 298 415 (1994) 109. Y Martin Bioska, J J Baeza Baeza, G Ramis Ramos J. Pharm. Belg. 50 362 (1995) 110. S Carda Broch,M C Garcia Alvares-Coque, E F Simo Alfonso, J S Esteve Romero Anal. Chim. Acta 353 215 (1997) 111. I Perez-Martinez, S Sagrado,M J Medina Hernandez Anal. Chim. Acta 304 195 (1995) 112. S B Thomas, S J Albazi J. Liquid Chromatogr. Relat. Technol. 19 977 (1996) 113. S Yang,M G Khaledi J. Chromatogr. A 692 311 (1995) 114. E Bonet Domingo,M J Medina Hernandez, M C Garcia Alvares- Coque J. Pharm. Biomed. Anal. 11 711 (1993) 115. E Bonet Domingo,M J Medina Hernandez, G Ramis Ramos, M C Garcia Alvares-Coque Analyst 117 843 (1992) 116. S Torres Cartas, M C Garcia Alvares-Coque, R M Villanueva Camanas Anal. Chim. Acta 302 163 (1995) 117. M Catala-Icardo, M J Medina Hernandez,M C Garcia Alvarez- Coque J. Liquid Chromatogr. 18 2827 (1995) 118. S Carda Broch, J S Esteve-Romero, M C Garcia Alvares-Coque Analyst 123 301 (1998) 119. I Rapado-Martinez, M C Garcia Alvares-Coque, R M Villanueva Camanas Analyst 121 1677 (1996) 120. I Perez-Martinez, S Sagrado,M J Medina Hernandez J. Liquid Chromatogr. Relat. Technol. 19 1957 (1996) 121. J H Aiken, C W Huie, J A Terzian J. Chromatogr. Biomed. Appl. 584 181 (1992) 122. I Carretero, J M Vadillo, J J Laserna Analyst 120 1729 (1995) 123. D Haupt, G Pettersson, D Westerlund J. Biochem. Biophys. Methods 25 273 (1992) 124. Y Qu, P Hu, P Zhu J. Liquid Chromatogr. 14 2755 (1991) 125. A G Mwalupindi, I M Warner Anal. Chim. Acta 306 49 (1995) 126. M G Khaledi, J G Dorsey Anal. Chem. 57 2190 (1985) 127. J Miura Fr. J. Anal. Chem. 344 294 (1992) 128. J Miura, M Yoshitome, I Goto, Y Nakamura, H Watanabe Anal. 129. M P San Andres, S Vera, M L Marina J. Chromatogr. A 685 271 130. M Sato, H Yoshimura, H Obi, S-i Hatakeyama, E Kaneko, 131. S Muralidharan, H Ma, H Freiser, in Pittsburgh Conference on Sci. 9 255 (1993) (1994) H Hoshino, T Yotsuyanagi Chem. Lett. 203 (1996) Analytical Chemistry and Applied Spectroscopy (Abstracts oReports), New Orleans, 1998 p. 1004 132. N-Q Li, M-H Zhu Acta Chim. Sinica 52 186 (1994); Ref. Zh. Khim. 21 B 2416 (1994) 133. N-Q Li, M-H Zhu Acta Chim. Sinica, 52 12 (1994); Ref. Zh. Khim. 10 B 2336 (1995) 134. H Ding, J Wang, J G Dorsey, J A Caruso J. Chromatogr. A 694 425 (1995) 135. A Sanz-Medel, M R Fernandez de la Campa,M L Fernandez, R Pereiro, B Fairman ICP Inf. Newslett. 21 419 (1995) 136. E Segova, J M Marchante, J M Gonzalez,M L Fernandez, J E Sanchez-Uria, A Sanz-Medel ICP Inf. Newslett. 21 724 (1996) 137. M P Dziewatkoski, in Pittsburgh Conference on Analytical Chem- istry and Appied Spectroscopy (Abstracts of Reports), New Orleans, 1998 p. 1089 138. Z Mester, P Fodor J. Chromatogr. A 756 292 (1996) 139. D T Heitkemper, J Creed, J A Caruso, F L Fricke Annu. Rep. 140. I P Alimarin, E M Basova, T A Bol'shova, V M Ivanov Zh. Anal. 141. E M Basova, T A Bol'shova,O A Shpigun, V M Ivanov Zh. Anal. Anal. At. Spectrosc. 4 279 (1989) Khim. 45 1478 (1990) a Khim. 48 1094 (1993) a

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (12) 1001 ± 1020 (1999) Reactions of adamantanes in electrophilic media I K Moiseev, N V Makarova,MN Zemtsova Contents I. Introduction II. Substitution reactions in the adamantane series in electrophilic media III. Halogenoadamantanes IV. Synthesis of hydroxyadamantanes and their nitrates V. Synthesis of adamantanecarboxylic acids VI. Aminoadamantanes VII. Aryladamantanes Abstract. Published data on the synthesis of adamantane deriva- tives bearing various functional groups (halogen, hydroxy, nitrato, amino, carboxy and others) are systematised and gener- alised. The bibliography includes 236 references. I. Introduction Synthesis, reactivity and application of adamantane derivatives have been considered in several monographs and reviews.1 ±4 These compounds are used both in the synthesis of pharmaceut- icals 5, 6 and for the preparation of novel polymers.7, 8 It was found that introduction of an adamantyl fragment into medical com- pounds enhances their biological activities owing to the higher fat- solubility.6, 9 Some adamantane derivatives exhibit antiviral activities, e.g., 1-aminoadamantane and 1-ethylaminoadaman- tane hydrochlorides, which are the active principles of antiviral Mydantan and Remantadinum manufactured by the pharma- ceutical plant in Olaine, Latvia.Nucleophilic substitution reactions of hydrogen are most widely used in the synthesis of adamantane derivatives. Never- theless, no special consideration of the reactions of adamantane and its derivatives in electrophilic media has been given before even in the most comprehensive monographs.10, 11 The present review systematises and generalises published data on the synthesis of functionally substituted adamantanes using reactions in electrophilic media.II. Substitution reactions in the adamantane series in electrophilic media Hydride transfer reactions in organic chemistry are most compre- hensively revealed in monographs.12, 13 In these reactions, alumi- nium halides, BF3, sulfuric acid, carbenium ions, protonated quinones and azobenzenes and some other compounds are commonly used as acceptors of the hydride ion. I K Moiseev, N V Makarova,MN Zemtsova Samara State Technical University, Galaktionovskaya ul. 141, Samara 443010. Fax (7-846) 232 21 22 (I K Moiseev).Tel. (7-846) 299 67 75 (M N Zemtsova) Received 31 March 1999 Uspekhi Khimii 68 (12) 1102 ± 1121 (1999); translated by S V Chapyshev #1999 Russian Academy of Sciences and Turpion Ltd UDC 547.51 1001 1001 1004 1011 1013 1014 1016 The reactions of adamantane in electrophilic media should be considered as nucleophilic substitution at the saturated carbon atom.10, 11 These reactions for adamantane and its derivatives can follow only the SN1 mechanism, i.e., via adamantyl cations, which then react with nucleophiles.14 ± 22 The reaction rates are mostly controlled by electronic factors, viz., electron-donating properties of substituents, which serve as `internal' nucleophiles thus delocalising the positive charge on the central carbon atom, electron-withdrawing properties of the substituents in the leaving groups, which decrease the basicity of these groups and electro- philic properties of a medium.Only a few examples describing the nucleophilic substitution of hydrogen are known. These data mainly relate to the substitu- tion of the hydride ion in polyhedranes (adamantane, proto- and homoadamantane, diamantane, bicyclo[3.3.1]nonane, etc.). The nucleophilic substitution of hydrogen in adamantane (1) has successfully been performed in electrophilic media (sulfuric acid, nitric acid, bromine), which, on the one hand, facilitated the hydride ion abstraction and, on the other hand, served as sources of nucleophilic species (HSO¡4 , ONO¡2 and Br7).Thus the bromination of adamantane (1) in liquid bromine occurs selectively at the bridgehead carbon atom yielding only monobromoadamantane 2.14 ± 16 The reaction mechanism presumably involves the intermedi- ate formation of the adamantyl cation 3.16 Br+ Br2 + Br 7HBr 7Br7 2 3 1 The reaction of adamantane (1) with sulfuric acid leads to adamantanone (7).17 ± 19 The reaction scheme includes the oxida- tion of 1 with 96% sulfuric acid to 1-hydroxyadamantane (4), the generation of the adamantyl cation 3 and the 1,2-hydrogen shift resulting in an equilibrium between 1-hydroxyadamantane (4) and 2-hydroxyadamantane (6) and the oxidation of 2-hydroxy- adamantane (6) to adamantanone (7). The latter can also be formed, along with adamantane 1, due to the intermolecular transfer of the hydride ion from 2-hydroxyadamantane (6) to the cation 3.+ H+ H2SO4 H2O OH + 1 H2O H+ 5 3 41002 OH O H2SO4 7 6 OH 7+1 + + 6 3 The reaction of 1-methyladamantane (8) with 96% sulfuric acid at 80oC leads to a complex mixture of products in which 5-methyladamantan-2-one (9) is the major component.19 O H2SO4, 80 8C Me Me 8 9 (78%) In systems containing a strong mineral acid (H2SO4, HNO3) and trifluoroacetic anhydride, the role of the hydride ion acceptor is probably played by the cations HSOá3 or NOá2 , which are generated in situ from the corresponding acids and weakly nucleophilic trifluoroacetic anhydride.20 Thus the reactions of adamantane (1) and aryladamantanes 10a ± e in such systems yield the corresponding trifluoroacetoxy-substituted adamantanes 11a ± e.20 H2SO4, HNO3 H OCOCF3 R R (CF3CO)2O 11a ± e 10a ± e R=4-NO2C6H4 (a), 4-MeSO2C6H4 (b), 3-NO2-4-MeC6H3 (c), 4-BrC6H4 (d), 4-MeO2CC6H4 (e).The reaction of adamantane (1) with concentrated nitric acid affords 1-adamantyl nitrate (12) and adamantane-1,3-diyl dini- trate (13).21 3HNO3 NOá +H3O++2ONO¡22 ONO¡2ONO2 NOá23 1 12+ ONO2H ONO2 ONO2 HNO2 HONO2 7H+ HNO3 ONO2 H+ 13 A mechanism of the formation of the nitrate 12 involving the stage of generation of the adamantyl cation 3 has been sug- gested 22, 23 based on studies of the behaviour of adamantane (1) in nitric acid and its mixtures with acetic acid.Exploration of the behaviour of adamantane (1) in electro- philic media made possible the development of the methods of synthesis of 1-bromoadamantane (2), 1-hydroxyadamantane (4) and nitrate 12, which are convenient starting materials for the preparation of various functionally substituted adamantanes. It should be noted that the latter of these three compounds is the most readily available. The functionalisation of adamantane has been achieved using sulfuric and nitric acids 24 or bromine as electrophilic media. The reactions of adamantane (1) and its derivatives 14 with nucleophiles such as urea, urethane or nitriles in concentrated nitric acid or in a mixture ofHNO3 and AcOH afford functionally substituted adamantanes 15.24 I K Moiseev, N V Makarova,MN Zemtsova O HNO3 +R2X R1 NHCR2 R1 AcOH 15 14 R1=H, Me, Et: X=NH2CO; R2=NH2, OEt; X=CN, R2=(CH2)2OMe.It has been shown 18, 19 that functionalisation of adamantane and its derivatives can be performed in liquid bromine, which also functions as a Lewis acid. As an example, synthesis of compounds 16 ± 18 using this method can be mentioned. By comparison with adamantane (1), bromoadamantane (2) reacts faster in these reactions.25, 26 R2CN, Br2 NHC(O)R2 16 (R2=Ph, Me) MeOH, Br2 R1 OMe 17 R1 = H (1), Br (2) R3CO2H, Br2 CO2R3 18 (R3=H, Me) Yurchenko et al.27, 28 have shown that adamantane (1) reacts with iodine monobromide in CCl4 to give a mixture of 1-bromo- adamantane (2) and 1,3-dibromoadamantane (19).d+ d7 IBr Br7 I Br Br Br Br 3 1 7HI,7Br7 19 2 The authors believe that a dipole Id+±Brd7 with a partial positive charge on the iodine atom is responsible for the hydride ion abstraction from the adamantane molecule. Adamantyl cation 3 thus formed then reacts with the bromide anion. The dibromide 19 is formed from the bromide 2 according to the same mecha- nism. UV-Spectral studies 29 showed that 1-bromoadamantane (2), 1-bromo-3-methyladamantane (20), 1-bromo-3,5-dimethylada- mantane (21) and 1-bromo-3,5,7-trimethyladamantane (22) form molecular complexes with bromine, which have linear structures. Presumably, complexes of the p,s-type with charge transfer from the lone pair of the bromine atom in bromoadamantanes to the molecule of Br2 are realised in these cases.As a result, the polarisation of the C± Br bond is increased. Me Me Me Br Me Br Me Br 21 20 22 Me It is supposed 30 that the increase in the polarity of the C± Br bond in bromoadamantanes is associated with localisation of the positive charge in the centre of the adamantane cage rather than on the carbon atom of the C± Br bond. This is possible owing to the overlap of the inwardly directed, hybridised orbitals of the bridgehead carbon atoms in adamantanes. Complexation facili- tates nucleophilic reactions at the carbon atom of 1-bromo- adamantane (2). It has also been shown 22 that the reactions that occur efficiently in liquid bromine, occur also in neutral solvents in the presence of molecular bromine, if its concentration is sufficient for the complexation.The kinetics of the reactions of adamantane (1) and its derivatives and homologues with nitric acid in dichloromethaneReactions of adamantanes in electrophilic media Table 1. The rate constants for the reactions of adamantane derivatives with nitric acid in dichloromethane a (Ref. 31). Compound (1) Me (8) Me (23) Me Me Me (24) MeD Me Me (25) Me CH2CO2Me (26) CO2Me (27) ONO2 (12) C(NO2)3 (28) a Initial concentration of HNO3 was 4 mol litre71. have been studied.31 Sequential methylation of the bridgehead positions of adamantane results in monotonic decrease in reac- tivity (Table 1), which is typical of the reactions involving the intermediacy of carbocations.The logarithms of the rate con- stants (keff) for these reactions calculated taking into account the statistical factor correlate fairly well (r=0.993) with the loga- rithms of the rate constants for the solvolysis of the corresponding bromides (in 80% aqueous ethanol at 75 8C),32 which suggests that both types of reactions have similar mechanisms. Introduc- tion of electron-withdrawing substituents into adamantane decreases the reactivities of its derivatives (Table 1). A correla- tion between the logarithms of the rate constants and the s* Taft constants is observed (r=0.99, r*=70.77) for adamantanes substituted by the CH2CO2Me, CO2Me, ONO2 and C(NO2)3 groups. The sign of the reaction constant indicates that these substituents destabilise the positively charged intermediate spe- cies.The rather high absolute value of the r* constant suggests that transition states in the reactions of adamantanes with nitric acid are highly polar. The kinetic isotope effect for the reactions of 1,3,5-trimethyl- adamantane (24) and 7-deuterio-1,3,5-trimethyladamantane (25) is 4.40.3.31 This suggests that the dissociation of the C7H bond at the tertiary carbon atom is the rate-determining step. It is worth mentioning that the kinetic isotope effect for the reactions of adamantane (1) with nitronium salts is 1.86 (Ref. 33) and that for Temperature Concentra- /8C tion /mol litre71 0.025 0 0.025 0 0.025 0 0.025 0 0.05 0.05 0.05 0.05 0 10 20 30 0.05 20 0.05 20 0.025 20 0.025 20 0.025 20 103 keff /s71 15.00.8 8.70.4 4.60.2 1.780.09 1.980.09 4.20.1 8.00.4 1.210.05 1.810.1 412 3.90.4 0.240.01 0.0410.002 1003 the oxidation of isobutane with a mixture ofH2SO4 andHNO3 is 2 (Ref.34). The activation energy and entropy for the reaction of 1,3,5- trimethyladamantane (24) are equal to 437 kJ mol71 and 14724 J mol71 K71, respectively. Judging from these data, it is reasonable to assume that the reaction of adamantane (1) with nitric acid involves the inter- mediate formation of the adamantyl cation 3. The rate-determin- ing step of this process is obviously the abstraction of the hydride ion from the tertiary carbon atom by an electrophile, which is most likely the nitronium ion.The adamantyl cation 3 thus formed is further stabilised by adding a nucleophile.31 The reactions of saturated hydrocarbons with sulfuric or nitric acids or their mixtures are well known. The mechanism of activation of the C7H bonds in adamantane or, in the other words, the mechanism of the hydride ion transfer is discussed in Refs 35 ± 37. In adamantane and its homologues, this activation can be realised by three routes. Route a Nu 3 7HE H + E Nu E+ 1 E A 7H+ Route b Nu E+ 1 3+HE Nu Route c Nu 7e 7e 3 1 Nu H+. 7H+ B Of these three, no route can be ruled out from consideration, since each of them leads to the adamantyl cation 3. Route a implies realisation of a transition state A with a five-coordinate carbon atom.Such a transition state is used 35 to describe hydride ion transfer in the reactions of alkanes with carbocations in sulfuric acid or with nitronium salts. The nitration of alkanes according to Olah,38 viz., with nitronium salts in a mixture of dichloromethane and sulfolane at 25 8C in the dark in the absence of moisture, giving rise to a mixture of nitroalkanes and nucleophilic substitu- tion products also invokes the realisation of the transition state with a five-coordinate carbon atom. However, the question whether the transition state A is converted into the final products by proton abstraction or by the elimination of HNO2 and subsequent reaction of the cation 3 with a nucleophile still remains open. The route b has already been considered in detail for the reactions of adamantane (1) in such electrophilic media as liquid bromine, sulfuric and nitric acids.In these cases, the initially formed adamantyl cation 3 reacts further with HSO¡4 , ONO¡2 , Br7or other nucleophilic species. The reactivity of adamantane in these electrophilic media decreases in the following order: H2SO4>HNO3>Br2. The reactions of adamantane (1) through a stage of single- electron transfer (route c) are also possible. However, this is considered in a limited number of publications (see, for example, Ref. 36). The anodic oxidation of alkyladamantanes in dry acetonitrile yields 1-acetamido-3-alkyladamantane 29, while the anodic frag- mentation of the same compounds leads to 1-acetamidoadaman- tane (30).The ratio of the reaction products depends on the nature of the alkyl substituents.361004 NHCOMe R +.MeCN 7e 29 R R NHCOMe 30 R=H, Et, Pri, But. Later, the decomposition of 1,10-azoadamantane (31) into adamantyl radical B involving the stage of the formation of [AdN=NAd]+. (Ad is 1-adamantyl) (route c) was described.37 Thus the reaction of 1,10-azoadamantane (31) with thianthre- niumyl perchlorate in acetonitrile affords 1-acetamido- adamantane (30) in 90% yield. [AdN NAd]+. Th+ Th+.+AdN NAd 31 [AdN NAd]+. Ad.+Ad++N2 3 B Th+Ad+ Th+.+Ad. + Ad++MeCN AdN CMe + H2O AdN CMe 7H+ AdNHCOMe 30 S ClO¡4 .Th+.= +. S III. Halogenoadamantanes Halogenoadamantanes can be obtained from both adamantane itself and its derivatives bearing hydroxy, nitroxy and other nitrogen-containing functional groups or halogen atoms. After the first reports 14, 15 on the synthesis of halogenoada- mantanes from adamantane (1), several publications appeared in the literature, describing methods of introduction of fluorine, chlorine, bromine and iodine atoms into the adamantane mole- cule, which involve the intermediate formation of the adamantyl cation 3. Using these methods, four halogen atoms can be introduced in the bridgehead positions of adamantane. 1-Fluoroadamantane (33) has been obtained by the reaction of adamantane (1) with a mixture of poly(hydrogen fluoride) and pyridine (30% Py, 70% HF).39 The fluorination is carried out in the presence of nitronium tetrafluoroborate.The complex 32 with the five-coordinate carbon is assumed to be formed in the attack of the nitronium cation on adamantane, which then decomposes to give 1-fluoroadamantane (33). F + H NO2 NOá2 BF¡4BF¡4 33 1 32 1-Fluoroadamantane (33) can also be prepared by the reaction of adamantane (1) with xenon fluoride 40 or with ClF3.41 In the latter case, 1,3-difluoro- and 1,3,5-trifluoro-adamantanes were formed together with the 1-fluoroadamantane (33). The reaction presumably involves the intermediacy of a complex with five- coordinate carbon and the adamantyl cation.41 + H F7 33 3 1+ClF3 7F7,7ClF 7HF F I K Moiseev, N V Makarova,MN Zemtsova Sulfur tetrafluoride reacts with adamantane in the presence of sulfur monochloride to give a mixture of 1-fluoro- (33) and 1,3- difluoro-adamantanes (34).42 F 100 8C, 6 h 1+SF4+S2Cl2 F+33 34 The fluorination of 1,3-dimethyladamantane (23) over CoF3 affords a mixture of perfluorotrimethylbicyclo[3.3.1]nonane and perfluoro-1,3-dimethyladamantane (5% ± 10%).43 The fluorina- tion of 1,3-difluoro-5,7-dimethyladamantane, prepared by the reaction of 1,3-dihydroxy-5,7-dimethyladamantane with SF4 (125 8C, 4 h), over CoF3 also leads to a mixture of perfluorinated derivatives.43 1-Fluoro- (33), 1,3-difluoro- (34), 1,3,5-trifluoro- and 1,3,5,7- tetrafluoro-adamantanes have been obtained by the action of sulfur tetrafluoride on adamantane (1).44 Much more attention is paid to the problem of the adaman- tane chlorination.Molecular chlorine, organic and mineral acid chlorides, AlCl3, FeCl3 and others are used as chlorinating agents. 1-Chloroadamantane (35) can be obtained either by the treatment of adamantane (1) with a mixture of iodine and sulfuric acid in dichloromethane followed by passing of molecular chlorine through the reaction mixture 45 or by the reaction of adamantane (1) with molecular chlorine adsorbed on SiO2.46 1) I2, H2SO4, CH2Cl2 2) Cl2 1 Cl 35 Acyl chlorides RCOCl (R=Me, Prn, Pri, Ph, cyclohexyl) in the presence of AlCl3 are also rather efficient chlorinating agents. Thus the reaction of adamantane (1) with acetyl chloride yields 1,3-dichloroadamantane (36).47 MeCOáAlCl¡ MeCOCl +AlCl3 4 1+MeCO+ 3+MeCHO Cl 3+AlCl¡4 Cl 7AlCl3 36 1,3,5-Trichloroadamantane (37) is obtained by the reaction of adamantane (1) with thionyl chloride in the presence of AlCl3 at 75 8C.48 Cl AlCl3 Cl Cl 1+SOCl2 37 In this reaction, adamantane sulfinyl chloride (38) is formed as the minor product probably due to the nucleophilic attack by the lone electron pair of the sulfur atom of thionyl chloride on the carbenium ion 3.Cl+ SOCl2 3 1 SOCl 7HCl 7Cl+ 38 Adamantane (1) reacts with sulfuryl chloride in sulfolane at 60 8C to give 1-chloro- (35) and 2-chloro-adamantanes (39) in 97.5% and 2.5% yield, respectively.49 1-Chloroadamantane (35) is also obtained by the reaction of adamantane with phosphorus trichloride in trifluoroacetic acid.50Reactions of adamantanes in electrophilic media Cl SOCl2 + Cl 39 35 PCl3 1 35 Unlike pure fluorosulfonic acid, which reacts with adaman- tane (1) to give 1-hydroxyadamantane (4),51 chlorosulfonic acid halogenates adamantane at 75 8C [the ratio of reactants 1 : (8 ± 10)] providing 1-chloroadamantane (35) in 60% yield.52 Synthesis of 1-chloro-, 1,3-dichloro- and 1,3,5-trichloro- adamantanes by the reaction of adamantane (1) with chlorosul- fonic acid has been described.53 When the reaction mixture is kept below 5 8C for less than 1 h, 1-chloroadamantane (35) is formed in 18.7% yield.1,3-Dichloroadamantane (36) is obtained in 80.7% yield after 50 ± 60 h, and 1,3,5-trichloroadamantane (37) is obtained in 70% yield after 140 ± 240 h of the reaction.Either 1-chloro-3,5-dimethyl- or 1,3-dichloro-5,7-dimethyladamantane is formed in the chlorination of 1,3-dimethyladamantane (23) depending on the reaction conditions. 1,3,5-Trimethyladaman- tane (24) gives 1-chloro-3,5,7-trimethyladamantane. 1,3,5,7-Tetrachloroadamantane (40) has been obtained by heating 1,3,5-trichloroadamantane (37) with chlorosulfonic acid for 150 h at 70 8C.54 Cl HSO3Cl Cl Cl 37 Cl 40 Chlorination of adamantane (1) and its homologues can be accomplished with alkyl halides in the presence of Lewis acids. Thus the treatment of 1-ethyladamantane (41) with a mixture of tert-butyl bromide or tert-butyl chloride and AlCl3 affords 1-chloro-3-ethyl-, 1,3-dichloro-5-ethyl- and 1,3,5-trichloro-7- ethyl-adamantanes.55 ButCl Et AlCl3 41 Cl Cl Cl Cl Cl + Et Et + Et Cl A mixture of 1,3-dichloro- (36), 1,3-dibromo- (19) and 1-bromo-3-chloro-adamantanes was obtained by refluxing ada- mantane (1) in CCl4 in the presence of boron tribromide and aluminium bromide.56 Table 2.Chlorination of adamantane (1) in the presence of FeCl3 and SbCl5.59, 60 Solvent FeCl3 Tempera- ture /8C Time /h Total yield of 35 and 39 (%) 78 CCl4 6.5 54 40 CH2Cl2 7 10 48 CH2Cl2, CCl4 4 64 ± 72 7 724 7 733 ± 6 8 ± 9 2.5 4.5 7 1005 Cl Br Br Br2, BBr3 1 + Cl Br + AlBr3, CCl4, 1 h 36 19 Cl A mixture of 1-chloro- (35) and 1,3-dichloro-adamantanes (36) was formed in the reaction of adamantane (1) with CCl4 in the presence of AlCl3 at 20 8C.57 1,3,5,7-Tetrachloroadamantane (40) was prepared in 55% yield by refluxing adamantane (1) with AlCl3 in CCl4.58 The reaction of adamantane (1) with AlCl3 in CCl4 for 4 h at 20 8C affords 1,3-dichloroadamantane (41) in 49% yield.59 The effects of solvents and Lewis acids on the chlorination of adamantane (1) have been studied in detail.59, 60 The yields and the ratio of 1-chloro- (35) and 2-chloro-adamantanes (39) obtained in these reactions in the presence of FeCl3 or SbCl5 are listed in Table 2.The yields of 1-chloroadamantane (35) are higher when SbCl5 is used as a catalyst. The ratio of the isomers is influenced by the nature of the solvent and the Lewis acid.The mechanisms of the halogenation reactions of adamantane considered above are probably not purely ionic, since 2-chloro- adamantane (39) can be formed exclusively by the free-radical mechanism. Two different mechanisms can be suggested for the reactions in the presence of FeCl3: the route a involving direct reaction of the radical B with FeCl3 and providing 1-chloroadamantane (35) and the route b implying the intermediacy of the cation 3. a 35 1+FeCl3 Cl7 b 7HCl,7FeCl2 3 35 B (a) FeCl3 (7FeCl2); (b) FeCl3 (7FeCl2,7Cl7). The reaction of adamantane (1) with SbCl5 can also be described by a scheme including a stage of the adamantyl radical B formation. 1+ SbCl5 Ad.+HCl+Sb.Cl4 BAdCl+Sb.Cl4 Ad.+SbCl5 B 1+ Sb.Cl4 Ad.+HCl+SbCl3 AdCl+ SbCl3 Ad.+Sb.Cl4 B The chlorination of adamantane (1) in the presence of AlCl3 and AgSbF6 has been studied by Olah et al.61 A mixture of 1-chloro- (35) and 2-chloro-adamantanes (39) (in 63.2% and 2.9% yields, respectively) is formed when the ratio of the reactants Cl2 : AlCl3 :AdH is 1 : 0.1 : 1 (21 h, 25 8C).If the ratio SbCl3 Time /h Ratio of 35 and 39 Total yield of 35 and 39 (%) Ratio of 35 and 39 (11 ± 15) : 1 (13 ± 15) : 1 (11 ± 16) : 1 31 : 1 (35 ± 39) : 1 (36 ± 42) : 1 22 : 1 21 : 1 22 : 1 7 ± 9 14 ± 20 66 ± 79 24 40 ± 42 68 ± 79 52 75 76 11 : 1 (12 ± 15) : 1 720 : 1 28 : 1 742 : 1 (27 ± 42) : 1 (23 ± 30) : 1 34 360.5 13256.51006 Cl2 : AgSbF6 :AdH is 5 : 1 : 10 (21 h, 715.5 8C), the yields of 1-chloro- (35) and 2-chloro-adamantanes (39) are 62.4% and 1.2%, respectively.In addition, 1-hydroxyadamantane (4) is formed in 36.3% yield. On the other hand, the ratio of 1-chloroadamantane (35) to 2-chloroadamantane (39) in photo- chemical chlorination was 1 : (0.6 ± 2.1).62 It is known that triphenylmethyl cation is an effective acceptor of the hydride ion (i.e., this serves as a `soft' organic acid in terms of the HSAB principle). Using this cation, it is possible to generate carbenium ions from hydrocarbons and functionalise them by the reactions with appropriate nucleophiles. Triphenylmethyl cation can be used in such reactions as a catalyst.In this case, triphenyl- methane derivatives serve as the sources of nucleophiles which react with the newly formed carbenium ions and also regenerate the triphenylmethyl cations. R+X7+Ph3CH, RH+Ph3C+X7 RY+Ph3C+X7 R+X7+Ph3CY X7 is a weakly nucleophilic anion. This type of functionalisation has been studied for various hydrocarbons, including adamantane.63 AdH+Ph3CY AdY+Ph3CH Y=OH, Cl, OMe, O(CH2)2OAc. The reaction was carried out in a mixture of CH2Cl2 and MeNO2 at 70 8C using Ph3COSO2CF3 (0.05 mol per mol of AdH) as a precursor of triphenylmethyl cation. The results obtained are listed in Table 3. Table 3. Functionalisation of adamantane (1) by triphenylmethane derivatives.63 Yields (%) a Ratio Y Time Ph3CY: /h :AdH AdY Ph3CH AdOH AdH (4) (1) 99% purity.69 A complex mixture of products comprising 1,3,5- 80 OH Cl OMe 59(57) 78(40) 31(20) 50(40) b 1492.5 3.5 1 : 1 1 : 1 1.25 : 1 AcO(CH2)2O 1.25 : 1 0 12 15 87 71 40 10 20 10 a Determined by GLC, the isolated yields are given in parentheses; b after deacylation.As can be seen from the Table 3, such functional groups as Cl and OH are readily introduced into adamantane and these exchange reactions can find synthetic application. The yields of the ethers are substantially lower, probably because of compet- itive disproportionation of trityl ethers. Several studies have been devoted to the preparation of bromoadamantanes. As has been mentioned above, 1-bromoada- mantane (2) was synthesised in 78% yield by the reaction of adamantane (1) with bromine in the presence of AlCl3.15 The synthesis of the bromide 2 by the oxidative bromination of adamantane (1) with the stoichiometric amount of bromine or an alkali metal bromide in trifluoroacetic acid in the presence of nitrogen-containing oxidants such as alkali nitrates or nitrites or NO2 has been proposed as an alternative.This affords 1-bromo- adamantane (2) in quantitative yield.64 Anhydrous AgSbF6 in CH2Cl2 catalyses the bromination of adamantane. At 745.5 8C, 1-bromoadamantane (2) is obtained in 54.1% yield.65 A strongly polarised complex of Br2 with AgSbF6 is supposed to be the brominating species in this reaction.65 I K Moiseev, N V Makarova,MN Zemtsova d7 d+ Ag+SbF¡ Br Br Br2+AgSbF6 6 .1-Hydroxy- (4), 2-bromo- (42) and 1-fluoro-adamantanes (33) are also formed in this reaction. Br2 Br 2 H2O + 4 SbF¡6SbF5 33 Br + Br2 AdH SbF¡6 42 A two-electron three-centre transition state of the carbenium type is believed 65 to be realised in this reaction. + H Br R3C Synthesis of 1,3-dibromoadamantane (19) by the reaction of adamantane (1) with bromine in the presence of AlBr3 (78% yield) or in the absence of a catalyst in autoclave at 150 8C for 22 h (12.4% yield) has been described.15, 66 ± 70 A detailed study 66 of the reaction of adamantane (1) with bromine in the presence of a mixture of AlBr3 and BBr3 has been performed. It was shown that the composition of the reaction products depends on the ratio of the catalysts used.Later, 1,3- dibromoadamantane (19) was obtained in 97% yield by the bromination of adamantane (1) in the presence of Fe and small amount of water.68 The reaction of adamantane (1) with bromine in the presence of Fe in Freon-113 (1,1,2-trichloro-1,2,2-trifluoroethane) or in hexafluorobenzene affords dibromide 19 in 97.3% yield and of trichloro- (37, 27.3%), 1-bromo-3,5-dichloro- (16.6%), 1,3- dibromo-5-chloro- (27.4%), 1,3,5-tribromo- (12.0%), 1-bromo- 3,5,7-trichloro- (10.8%) and 1,3-dibromo-5,7-dichloro-adaman- tanes (5%) were obtained in the presence of AlCl3.69 The treatment of adamantane (1) with bromine at 100 8C under 27.2 atm also yields 1,3-dibromoadamantane.70 1,3,5,7-Tetrabromoadamantane (43) has been obtained in 75% yield by heating 1,3,5-tribromoadamantane (44) with bro- mine and AlBr3 in an autoclave at 150 8C for 2 h.16 Br Br2, AlBr3 Br Br Br 44 Br Br Br2, AlCl3 Br 43 1 Synthesis of 1,3,5,7-tetrabromoadamantane (43) by bromina- tion of adamantane (1) has also been performed under relatively mild reaction conditions (an equimolar amount of AlCl3, 30 min at 20 8C and then 3 h at 60 8C).71 1,3,5-Tribromoadamantane (44) has been synthesised 72 from adamantane (1) under the conditions used earlier 15 for the preparation of 1,3-dibromoadamantane (19).1,3-Dibromo- (19) and 1,3,5,7-tetrabromo-adamantanes (43) have been prepared from salts of the corresponding acids by the Hunsdiecker reaction.73, 74Reactions of adamantanes in electrophilic media Interestingly, 1,3,5,7-tetrabromoadamantane (43) has also been synthesised by the reaction of 2-chloroadamantane (39) with bromine in the presence of the equivalent amount of AlBr3.75 No product containing the chlorine and bromine atoms together in the same molecule was detected in this reaction.The methods for direct iodination of adamantane (1) are unknown. Iodoadamantanes can be obtained by a substitutive iodination of chloroadamantanes or hydroxyadamantanes. These methods are considered below. Halogenated adamantanes can also be synthesised from adamantanes bearing other functional groups. A possibility for halogen exchange in halogenoadamantanes upon their treatment with appropriate hydrogen halides has been demonstrated.76, 77 HBr Br HF 2 F HBr HCl 33 HCl Cl HF 35 1-Fluoroadamantane (33) has been obtained by the treatment of 1-chloro-, 1-bromo- or 1-iodoadamantane with a mixture of poly(hydrogen fluoride) and pyridine in the presence of nitronium tetrafluoroborate or sodium nitrate.78 PyH+(HF)nF7 F X NaNO3 (or NO2BF4) 33 X=Cl, Br, I.1-Fluoro- (33), 1,3-difluoro- (34) and 1,3,5,7-tetrafluoro- adamantanes (45) have also been synthesised by the reactions of the corresponding bromides 2 and 43 with ZnF2 or AgF.79 ZnF2 Br F cyclo-C6H12 2 33 Br F AgF Br Br F F Br 43 F 45 1-Bromo- (2) and 1-iodo-adamantanes (46) have been obtained by the reaction of 1-chloroadamantane (35) with gaseous HBr or HI in CCl4 or CH2Cl2 in the presence of FeBr3.80 The latter was prepared in situ from Fe(CO)5 and Br2.HX 35 X FeBr3 2, 46 X=Br (2), I (46). Table 4. Reaction conditions and yields of halogenoadamantanes.81 Solvent Substrate Temperature /8C 1-Bromoadamantane (2) CH2 I2 CHCl3 1,3,5,7-Tetrabromoadamantane (43) CH2 I2 CHCl3 20 65 80 20 A similar procedure has been proposed 81 to exchange the bridgehead halogen atoms in bromoadamantanes. The method is based on the generation in situ of an aluminium halide catalyst by the reaction of an excess of aluminium with bromine in a halogenated solvent. The results obtained are summarised in Table 4. In CH2I2, exchange of bromine by iodine in 1-bromoa- damantane occurs at room temperature in several minutes and is accompanied by the formation of adamantane (30%).Using this method, 1,3,5,7-tetraiodoadamantane was obtained in 83% yield. 1,3,5,7-Tetraiodoadamantane (47) is formed from 1,3,5,7- bromoadamantane (43) in 75% yield.82 I Br Br I I Br Br 43 I 47 The replacement of the fluorine atoms in 1,3,5-trifluoro-7- trifluoromethyl- (48), 2,2,5,7-tetrafluoro- (49), 2-trifluoromethyl- 5,7-difluoro- (50) and 1,3,5,7-tetrafluoroadamantane (45) by the bromine atoms under the action of BBr3 has been studied.83 The exothermic reactions of compounds 48, 49 and 50 with BBr3 begin at room temperature, whereas the exchange of the fluorine atoms in 1,3,5,7-tetrafluoroadamantane (45) by the bromine atoms requires heating. CF3 CF3 BBr3 Br Br F F 20 8C F 48 Br F F BBr3 F F F Br 20 8C F 49 Br CF3 CF3 BBr3 Br F 20 8C Br Br F 50 F BBr3 Br F Br F 70 8C F 45 43 Br Only the bridgehead fluorine atoms were found to be exchangeable.The trifluoromethyl groups are completely inert towards the action of neat BBr3. However, the addition of a catalytic amount of AlBr3 allows the exchange of the fluorine atoms in the trifluoromethyl groups by bromine. The reactions were carried out with a great excess of BBr3, which serves simultaneously as a reagent and as a solvent.84 Product Time /min 1-Iodoadamantane (46) 47 1-Chloroadamantane (35) 89 1,3,5,7-Tetraiodoadamantane (47) 83 1,3,5,7-Tetrachloroadamantane (40) 88 0.5 1.5 15 1200 1007 Yield (%)1008 CBr3 CF3 BBr3, AlBr3 Br Br CBr3 CF3 Br Br As reported by Cuddy et al.,85 2,2-dibromoadamantane (52) isomerises in the presence of AlBr3.However, this dibromide can be obtained by the reaction of 2,2-difluoroadamantane (51) with BBr3. No isomerisation of the reaction product was also observed in the reaction of 2,2,5,7-tetrafluoroadamantane with BBr3. 86 Br Br FF BBr3 51 52 1-Hydroxyadamantane (4) or its nitrate 12 are widely used as starting compounds in the synthesis of halogenoadamantanes. Thus 1-fluoroadamantane (33) was obtained by the reaction of 1- hydroxyadamantane (4) with boron trifluoride diethyl ether- ate.87 ± 89 BF3 . Et2O 3 F OH 33 4 By analogy with the halogen substitution, pyridinium poly(hydrogen fluoride) was employed for the replacement of the adamantane hydroxy group by the fluorine atom.90, 91 1-Fluoro- and 2-fluoro-adamantanes are formed from the corre- sponding alcohols in virtually quantitative yields.For the prepa- ration of adamantanes with other halogen atoms, this reagent is used in combination with the appropriate metal or ammonium halide. MX 4 X PyH+(HF)nF7 M=Na, K, NH4; X=Cl, Br, I. A convenient procedure for the replacement of the hydroxy group by iodine is the use of a the mixture of Me3SiCl and NaI.92 For example, 1-iodo- (46, yield 80%) and 1-hydroxy-3-iodo-5,7- dimethyl-adamantanes (53) have been prepared by this procedure. In the latter case, only one hydroxy group can be substituted even with an excess of the halogenating reagent.OH OH Me3SiCl I Me OH Me NaI, MeCN Me 53 (60%) Me The replacement of the hydroxy group in compound 4 by a halogen can be performed by the reactions with halogens, alkali metal halides or hydrohalic acids in sulfuric acid. Thus 1-bromo- adamantane derivatives have been obtained by the reactions of 1-hydroxyadamantane (4) and its derivatives with bromine in sulfuric acid.93, 94 1,3-Dibromo-5,7-dimethyladamantane has been prepared by the reaction of 1,3-dihydroxy-5,7-dimethyladamantane (54) with bromine in oleum.95 ± 97 Me Me NaBr Br Me OH Me H2SO4, SO3 HO 54 Br I K Moiseev, N V Makarova,MN Zemtsova The hydroxy groups in the hydroxy ketone 55 and hydroxy carboxylic acids 56 and 57 have successfully been replaced by halogens.98 ± 101 OH X a or b 55 O O (a) HBr, X=Br (83%); (b) SOCl2, X=Cl (90%).X OH a or b COOH COOH 56 COOH COOH a or b X HO 57 (a) PBr5, X=Br; (b) KI, H3PO4; X = I. The replacement of the hydroxy group in 2-hydroxyadaman- tane-1-carboxylic acid (56) by halogens presumably follows the SN2 mechanism. 1,3-Diiodoadamantane has been obtained by the reaction of 1,3-dihydroxyadamantane with hydroiodic acid.102 Yet another method for the introduction of the halogen atoms into adamantanes is the reactions of adamantyl nitrates with hydrohalic acids or their salts in sulfuric acid. Thus the nitrato groups in 1-adamantyl nitrate (12) and adamantane-1,3-diyl dinitrate (13) have successfully been replaced by the chlorine or bromine atoms using the reaction of these adamantanes with potassium (sodium) chloride or bromide in 98% sulfuric acid.103 + X ONO2H ONO2 + MX H2SO4 H2SO4 12 M=Na, K; X=Cl, Br.1-Bromo-(2),104 1,3-dichloro- (36) and 1,3-dibromo-adaman- tanes (19)105 were obtained in a similar manner. Under the action of halogenating agents on adamantanes bearing Alk, Ph, CO2H, CF3, C(NO2)3, NO2 , NH2 and other groups either the adamantane hydrogen atom in the 3-position or the substituent is replaced depending on the nature of the substituent. No problem arises in the bromination of 1-bromo-3- methyl-,106 1-methyl- (8),107, 108 1-ethyl-,109 1-tert-butyl-,110 1,3- dimethyl- (23),108 1,3,5-trimethyl- (24) 111 and 1-ethyl-3-methyl- adamantanes with bromine.106 The yields of the corresponding bromoadamantanes are almost quantitative.The bromination of 2-(1-adamantyl)-2-bromopropane (58) is worth a more detailed consideration, since the structure of the product obtained in this reaction has not initially been established. It was supposed that 2-(3-bromo-1-adamantyl)-2-bromopropane is formed in the bromination of compound 58 with bromine at 70 ± 80 8C for 2 h.112 Br2 C(Me)2Br C(Me)2Br 58 BrReactions of adamantanes in electrophilic media However, further studies 113, 114 have shown that it was the side chain that underwent bromination of 1-isopropyladamantane (59) yielding 2-(1-adamantyl)-1,3,3-tribromoprop-1-ene (60), and compound 58 is brominated to give tetrabromo-substituted derivative 61.Br2 CHBr CHMe2 CHBr2 59 60 Br2 CHBr CBrMe2 CHBr2 Br 61 58 Analogous products were obtained in the bromination of 1-isopropyl-3,5,7-trimethyladamantane and 2-bromo-2-(3,5,7- trimethyl-1-adamantyl)propane.113 A mechanism involving the intermediate formation of cations 62, 63 and the tribromide 60 has been proposed 113 to account for the reaction of 2-(1-adaman- tanyl)-2-bromopropane (58) with bromine. Br2 CH2 + 58 CMe2 7H+ 7Br7 Me 62 Br 2Br2 + 60 CH2Br 72H+,73Br7 C(CH2Br)2 7H+ Me 63 Judging from the ratio of the rate constants for solvolysis of 2-(1-adamantyl)-2-bromopropane (58) and 1-bromo-3-isopropyl- adamantane (104 : 1), an assumption was made 112 that the rates of bromination of the side chain and nucleus in 1-isopropyladaman- tane (59) should also differ by four orders of magnitude.However, since only the nucleus of 1-isopropyladamantane is brominated in the reaction, a complex reaction scheme including several rear- rangements of carbenium ions was proposed to account for this contradiction. In fact, solvolysis of bromides and bromination involving the saturated carbon atom occur most likely by different mecha- nisms.113 The rate-determining stage in the solvolysis is the formation of a carbenium ion, whereas the rate of the bromina- tion is determined by a stage which precedes the ionisation, e.g., the formation of a complex R+7H77Br+7Br7.The energy of such a complex is lower if the adamantane nucleus rather than side chain takes part in the complex formation. The reaction of 2-methyladamantane (64) with a 10% solution of bromine in CCl4 leads to 2-bromo-2-bromomethyladamantane (65), while a mixture of exo- and endo-2-bromo-4-(dibromome- thylidene)adamantanes (66) is formed upon refluxing 2-methyl- adamantane (64) with bromine for 4 ± 5 h. A scheme of the formation of compound 66 has been proposed.115 CH2Br Br2 Br CCl4 Me 65 CBr2 64 Br2 D 66 Br A mixture of 1-bromo-2-methyl-, 1-bromo-4-methyl- and 1-bromo-3-methyl-adamantanes was formed upon treatment of 2-methyladamantane with bromine.116 1009 Table 5. The rate constants for the solvolysis of alkylbromoadamantanes in 80% ethanol.Compound 105k krel DH= DS= /kcal mol71 /cal mol71 deg71 /s71 8.76 6.05 8.42 1.00 23.1 (22.5) (1.0) (23.63) 0.69 23.6 (15.9) (0.7) (24.36) 0.96 23.3 (22.5) (1.0) (24.5) 1.36 24.5 (32.1) (1.4) (24.0) 0.47 24.5 710.1 (79.6) a 79.6 (78.1) a 79.4 (77.0) a 75.4 (77.8) a 77.4 (6.93) 0.317 7 (0.31) 1-Bromoadamantane (2) 1-Bromo-3-methyl- adamantane (20) 1-bromo-3-ethyl- adamantane 1-Bromo-3-isopropyl- 11.9 adamantane 1-Bromo-3,5-dimethyl- 4.12 adamantane (21) 1-Bromo-3,5,7-tri- methyladamantane (22) Note. According to the data in Ref. 117 (70 8C). Data from Ref. 109 (Et3N, 75 8C) are given in parentheses. aDS= values at 60 8C. The effect of the nature and number of alkyl substituents on the stability of the adamantyl cations can be evaluated by an analysis of the rate constants for the solvolysis of the correspond- ing alkylbromoadamantanes (Table 5).117 As expected, the more the number of methyl groups in 1-bromoadamantane, the lower the rate of solvolysis.However, this is increased in the series 3-methyl-, 3-ethyl- and 3-isopropyl- substituted 1-bromoadamantanes (6.0561075, 8.4261075 and 11.961075, respectively). Both the aromatic ring and the adamantane nucleus are expected to be susceptible in the bromination of aryladamantanes. 1-Phenyladamantane (67) reacts with bromine under reflux to give a mixture of products from which no individual compound was isolated;118 however, 1-(4-bromophenyl)adamantane (68) was obtained at 0 8C, and 1-(3,4-dibromophenyl)adamantane (69) as the product of substitution of two hydrogen atoms in the aromatic ring was prepared at 25 8C.119 0 8C Br 68 Br2 Br 67 25 8C Br 69 The introduction of the carboxy group into position 3 of the adamantane nucleus in 1-phenyladamantane does not affect the character of the phenyl ring substitution.On the other hand, if electron-donating substituents are present on the phenyl ring, the corresponding dibromides are formed even at 0 8C.119 Br Br2 R2 R2 0 8C R1 R1 Br R1=H, CO2H; R2=OMe, OH, Me. The bromination of 4-(1-adamantyl)acetanilide (70) with bromine in acetic acid leads to 4-(1-adamantyl)-2-bromo- acetanilide (71).1201010 Br2, AcOH NHC(O)Me 70 NHC(O)Me 71 (90%) Br When the nucleophilicity of the aromatic ring is decreased due to the introduction of electron-withdrawing substituents, only the adamantane nucleus of aryladamantanes is brominated.Thus 1-(4-nitrophenyl)adamantane reacts with bromine to give 1-bromo-3-(4-nitrophenyl)adamantane.110, 121 Br Br2 NO2 NO2 Adamantanes bearing electron-withdrawing groups such as CO2H, COMe, C(NO2)3, CF3, etc., are brominated only in the presence of catalysts. Thus 3-bromoadamantane-1-carboxylic acid (72) was obtained upon treatment of adamantane-1-carbox- ylic acid (73) with bromine in the presence of AlBr3.122 Br Br2 CO2H CO2H AlBr3 73 72 The acid 73 can also be brominated in liquid bromine in the presence of water or acetic acid.123 (3-Chloro-1-adamantyl)acetic acid was prepared by the reaction of (1-adamantyl)acetic acid with an alkali metal chloride in concentrated sulfuric acid.124, 125 (3-Bromo-1-adamantyl)acetic acid is formed in 95% yield on treating (1-adamantyl)acetic acid with bromine at 20 8C for 18 h followed by refluxing the reaction mixture for 6 h.(3-Bromo-5- methyl-1-adamantyl)- and (3-bromo-5,7-dimethyl-1-adamantyl)- acetic acids were obtained under the same conditions.126 Adamantane-2-carboxylic acid is brominated with Br2 in the presence of AlBr3 to give 5,7-dibromoadamantane-2-carboxylic acid.127 1,3-Bis(trifluoromethyl)adamantane (74) reacts with anhy- drous hydrogen fluoride in the presence of SF4 yielding 1,3- bis(trifluoromethyl)-5,7-difluoroadamantane (75).128 CF3 CF3 HF, SF4 F CF3 CF3 F 74 75 Halogenation of adamantanes bearing such functional groups as NO2, NH2, CH2NH2, NHCOMe, NHCO2Me, etc., under the action of various halogenating agents has been studied in detail.The fluorination of aminoadamantanes 76 and 77 with a mixture of liquid hydrogen fluoride and SF4 affords several products.129 NH2 1) SF4, HF 2) HCl 76 F NH3Cl NH3Cl F + F + F F F F I K Moiseev, N V Makarova,MN Zemtsova CN CH2NH3Cl CH2NH2 1) SF4, HF 2) HCl + F F 77 F F 2-Dimethylaminoadamantane (78) reacts with HF and SF4 much more selectively to give difluoride 79 in 99.3% yield. 1) SF4, HF NMe2 . HCl NMe2 2) HCl F F 79 78 2-Aminoadamantane hydrochloride (80) is converted into 1,3,5-trifluoro- and 1,3,5,7-tetrafluoro-adamantanes (45) under the action of SF4 in liquid hydrogen fluoride.129 No reaction occurs in the absence of SF4.NH3Cl SF4, HF F F +45 F 80 The reactions of 1-acetamidoadamantanes 81 with HCl have been studied.130 The replacement of the acetamido group or/and its hydrolysis can occur depending on the nature of substituents in the adamantane nucleus. Cl R=Me Me Cl NH2 NHCOMe HCl R=OH + Cl OH R 36 81 NH2 R=CO2H CO2H A mechanism of this reaction was proposed. After protona- tion of the carbonyl group, two different reaction pathways are possible: (1) elimination of acetamide from an intermediate 82 affords the adamantyl cation, which is then attacked by the chloride anion yielding chloroadamantane; (2) hydrolysis of the acetamidoadamantane gives rise to aminoadamantane hydro- chloride.OH H+ NHCMe R NHCOMe R + 82 HCl, H2O R NH3Cl Cl7 + Cl R R The reactions of urethanes 83 with hydrochloric acid can occur as the replacement of the methoxycarbonylamino group by chlorine.130Reactions of adamantanes in electrophilic mediaNH3Cl R=OH NHCO2Me OH 36+ HCl Cl R R=Ph 83 Ph Treatment of 1-nitroadamantane with bromine in the pres- ence of an equimolar amount of AlBr3 was shown 131 to afford 1,3- dibromoadamantane (19) in 26%± 30% yield. A method for the preparation of phosphorus-containing adamantanes by the reactions of hydroxy- and nitroxy-substi- tuted adamantanes with phosphorus trichloride or acid chlorides of tervalent phosphorus has been developed.132 ± 134 From the data presented above one can conclude that the introduction of the halogen atoms into adamantanes can be performed by various halogenating agents, viz., halogens, hydro- gen halides, thionyl chloride, chlorosulfonic acid, alkali metal halides in sulfuric acid, a mixture of nitronium tetrafluorobrate with pyridinium poly(hydrogen fluoride), etc.The choice of the reagent is dictated by the nature of leaving group. The NO2BF4 ± PyH+(HF)nF7 system is the most versatile reagent, which can be employed for the replacement of the hydrogen and halogen atoms and hydroxy groups in adamantanes by fluorine atoms.From the practical standpoint, the most convenient method for the introduction of a halogen into adamantanes consists in the treatment of hydroxy- and nitroxy-adamantanes with hydrohalic acids or their salts in concentrated sulfuric acid. IV. Synthesis of hydroxyadamantanes and their nitrates Adamantane (1) and 1-bromoadamantane (2) are most widely used as starting compounds in the synthesis of hydroxyada- mantanes and their nitrates. Preparation of adamantan-1-ol (4) by the hydrolysis of 1-bromoadamantane (2) was described for the first time by Landa et al.14 This compound was also obtained by the reaction of 1-bromoadamantane (2) with silver nitrate in aqueous THF.15 AgNO3, THF,H2O OH Br 4 2 The reactions of 1-bromoadamantane (2) with silver nitrate in different solvents (benzene, phenol, dichloromethane, diethyl ether) was studied; it was shown that 1-adamantyl nitrate is formed in quantitative yield if the reaction is carried out in ether.135 Since the increase in the number of bromine atoms in adamantanes makes them more resistant to hydrolysis, 1,3- dihydroxy- (84), 1,3,5-trihydroxy- (85) and 1,3,5,7-tetrahydroxy- adamantanes (86, 84% yield) were obtained by the reaction of the corresponding bromides with Ag2SO4 in concentrated sulfuric acid.136, 137 OH OH OH OH HO OH OH 86 84 HO HO 85 Hydrolysis of 1,3-dibromoadamantane (19) in an autoclave at 190 ± 200 8C in alkaline, neutral and acidic media affords 1,3- dihydroxyadamantane (84) in quantitative yield.138 When the reaction is carried out at lower temperature, 1-bromo-3-hydroxy- 1011 adamantane could be isolated.1-Hydroxy-3-methyladamantane was obtained from 1-bromo-3-methyladamantane (20).109 3-Bromoadamantane-1-carboxylic acid (72) undergoes the replacement of the bromine atom by the hydroxy group in 3 M aqueous alkali.122 CO2H CO2H NaOH Br OH 72 The rate constants for the hydrolysis of 1-bromoadamantane (2), 3-bromoadamantane-1-carboxylic acid (72) and 1,3-dibro- moadamantane (19) (100 8C, 74% aqueous dioxane) were found 118 to be equal to 7.9861074, 1.361075 and 2.461076 s71, respectively.Hydroxyadamantanes were also prepared by the reactions of adamantane and its derivatives with sulfuric, nitric, fluorosulfonic and trifluoroacetic acids.Treatment of adamantane (1) with a mixture of Br2 and nitric acid affords 1-hydroxyadamantane (4).139 This compound was also obtained by the action of 69% nitric acid on 1-bromoada- mantane (2).140 a or b OH R 4 (a) HNO3, Br2, R=H; (b) HNO3, R=Br. A one-step method for the preparation of adamantan-1-ol (4) from adamantane (1) by the reaction with a mixture of trifluoro- acetic acid and ButOH (55 8C, 12 h, the ratio AdH:ButOH: :CF3CO2H is 1 : 2.5 : 17) in dichloromethane was pro- posed.141, 142 A reaction scheme including the intermediate for- mation of the adamantyl cation 3 was suggested. H2O Me3C+ CF3COO7 4 3 OCOCF3 1 The use of the system trifluoroacetic acid ± 56% nitric acid made possible the preparation of 1-aryl-3-hydroxyadamantanes by the replacement of hydrogen in aryladamantanes.In the case of reactive aryl substituents, the oxidation is accompanied by the nitration of the aromatic ring.20 The reaction conditions and the yields of the corresponding alcohols are listed in Table 6. Treatment of adamantane with 100% nitric acid affords 1-hydroxyadamantan-4-one in 77% yield.98 Hydroxyadamantanes 87 were obtained by the reactions of adamantane (1) and its homologues with fluorosulfonic acid.51 R2 R2 FSO3H OH R1 R1 87 R1 R1 R1=R2=H;R1=Me, R2=H;R1=R2=Me. 1-Hydroxyadamantane (4) is a reaction product of adaman- tane (1) with the system bis(trimethylsilyl) peroxide ± trifluoro- methanesulfonic acid.143 Adamantanes with the hydroxy group at the bridgehead carbon atom can be obtained by the isomerisation of 2-alkyl-2- hydroxyadamantanes on treating them with concentrated sulfuric acid.Thus 2-hydroxy-2-methyladamantane (88) is converted into tertiary alcohols 89 ± 91 (a mixture of Z- and E-isomers) under the action of 96% H2SO4.144 The composition of the reaction products depends on the temperature and time of the reactions.1012 Table 6. Synthesis of hydroxyadamantanes.20 Substrate a AdC6H4Me-4 (2.4 : 1) AdC6H3Me2-3,4 (3.6 : 1) AdC6H4NO2-4 (2.4 : 1) AdC6H4Br-4 (1.2 : 1) AdC6H3Me-4-NO2-3 (3.6 : 1) AdC6H4CO2H-4 (1.2 : 1) a The ratio HNO3 : substrate is given in parentheses. H2SO4 50 8C, 5 min OHMe 88 0 8C, 1 min At 25 8C, 2-hydroxy-2-methyladamantane (88) gives a mix- ture of alcohols 90 and 91, while at 45 8C (10 min), 1-hydroxy-3- methyladamantane (89) is the major reaction product.19 In trifluoroacetic acid, 2-hydroxy-2-R-adamantanes isomerise into Z- and E-isomers of 1-hydroxy-4-R-adamantanes.145 CF3CO2H RHO Thus the isomerisation of 2-hydroxy-2-methyladamantane (88) leads to a 1 : 1 mixture of Z- and E-isomers of 1-hydroxy-4- methyladamantane in 90% overall yield.The well-known method for the preparation of alcohols such as deamination of amines is also used in the synthesis of hydroxyadamantanes.146 ± 151 Thus the deamination of 1,3-diami- noadamantane (92) and 1-amino-3-hydroxyadamantane (93) by the action of NaNO2 in acetic acid was shown to result in the formation of 1,3-dihydroxyadamantane 84.146 I K Moiseev, N V Makarova,MN Zemtsova NH2 t /h Products Yield (%) NH2 OH 92 NaNO2 OH 2 2 HO C6H3Me-4-NO2-30 NH2 84 OH C6H3Me-4-NO2-3 62 93 3 2 HO C6H2Me2-3,4-NO2-56 Besides the expected 1-hydroxyadamantane (4), adamantane (1) was identified as the reaction product in the deamination of 1-aminoadamantane (80).149 2 6 HO C6H4NO2-4 0 The deamination of 1-aminoadamantane (80), including that 2 3 HO C6H4Br-4 2in the presence of benzene and anisole, has been studied in detail.150 It was expected that the adamantyl cation 3 would alkylate the arenes, react with external nucleophiles or rearrange.ArH 3 5 HO C6H3Me-4-NO2-38 AcO7 OAc Ar 7H+ 94 + 2 1 HO C6H4CO2H-4 1H2O 3 OH 7H+ 4 However, no aryladamantane was found among the reaction products when the deamination was carried out in the presence of benzene or anisole (Table 7).148 OH Table 7.Deamination of 1-aminoadamantane (80).148 Me Yield (%) Reaction conditions Reagent 89 Me Me 4 94 Solvent T /8C H2SO4 OH + 3.01.0 90 (22%) HO 91 (47%) AcOH, H2O HCl, H2O AcOH, PhH 85 85 refluxing AcOH, PhOMe the same NaNO2 NaNO2 n-C5H11ONO n-C5H11ONO 97.01.0 100 40.00.5 59.00.5 100 The deamination of 2-amino-2-hydroxy- (95) and 2-amino- 1,3-dihydroxy-adamantanes (96) is accompanied by a carbon skeleton rearrangement leading to the formation of tricyclic ketones 97 and 98.151 O OH OH R NH2 HNO2 95 97 (92%) OH O NH2 HNO2 OH OH 98 96 Hydroxyadamantanes are also synthesised from the corre- sponding nitrates, which, as was mentioned above, are prepared by the treatment of adamantanes with nitric acid.21 Later, the reactions of substituted adamantanes with nitric acid in acetic anhydride was studied.152 In these reactions, 1-chloro- (35) and 1-bromo-adamantanes (2) produce nitrate 12 due to the replace- ment of the halogens.As regards adamantane-1-carboxylic acidReactions of adamantanes in electrophilic media (73), 1-(trinitromethyl)adamantane (28) and the nitrate 12, they undergo the replacement of the hydrogen atom in the 3-position. R R HNO3 ONO2 Ac2O R=CO2H, C(NO3)3, ONO2. Adamantanes with nitrogen-containing functional groups like NHCO2Me and NO2 react with nitric acid through the replace- ment of these groups by the nitrato group.NHCO2Me ONO2 HNO3 ONO2 HNO3 ONO2 NO2 Upon treatment with a mixture of nitric acid and acetic anhydride, urethanes give nitrourethanes and 1-aminoadaman- tane (80) gives 1-nitroxy-3-nitroadamantane.152 A method for the preparation of alkyladamantanols by the hydrolysis of the corresponding nitrates in 25%± 30% nitric acid has been described.153 R1 R1 R1 H3O+ HNO3 R2 R2 R2 OH ONO2 R3 R3 R3 R1=Me: R2=R3=H; R2=Me, R3=H; R2=R3=Me; R1=Et, R2=R3=H. The preparation of hydroxyadamantanes from their nitrates is the most convenient method, taking into account its simplicity and availability of the starting materials. Using this method, 1-hydroxyadamantane (4), 1,3-dihydroxyadamantane (84), 3-hydroxyadamantane-1-carboxylic acid and many other com- pounds have been synthesised.Later, this method was extended to 2-substituted adaman- tanes.154 ± 156 V. Synthesis of adamantanecarboxylic acids The Koch ± Haaf reaction is widely used in the synthesis of carboxylic acids of the adamantane series. Adamantane (1), 1-bromoadamantane (2), 1-hydroxyadamantane (4) and its nitrate 12 are used as starting compounds. Adamantane-1-carboxylic acid (73) was obtained by the reaction of 1-bromo- (2) or 1-hydroxyadamantane (4) with formic acid in sulfuric acid 15, 157 or of adamantane (1) with formic and sulfuric acids in the presence of tert-butyl alco- hol.158, 159 As was shown by Langhals et al.,160 the maximum yield (84%) of adamantane-1-carboxylic acid (73) was achieved when the ratio AdOH (4) :HCOOH:H2SO4 was 1 : 1 : 24.The yield is decreased in a deficiency of formic acid. Adamantane-1-carboxylic acid (73) was also obtained from adamantane (1) in 20% oleum.161, 162 The reaction is assumed 161 to involve the intermediate formation of the adamantyl cation 3. SO3+H2SO4 HSOá3 +HSO¡4 1013 + +HSOá +H2SO3 3 3 1 3+HCO2H CO2H +H+ 73 To synthesise adamantanecarboxylic acids from adamantane, the latter is made to react with CO in sulfuric acid or in oleum (autoclave, 90 ± 160 8C).163, 164 This provides a mixture of ada- mantane-1-carboxylic acid (73) and adamantane-1,3-dicarboxylic acid in the ratio of 1 : 6.(1-Adamantyl)acetic acid is synthesised from 1-bromo- (2) or 1-hydroxy-adamantanes (4) and dichloroethylene in 80% ± 100% H2SO4 in the presence of BF3 at 0 ± 15 8C.165 The preparation of adamantylacetic acids from functionally substituted adamantanes and adamantane homologues has also been described.166 R1 CH2CO2H CCl2 CH2 R2 R2 R4 R4 H2SO4, BF3 R3 R3 R1=Br, OH, OAc; R2, R3, R4=H, Me. The reactions of adamantane and its derivatives with tri- chloroethylene in the presence of 90% sulfuric acid yield the corresponding a-chloroacetic acids.167 CHClCO2H X H2SO4, H2O R1 R3 R1 R3 +CHCl CCl2 D R2 R2 X=H, Br, OAc; R1, R2=H, Me; R3=H, Br, CH2CO2H. 3-Alkyladamantane-1-carboxylic acids are obtained from alkyladamantanes in sulfuric acid in the presence of tert-butyl alcohol and 95% formic acid.Using this approach, 3-methyl-, 3,5-dimethyl- and 3,5,7- trimethyl-adamantane-1-carboxylic acids have been synthesised in 72%, 49.8% and 38.5% yields, respectively.111 Also, the synthesis of 3-methyl-,106 3-methyl-5-ethyl- and 3-bromomethyl- adamantane-1-carboxylic acids 107 as well as of 3-carboxymethyl- 1-adamantyl-, 3-carboxymethyl-5-methyl-1-adamantyl- and 3-carboxymethyl-5,7-dimethyl-1-adamantyl-acetic acids 126 has been described. Adamantane-2-carboxylic acid was synthesised from 2-hydroxyadamantane (6).168, 169 Adamantanecarboxylic acids bearing the trifluoromethyl groups were also prepared by the Koch reaction.170 CF3 CF3 HCO2H F3C CO2H CF3 CF3 CF3 CO2H HCO2H CH2CF3 HNO3, H2SO4, SO3 F3CCH2 The reactivities of the trifluoromethyl- and trifluoroethyl- substituted adamantanes were shown 171 to be lower than those of non-fluorinated adamantanes due to the more difficult hydride ion abstraction from the former with the generation of the corresponding carbenium cations.Therefore, the preparation of 3-fluoromethyladamantane-1-carboxylic acid from 1-trifluoro- methyladamantane required that the reaction of the latter with CO was carried out in a mixture of nitric and sulfuric acids.1711014Treatment of adamantane (1) or 2,2-difluoroadamantane (51) with a mixture of formic acid, SF4 and HF in dichloromethane yielded trifluoromethyl-containing adamantanes (e.g., compound 99) rather than the corresponding carboxylic acids.These are formed as the result of reactions of intermediate carboxylic acids with SF4.147 CF3 F F HCO2H, SF4, HF F 99 51 F Studies of the reaction of 2-fluoroadamantane-2-carboxylic acid (100) with SF4 and HF were undertaken to clarify the mechanism of these transformations.148 This reaction afforded 4,4-difluoro-1-trifluoromethyladamantane (99), thus supporting the reaction scheme proposed earlier.147 F CO+ F CO2H F SF4, HF + 100 F F F F F F 99 + CO2H Dicarboxylic acids of the adamantane series were also syn- thesised by the Koch ± Haaf reaction. Thus adamantane-1,3- dicarboxylic acid was obtained from 1,3-dibromoadamantane (19) or 3-bromoadamantane-1-carboxylic acid (72).16, 122 Atwo-step synthesis of adamantane-1,3-dicarboxylic acid and (3-carboxymethyl-1-adamantyl)acetic acid from adamantane (1) or the corresponding monocarboxylic acids has been pro- posed.172 ± 175 Adamantane-1,3-dicarboxylic acid (101) or (3-carboxy- methyl-1-adamantyl)acetic acid (102) can be prepared from the corresponding monocarboxylic acids. In these cases, the reactions are carried out in a mixture of nitric and sulfuric acids with formic acid or CO and dichloroethylene as the carboxylating reagents for the former and the latter, respectively.173CO2H CO2H HCO2H, H2SO4, HNO3 CO2H 101 (67%) 73 CH2CO2H CH2CO2H CH2 CCl2, H2SO4, HNO3 CH2CO2H 102 (100%) Using regression analysis, it was established that the best yields of the acids 101 (94%) and 102 (66%) starting from adamantane (1) are achieved with a mixture of nitric and sulfuric acids in 60% oleum.174 Dicarboxylic acids of the adamantane series, like the acid 102, are of interest as starting materials in the synthesis of polyamides and polyesters.176 ± 178 Adamantanecarboxylic acids can be prepared from the corresponding nitrates.Thus adamantane-1-carboxylic acid (73) has been obtained in 94% yield from 1-adamantyl nitrate (12) and formic acid in sulfuric acid.104, 179 The preparation of carboxylic acids 103 from aryl- and alkyl- substituted adamantyl nitrates 104 has been described.180 ± 182 I K Moiseev, N V Makarova,MN Zemtsova R1 R1 a or b R2 R2 (CH2)nCO2H ONO2 103 104 R3 R3 R1, R2, R3=H, Alk, Ar; (a) HCO2H, H2SO4 (d=1.84 g cm73), n=0; (b) CH2=CCl2, H2SO4, n=1.Using this approach, the corresponding dicarboxylic acids have been obtained from nitrates of 3-hydroxyadamantane-1- carboxylic and (3-hydroxy-1-adamantyl)acetic acids, however, more concentrated sulfuric acid was required in this case.183 A mixture of 1-formyladamantane and ethyladamantane-1- carboxylate results from the formylation of adamantane in CH2Br2 in an atmosphere of CO in the presence of superacid systems as catalysts.184 VI. Aminoadamantanes The great interest in the synthesis of amino-substituted adaman- tanes is caused first of all by the discovery that 1-amino- adamantane hydrochloride (80) is active against influenza virus. Amines of the adamantane series are obtained by the reactions of 1-bromoadamantane (2), 1-hydroxyadamantane (4) or 1-adamantyl nitrates with ammonia, amines, amides or nitriles of carboxylic acids, urea, etc.NH3 or NH2R2 NHR2 H2NC(O)NH2 NHC(O)NH2 R1 R2CONH2 NHCOR2 R1=Hal, OH, ONO2; R2=H, Alk, Ar. Oleum, concentrated H2SO4 or HNO3 are used as an electro- philic medium. N-Adamantylamides are converted into the corresponding amines by alkaline or acid hydrolysis. 1-Amino- adamantane is synthesised from 1-bromoadamantane (2) under the action of ammonia at 200 8C for 6 h.185 The preparation of compounds of the general formula AdNR1R2 [R1, R2=H, Ph, Alk(C1±C12), AlkO, cyclo-Alk, HetAr, etc.] has been described.186 Thus 1-dodecylamino- adamantane was obtained by the reaction of 1-bromo- adamantane (2) with dodecylamine for 8 h.1-Aminoadamantane can be synthesised under fairly mild conditions with the use of amides, e.g., formamide. The reaction of 1-bromoadamantane (2) with formamide is carried out in concentrated H2SO4.187, 188 N-(1-Adamantyl)formamide thus formed is converted into 1-aminoadamantane under the action of 10% NaOH or 20% HCl. This method served as the basis for the manufacture of 1-aminoadamantane hydrochloride (80) (Amantadine, Mydantan and Symmetrel).189 The reactions of 1-hydroxyadamantane with alkylamides in trifluoroacetic acid have been proposed for the preparation of N-[1-(1-adamantanyl)alkyl]amides.190 N-Alkyl-1-aminoadamantanes, including 1-methylamino- adamantane hydrochloride (105), were obtained from the bro- mide 2 and N-alkylacetamides.191 In these reactions, silver sulfate was used as an electrophilic catalyst.Reactions of adamantanes in electrophilic media Ag2SO4 +MeCONHMe Br H2SO4 2 HCl + N(Me)COMe NH2Me Cl7 105 The synthesis of aminoadamantanes by the reactions of urea or its alkyl derivatives with 1-halogenoadamantanes has been described in some patents.192 ± 194 As a rule, the reactions are carried out at high temperature.180 8C 2+NH2C(O)NH2 NH2 Br NH2 D +NH2C(O)NH2 Et Et Me Me D +NH2C(O)NH2 Me Me Cl NH2 Stetter et al.195 were the first to have studied the reactions of 1-bromo- (2) and 1-hydroxyadamantane (4) with acetonitrile in the presence of acid catalysts (the Ritter reaction).1-Acet- amidoadamantane was obtained by these reactions in 61% yield. Diethylene glycol was used as a medium for the alkaline hydrolysis of 1-acylaminoadamantane.196 NaOH MeCN, H2SO4 NHCOMe NH2 1The reaction of hydrocyanic acid with adamantane (1) in sulfuric acid in the presence of a hydride ion acceptor leads to 1-formamidoadamantane. The influence of the reaction condi- tions on the yield of this product has been studied by Haaf.197 The reaction is carried out at 21 ± 26 8C in an excess of 96%± 100% H2SO4 using the molar ratio adamantane : accep- tor :HCN=0.1 : 0.4 : 1.7. HCN, H2SO4 NHC(O)H RX 1 Time /h Solvent X R Yield (%) 1.5 2.0 1.0 1.5 n-hexane the same cyclohexane CCl4 OH OH OH Cl But EtCHMe Bui But 78 75 60 42 In the reaction of adamantane with acetonitrile under similar conditions, 1-acetamidoadamantane was obtained in 36% yield.197 Nitriles of not only aliphatic monocarboxylic acids, but also of saturated dicarboxylic acids (maleic, adipic) and unsaturated monocarboxylic acids, e.g., acrylic, have been used in the reactions with 1-bromoadamantane (2).198 H2SO4 NHC(O)(CH2)nCN 2+NC(CH2)nCN n=1, 2.H2SO4 CHCN NHC(O)CH CH2 2+CH2 1015 Adamantyl cation can be generated in the Ritter reaction using NOPF6.199 The reactions of 1-bromoadamantane (2) with acetonitrile and propionitrile in the presence of NOPF6 and subsequent hydrolysis resulted in N-adamantylamides. Alcohols react nicely with acetonitrile in the presence of H2SO4.Thus 3-ethyl-1-hydroxy-5,7-dimethyladamantane in the reaction with acetonitrile in sulfuric acid gives the corresponding acetamide 107.200 Et Et H2SO4 Me Me NHAc OH+MeCN Me Me 107 106 The preparation of 3-aminoadamantane-1-carboxylic acid (108) from adamantane-1-carboxylic acid (73) has been described.201 CO2H CO2H HCl H2SO4 NHAc 73 +MeCN NH2 108 Diamines are of particular importance for the chemistry of polymers. Methods for the synthesis of diaminoadamantanes have been reported in several patents.102, 202, 203 Thus 1,3-diami- noadamantane (92) was prepared by the ammonolysis of 1,3- diiodoadamantane. The reaction is carried out at 260 8C under a pressure above 1000 atm.102 1,3-Diaminoadamantane (92) is also synthesised by the reactions of hydrocyanic acid with 1,3- dichloro-, 1,3-dibromo- or 1,3-dihydroxy-adamantanes in sulfu- ric acid.202 R NHC(O)H NH2 hydrolysis H2SO4 R NHC(O)H NH2 HCN 92 R=Cl, Br, OH NHAc Diaminodimethyladamantane (109) is obtained by the reac- tions of 1,3-dichloro-and 1,3-dibromo-5,7-dimethyl-adamantanes with acetonitrile.203 X hydro- lysis H2SO4 109 +MeCN Me Me NHAc X Me Me X=Cl, Br.Nitrates of the corresponding alcohols can also be used as starting compounds in the synthesis of aminoadamantanes. Thus 1-acetamidoadamantane is formed in the reaction of 1-adamantyl nitrate (12) with acetonitrile.204 It was later shown that adamantane (1) and its homologues can be functionalised with urea in a mixture of acetic acid and HNO3.205 The isolation of only the N-alkylation products of urea provides evidence in favour of the intermediacy of carbocations.R R HNO3, AcOH NHC(O)NH2 (NH2)2CO R=H, Me, Et. Carboxylic acids can also serve as starting compounds in the synthesis of the corresponding aminoadamantanes.206 Thus adamantane-1-carboxylic acid (73) in 100% sulfuric acid produ- ces the adamantyl cation (3), which then reacts withHCNorRCN1016 providing N-adamantylformamide or N-adamantylacetamide in 56% and 51% yields, respectively. + CO+ N CH NHCOH HCN 73 3 VII. Aryladamantanes Halogenoadamantanes, adamantanols and their O-nitrates are used for the alkylation of aromatic compounds with adamantane.The reaction of bromoadamantane (2) with benzene has been used by Stetter et al.15 to prepare phenyladamantane (67). The synthesis of some other adamantane-substituted aromatic com- pounds was also described.207, 208 Either adamantane with several phenyl groups or benzene bearing up to three adamantyl substituents are formed in the reaction of 1-bromoadamantane (2) with benzene in the presence of Lewis acids, depending on the ratio of the reactants.209, 210 1,3,5-Triadamantylbenzene (110) has been synthesised by the reaction of benzene or 1-phenyladamantane (67) with 1-bromo- adamantane (2) in liquid SO2 in the presence of AlCl3.210 A small amount of p-diadamantylbenzene was also isolated. Ad PhH 2 Ad PhAd 110 Ad 67 1,3,5-Triphenyladamantane and 1,3,5,7-tetraphenyladaman- tane have been synthesised by the reaction of 1-bromoadaman- tane (2) with benzene in the presence of tert-butyl bromide and a catalytic amount of AlCl3.209 It was shown 211 that the reaction of 1-bromoadamantane (2) with toluene in the presence of water resulted only in 4-(1- adamantyl)toluene, while a mixture of 4-(1-adamantyl)toluene and 3-(1-adamantyl)toluene was obtained in the presence of ZnCl2 or FeCl3.1-Bromoadamantane (2) reacts also with naphthalene, 2-methylnaphthalene, biphenyl, anthracene, phenanthrene and fluorene to give the corresponding substituted adamantanes.212 The reactions of naphthalene and its derivatives with 1-bromoadamantane (2) were the subject of several studies.213 ± 219 Thus the reaction of 1-bromoadamantane (2) with naphtha- lene in the presence of FeCl3 at 60 ± 140 8C leads to 2-(1- adamantyl)naphthalene.The second molecule of adamantane is introduced either in the 6- or 7-position.215 H Ad + 2, FeCl3 Ad 7H+ +H+ According to Stewart ± Breigleb models of 1- and 2-ada- mantylnaphthalene, the molecule of the former is much more sterically hindered due to the spatial interaction of the hydrogen at the C(2) atom of the adamantane fragment and the H(8) atom of naphtalene. Attempts to obtain 2-(1-adamantyl)naphthalene by the iso- merisation of 1-(1-adamantyl)naphthalene were unsuccessful. 2-(1-Adamantyl)naphthalene has been synthesised by the reac- tion of 1-hydroxyadamantane (4) with naphthalene in the pres- ence of phosphoric anhydride.216 N-Acetyl-1-aminonaphthalene is alkylated under these conditions with hydroxyadamantane to give the corresponding 4-adamantyl-1-aminonaphthalene.216 Phenol, anisole, catechol, 2,6- and 3,4-dimethylphenols, o-, m- and p-cresols, diphenyl ether and others were also reacted with 1-bromoadamantane (2).220 ± 227 I K Moiseev, N V Makarova,MN Zemtsova Thus phenol, 2,6- and 3,4-dimethylphenols and catechol react with 1-bromoadamantane (2) in the absence of a catalyst.222 OH OH OH OH 2 Ad OH OH OH R1 R1 R1 Ad R1 2 + R2 R2 R2 Ad R2 R1, R2=Me; R1=H, Me.2,6-Di(tert-butyl)phenol and 2-methoxyphenol react with 1-bromoadamantane (2) at 130 ± 140 8C to form 4-adaman- tylphenols. The alkylation of m-aminophenol with 1-bromo- adamantane (2) leads to adamantane-substituted m-aminophenol.223 4,40-Diadamantyldiphenyl ether is formed upon heating of 1-bromoadamantane (2) with diphenyl ether in carbon disulfide in the presence of aluminium chloride.In the presence of zinc chloride and at elevated temperature, the oligomer 111 becomes the major product.207, 223 O O 3 111 The possibility of alkylation of phenylchlorosilanes with 1-bromoadamantane (2) has been demonstrated.228 SiCl3 Ad 2 SiCl3 SiCl3 Ad Ad It should be noted that the reaction of 1-bromoadamantane (2) with methyl(phenyl)dichlorosilane results in a mixture of mono- and polyadamantylbenzenes.228 1,3-Dichloro- and 1,3-dibromo-adamantanes react with aryl(alkyl)chlorosilanes at 160 ± 200 8C in the presence of ZnCl2 to give 1,3-di(alkylchlorosilylalkyl)phenyladamantanes 112.229 1-(Alkylchlorosilylalkyl)phenyladamantanes are also formed in appreciable yields.R R X +PhR X 112 X=Cl, Br; R=CH2SiMe3, C2H4SiMe3, C2H4SiMeCl2, C2H4SiCl3, CH2SiMe2Cl, CH2SiCHCl2, CH2SiCl3. 1,3-Dibromoadamantane was found to be more reactive than its 1,3-dichloro-analogue and aryl(alkyl)chlorosilanes can be ar- ranged according to their reactivity in the following order: PhCH2SiMe3>PhC2H4SiMe3>PhC2H4SiMeCl2 >PhC2H4SiCl3 >PhCH2SiCH2Cl>PhCH2SiMeCl2>PhCH2SiCl3. Mono- (12) and dinitrate 13 were also used as the alkylating reagents. The nitrate 12 readily reacts with benzene, toluene, ethylbenzene, p- and m-xylenes and anisole yielding the corre- sponding mono- and di-substituted adamantanes.230 The prepa- ration of 1,3-diaryladamantanes from the dinitrate 13 was also described.230Reactions of adamantanes in electrophilic media The nitrates 113 were used in the synthesis of diadamantyl- substituted arenes.180 R1 R1 PhH + R2 ONO2 R2 113 R1 R1 R1 Ph R2 R2 R2 R1=Me, Et; R2=H, Me.Reactions of 1-adamantyl nitrate (12) with such heterocyclic compounds as thiophene, 2-nitrothiophene, thiophene-2-carbox- ylic acid, furan and 2-furoic acid have been described.231 The alkylation was carried out in concentrated sulfuric acid at75 8C; the corresponding 1-hetaryladamantanes were obtained in 40%± 60% yields. Data concerning the relative reactivities of 1-bromoadaman- tane (2), 1-hydroxyadamantane (4) and its nitrate 12 in the reaction with isopropylbenzene have been reported.232 The high- est yield of p-adamantylisopropylbenzene was obtained from adamantyl nitrate.In recent years, an efficient procedure for the introduction of the adamantyl substituent into the aromatic ring of crown ethers by their direct alkylation with hydroxyadamantane in the presence of boron trifluoride diethyl etherate as a catalyst has been developed.233 1-Hydroxyadamantane (4) has been used as an alkylating reagent for 1,2,4-triazoles and 5-R-tetrazoles.234 ± 236 Its reactions with tetrazoles in 85% sulfuric acid gave the corresponding 2-(1- adamantyl)tetrazoles 114, while a mixture of 2-(1-adaman- tyl)tetrazoles 114 and 1-(1-adamantyl)tetrazoles 115 was formed in 94% sulfuric acid.R N a N N R N N 114 R OH+ HN N N N b 114+ N N N 115 (a) 85% H2SO4; (b) 94% H2SO4. The studies of the reactions of adamantanes in electrophilic media have expanded the general concept of the mechanism of C7Hbond activation and made it possible to develop preparative methods for the synthesis of halogen-, hydroxy- and nitroxy- substituted adamantanes. The latter are of particular interest as starting materials. Among the methods of adamantane function- alisation, the reactions with urethanes, amides and nitriles in nitric acid or its mixtures with AcOH or Ac2O are worth especial attention. Some of these reactions have formed the basis for technological processes in the manufacture of pharmaceuticals Remantadinum and Mydantan.References 1. V V Sevost'yanova, M M Krayushkin, A G Yurchenko Usp. Khim. 39 1721 (1970) [Russ. Chem. Rev. 39 817 (1970)] 2. N V Averina, N S Zefirov Usp. Khim. 45 1077 (1976) [Russ. Chem. Rev. 45 1077 (1976)] 3. E I Bagrii Neftekhimiya 35 214 (1995) 4. M-G A Shvekhgeimer Usp. Khim. 65 603 (1996) [Russ. Chem. Rev. 66 (1996)] 1017 5. G N Pershin, N S Bogdanova Khimioterapiya Virusnykh Infektsii (The Chemistry of Virus Infections) (Moscow: Meditsina, 1973) 6. A N Grinev Khim.-Farm. Zh. 10 (1) 26 (1976) a 7. A P Khardin, S S Radchenko Vysokomolekulyarnye Soedineniya na Osnove Poliedranovykh Uglevodorodov (High-Molecular Compounds Based on Polyhedral Hyfrocarbons) (Volgograd: Volgograd Polytechnic Institute, 1981) 8.A P Khardin, S S Radchenko Usp. Khim. 51 480 (1982) [Russ. Chem. Rev. 51 272 (1982)] 9. V Yu Kovtun, V M Plakhotnik Khim.-Farm. Zh. 21 931 (1987) a 10. R C Fort Adamantane. The Chemistry of Diamond Molecules (New York: Marcel Dekker, 1976) 11. E I Bagrii Adamantany, Poluchenie, Svoistva, Primenenie (Adaman- tanes, Production, Properties, Applications) (Moscow: Nauka, 1989) 12. Z N Parnes, D N Kursanov Reaktsii Gidridnogo Peremeshcheniya v Organicheskoi Khimii (Reactions of Hydride Transfer in Organic Chemistry) (Moscow: Nauka, 1969) 13. E S Rudakov Reaktsii Alkanov s Okislitelyami, Metallo-Komplek- sami i Radikalami v Rastvorakh (Reactions of Alkanes with Oxidants, Metal Complexes and Radicals in Solutions) (Kiev: Naukova Dumka, 1985) p.247 14. S Landa, SÏ Kriebel, E Knobloch Chem. Listy 48 61 (1954) 15. H Stetter,M Schwarz, A Hirschhorn Chem. Ber. 92 1629 (1959) 16. H Stetter, C Wulff Chem. Ber. 93 1366 (1960) 17. H W Geluk, J L M A Schlatmann Tetrahedron 24 5361 (1968) 18. H W Geluk, J L M A Schlatmann Tetrahedron 24 5369 (1968) 19. H W Geluk, J L M A Schlatmann Recl. Trav. Chim. Pays-Bas 88 13 (1969) 20. E A Shokova, V V Kovalev, O A Fedorova, I V Knyazeva, in Perspektivy Razvitiya Khimii Karkasnykh Soedinenii i ikh Primenenie v Narodnom Khozyaistve (Tez. Dokl. Vsesoyuz. Konf.), Kuibyshev, 1989 [Perspectives of Development of the Chemistry of Cage Compounds and Their Use in National Economy (Abstracts of Reports of the All- Union Conference), Kuibyshev, 1989] p.155 21. I K Moiseev, P G Belyaev, N V Barabanova, O P Bardyug, E N Vishnevskii, N I Novatskaya, E L Golod, B V Gidaspov Zh. Org. Khim. 11 214 (1975) b 22. I K Moiseev, P G Belyaev, G V Plokhikh, R I Doroshenko Izv. Vyssh. Uchebn. Zaved., Khim. Khim. tekhnol. 25 1185 (1982) 23. USSR P. 524 791; Ref. Zh. Khim. 9 O 103P (1978) 24. Yu N Klimochkin, I K Moiseev, E I Boreko, G V Vladyko, L V Korobchenko Khim.-Farm. Zh. 23 418 (1989) a 25. V F Baklan, A N Khil'chevskii, V P Kukhar' Zh. Org. Khim. 20 2238 (1984) b 26. V F Baklan, A N Khil'chevskii, V P Kukhar' Zh. Org. Khim. 23 2381 (1987) b 27. A G Yurchenko, V M D'yakovskaya, V F Baklan Zh.Org. Khim. 20 2239 (1984) b 28. A G Yurchenko, N I Kulik, V P Kuchar, V M Djakovskaja, V F Baklah Tetrahedron Lett. 27 1399 (1986) 29. A A Kisilenko, V F Baklan, A N Khil'chevskii, V P Kukhar' Zh. Obshch. Khim. 57 2659 (1987) c 30. P von R Schleyer, R C Fort Jr ,W E Watts,M B Comisarov, G Olah J. Am. Chem. Soc. 86 4195 (1964) 31. Yu N Klimochkin, I K Moiseev Zh. Org. Khim. 24 557 (1988) b 32. R S Fort, in Carbonium Ions Vol. 4 (EdsG A Olah, P von R Schleyer) (New York: Wiley, 1973) p. 1798 33. R D Bach, J W Holubka, R C Badger, S J Rajan J. Am. Chem. Soc. 101 4416 (1979) 34. E S Rudakov, N P Belyaeva, V V Zamashchikov, L N Arzamaskova Kinet. Katal. 15 45 (1974) d 35. G A Olah Angew. Chem., Int. Ed. Engl. 12 173 (1973) 36.G J Edwards, S R Jones, J M Mellor J. Chem. Soc., Chem. Commun. 816 (1975) 37. Dong Hak Bae, P S Engel, A K M Mansurul Hoque, D E Keys, Wang-Keun Lee, R W Shaw, H J Shine J. Am. Chem. Soc. 107 2561 (1985) 38. G A Olah, H C Lin J. Am. Chem. Soc. 93 1259 (1971) 39. G A Olah, J G Shih, Brij P Singh, B G B Gupta J. Org. Chem. 48 3356 (1983) 40. A T Podkhalyuzin, M P Nazarova Zh. Org. Khim. 11 1568 (1975) b 41. K R Brower J. Org. Chem. 52 798 (1987) 42. B V Kunshenko, N N Muratov, A I Burmakov, J M Jagupol J. Fluor. Chem. 22 105 (1983)1018 43. R E Moore, G L Driscoll J. Org. Chem. 43 4978 (1978) 44. A P Khardin, A D Popov, P A Protopopov, A O Litinskii, in Tez. Dokl. Nauch. Konf. po Khimii Poliedranov, Volgograd, 1976 (Abstracts of Reports of the Scientific Conference on the Chemistry of Polyhedranes, Volgograd, 1976) p.49 45. Czech. SSR P. 235 186; Ref. Zh. Khim. 18 O 60P (1987) 46. A G Gonzalez, G La Fuente, V J Trujillo An. Quim. Publ. Real. Soc. Esp. Quim. C81 38 (1985); Ref. Zh. Khim. 14 Zh 93 (1986) 47. I Tabushi, J Hamuro, R Oda J. Chem. Soc. Jpn., Pure Chem. Soc. 89 794 (1968) 48. H Stetter,M Krause, W-D Last Chem. Ber. 102 3357 (1969) 49. I Tabushi, Z Yoshida, Y Tamaru Tetrahedron 29 81 (1973) 50. E A Shokova, E V Erokhina, V V Kovalev Zh. Obshch. Khim. 66 1666 (1996) c 51. B M Lerman, Z Ya Aref'eva, G A Tolstikov Zh. Org. Khim. 7 1084 (1971) b 52. B M Lerman, Z Ya Aref'eva, A R Kuzyev, G A Tolstikov Izv. Acad. Nauk SSSR, Ser. Khim. 894 (1971) e 53.USSR P. 330 150; Ref. Zh. Khim. 1 N 85P (1973) 54. USSR P. 362 803; Ref. Zh. Khim. 22 N 112P (1973) 55. US P. 3 485 880; Ref. Zh. Khim. 3 N 124P (1971) 56. US P. 3 626 017; Ref. Zh. Khim. 16 N 81P (1972) 57. H Stetter,M Krause, W-D Last Angew. Chem. 80 970 (1968) 58. R D Bach, R C Badger Synthesis 529 (1979) 59. P Kovacic, Ju-Hua Chen Chang J. Org. Chem. 36 3138 (1971) 60. P Kovacic, Ju-Hua Chen Chang J. Chem. Soc., Chem. Commun. 1460 (1970) 61. G A Olah, R Renner, P Schilling, Y K Mo J. Am. Chem. Soc. 95 7686 (1973) 62. P von R Schleyer, R D Nicholas J. Am. Chem. Soc. 83 2700 (1961) 63. A F Bochkov, B E Kalganov Izv. Akad. Nauk SSSR, Ser. Khim. 2560 (1987) e 64. A V Cheprakov, D I Makhon'kov, I P Beletskaya Izv. Akad. Nauk SSSR, Ser.Khim. 698 (1987) e 65. G A Olah, P Schilling J. Am. Chem. Soc. 95 7680 (1973) 66. G L Baughman J. Org. Chem. 29 238 (1964) 67. USSR P. 852 851; Ref. Zh. Khim. 9 N 107P (1982) 68. I R Likhotvorik, N L Dovgan', G I Danilenko Zh. Org. Khim. 13 897 (1977) b 69. A I Rakhimov, A A Ozerov, A N Popov Zh. Org. Khim. 22 540 (1986) b 70. G W Smith, H D Williams J. Org. Chem. 26 2207 (1961) 71. A P Khardin, I A Novakov, S S Radchenko Zh. Org. Khim. 9 429 (1973) b 72. R E Delimarskii, V N Rodionov, A G Yurchenko Ukr. Khim. Zh. 54 437 (1988) 73. V Prelog, R Seiwerth Chem. Ber. 74 1769 (1941) 74. H Stetter, O E BoÈ nder,W Neumann Chem. Ber. 89 1922 (1956) 75. L S Simonenko, A M Kotlyarov, S S Novikov Izv. Akad. Nauk SSSR, Ser. Khim. 1125 (1971) e 76.Yu I Srebrodol'skii Vestn. Kievsk. Politekhn. Inst., Ser. Khim. Mashinostr. Tekhnol. 15 (1966) 77. M Namavari, N Satyamurthy,M E Phelps, J R Barrio Tetrahedron Lett. 31 4973 (1990) 78. G A Olah, J G Shih, Brij P Singh, B G B Gupta Synthesis 713 (1983) 79. K S Bhandari, R E Pincock Synthesis 655 (1974) 80. K B Yoon, J K Kochi J. Chem. Soc., Chem. Commun. 1013 (1987) 81. J W McKinley, R E Pincock,W B Scott J. Am. Chem. Soc. 95 2030 (1973) 82. G P Sollott, E E Gilbert J. Org. Chem. 45 5405 (1980) 83. AE Sorochinskii,AMAleksandrov,VP Kukhar',AP Krasnoshchek Zh. Org. Khim. 15 445 (1979) b 84. A E Sorochinskii, A M Aleksandrov, V P Kukhar' Zh. Org. Khim. 15 1555 (1979) b 85. B D Cuddy, D Grant, A Karim, M A McKervey, E J F Rea J.Chem. Soc., Perkin Trans. 1 2701 (1972) 86. A E Sorochinskii, A S Tarasevich, A M Aleksandrov, V P Kukhar' Zh. Org. Khim. 17 2339 (1981) b 87. B A Kazanskii, E A Shokova, S I Knopova Dokl. Akad. Nauk SSSR 211 346 (1973) f 88. E A Shokova, S I Knopova, Nguen Dang K'en, B A Kazanskii Neftekhimiya 13 585 (1973) I K Moiseev, N V Makarova,MN Zemtsova 89. E A Shokova, S I Knopova, Nguen Dang K'en, B A Kazanskii Izv. Akad. Nauk SSSR, Ser. Khim. 1429 (1973) e 90. G A Olah, J T Welch, Y D Vankar, M Nojima, I Kerekes, J A Olah J. Org. Chem. 44 3872 (1979) 91. G A Olah, J Welch Synthesis 653 (1974) 92. G A Olah, A Husain, Brij P Singh, A K Mehrotra J. Org. Chem. 48 3667 (1983) 93. M R Peterson, G H Wahl J. Chem. Soc., Chem. Commun. 1552 (1968) 94.V F Baklan, A N Khil'chevskii, V P Kukhar' Zh. Org. Khim. 33 557 (1997) b 95. US P. 3 819 721; Ref. Zh. Khim. 5 N 128P (1976) 96. US P. 3 666 806; Ref. Zh. Khim. 6 N 153P (1973) 97. US P. 3 78 4615; Ref. Zh. Khim. 20 N 134P (1974) 98. H W Geluk Synthesis 374 (1972) 99. M L Bagal, T K Klindukhova, V I Lantvoev Zh. Org. Khim. 11 1645 (1975) b 100. V I Lantvoev Zh. Org. Khim. 12 2516 (1976) b 101. G I Danilenko, L S Belash Vestn. Kievsk. Politekhn. Inst., Ser. Khim. Mashinostr. Tekhnol. 56 (1976) 102. Fr. P. 1 345 138; Chem. Abstr. 60 14 407 (1964) 103. L P Malevich, N V Stulin, V V Gorshkov, in Sintez i Svoistva Biologicheski Aktivnykh Soedinenii (Mezhvuzovskii Sbornik) [The Synthesis and Properties of Biologically Active Compounds (Itercolligeate Collection)] (Kuibyshev: Kuibyshev Polytechnic Institute, 1984) p.89 104. N V Stulin, A V Yudashkin, A K Shiryaev, I K Moiseev, A S Petrov Khim.-Farm. Zh. 18 595 (1984) a 105. USSR P. 433 120; Ref. Zh. Khim. 20 N 96P (1975) 106. F N Stepanov, V F Baklan, S D Isaev Zh. Org. Khim. 1 280 (1965) b 107. F N Stepanov, V F Baklan Zh. Obshch. Khim. 34 579 (1964) c 108. K Gerzon, E V Krumkalns, R L Brindle, F J Marshall, M A Root J. Med. Chem. 6 760 (1963) 109. C A Grob, W Schwarz, H P Fisher Helv. Chim. Acta 47 1385 (1964) 110. W Fischer, C A Grob, H Katayama Helv. Chim. Acta 59 1953 (1976) 111. H Koch, J Franken Chem. Ber. 96 213 (1963) 112. D J Raber, R C Fort, E Wiskott, C W Woodworth, P von R Schleyer, J Weber, H Stetter Tetrahedron 27 3 (1971) 113.F N Stepanov, A G Yurchenko, Z N Murzinova, V F Baklan, A I Erisheva Zh. Org. Khim. 8 1837 (1972) b 114. B P Raguel', Ya Yu Polis, V D Shatts,M P Gavars, in Tez. Dokl. Nauch. Konf. po Khimii Organicheskikh Poliedranov, Volgograd, 1981 (Abstracts of Reports of the Scientific Conference on the Chemistry of Organic Polyhedranes, Volgograd, 1981) p. 87 115. J R Alford, D Grant,M A Mc Kervey J. Chem. Soc., C 880 (1971) 116. T N Dubinina, V D Popov, G Yu Stepanova, G I Danilenko Vestn. Kievsk. Politekhn. Inst., Ser. Khim. Mashinostr. Tekhnol. 56 (1974) 117. R C Fort, P von R Schleyer J. Am. Chem. Soc. 86 4194 (1964) 118. E I Dikolenko, V P Ishchenko, in Khimiya i Perspektivy Primene- niya Uglevodorodov Ryada Adamantana i Rodstvennykh Soedinenii (Tez.Dokl. Ukr. Respubl. Konf.), Kiev, 1974 [The Chemistry and Application of Hydrocarbons of the Adamantane Series and Related Compounds) (Abstracts of Reports of Ukrainian Republican Conference), Kiev, 1974] p. 79 119. E I Dikolenko, G N Danilenko Vestn. Kievsk. Politekhn. Inst., Ser. Khim. Mashinostr. Tekhnol. 47 (1976) 120. H Stetter, J Weber, C Wulff Chem. Ber. 97 3488 (1964) 121. F N Stepanov, E I Dikolenko, G I Danilenko Zh. Org. Khim. 2 640 (1966) b 122. H Stetter, J Mayer Chem. Ber. 95 667 (1962) 123. V F Baklan, A N Khil'chevskii, L S Sologub, V P Kukhar' Zh. Org. Khim. 28 2098 (1992) b 124. US P. 3 897 479; Ref. Zh. Khim. 10 N 115P (1976) 125. US P. 3 966 800; Ref. Zh. Khim. 7 O 396P (1977) 126.K Bott Chem. Ber. 101 564 (1968) 127. H Stetter, V Tillmanns Chem. Ber. 105 735 (1972) 128. A M Aleksandrov, A E Sorochinskii, A P Krasnoshchek Zh. Org. Khim. 15 336 (1979) b 129. A E Sorochinskii, AMAleksandrov, V P Kukhar', V F Gamaleya, A P Krasnoshchek Zh. Org. Khim. 15 2480 (1979) bReactions of adamantanes in electrophilic media 130. F N Stepanov, Yu I Srebrodol'skii Vestn. Kievsk. Politekhn. Inst., Ser. Khim. Mashinostr. Tekhnol. 6 (1966) 131. M L Bagal, V I Lantvoev Zh. Org. Khim. 6 1347 (1970) b 132. R I Yurchenko, V V Miroshnichenko Zh. Obshch. Khim. 62 467 (1992) c 133. R I Yurchenko, L P Peresypkina, N A Fokina Zh. Obshch. Khim. 62 1912 (1992) c 134. R I Yurchenko, L P Peresypkina, V V Miroshnichenko, A G Yurchenko Zh.Obshch. Khim. 63 1534 (1993) c 135. Ya Yu Polis, I G Strazdyn', in Tez. Dokl. Nauch. Konf. po Khimii Poliedranov, Volgograd, 1976 (Abstracts of Reports of the Scientific Conference on the Chemistry of Polyhedranes, Volgograd, 1976) p. 21 136. H Stetter,M Krause Tetrahedron Lett. 19 1841 (1967) 137. H Stetter,M Krause Liebigs Ann. Chem. 717 60 (1968) 138. B K Kruptsov, in Tez. Dokl. Nauch. Konf. po Khimii Organiches- kikh Poliedranov, Volgograd, 1981 (Abstracts of Reports of the Scientific Conference on the Chemistry of Organic Polyhedranes, Volgograd, 1981) p. 61 139. L VodicPka, J Janku, J Burkhard Collect. Czech. Chem. Commun. 43 1410 (1978) 140. Czech. SSR P. 161 526; Ref. Zh. Khim. 24N 110P (1977) 141. E A Shokova, V V Kovalev, Yu N Luzikov, P A Sharbatyan Neftekhimiya 21 271 (1981) 142.V V Kovalev, I V Knyazeva, E A Shokova Zh. Org. Khim. 22 776 (1986) b 143. G A Olah, T D Ernst, C B Rao, G K S Prakash New. J. Chem. 13 791 (1989) 144. M A McKervey, J R Alford, J F McGarrity, E J F Rea Tetrahedron Lett. 5165 (1968) 145. V V Kovalev, A K Rozov, E A Shokova, in Perspektivy Razvitiya Khimii Karkasnykh Soedinenii i ikh Primenenie v Narodnom Khozyaistve (Tez. Dokl. Vsesoyuz. Konf.), Kuibyshev, 1989 [Perspectives of Development of the Chemistry of Cage Compounds and Their Use in National Economy (Abstracts of Reports of the All-Union Conference), Kuibyshev, 1989] p. 61 146. I K Moiseev, V P Konovalova, S S Novikov Izv. Akad. Nauk SSSR, Ser. Khim. 2378 (1973) e 147.K Gerzon, D J Tobias, R E Holmes, R E Rathbun, R W Kattau J. Med. Chem. 10 603 (1967) 148. B Cuddy, D Grant, M McKervey J. Chem. Soc., Chem. Commun. 27 (1971) 149. B A Kazanskii, E A Shokova, S I Knopova Izv. Akad. Nauk SSSR, Ser. Khim. 2645 (1968) 150. J Sabo, J Hiti, R C Fort Jr J. Org. Chem. 39 250 (1974) 151. J H Bayless, A T Jurewicz, L Friedman J. Am. Chem. Soc. 90 4466 (1968) 152. I K Moiseev, Yu N Klimochkin,MN Zemtsova, P L Trakhtenberg Zh. Org. Khim. 20 1435 (1984) b 153. I K Moiseev, E I Bagrii, Yu N Klimochkin, T N Dolgopolova, P L Trakhtenberg,M N Zemtsova Izv. Akad. Nauk SSSR, Ser. Khim. 2141 (1985) e 154. Yu N Klimochkin,M V Leonova, I K Moiseev Zh. Org. Khim. 33 381 (1997) b 155. Yu N Klimochkin,MV Leonova, I K Moiseev, AMAleksandrov Zh.Org. Khim. 33 387 (1997) b 156. Yu N Klimochkin,M V Leonova, I K Moiseev Zh. Org. Khim. 34 46 (1998) b 157. H Stetter, E Rauscher Chem. Ber. 93 1161 (1960) 158. H Koch, W Haaf Org. Synth. 44 1 (1964) 159. H Koch, W Haaf Angew. Chem. 72 628 (1960) 160. H Langhals, J Mergelsberg, C RuÈ chardt Tetrahedron Lett. 22 2365 (1981) 161. Ya Yu Polis, B P Raguel', in Tez. Dokl. Nauch. Konf. po Khimii Poliedranov, Volgograd, 1976 (Abstracts of Reports of the Scientific Conference on the Chemistry of Polyhedranes, Volgograd, 1976) p. 17 162. USSR P. 583 999; Ref. Zh. Khim. 17 N 100P (1978) 163. US P. 3 250 805; Ref. Zh. Khim. 15 N 104P (1967) 164. Fr. P. 1 353 906; Chem. Abstr. 61 593 (1964) 165. K Bott Angew.Chem. 77 967 (1965) 166. K Bott, H Hellman Angew. Chem. 78 932 (1966) 167. K Bott Angew. Chem. 79 943 (1967) 168. S Landa, J Burkhard, J Vais Z. Chem. 7 388 (1967) 1019 169. J Burkhard, J Vais, S Landa Z. Chem. 9 29 (1969) 170. A P Khardin, A S D'yachenko, P A Protopopov, in Tez. Dokl. Nauch. Konf. po Khimii Organicheskikh Poliedranov, Volgograd, 1981 (Abstracts of Reports of the Scientific Conference on the Chemistry of Organic Polyhedranes, Volgograd, 1981) p. 59 171. A P Khardin, L N Butenko, A D Popov, P A Protopopov, in Tez. Dokl. Nauch. Konf. po Khimii Organicheskikh Poliedranov, Volgograd, 1981 (Abstracts of Reports of the Scientific Conference on the Chemistry of Organic Polyhedranes, Volgograd, 1981) p. 69 172. L N Butenko, V E Derbisher, A P Khardin, in Tez.Dokl. Nauch. Konf. po Khimii Organicheskikh Poliedranov, Volgograd, 1981 (Abstracts of Reports of the Scientific Conference on the Chemistry of Organic Polyhedranes, Volgograd, 1981) p. 54 173. A P Khardin, A I Val'dman, L N Butenko, B I Panfilov, D I Val'dman, in Tez. Dokl. Nauch. Konf. po Khimii Organicheskikh Poliedranov, Volgograd, 1981 (Abstracts of Reports of the Scientific Conference on the Chemistry of Organic Polyhedranes, Volgograd, 1981) p. 38 174. A P Khardin, L N Butenko, V E Derbisher, R A Kosenkov Izv. Vyssh. Uchebn. Zaved., Khim. Khim. Tekhnol. 20 495 (1977) 175. S S Novikov, A P Khardin, L N Butenko, I A Novakov, S S Radchenko Izv. Akad. Nauk SSSR, Ser. Khim. 2597 (1976) e 176. USSR P.483 393; Ref. Zh. Khim. 6 N 129P (1977) 177. Jpn. P. 48-28 906; Ref. Zh. Khim. 12 N 124P (1974) 178. Jpn. P. 48-28 905; Ref. Zh. Khim. 12 N 123P (1974) 179. I K Moiseev, R I Doroshenko Zh. Org. Khim. 19 1117 (1983) b 180. Yu N Klimochkin, A I Matveev, V K Appolonov, in Sintez i Svoistva Biologicheski Aktivnykh Soedinenii (Mezhvuzovskii Sbornik) [The Synthesis and Properties of Biologically Active Compounds (Itercolligeate Collection)] (Kuibyshev: Kuibyshev Polytechnic Institute, 1984) p. 93 181. I K Moiseev, A K Shiryaev, Yu N Klimochkin, A I Matveev Zh. Org. Khim. 24 1410 (1988) b 182. I K Moiseev, E I Bagrii, Yu N Klimochkin, T N Dolgopolova, M N Zemtsova, P L Trakhtenberg Izv. Akad. Nauk SSSR, Ser. Khim. 2144 (1985) e 183. I K Moiseev, N V Stulin, A V Yudashkin, Yu N Klimochkin Zh.Obshch. Khim. 55 1655 (1985) c 184. I S Akhrem, I M Churilova, S V Vitt,A V Orlinkov,M E Vol'pin Izv. Akad. Nauk, Ser. Khim. 512 (1997) e 185. Austr. P. 293 343; Ref. Zh. Khim. 8 N 221P (1971) 186. US P. 3 391 142; Ref. Zh. Khim. 20 N 394P (1969) 187. Swiss P. 447 146; Ref. Zh. Khim. 9 N 297P (1969) 188. Czech. P. 126 397; Chem. Abstr. 70 46 961 (1969) 189. USSR P. 408 546; Ref. Zh. Khim. 20 O 46P (1977) 190. E Shokova, T Mousoulou, Y Luzikov, V Kovalev Synthesis 1034 (1997) 191. Br. P. 1 006 885; Ref. Zh. Khim. 16 N 478P (1967) 192. Fr. P. 1 294 371; Ref. Zh. Khim. 24 N 268P (1970) 193. Czech. P. 146 405; Ref. Zh. Khim. 19 O 86P (1975) 194. Zayavka 2318461 FRG; RZhKhim. 13 O 77P (1975) 195.H Stetter, J Mayer,M Schwarz, K Wulff Chem. Ber. 93 226 (1960) 196. Irish P. 27 380; Ref. Zh. Khim. 3 N 335P (1970) 197. W Haaf Chem. Ber. 97 3234 (1964) 198. T Sasaki, S Eguchi, T Toru Bull. Chem. Soc. Jpn. 41 236 (1968) 199. G A Olah, B G B Gupta J. Org. Chem. 45 3532 (1980) 200. US P. 3 450 761; Ref. Zh. Khim. 12 N 559P (1970) 201. Swiss P. 459 186; Ref. Zh. Khim. 2 N 351P (1970) 202. US P. 3 419 611; Ref. Zh. Khim. 13 N 270P (1970) 203. US P. 3 523 137; Ref. Zh. Khim. 10 N 121P (1971) 204. I K Moiseev, R I Doroshenko, V I Ivanova Khim.-Farm. Zh. 10 (4) 32 (1976) a 205. Yu N Klimochkin, I K Moiseev Zh. Org. Khim. 23 2025 (1987) b 206. W Haaf Chem. Ber. 96 3359 (1963) 207. O F Kozlov, A G Yurchenko Vestn. Kievsk. Politekhn. Inst., Ser. Khim. Mashinostr. Tekhnol. 49 (1977) 208. R Perkins, Sh Bennett, E Bowering, J Burke, K Reid, D Wall Chem. Ind. 790 (1980) 209. H Newman Synthesis 692 (1972) 210. W Rundel Chem. Ber. 99 2707 (1966) 211. V A Topchii, I V Makhiboroda, A G Zhukov, A G Yurchenko Zh. Org. Khim. 27 108 (1991) bI K Moiseev, N V Makarova,MN Zemtsova 1020 212. N N Skhirtladze, L D Melikadze, Z A Dzpmukashvili, in Khimiya i Perspektivy Primeneniya Uglevodorodov Ryada Adamantana i Rodstvennykh Soedinenii (Tez. Dokl. Ukr. Respubl. Konf.), Kiev, 1974 [The Chemistry and Application of Hydrocarbons of the Adamantane Series and Related Compounds) (Abstracts of Reports of Ukrainian Republican Conference), Kiev, 1974] p. 63 213. USSR P. 376 347; Ref. Zh. Khim. 2 N 157P (1974) 214. A G Zhukov, V V Topchii, A G Yurchenko Vestn. Kievsk. Politekhn. Inst., Ser. Khim. Mashinostr. Tekhnol. 16 (1986) 215. AGZhukov,VA Topchii,V P Sungurova, S I Bass,AGYurchenko Zh. Org. Khim. 26 350 (1990) b 216. A G Yurchenko, B G Rorog, A B Khotkevich, V A Topchii, in Tez. Dokl. Nauch. Konf. po Khimii Organicheskikh Poliedranov, Volgograd, 1981 (Abstracts of Reports of the Scientific Conference on the Chemistry of Organic Polyhedranes, Volgograd, 1981) p. 84 217. B G Rarog, V A Topchii, A G Yurchenko Vestn. Kievsk. Politekhn. Inst., Ser. Khim. Mashinostr. Tekhnol. 25 (1981) 218. A G Zhukov, B G Rarog, V A Topchii, A G Yurchenko Ukr. Khim. Zh. 56 1119 (1990) 219. B G Rarog, V A Topchii, A G Yurchenko Zh. Org. Khim. 17 553 (1981) b 220. F N Stepanov, G I Danilenko, E M Dikolenko,M I Novikova Vestn. Kievsk. Politekhn. Inst., Ser. Khim. Mashinostr. Tekhnol. 6 (1969) 221. US P. 3 883 603; Ref. Zh. Khim. 7 N 245P (1976) 222. I A Khardina, S S Radchenko, in Tez. Dokl. Nauch. Konf. po Khimii Organicheskikh Poliedranov, Volgograd, 1981 (Abstracts of Reports of the Scientific Conference on the Chemistry of Organic Polyhedranes, Volgograd, 1981) p. 64 223. O F Kozlov, M A Novikova, M I Filippova Vestn. Kievsk. Politekhn. Inst., Ser. Khim. Mashinostr. Tekhnol. 15 (1982) 224. A G Zhukov, V A Topchii, V P Sungurova, in Tez. Dokl. Nauch. Konf. po Khimii Poliedranov, Volgograd, 1976 (Abstracts of Reports of the Scientific Conference on the Chemistry of Polyhedranes, Volgograd, 1976) p. 40 225. N I Miryan, A G Yurchenko, E I Kirienko Ukr. Khim. Zh. 56 183 (1990) 226. USSR P. 1 641 801; Ref. Zh. Khim. 22 N 73P (1991) 227. V A Sokolenko, L N Kuznetsova, N F Orlovskaya Izv. Akad. Nauk, Ser. Khim. 505 (1996) e 228. B I No, Yu V Popov, V V Son, in Tez. Dokl. Nauch. Konf. po Khimii Organicheskikh Poliedranov, Volgograd, 1981 (Abstracts of Reports of the Scientific Conference on the Chemistry of Organic Polyhedranes, Volgograd, 1981) p. 115 229. B I No, Yu V Popov, N N Mamutova,M V Troitskaya, in Tez. Dokl. Nauch. Konf. po Khimii Organicheskikh Poliedranov, Volgograd, 1981 (Abstracts of Reports of the Scientific Conference on the Chemistry of Organic Polyhedranes, Volgograd, 1981) p. 117 230. I K Moiseev, R I Doroshenko Zh. Org. Khim. 18 1233 (1982) b 231. USSR P. 1 122 658; Ref. Zh. Khim. 11 N 208P (1985) 232. B G No, G M Butov, S M Ledenev, I A Ryabukhina, in Perspektivy Razvitiya Khimii Karkasnykh Soedinenii i ikh Primenenie v Narodnom Khozyaistve (Tez. Dokl. Vsesoyuz. Konf.), Kuibyshev, 1989 [Perspectives of Development of the Chemistry of Cage Compounds and Their Use in National Economy (Abstracts of Reports of the All-Union Conference), Kuibyshev, 1989] p. 109 233. N V Averina, V V Samoshin, N S Zefirov Zh. Org. Khim. 32 845 (1996) b 234. I V Bryukhankov,M S Pevzner, E L Golod Zh. Org. Khim. 28 1545 (1992) b 235. A L Kovalenko, I V Bryukhankov Zh. Org. Khim. 30 1698 (1994) b 236. V V Saraev, E L Golod Zh. Org. Khim. 33 629 (1997) b a�Pharm. Chem. J. (Engl. Transl.) b�Russ. J. Org. Chem. (Engl. Transl.) c�Russ. J. Gen. Chem. (Engl. Transl.) d�Kinet. Catal. (Engl. Transl.) e�Russ. Chem. Bull. (Engl. Transl.) f�Dokl. Chem. Technol., Dokl. Chem. (En

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (12) 1021 ± 1039 (1999) Non-isothermal regimes and chemical control of branching-chain processes V V Azatyan Contents I. Introduction II. The main differences between thermal and chain ignition III. Specific features of the energetics of branching-chain processes IV. Temperature dependence of the specific rate of branching-chain processes V. The governing role of the branching-chain mechanism in the ignition and combustion of gases under atmospheric and higher pressures VI. Specific features of heterogeneous reactions determining the dependence of combustion characteristics on the surface properties at high pressures VII. The role of chain mechanism in the transition from deflagration combustion to detonation VIII. Thermal modes of chain combustion in the low-pressure region IX.Conclusion Abstract. Non-isothermal modes of branching-chain reactions are analysed. Characteristic features of these reactions caused by the simultaneous action of the reaction chain avalanche and self- heating are considered in terms of concepts of the modern theory. New data are presented indicating that chain branching is the predominant factor in gas combustion processes not only at very low pressures, as it was assumed earlier, but also at atmospheric and higher pressures. The specific character of the temperature dependence of the rate of chain reactions and data dealing with the discovery of a special chain combustion mode, chain thermal explosion, are considered. The efficiency of chemical control of chain combustion in any mode including explosion, flame propagation and its transition to detonation by virtue of highly efficient non-corrosive inhibitors is demonstrated.The bibliography includes 135 references. I. Introduction The significance of the discovery of branching-chain processes (BCP) and chain combustion 1±3 for the development of chemistry is well known. A special place in the variety of the kinetic features of this frequently encountered class of reaction belongs to the combustion mode, which differs fundamentally from the `purely thermal' combustion both in phenomenological features and the factors determining this mode. In recent years, it was shown that the branching-chain mechanism (BCM) predominates in most gas combustion processes not only at very low pressures, as had been considered until recently, but also at atmospheric pressure and at higher pressures. Combustion in air of virtually any hydrogen- containing compound (including hydrocarbon fuel), inorganic gaseous hydrides, and many pyrophoric compounds containing V V Azatyan Institute of Structural Macrokinetics and Material Science, Russian Academy of Sciences, 142432 Chernogolovka, Moscow Region, Russian Federation.Fax (7-095) 962 00 25. Tel. (7-095) 962 80 45. E-mail: azatyan@ism.ac.ru Received 15 March 1999 Uspekhi Khimii 68 (12) 1122 ± 1141 (1999); translated by Z P Bobkova #1999 Russian Academy of Sciences and Turpion Ltd UDC 541.124.7 : [546.21+546.11]+621/43.019.2 no hydrogen follows a BCM. Branching-chain decomposition of, for example, nitrogen trichloride is known.3 Until recently, the theory of BCP has considered only isothermal chain combustion processes which take place at pressures factors of tens or hundreds lower than atmospheric pressure, mainly in the vicinity of the first ignition limit (P1) (Fig.1). Under isothermal conditions, no self-heating or convec- tion is involved; therefore, chemical mechanisms play the crucial role. Thus, this approach proved fruitful in identification of the chemical factors and elucidation of their role in chain combustion. However, chain combustion processes at pressures markedly higher than P1 are exothermic and proceed at high rates; hence, they are accompanied by a substantial increase in the temperature of the reaction mixture. The combined influence of two funda- mentally different factors, the avalanche multiplication of the intermediate active species and self-heating, both accelerating the process, determines the specific character of virtually all features of branching-chain combustion, which differ fundamentally from those of `purely thermal' combustion caused by self-heating and heat transfer. Unlike the theory of BCP, the scope of combustion P/ Torr 1 atm 800 600 400 1 200 2 520 440 400 480 Figure 1.Region of self-ignition of a stoichiometric mixture of H2 with O2 in a reactor 7.4 cm in diameter with the surface coated with KCl.3, 6 (1 ), (2) and (3) are the first, second and third ignition limits, respectively. 1021 1022 1024 1025 1027 1032 1033 1034 1037 3560 T/ 8C1022 theory has actually been limited until recently to the latter type of combustion, this chemical process being represented as a simple one-step or, rarely, two-step reaction.Only quite recently was systematic investigation into the specific features of non-isother- mal BCP started. This review covers the regularities of BCP under non-isother- mal conditions, brought about by self-heating of the reacting gas, and the role of chain mechanism over a wide range of pressures including the pressure region important for practical purposes. The data presented bring to light the governing role of chain mechanism under atmospheric and higher pressures in any mode of gas combustion; they demonstrate the specific pattern of the temperature variation of the BCP rate, disclose the chain thermal explosion mode and reveal the important role of BCM in the origination of detonation.Efficient methods and means for the chemical control of ignition, developed combustion, explosion and deflagration-to-detonation transition are described. New aspects of the relevant sections of the combustion theory and chemical kinetics are considered. II. The main differences between thermal and chain ignition Ignition and developing combustion are progressively self-accel- erated chemical reactions. The kinetics of these reactions are characterised by not only increasing rates but also increasing accelerations (see, for example, Refs 3 ¡¾ 5): (1) d2W dt2 > 0, dWdt > 0; whereWis the reaction rate and t is time. Among other factors, ignition and combustion are governed by the positive feedback between the reaction rate and self- heating.Thermal ignition, caused by this factor alone, occurs if the rate of heat evolution (q+) prevails over the rate of heat removal (q7) and if the increase in the heat evolution rate upon rise of the temperature is more pronounced than the increase in the heat removal rate, i.e. the following conditions should be satisfied simultaneously:3¡¾5 (2) q+5q7, (3) dq dTa 5ddqT¢§ , where (4) q+=WQ, Q is the heat of the reaction, (5) q7=aSOT ¢§ T0U , V a is the heat removal coefficient, S and V are the surface area and the volume of the reaction chamber, respectively, and (T7T0) is the difference between the temperatures of the reaction mixture and the reactor walls.3¡¾6 The equality signs in relations (2) and (3) refer to the critical condition of thermal ignition.In the theory of combustion, a chemical process is usually represented as a simple reaction, the rate of which depends on the temperature according to the Arrhenius law. Since the reaction rate (and hence, the rate of heat evolution), unlike the rate of heat removal, increase non-linearly with temperature, inequality (3) starts to be fulfilled, in addition to relation (2), after a definite temperature has been attained during heating. Therefore, accu- mulation of thermal energy in the reaction system becomes progressive, which results in an accelerated rise in temperature. Correspondingly, the chemical process becomes progressively self-accelerated, i.e. thermal ignition takes place. Branching-chain ignition, unlike the thermal one, is caused by avalanche-like multiplication of active intermediate products �¢ free atoms, radicals and, in some cases, excited species �¢ upon V V Azatyan fast reactions with the starting reactants and with one another.1 ¡¾8 This type of multiplication takes place if the rate of generation of active species, chain carriers (CC), is greathan the rate of their decay. The rate of BCP is (6) W a ¢§daBa dt a w0 a kbnaBa, where [B] and n are the concentrations of the starting reactant and the CC, respectively, w0 is the rate of chain initiation, i.e.the rate of appearance of CC in reactions involving only the starting molecules, and kb is the effective rate constant for chain branch- ing (the rate-determining step of the process).1, 3, 8 The change in the CC concentration is described by the equation (7) dn dt a w0 a Of ¢§ gUn, where f and g are the branching and chain termination rates for unit concentration of CC (specific rates with respect to CC), equal to (8) f=2kb[B], (9) g=khet+khom[B][M] (khet and khom are the effective rate constants for the heteroge- neous and termolecular termination, respectively, and [M] is the concentration of the gas mixture).1¡¾3 The difference ( f7g) is often denoted by j.It follows from Eqn (6) that at initial stages of the process, when the decrease in the concentration of the starting compounds is still insignificant but the w0 value can already be neglected, the following relations hold: W*n, dWdt ddnt , (10) d2W dt2 dd2tn2 because the relative change in the concentration [B] is much smaller than the relative change in n.It is obvious that for f >g , the concentration of CC increases with time with progressive self-acceleration, i.e. (11) d2n dt2 > 0. The same follows from Eqn (7). Correspondingly, the rate of variation of the concentrations of the starting compounds in their reactions with active intermediate species, i.e. the rate of the process as a whole, increases progressively. This is indicated by relation (10) provided that inequality (11) holds. The attainment of a high reaction rate over a short period of time as a result of progressive self-acceleration is ignition. Thus, chain self-acceler- ation is caused by the positive feedback between the reaction rate and the CC concentration [see Eqns (6) and (7)]. Self-acceleration passes to the chain ignition mode in the case where multiplication of active intermediate species becomes avalanche-like because the multiplication rate exceeds the decay rate.Self-heating is not necessary for chain ignition. During progressive self-acceleration, high reaction rates can be attained; being exothermic, the reaction is accompanied by substantial light and heat evolution when the reactant concentrations are sizeable. This is a consequence of chain ignition alone. It can be seen from Eqn (7) that, when f<g, the process occurs from the very beginning in a qualitatively different kinetic mode in which the increase in n (and, hence, in W) is retarded d2n dt2 < 0.Non-isothermal regimes and chemical control of branching-chain processes In this mode, as the concentration of CC approaches the steady state value nst= w0 g ¢§ f, the right-hand side of Eqn (7) and, correspondingly, dn/dt diminish and approach zero, which implies a complete cessation of increase in n andW.Since the rate of reaction between valence- saturated compounds w0 is very low, it follows from the last- mentioned expression that under conditions where chain termi- nation substantially prevails over chain branching (i.e. the g value noticeably exceeds f), the steady-state values of n and W are also extremely low. Thus, the relation (12) f 5g is the condition for chain ignition. The equality sign determines the critical condition for transition from the steady-state mode, in which the reaction rate is extremely low, to the chain ignition and combustion mode.For initial stages, when the variation of the f and g values during an isothermal reaction can yet be neglected, integration of Eqn (7) gives 1, 3, 7 (13) W a w0 f ¢§ g faexpO f ¢§ gUta ¢§ 1g. It can be seen from this expression [see also Eqn (7)] that at f>g, the rate of the process increases with progressive self-acceleration, whereas at f<g, the increase in the reaction rate becomes less pronounced until a steady-state rate is attained. It is noteworthy that, to consider the f and g values invariant with time, it is not enough to assume the process to be isothermal and neglect only the consumption of the initial compounds, as has been done until recently (see below). Heterogeneous propagation of reaction chains is not taken into account in Eqn (13) either.9 Since f and g depend on the initial reactant concentrations, temperature and the reactor geometry in different ways, the ratio of these parameters is governed by the reaction conditions.It has been found 1¡¾3, 6 that the specific patterns of the dependence of f and g on [B] and [M] and on the temperature stipulate the existence of two limiting pressures of ignition, which are both much lower than atmospheric pressure. For most combustible gases, these limiting values coalesce as the temperature diminishes, thus forming an ignition peninsula 1¡¾ 3,6 ¡¾9 (see Fig. 1). If the chains branch upon reactions of CC not only with the initial molecules but also with one another (non-linear branching), the ignition condition can be written as follows: (14) f5g72(w0k+)1/2, where k+ is the effective rate constant for non-linear branching.3 When there is no external source of atoms or radicals, the second term in the right-hand part of the equation is much smaller than the first one and the ignition is actually determined by relation (12).Therefore, expression (12) is used in this paper, although the discussion equally refers to expression (14). If the quadratic chain termination (e.g., recombination of the (15) CC) is taken into account in addition to the linear chain termination, characterised by the parameter g, the rate of variation of n in the one-center consideration can be written as dn dt =w0+(f7g)n7k7 n2, where k7 is the rate constant for the quadratic termination. The solutions of Eqn (7) at f >g are unstable (Lyapunov instability), while the solutions of non-linear equation (15) are stable at any ratio of f to g (when f > g, the stability is non- asymptotic).Proceeding from this fact, some researchers (see, for example Refs 10 ¡¾ 15) who, unlike some other authors,1 ¡¾9 define ignition as a kinetic mode described by only unstable solutions of equations, concluded that chain self-ignition requires self-heating, allowing for which accounts for the instability of the solutions. 1023 Thus they disclaimed one of the most important statements of the BCP theory, viz. the assumption that self-heating is not obligatory for chain self-ignition to occur. According to this view, it was proposed to elucidate the ignition conditions resorting to analysis of the character of solutions of the set of kinetic equations together with the heat balance equation and regarding the ignition limit as the condition of coalescence of stable and unstable singular points (bifurcation) on the concentra- tion ¡¾ temperature phase plane.However, this definition of ignition and this approach to the determination of its limit do not correspond to the essence of this phenomenon and, therefore, contradict experimental data. Experiments show that chain ignition is possible even when the reaction system is monotoni- cally cooled provided that inequality (12) holds. This can be illustrated in relation to a model BCP �¢ combustion of H2 with O2 above the first ignition limit.16 When the pressure is low (*0.1 Torr) and only slightly exceeds P1 (by 5%¡¾ 8%), burning lasts for 10 ¡¾ 15 s after an induction period.In addition to the monotonic decrease in the pressure, caused by the reduction of the number of gas moles in the reaction, the specially arranged monotonic decrease in the temperature immediately after admis- sion of the mixture in the reactor can also be observed during this period. Ignition was detected as the appearance of chemilumines- cence after the induction period and the progressive acceleration of the consumption of the reactants. It has been shown 17 ¡¾ 19 that equations such as (15) (or, in a more general case, a set of equations wi negative quadratic terms corresponding to the concentrations of CC), although all of their solutions are stable, describe two qualitatively different isother- mal modes of BCP: progressive self-acceleration (d2W/dt2>0) up to high rates, i.e.ignition, and the extremely slow reaction with establishment of a steady-state rate (d2W/dt2<0). It was also demonstrated that transition from one isothermal mode to the other following variation of the initial conditions has a critical character and is determined by the equality f=g (16) both with and without allowance for the quadratic termination. If the rate constants appearing in f and g are known, the critical conditions of ignition, for example, the first and second limits calculated from expression (16) taking account of (8) and (9), coincide with the experimental values (e.g., see Refs 3, 8, 20 ¡¾ 23). This confirms once again the statement that self-heating is not a necessary condition for chain ignition.Thus, when the notion of ignition is defined correctly, kinetic equations permit one to describe its origin without self-heating. Strict mathematical analysis of differential equations of ignition and combustion in terms of the above definition of this phenom- enon as a kinetic mode characterised by progressive self-acceler- ation has been carried out.24, 25 The quantitative theory of the first and second ignition limits underlie precision methods for deter- mination of rate constants.20, 23, 26 ¡¾ 32 When reactions (most of all, bimolecular) of CC with one another and heterogeneous develop- ment of reaction chains are taken into account, the theory also describes quantitatively the kinetics of isothermal chain combus- tion near the first limit, isothermal critical phenomena within the ignition region, and hystereses of various kinetic characteristics over a broad pressure region.9, 33 ¡¾ 36 Azatyan 37 has considered the difference between thermal and chain self-acceleration from the molecular-kinetic viewpoint, i.e.in terms of the mechanism assuming an increase in the number of species able to overcome the energy barrier to the reaction during the process (Fig. 2). The values marked on the abscissa axis are activation energies for the reactions of the initial molecules with each other (E1) and with active intermediate species (E2). The areas under the Boltzmann curves of the distribution of species over energy, to the right of E1 and E2 correspond to the number of the initial molecules whose energy exceeds the energy needed for them to enter into the reactions.As the temperature increases, the distribution curve shifts to the right (see Fig. 2); thus, the number1024 dNB dE 1 2 E E1 E2 Figure 2. Equilibrium energy distribution of the initial molecules at temperatures T1 (1) and T2 (2) (T1<T2); E1 and E2 are the activation energies of reactions of the initial molecules with one another and with chain carriers, respectively. of molecules of the initial compound able to overcome the energy barriers E1 and E2 increases. In the case of BCP, the number of species able to overcome the energy barrier and enter into the reaction increases even if no rise in temperature occurs in the reaction mixture (the distribution function does not shift towards higher energies).This is due to multiplication of free atoms and radicals in their reactions with the starting compounds, which require overcoming the barrier E2 , which is much lower than E1 . The area under the distribution curve on the right of E2 (the fraction of initial molecules capable of reacting with free atoms and radicals) is quite large even at relatively low temperatures. Since reactions of atoms and radicals with the initial molecules not only give final products but also result in regeneration and multiplication of these atoms and molecules, the process self- accelerates even without self-heating. If f >g, progressive self- acceleration is observed, i.e. ignition takes place. Thus, chain ignition and combustion are due to chemical factors; they differ fundamentally from the self-heating factors and can alone ensure the progress of the reaction in this mode.In the chain combustion accompanied by self-heating, the distribution curve also shifts towards higher temperatures; there- fore, the number of species capable of reacting increases due not only to the chain avalanche but also to an additional increase in the number of reactive substances (both intermediate and initial), caused by the rise in temperature. This enhances additionally the chain avalanche. The specific features of combustion caused by the joint effect of the chain avalanche and self-heating are considered in Sections V and VIII. When the initial concentra- tions of reactants increase above the region of the first limit, self- heating which accompanies chain combustion becomes more and more pronounced, its influence on chain combustion being largely determined by the specific character of energy liberation during these processes.Thus, the phenomena of chain and thermal ignition are caused by fundamentally different factors; the processes determining the critical conditions for their origination and the mechanisms of self-acceleration are also dissimilar. Self-heating is not a necessary condition for chain ignition. III. Specific features of the energetics of branching-chain processes In simple exothermic reactions, the energy equal to the difference between the chemical energies of the initial and final components is converted only to the kinetic energy of particles, i.e. it is liberated as heat.On the contrary, the greater part of the energy evolved during the developing BCP combustion is converted into chemical energy of active intermediates. This is a crucial factor determining the kinetic features of BCP. V V Azatyan Table 1. Energy characteristics of the elementary reactions of a chain cycle. DH E/kJ mol71 /kJ mol71 Number Reaction of step 69.9 21.5 43.5 21.5 III III II H+O2=OH+O OH+H2=H2O+H O+H2=OH+H OH+H2=H2O+H 66.9 760.7 8.4 760.7 A 2H2+O2+H2=2H2O+2H 746.1 Let us consider the liberation of energy in a developing chain combustion over one chain cycle { for the model reaction of H2 with O2 using a procedure similar to that used in a previous study. 3 In this particular case, a chain cycle consists of the elementary reactions presented in Table 1; the Table also presents the respective changes in the enthalpy (DH) and the activation energies (E ).The DH values were calculated from thermochemical data; 38 the E values correspond to the recom- mendations given in the literature.39, 40 It can be seen that each chain cycle yields two additional chain carriers, free hydrogen atoms, in addition to two molecules of the final product. If the overall reaction 2H2+O2=2H2O (B) followed a molecular pathway, it would give off 477.8 kJ mol71 of heat. Meanwhile, the real process A, which follows a BCM, affords only 46.1 kJ mol71 per branching-chain cycle, i.e. less than 10% of the heat of the overall process B. As noted above, in conventional exothermic reactions, the difference between the chemical energies of the starting com- pounds and the reaction products is converted almost entirely into heat (to within a very small fraction of energy dissipated as radiation).Conversely, in a developing branching-chain combus- tion process, a substantial portion of the difference between the energies of the initial and final components is spent for the formation of free atoms and radicals, i.e. it is accumulated in the system as chemical energy of unpaired spins. This type of energy is the most rational for self-acceleration of the process not only due to the very high chemical reactivity of species with a free valence but also because they can be regenerated in reactions with the initial valence-saturated reagents. Indeed, the fraction of heat evolved during combustion and used to increase the number of the initial molecules able to overcome the energy barrier to a molecular reaction is extremely low bacause the activation energy is high.The rest of the thermal energy is dissipated or wasted in heating the final products. Meanwhile, the reactions of CC characterised by low energy barriers involve a much greater percentage of the initial molecules; in addition the free-valence energy (in the form of high reactivity) is actually transferred as a baton to the newly formed atoms and radicals and even provides for their multiplication. Only when the content of the initial reactants in the gas mixture decreases and the final stage of combustion, in which recombination becomes more significant, is approached, does energy evolve as heat.Of course, at the initial stages of the process when the content of reactants is still high, chain carriers also recombine partially, simultaneously with their participation in chain propagation (see Table 1, steps I ± III). The role of chain termination by the route representing back reaction with respect to step I is especially significant in the combustion of hydrogen-containing compounds even at low pressures.9 The percentage of energy evolved in the chain termination, which ensures self-heating of the reaction mixture, depends on the rate of heterogeneous recombination, the heat of which is {A chain cycle is the repeating set of several elementary reactions of CC including the CC regeneration.3, 8, 31Non-isothermal regimes and chemical control of branching-chain processes extracted by the reactor walls.For example, at very low pressures (near the first limit), chain termination is heterogeneous, the energy evolved being almost entirely absorbed by the reactor walls. In the above example of energy evolution in a chain cycle, only the indicated *10% of the energy of overall process B is evolved in the bulk of the gas phase. At low pressures, virtually all the heat evolved is extracted by the reactor walls through conductive heat transfer, because the duration of combustion is hundreds or thousands times longer than the characteristic time of heat removal. This accounts for the fact that virtually no self- heating is observed near the first limit. The numbers of free atoms and radicals diminish upon recombination; nevertheless, the concentrations of these species in the area of branching-chain combustion are many orders of magnitudes higher than the thermodynamically equilibrium values and even than those concentrations which would exist without chain branching.This is confirmed by direct measure- ments of concentrations of atoms and radicals during chain combustion processes of various types (see, for example, Refs 41 ¡¾ 47). The overall reaction A of a chain cycle shows that only one third of the consumed dihydrogen is converted into atomic hydrogen. The rest of dihydrogen is used in the formation of H2O. It is due to this fact that the free energy (G) decreases to a degree which permits accumulation of atomic hydrogen (and other CC), bringing about a local increase in G.The overall change in the free energy G is negative; therefore, combustion proceeds spontaneously. In view of the foregoing, it becomes clear that the statements that half the dihydrogen in an oxygen ¡¾ hydrogen flame is converted into atomic hydrogen, which can be found in the literature (see, for example, Ref. 48), are at variance with the thermodynamic prohibition; indeed, realisation of this process would require absorption of a large amount of heat (DH^200 kJ per mole of O2) and a nearly equal increase in the free energy (the entropy change is insignificant in this case). This means that such a process cannot occur spontaneously. It is also obvious that the number of free hydrogen atoms cannot exceed the number of H2O molecules formed, even if they did not recombine at all (this can also be seen from the final stoichiom- etry of the chain cycle).Therefore, concentrations of hydrogen atoms (such as those presented in Fig. 136 in a monograph 41) much greater than the total concentration of H2O in the flames of stoichiometric (or enriched in the combustible component) H2 ¡¾ O2 mixtures are impracticable. Since the concentrations of excited species in branching-chain combustion are much higher than the equilibrium values, the radiation of such flame is much more intense than that in a non- chain combustion. In some processes, chain branching involves species in not only vibrationally 49 ¡¾ 55 but also electronically 53, 54 excited states. The inverted population of energy levels corre- sponding to a branching-chain mechanism is used in chemical lasers.56 ¡¾ 58 Thus, high concentrations and high reactivity of reactive intermediate species as well as their ability to be regenerated and multiplied in reactions with the initial compounds determine the advantage of the BCM over a non-chain mechanism.This advantage is realised not only at low pressures. It has been noted in studies,59, 60 devoted to the elucidation of the mutual influence of chain branching and self-heating, that, when f >g, the W, dn/dt and dW/dt values are related to n through positive feedback [see also Eqns (7) and (10)]. The branching rate constant (kb), which is related to the kinetic coefficient f [see Eqn (8)], increases with temperature according to the Arrhenius law. The temperature dependence of g is much weaker than that for f.Moreover, as noted above, in the case of termolecular chain termination, g decreases with increase in temperature. Therefore, an increase in temperature sharply increases the difference f7g, which appears in Eqn (7) as a positive factor in n; thus, the positive feedback of n with dn/dt, W, and dW/dt values is enhanced. The progressive growth of W 1025 and dW/dt brings about a similar progressive increase in the rate of heat evolution (q+) and a rise in temperature. Thus, self-heating intensifies the chain avalanche and the process acceleration, and the enhancing chain avalanche, in turn, accelerates self-heating. This non-additive combined action of chain branching and self- heating accounts for the specific character of non-isothermal BCP including the specific temperature dependence of the chain combustion rate.IV. Temperature dependence of the specific rate of branching-chain processes The kinetic features of a BCP have been considered in substan- tially fewer studies than the first and second ignition limits. The temperature dependence of the BCP rate has received even less attention. On the basis of experimental data referring to very low pressures, it has been found 41 that the rates of combustion of H2 and CO in oxygen are proportional to the rates of the steps OH+H2=H2O+H, OH+CO=CO2+H. Relying on this fact, it was claimed that temperature variations of these and some other branching-chain processes `should evidently obey the Arrhenius law'. In another study,55 this statement has been extended and it has been suggested that the rates of the rate- determining step and the chain process as a whole follow identical patterns of temperature variation.The fact that the chain propagation reactions (including the rate-determining step) con- stitute a branching chain is, however, ignored. This means that the temperature dependence of the reaction rate is determined not only by the temperature dependences of the rate constants of these steps themselves but, to a larger extent, by the temperature variation of the concentrations of chain carriers. Therefore, attempts to describe the temperature variation of the BCP rate by an Arrhenius function give different activation energies for the same chain process (depending on the process conditions);61 ¡¾ 66 in addition, the resulting values are usually very high (in some cases, >600 ¡¾ 700 kJ mol71).This is not only much greater than the activation energy values typical of reactions involving free atoms or radicals but also exceeds the activation energies of reactions of the initial valence-saturated molecules with one another and even the energies needed to cleave the bonds in these molecules. Obviously, these `activation energies' do not correspond to the real energy barriers to the reactions and only point to the fact that the temperature variation of the specific rate of a BCP deviates crucially from the Arrhenius law. This also follows from the unnaturally great magnitudes of `pre-exponential factors' for the rate constants (see, for example, Refs 61, 65).Vedeneev et al .,15 who considered the role of self-heating in the ignition of H2 and O2 , transformed the exponential quantity of the Arrhenius function for the rate constant using the known 5 `exponential quantity expansion' technique. Subsequently, the resulting new function was expanded into the Taylor series with respect to the temperature, the series being restricted to the linear term. Meanwhile, the nonlinear character of the temperature dependence of the reaction rate is the most important feature of combustion; when it is neglected, the notion of combustion itself loses its meaning (see, for example, Ref. 5). The temperature dependence of the BCP rate has been presented 1, 7, 41 by the following expression: (17) W a 1 ¢§ BexpO¢§E1=RTU , AexpO¢§E=RTU where A and B are temperature-independent factors and E and E1 are effective activation energies.However, this formula is inappli- cable to a real process. Indeed, if Bexp(7E1/RT)>1, then W<0, which has no physical meaning. In addition, at a particular finite temperature corresponding to the equality1026 ¢§ E1 a 1, RT B exp the rate of the process becomes infinitely great, even at the beginning of the reaction, which has no physical meaning either. The drawbacks of Eqn (17) are due to the fact that it has been derived using the following formula for the length of branching reaction chains (nb):1, 3, 7 (18) a nb a b ¢§ d , (19) where a, b and d are the probabilities of chain propagation, termination and branching, equal to the ratio of the effective rate of the corresponding step to the sum of the rates of all these steps.When deriving Eqn (17), Semenov 1, 7 substituted the nb value expressed by Eqn (18) into the known relation W=w0n, where n is the chain length, and also took into account the temperature dependence of the rate constants included in the a, b and d values. However, we have shown 59 that Eqn (18), which has been used also in other studies,41, 65, 67 does not correspond to the real process. For example, for d>b, i.e. under conditions correspond- ing to ignition and combustion, it gives inevitably a negative value for the chain length. The b7d difference, appearing in the expression for the flame propagation rate, has resulted (when d>b) in imaginary values for this rate.67 In our study, 59 a fundamentally different expression for nb was proposed nb a Oa=bUaOadh=bUm ¢§ 1a .(20) adh=b ¢§ 1 Here h is the average number of new branches for one branching step,m=kb[B]a1dt, (21) where a1 is the probability that at least two of the active sites formed in the multiplication step enter into chain continuation reactions. Unlike expression (18), Eqn (20) affords real, positive values of the chain length for any ratio of the branching and termination probabilities. This equation also shows that nb (the length of the growing chain) increases with time, the kinetics of this increase being dependent on the ratio of probabilities of the reaction chain branching and termination, which is consistent with the essence of this phenomenon. Under conditions where the inequality adh>b holds, the increase in nb and, according to Eqn (19), also the increase in W proceed with progressive acceleration up to fairly high values.However, when adh<b, the increase in the nb andW values slows down until steady-state values are attained. The critical point for the transition from one kinetic mode to the other is determined by the equality: (22) had=b. It can be seen from Eqn (21) that the exponential quantity m includes the kb value, which is an Arrhenius function of temper- ature. Therefore, in the combustion mode, i.e. for adh>b, the temperature dependences of nb and Wb contain a positive Boltzmann factor in the exponential quantity (the double expo- nent law) and, thus, they differ fundamentally from the Arrhenius law.When deriving the analytical expression for the variation of the BCP rate as a function of temperature,59 we took into account the fact that the contribution of w0 to the rate of variation of CC concentrations in a propagating combustion process is insignif- icant, except for the very early steps of the process.This provides grounds for neglecting the w0 value in Eqns (6) and (7) for the time V V Azatyan (23) t>t0^2.5/j. Integration of the equation obtained in this way from Eqn (7) dn dt a Of ¢§ gUn, (24) f0 exp n a n0 exp t0 with allowance for the f value [see Eqn (8)] and for the temperature dependence of kb, gave the following expression for the temper- ature variation of the concentration of chain carriers:59 dt . ¢§ Eb ¢§g RT

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Russian Chemical Reviews 68 (12) 1041 ± 1064 (1999) Biosensors for the determination of environmental inhibitors of enzymes G A Evtugyn, H C Budnikov, E B Nikolskaya Contents I. Introduction II. Biosensors: general characteristics and functional features III. Determination of kinetic parameters of inhibition IV. Optimisation of operating conditions for inhibitor assays V. Cholinesterase-based biosensors VI. Conclusion Abstract. Characteristic features of functioning and practical application of enzyme-based biosensors for the determination of environmental pollutants as enzyme inhibitors are considered with special emphasis on the influence of the methods used for the measurement of the rates of enzymic reactions, of enzyme immobilisation procedure and of the composition of the reaction medium on the analytical characteristics of inhibitor assays.The published data on the development of biosensors for detecting pesticides and heavy metals are surveyed. Special attention is given to the use of cholinesterase-based biosensors in environ- mental and analytical monitoring. The approaches to the estima- tion of kinetic parameters of inhibition are reviewed and the factors determining the selectivity and sensitivity of inhibitor assays in environmental objects are analysed. The bibliography includes 195 references. I. Introduction Biochemical analytical methods have assumed great importance in the last years. They provide information about qualitative and quantitative composition of a subject under study on the basis of in vitro and in vivo studies of biochemical processes, which involve enzymes, antibodies, DNA, etc.The main advantage of biochem- ical methods is that the formation of analytical signals simulates the processes occurring in the living beings, which offers a unique possibility for rapid and selective detection of biologically active substances and elucidation of the mechanism of their effects on living organisms.1, 2 Enzymic methods of analysis relate to biochemical methods. Enzymes have long attracted the attention of analytical chemists as highly efficient protein catalysts.2, 3 By virtue of their unique properties, such as high selectivity and sensitivity towards specific substrates and functioning in aqueous media at ambient temper- atures,4, 5 enzymes appear to be ideal analytical reagents for the determination of biologically active substances by diverse kinetic G A Evtugyn,HC Budnikov Kazan State University, ul.Kremlevskaya 18, 420008 Kazan, Russian Federation. Fax (7-843) 238 04 12. Tel. (7-843) 231 54 16 E B Nikolskaya IM Sechenov Institute of Evolution Physiology and Biochemistry, Russian Academy of Sciences, prosp. MToreza 44, 194223 St. Petersburg, Russian Federation. Fax (7-812) 552 30 12 Received 13 April 1999 Uspekhi Khimii 68 (12) 1142 ± 1167 (1999); translated by R L Birnova #1999 Russian Academy of Sciences and Turpion Ltd UDC 543.866 1041 1042 1050 1051 1053 1061 methods. The development of enzymic methods of analysis was, to a certain extent, stimulated by the increasing demand for clinical and toxicological assays (e.g., determination of major metabo- lites, diagnostics of various diseases and dysfunctions of discrete organs of an organism on the basis of changes in the composition of biological fluids, etc.).The possibility of usage of enzymes for the determination of specific substrates and effectors has been at the centre of discussion since the late 1940's, i.e., since the beginning of the first systematic ex vivo studies of enzyme properties. However, the difficulties connected with the isolation of enzymes, their low stability in solution and high cost of analyses impeded the practical application of novel methods. As a result, until the early 1970's it was thought that enzymic methods ranked below routine physicochemical analyses.The situation changed radically owing to the progress in engineering enzymology and industrial biotechnology (viz., to large-scale production of a great variety of enzyme preparations, which sharply reduced their cost) and to the development of novel methods for stabilisation of enzymes including their fixation (immobilisation) on inert carriers.6 Of no less importance for the development of enzymic methods of analysis was the creation of special-purpose measuring devices called biosensors.2, 3 In those, enzymes are integrated into a single system comprising a device for detecting biochemical signals (a sensor or a transducer) and a reaction medium containing essential stabilisers, effectors and other low-molecular-weight components required for their steady functioning.With the advent of biosensors, the maintenance of optimal experimental conditions and measurement of analytical signals have become routine laboratory procedures available for users having no special biochemical training or skills. It is the manufacture of biosensors that has opened the way for commercial application of enzymes for the determination of biologically active compounds. In 1996, the world's market of biosensors including those designed for clinical assays of glucose, urea, lactic acid and other essential metabolites (*70% of the total sales) amounted to US $ 508 million.7 The commercial success of medical biosensors is not exclu- sively due to their demand and considerable investment into scientific research.At a sufficiently high lower level of metabolite assays of the order of 1074 mol litre71, high selectivity is the main requirement for the method. Thus, the advantages of enzymes as analytical reagents are especially well realised in biosensors. The situation with enzymic monitoring of environmental pollutants is different in principle. Only cholinesterase-based biosensors designed for the detection of residual organophos-1042 phorus pesticides in water and agricultural products 8 and immu- noenzymic diagnostica for specific haptens have found commer- cial application. The reasons preventing practical application of enzymic methods in environmental monitoring are due to specific features of subjects under study and more strict requirements for analytical characteristics of determination of environmental pollutants in comparison with subjects of clinical analysis.9 As the requirements for the maximum permissible concen- trations of toxic agents become more rigorous, novel procedures are required for the determination of pollutants (e.g., dioxins and pesticides in drinking water and foods, radionuclides, etc.) at the level of 10712 mol litre71 (Ref.10). Such low concentrations can be detected only in those cases where the pollutant is an effector (viz., an inhibitor or an activator) of the corresponding enzymic reaction rather than its substrate. Besides, qualitative and quantitative compositions of the main components of those items that subject to environmental and analytical monitoring (e.g., natural and sewage waters, soils, sewage water sediments, etc.) are much more diverse than the compositions of biological media analysed with the use of medical biosensors.This imposes additional demands on the selectivity of biochemical signals and the methods used for their measurement. Many compounds, such as salts of alkaline and alkaline-earth metals, biogenic amines, phenols and surfactants, influence not only the rate but also the sensitivity of enzymic reactions with respect to inhibitors. Low temporal stability of specimens, high degree of mineralisation (which is especially apparent in the case of sewage water and silt extracts), the presence of microheteroge- neities (suspensions, emulsions), coloured compounds, etc., also restrict potential applications of biosensors.These limitations can be overcome through the use of special methodological approaches, since the use of biosensors for the analysis of environmental subjects differs in principle from their application for medical purposes.11 The present review summarises the published approaches to the development of biosensors designed for the determination of environmental pollutants as enzyme inhibitors and demonstrates the potentials of purposeful control over the selectivity and sensitivity of biosensors used for the assay of toxic substances. II. Biosensors: general characteristics and functional features The term biosensors refers to devices which contain a biological element (an enzyme or a cell culture, microorganisms, biological tissues, antibodies, etc.) and a means for measurement of the reaction (response) of this element to the alteration of the qualitative composition of the medium.3, 11 Some authors apply the term `biosensors' to the supramolecular level of organisation of a biological component and thus separate enzyme-based or immune sensors from microbial and tissue biosensors.12 A general scheme of a biosensor assembly is shown in Fig.1. Irrespective of its application, each biosensor consists of three elements, which determine its operational and analytical charac- teristics, viz., a sensor, a biological component itself and a reaction medium.In addition to the subject to be assayed (e.g., a substrate or an effector), the latter includes enzyme stabilisers, a supporting electrolyte, screening reagents, buffer components, etc. The choice of a method used for detecting analytical signals determines not only the convenience of operation but also the limit of detection of the substrate. Preference is given to versatile methods and detectors of similar type for measuring the rates of different enzymic reactions. Electrochemical biosensors are rep- resented by Clark's oxygen electrode, amperometric hydrogen peroxide sensors, pH-metric electrodes, field transistors and conductometric cells. In recent years, planar thin- and thick- filmed electrodes prepared from metals and carbonaceous materi- als which are photolithographed or screen-printed onto plastic or ceramic planar surfaces have received wide acceptance.13, 14 G A Evtugyn, H C Budnikov, E B Nikolskaya 123 Figure 1.A schematic diagram of a biosensor; (1) sensor (electrode); (2) biological element; (3) working solution, i.e., the reaction medium containing the analysed sample. The selectivity of signal detection with a biosensor is provided by the selection of operation conditions for the measurement (the electrode potential, the composition of the reaction medium, etc.) as well as by optimisation of the biosensor design. Thus the use of electron transfer mediators makes it possible to decrease the working potential of an electrode and to exclude side oxidation (reduction) of electroreactive admixtures present in the speci- men.15 The use of additional membranes diminishes the effect of the buffer capacity of the specimen and the pH of the working solution to the response of pH-metric biosensors.16 The general requirements for biosensors developed for envi- ronmental and analytical monitoring were formulated by the US Environmental Protection Agency (EPA).11, 17 They are as follows: low cost of assays (US $ 1 ± 15 per assay), the possibility of detection of toxic substances at the level of 1 mg litre71 ±1 mg litre71 and within the range of detectable concentrations of not less than two orders of magnitude, dura- tion of analysis of 1 ± 60 min, short-term (1 ± 3 h) training of laboratory personnel, application of zero or minimum supply voltage, specificity with respect to one or a group of related compounds, simplicity of the sampling procedure and portability of the device (which is designed to be easily moved by one man or transported by car).1. The choice of an enzyme The choice of biological elements (enzymes) to be incorporated into biosensors is usually made on the basis of a concept about the pathways of biochemical transformations of a particular com- pound for substrate assays or the mechanism of its toxic effect for inhibitor assays. Obviously, the great diversity of living organisms and well-characterised enzymes makes it possible in principle to detect any biologically active compound in various ways.A combination of several enzymes providing consecutive conver- sions of the substrate in a biosensor assembly adds new features. In each particular case, the accessibility of an enzyme (simplicity of the isolation procedure or commercial availability), its storage and immobilisation stability, efficiency of measurement proce- dures, detection limits and ranges of measurable substrate/ inhibitor concentration are taken into account. A change in the analytical characteristics of inhibitor assays within the same enzyme ± substrate system is attained by varying the sources of the enzyme, the method used for its immobilisation, the measure- ment conditions and the steps involved in the incubation proce- dure.The data concerning the application of enzymes for assaying environmental pollutants as inhibitors [with the exception of cholinesterase (ChE) biosensors, see below] published in 1993 ± 1998 are presented in Table 1. The results obtained in earlier studies are documented in several reviews.1, 9±11 It should be noted that far from every publication provides analytical characteristics of inhibitor assays, viz., detection limits or the lowest limits of measured concentrations, sensitivity of assays, range of detectable concentrations, etc. In many cases, biosensors are used to merely demonstrate the efficiency of an idea of determination of toxic substances with immobilised enzymes (see, e.g., Ref. 41) or for calculation of kinetic parameters of inhibition, usually, with reference to inhibition of free enzymes.Table 1.Some characteristics of methods used for inhibitor assays in the environment and the experimental parameters, viz., concentration range (c/mol litre71), limits of detection (LD/mol litre71) and inhibitor concentrations inducing a 50% decrease in the biosensor response (I50/mol litre71). Method Enzyme Oxidoreductases amperometric Cytochrome (cyclic voltamperometry) oxidase (EC 1.9.3.1) amperometric (linear voltamperometry) pulse amperometry, chronocoulometry UV/visible spectrophotometry amperometric (cyclic Tyrosinase voltamperometry) (EC 1.14.18.1) amperometric (Clark's electrode) amperometric (linear voltamperometry) amperometric (by quinone reduction current) amperometric (detection of stationary signals) amperometric (cyclic and pulse amperometry, chronocoulometry) Conditions of measurement and enzyme stabilisation the enzyme was incorporated into a carbon paste together with azolectin and cytochrome c from horse heart; phosphate buffer, pH 7.0 electrochemical polymerisation of pyrrol on a planar gold electrode with subsequent immobilisation with glutaraldehyde; phosphate buffer, pH 7.0 immobilisation by cross-linking with glutaraldehyde on the surface of a glassy carbon electrode in the presence of a mediator (1,2-naphthoquinone-4-sulfonate); phosphate buffer, pH 7.0 the enzyme, myoglobin and hemoglobin were incorporated into a sol-gel matrix the enzyme was incorporated into a carbon paste (4%); phosphate buffer, pH 7.0; catechol was used as a substrate sorption on paper with subsequent application of a dialysis membrane; hexane saturated with phosphate buffer, pH 7.0; catechol was used as a substrate the enzyme was immobilised on the electrode surface; phosphate buffer, pH 7.4; phenol was used as a substrate immobilisation in polypyrrol formed on the electrode surface by oxidation of substituted pyrrol; catechol, dopamine, L-DOPA and epinephrine were used as substrates; air-saturated phosphate buffer, pH 6.5 the enzyme was incorporated into a carbon paste prepared from mineral oil without a cross-linking reagent (immersion or planar electrodes); phenol was used as a substrate; phosphate buffer, pH 7.4 the enzyme was immobilised on the surface of a glassy carbon electrode by cross-linking with glutaraldehyde without the carrier; phosphate buffer, pH 7.0; potassium ferricyanide is used as a mediator Inhibitor cyanides atrazine benzoic acid nitrogen(II) oxide and carbon monoxide cyanides phenylthiourea N,N-diphenylthiourea ethylphenylthiourea benzylthiourea isopropylthiourea allylthiourea benzoic acid, thiourea, propyl gallate 3,4-dichlorophenol chloroisopropyl phenylcarbamate 3-chloroaniline atrazine cyanide hydrazine methylhydrazine dimethylhydrazine cyanides Ref.Parameters of determination of the inhibitor I50 LD c(0.5 ± 12)61076 7 7 18 561076 (see a) 7 7 19 161075 (see b) 161075± 561074 361076 (see c) 7 20 around 1076 7 7 21 161075 7 7 18 0.8061076 0.01361076 7 22 1.8361076 0.04561076 7 3.0861076 0.07861076 7 4.5861076 0.10461076 7 5.5561076 0.13761076 7 6.9161076 0.18161076 77 7 23 7 0.461076 7 7 24 261076 7 7 261076 7 7 461076 7 7 0.0261076 7 7 7 7 4.561074 25 7 7 5.561074 7 7 6.361074 7 7 26 around 1073± 1074 (see c)Table 1 (continured).Method Enzyme amperometric (detection Tyrosinase of stationary signals) (EC 1.14.18.1) amperometric (linear voltamperometry) the same amperometric, coulometric amperometric, Peroxidase photometric (EC 1.11.1.7) amperometric (cyclic voltamperometry) Conditions of measurement and enzyme stabilisation screen-printed carbon electrodes placed onto a ceramic mount (working area 165 mm); the enzyme was incorporated into a carbon paste.catechol was used as a substrate; phosphate buffer, pH 7.4 tyrosinase isolated from meadow mushroom (Agaricus campestris) carposomes was incorporated into a carbon paste; catechol was used as a substrate; phosphate buffer, pH 7.4 the enzyme was incorporated into a hydroxycellulose cryogel; the measurements were performed in chloroform, chlorobenzene and 1,2-dichlorobenzene containing tetraethylammonium perchlorate tyrosinase was immobilised on the surface of Clark's electrode together with a mediator (1,2-naphthoquinone-4-sulfonate) the enzyme was immobilised by incorporation into a polymeric film (Eastman Kodak 55D) on the surface of a Pt electrode; the measurements were performed in acetonitrile, methanol, acetone and butan-2-ol containing 1.5% of water and butanone peroxide; phenylenediamine was used as a substrate the enzyme was immobilised by incorporation into a polymeric film (Eastman Kodak 55D) on the surface of a glassy carbon electrode; the measurements were performed in acetonitrile, methanol and acetone containing 1.5% of water, butanone peroxide and dimethylferrocene Inhibitor diethyl dithiocarbamate benzoic acid 4,6-dinitro-o-cresol thiourea 2,4-D diethyl dithiocarbamate benzoic acid thiourea benzoic acid, in chloroform, in chlorobenzene, in 1,2-dichlorobenzene thiourea, in chloroform, in chlorobenzene, in 1,2-dichlorobenzene 2-mercaptoethanol, in chloroform, in chlorobenzene, in 1,2-dichlorobenzene diazinon dichlorvos thiourea, in methanol, in acetone, in butan-2-ol, in acetonitrile ethylenethiourea, in methanol, in acetone, in butan-2-ol, in acetonitrile methyl isocyanate, in methanol, in butan-2-ol 2-mercaptoethanol, in acetonitrile thiourea, in acetonitrile, in methanol, in acetone Ref.Parameters of determination of the inhibitor I50 LD c 9.261075 261076 7 27 161074 561076 7 261074 961076 7 461076 7 77 7 761076 7 7 461075 28 7 7 5.861075 7 7 6.261075 29 0.0561076 (1 ± 16)61076 7 161076 (10 ± 85)61076 7 561076 (10 ± 30)61076 7 0.561076 (1 ± 35)61076 7 261076 (5 ± 100)61076 7 161075 (25 ± 85)61076 7 161077 (0.5 ± 70)61076 7 0.0561076 (0.1 ± 35)61076 7 0.561076 (0.5 ± 70)61076 7161073 561076 7 30 261075 7.561078 7 31 ± 33 7 7 1.261074 7 7 0.561074 7 7 2.161074 7 7 19.461074 7 7 1.361074 7 7 0.961074 7 7 4.561074 7 7 10.461074 7 7 1.861074 7 7 3.761074 7 7 20.261074 34 7 7 2.461074 7 7 18.961074 7 7 3.561074Table 1 (continured).Method Enzyme amperometric (linear Peroxidase voltamperometry) (EC 1.11.1.7) conductometric (planar Alcohol oxidase Pt microelectrodes) (EC 1.1.3.13) amperometric (detection Catalase of stationary signals) (EC 1.11.1.6) Aldehyde dehydro- amperometric (linear genase (EC 1.2.1.5) voltamperometry) and diaphorase (EC 1.8.1.4) potentiometric Glucose oxidase (pH-sensitive ion-selective (EC 1.1.3.4) field transistor, ISFT) the same b-Glycosidase " (EC 3.2.1.21), mandelonitrile lyase (EC 1.11.1.7), peroxidase Conditions of measurement and enzyme stabilisation the enzyme was incorporated into a methyltriethoxysilane gel on the surface of a glassy carbon electrode modified with an Os-containing polymer; phosphate buffer under argon, pH 5.5, was used the enzyme was immobilised on the active surface of a field transistor by vapour of glutaraldehyde into a bovine serum albumin film and dextrin; Tris-buffer, pH 7.5; formaldehyde was used as a substrate the enzyme was immobilised in polyacrylamide gel; the membranes were fixed directly on the teflon membrane of Clark's electrode; the measurements were performed in a flow-injection regime; hydrogen peroxide was used as a substrate; phosphate buffer, pH 7.0 diaphorase was immobilised on a membrane prepared from a copolymer of styrylpyridine and vinyl alcohol (PVA-SbQ); aldehyde dehydrogenase was immobilised on the same mem- brane or added to the working solution; phosphate buffer pH 7.5; propionaldehyde was used as a substrate the enzyme was immobilised on the active surface of a field transistor by vapour of glutaraldehyde into a bovine serum albumin film the enzyme was immobilised on the active surface of a field transistor by vapour of glutaraldehyde into a bovine serum albumin film and dextrin; Tris-buffer, pH 7.5 the three enzymes were co-immobilised on the active surface of a field transistor by vapour of glutaraldehyde into a bovine serum albumin film; citrate buffer, pH 6.0 Inhibitor methyl isocyanate d Hg(II) fluorides cyanides cineb and maneb cineb Cu(II) Co Sn Pb Ni Cd Ag(I) cyanogenic glycosides (amygdalin) Parameters of determination of the inhibitor c(4 ± 40)61076 (1 ± 100)61076 10 e 100 e 0.15 e 7.4 e, 1.5 e 4.0 ± 3.0 f 4.2 ± 3.2 f 3.8 ± 2.0 f 3.5 ± 2.2 f 3.8 ± 2.0 f 2.0 ± 1.5 f (1 ± 100)61076 (1 ± 30)61075 Ref.I50 LD 561077 7 35 7 7 36 1 e 7 37 1.5 e 7 7 7 38, 39 7 7 7 7 40, 41 7 7 7 7 7 7 7 7 7 7 7 7 37 7 7 42Table 1 (continured). Enzyme Miscellaneous oxidoreductases L-glycero- phosphate oxidase, (Clark's electrode) pyruvate oxidase, lactate dehydro- genase alcohol oxidase glycerol 3-phos- phate oxidase choline oxidase, glutathion oxidase, sarcosine oxidase Hydrolytic enzymes Liver esterase (EC 3.1.1.1) Acid phosphatase (EC 3.1.3.2) Method amperometric amperometric (detection of stationary signals) photometric (by NADH absorption) amperometric (detection of stationary signals) potentiometric Inhibitor Conditions of measurement and enzyme stabilisation HgCl2 the enzyme was immobilised on the surface of Clark's electrode; Tris-buffer HgCl2 HgCl2 AgNO3 CdCl2 ZnCl2 Pb(CH3COO)2 CuSO4 Hg(II) the enzymes were immobilised by cross-linking using Cu(II) glutaraldehyde into a bovine serum albumin film or on Hg(II) activated nylon (Immobilon), on the surface of a planar ruthenium-coated (0.5%) carbon-paste electrode; Dulbecco's Hg(II) phosphate-buffered saline was used; V(V) incubation for 5 min Se(IV) Ni fluorides esterase and NAD-dependent alcohol dehydrogenase were co-immobilised on glass; the product of enzymic hydrolysis of the substrate (ethyl propionate), i.e., ethanol oxidised by alcohol dehydrogenase, was assayed in a flow-injection regime pesticides h, 46 the enzyme (tissue sections of potato tubers) was co-immobilised malathion, with glucose oxidase on activated nylon in the presence of methylparathion, polyazotidine; glucose 6-phosphate was used as a substrate; paraoxon, phosphate buffer pH 6.6; the measurements were performed in citrate buffer pH 6.0 at 30 8C aldicarb Hg(II) the enzyme was incorporated into a polyacrylamide gel on a dialysis membrane; 4-nitrophenyl phosphate was used as a substrate Parameters of determination of the inhibitor c7 7 2061076 7 7 0.0561076 7 7 161076 7 7 0.161076 7 7 1061076 7 7 1061076 7 7 5061076 7 7 2561075 7777777(0.8 ± 8)61076 4.4 ± 15 e (3.0 ± 2.0) e 0.7 ± 12 e (0.7 ± 8.0) e 6.2 ± 18.3 e (3.1 ± 2.0) e 46 ± 125 e (46 ± 125) e 7 Ref. I50 LD 43 0.05 g 7 44 0.05 g 7 0.5 g 7 1.0 g 7 20 g 7 0.015 g 7 0.10 g 7 7 7 45 3 e (1.5) e 7 0.5 e (0.3) e 7 5 e (1.5) e 7 40 e (40) e 77 47 1Table 1 (continured).Method Enzyme potentiometric Alkaline phosphatase (EC 3.1.3.1) amperometric (detection of stationary signals) fluorimetric Urease (EC 3.5.1.5) conductometric (planar Pt microelectrodes) potentiometric (pH-ISFT) the same " Hydrodynamic method.b Flow-through method. c Calibration curve. d Thiourea, thiocyanates and azides do not interfere with the determination up to concentration 161073 mol litre71. e mg litre71. f The values of apc=7log c (mol litre71) are given. gmg litre71. h The parameters for the listed pesticides were obtained using bienzyme and hybrid (in parentheses) biosensors. Inhibitor Conditions of measurement and enzyme stabilisation Hg(II) the enzyme was immobilised by cross-linking using glutaraldehyde into a bovine serum albumin film on the surface of a dialysis membrane; Tris-buffer pH 7.0, 34 8C; a-D-glucose 1-phosphate was used as a substrate sorption together with glucose oxidase on the surface of a planar ruthenium-coated carbon electrode; glucose 6-phosphate was used as a substrate; the rate of oxidation of hydrogen peroxide, the product of enzymic oxidation of glucose formed upon substrate hydrolysis, was measured in glycine buffer, pH 9.0; incubation for 10 min dichlorvos Hg(II) the enzyme was immobilised on a glass surface; the measurements were performed in a flow-type detector the enzyme was immobilised on the active surface of a field transistor Hg(II) by vapour of glutaraldehyde into a bovine serum albumin film in the presence of a stabilising agent (glycerol); Tris-buffer pH 7.4; urea was used as a substrate; the inhibition was studied in a kinetic regime by the rate of signal alteration after addition of the substrate Cu(II) Cd Co(II) Pb Sr Ag sorption on the surface of a field transistor coated with a polymeric film (Nafion); Tris-buffer, pH 7.0; urea was used as a substrate Hg(II) Cu(I) Hg(II) the enzyme was immobilised by glutaraldehyde on the active surface of a field transistor into a nitrocellulose film; Tris-buffer, pH 7.4; urea was used as a substrate Cu Co Cd Ni Pb Sn EDTA sorption on the surface of a field transistor coated with a film of poly(4-vinylstyrene); Tris-buffer, pH 7.0; the activity of the enzyme recovered following inhibition with CuCl2 was determined; urea was used as a substrate Parameters of determination of the inhibitor c11078± 1075 (see c) 0.5 ± 100 g (1 ± 50)61076 (2 ± 100)61076 (5 ± 200)61076 (1 ± 50)61075 (2 ± 500)61075 (1 ± 500)61074 7778.7 ± 3.5 f 6.7 ± 3.3 f 5.8 ± 2.8 f 5.8 ± 2.2 f 5 ± 3 f 4.8 ± 3.2 f 7 7 7 5 ± 500 e Ref.I50 LD7 7 47 661078 7 48 (see c) 7 7 49 7 7 50 7 7 7 7 7 7 7 7 7 7 0.261076 7 51 1.561076 7 361076 7 7 7 40 7 7 7 7 7 7 7 7 7 7 7 7 521048 This complicates comparison of experimental results obtained by different authors. The majority of publications devoted to the practical applica- tion of analytical enzymic methods deal with the studies of biosensors including peroxidase, tyrosinase and urease. Other enzymic systems are described in only a limited number of papers, although their potentials are rather high.Different organic compounds are determined with horserad- ish peroxidase, which is resistant to denaturing reagents and elevated temperatures and manifests high specific activity, and is a reliable and sensitive tool for determining the rates of enzymic reactions. The rate of peroxidase-based oxidation of different substrates can be used for determining the concentration of hydrogen peroxide; therefore, this enzyme is a common constitu- ent of two-enzyme biosensors designed for assaying substrates or effectors of oxidoreductases.2, 3 Tyrosinase (polyphenol oxidase) fulfils similar biochemical functions by oxidising various biogenic phenols.Peroxidase and tyrosinase inhibitors can block the active sites of enzymes (due to the complex formation with metal cofactors and uncoupling of an electron transfer chain) or manifest non- specific inactivating effects by disturbing the tertiary structures of proteins. For example, the results of Hg(II) and Fe(III) assays which employed peroxidase with photometric detection of sig- nals 53 ± 55 and phenylthiourea derivatives assays using an electro- chemical tyrosinase biosensor 22 (both processes were characterised by low detection limits) have made it possible to recommend these methods for direct assays of toxic compounds in environmental objects without sample concentration. The main advantage of peroxidase and tyrosinase in compar- ison with other enzymes consists in the possibility of using them as biosensor constituents for measurements in organic or aqueous- organic media,56 ± 58 including inhibitor assays.22, 29, 31 ± 33, 59, 60 This makes it possible to use biosensors as detectors in liquid chromatography 61 and to carry out inhibitor assays directly on organic extracts.On the whole, the stability of enzymes in organic solvents is lower than in aqueous media, however, definite approaches to their stabilisation do exist.62 In the case of inhibitor biosensors able to function in aqueous-organic and non-aqueous media, enzymes are incorporated into hydrophilic polymers based on polyaniline,61 a hydrophilic vinyl chloride and pyrrole copolymer Eastman AQ31 ± 34 and hydrogels based on hydroxycellulose 29 and methyltriethoxysilane.35 Electrical con- ductivity is achieved by adding 1%± 20% of a buffer solution or a tetraalkylammonium salt to the organic solvent.Kinetic param- eters of both the enzymic reaction and the enzyme ± inhibitor interaction depend on the nature of the solvent and the inhibitor. Usually, the presence of polar organic solvents miscible with water decreases the sensitivity of the biosensor to reversible inhibitors of tyrosinase and peroxidase, such as thiourea and dithiocarbamate derivatives. Nonpolar solvents have insignificant effects on the inhibition kinetics. Urease catalyses the hydrolysis of urea. High substrate specificity of the enzyme and ease of potentiometric or photo- metric determination of the hydrolysis product, viz., ammonia, provide the practical application of urease for the determination of residual heavy metal salts in environmental objects.Thus a flow system consisting of 26 glass capillaries with a total volume of 100 ml was developed for the determination of urease inhibitors in natural and sewage waters. Urease was immobilised on the inner surface of the gelatinised capillaries by physical sorption.63 A water sample and a substrate solution were consecutively pumped through the capillary tubes with a peristal- tic pump for 15 ± 240 and 20 min, respectively, and the urease activity was measured from the concentration of the ammonia formed using a potentiometric method.The activity of immobi- lised urease was preserved after continuous pumping of solutions for 8 h. The use of this system made it possible to detect the presence of carbaryl (161074 mol litre71) (4 h) and copper(II) salts (161077 and 161079 mol litre71) (1 and 4 h). A similar G A Evtugyn, H C Budnikov, E B Nikolskaya indicator system based on a urease reactor (a column filled with glass beads with immobilised urease) was used to evaluate the general contamination of waters with heavy metals.64 Ammonia concentration was measured by a photometric (with the Nessler reagent) or a potentiometric method. The detection limits of the majority of heavy metals were (1 ± 2)61076 mol litre71, those for iron were 1078± 1079 mol litre71. The selectivity of individual metal assays could not be reached. This system was tested in the monitoring of pollution of drinking water sources in the Moscow Region.Urease biosensors based on pH-sensitive field transistors possess enhanced analytical characteristics for the determination of metal ions. The progress in this area was achieved owing to both the use of ultrathin membranes contacting with the electrode and the coating of biosensors with ion-exchange polymers, which favour the concentration of cations on the electrode surface.51, 52 Microorganisms and sections of biological tissues incorpo- rated into biosensors may sometimes be regarded as sources of the corresponding enzymic activity. This can increase the stability of the response and reduce the cost of biosensors, since the enzymes are in a customary microenvironment in the presence of the necessary stabilising agents and effectors.The cells possess natural mechanisms, which protect enzymes against autooxida- tion and hydrolysis by proteolytic enzymes. Besides, inactivated enzymes are continuously reproduced owing to catabolic intra- cellular processes. Some applications of tissue sections of mush- room carposomes as a source of tyrosinase 28 and of potato as a source of acid phosphatase 46 are given in Table 1. Biosensors based on the use of complete biochemical cycles in which inhibition of a particular biochemical function is measured as an analytical signal offer radically new possibilities. Special mention should first be made of the use of biosensors for the assay of herbicides (algicides) and evaluation of general toxicity in respect of the respiratory activity of microorganisms.Biosensors designed for measuring concentrations of phyto- toxic chemical compounds incorporate chloroplasts as the bio- logical components (as whole cells of unicellular microalgae 65 ± 67 or within thylakoid membranes isolated from these cells 68). Illumination of chlorophyll generates a photocurrent measured as the current of oxidation of electron acceptors added to the solution instead of natural acceptors, viz., carbon dioxide. Ferricyanide,68 soluble ferrocene derivatives 65, 66 and diamino- durene 67 are used as mediators. The reduction of the photo- current is a measure of the herbicide effect of the toxicants.Thus in the case of Synechoccus sp. cells, the detection limit of sym- triazine-based pesticides is 2 mg litre71, that in the case of Chlorella vulgaris cells is *1 mg litre71. When thylakoid mem- branes isolated from spinach cell cultures were used, a 50% reduction of the photocurrent was observed at the following concentrations (mol litre71): 161076 (diurone), 961076 (atra- zine), 5.561074 (sodium nitrite), 1.761073 (sodium sulfite), 7.261073 [Cu(II) salts], 161073 (Pb), 1.861073 [Hg(II)], 261073 (Zn) and 461073 (Cd). These data suggest that biosensors based on chloroplasts and microalgae are promising tools, especially for the determination of residual herbicides. However, these biosensors cannot compete with enzymic ones in terms of sensitivity of metal ion assays (see Table 1).Biochemical oxygen demand (BOD) is measured with the aid of microbial biosensors designed for detection of toxic substances in aqueous media.69 ± 71 The magnitude of BOD can be established within 15 ± 40 min (cf. 5 ± 20 days in the titration method) by the rate of oxygen consumption in cells immobilised on the oxygen electrode membrane. The presence of toxic substances decreases the rate of oxygen consumption. In toxicity studies, a sample is introduced simultaneously with the nutrient (determination of respiratory activity) or after it (by a characteristic bend on the respiration curve).72 ± 74 Commercial systems for monitoring of the toxicity of sewage waters (e.g., Toxiguard,73 RODTOX72, 75) were developed based on some microbial biosensors.Microbial biosensors for toxicity studies allow the detection of differentBiosensors for the determination of environmental inhibitors of enzymes substances with a broad spectrum of toxic effects ranging from disturbance of transmembrane oxygen transport (surfactants, petroleum products) to uncoupling of oxidative phosphorylation and inhibition of oxidoreductases (phenols, cyanides, heavy metal ions). However, adaptive protective detoxification mechanisms decrease the sensitivity of microbial biosensors to individual groups of toxic agents in comparison with enzyme-based electro- des. Thus the detection limit of Cu(II) ions is 0.2 mg litre71 with the yeast-based electrode.1 In addition to respiratory activity, the intensity of lumines- cence of luciferase microorganisms (a biosensor version of the Microtox test) can serve as the measure of toxicity.76 The use of an electrode containing an immobilised revertant strain of Salmo- nella for the screening of chemical compounds possessing muta- genic activity has been described.77 2.Effect of enzyme immobilisation Immobilisation of enzymes usually causes significant alterations in the analytical characteristics of inhibitor assays. The main reasons for the change in the sensitivity of enzymes to inhibitors upon their transition from the native to the immobilised state are as follows:3, 5 (1) diffusion-dependent retardation of the reagent (substrate, inhibitor, reaction product) transfer to the membrane including steric hindrances in the accessibility of the active site of an enzyme for high-molecular-weight inhibitors and/or substrates (e.g., proteolytic enzymes), (2) changes in the enzyme micro- environment (buffer capacity, dielectric conductivity, ionic com- position) and concomitant changes in electrostatic interactions and activity coefficients of the reagents.There is no unequivocal evidence so far that the change in the enzyme affinity for a substrate or an inhibitor upon immobilisa- tion is due to the changes in the tertiary or quaternary structure of the protein. Physical sorption or covalent cross-linking of protein globules with bifunctional reagents results in modification of very few surface functional groups.Since active and allosteric sites occupy a small part of the protein globule surface, any change in the structure of protein globules upon immobilisation is hardly likely. Optimisation of the immobilisation procedure is aimed at retaining the activity of immobilised enzymes and implies mild, sparing treatment in order to preserve the microenvironment of the enzyme active site. This allows one to consider the enzyme- containing film on the biosensor surface as a solid solution where the native protein is uniformly distributed and `fixed' in a homogeneous medium, which is denser and more viscous than water. Despite its obvious arbitrary character, in many cases this model describes adequately the behaviour of biosensors incorpo- rating an immobilised enzyme.Diffusion-dependent retardation of the reagent transfer is the main reason for discrepancy in the results of inhibitor assays employing native and immobilised enzymes. Depending on the ratio of the rates of transfer of the substrate into the membrane and its enzymic conversions, one can distinguish two regimes of biosensor functioning 3, 78 which differ in stationary distribution of the substrate throughout the membrane. Kinetic regime. The rates of transfer and diffusion of a substrate from the solution into the membrane and its diffusion into the membrane are higher than the rate of its enzymic conversion. The reactants are uniformly distributed throughout the membrane and the active sites of enzymes are saturated with the substrate.In this case, the rate of the formation of the reaction product (and, consequently, the response of the biosensor) is determined exclusively by the kinetics of the enzymic reaction. This regime is particularly suitable for detection of enzyme effectors. However, any decrease in the enzymic activity caused by desorption, spontaneous inactivation or microbial degradation influences the biosensor response. The kinetic regime is observed in the case of low specific activities of immobilised enzymes, small thickness of the membrane and lack of the substrate transfer retardation at the membrane ± solution interface. 1049 Diffusion-dependent regime. If the activity of an enzyme is high, its diffusion-controlled transfer does not fully compensate for the decrease in the substrate concentration in the enzymic reaction.As a result, the membrane incorporating the enzyme is not saturated and part of the enzyme active sites are unoccupied. The electrode response is steady and remains at this level despite the variations in the activity of the immobilised enzyme unless the reaction regime is changed. The diffusion-dependent regime is observed in the case of high specific activity of the enzyme, retardation of the substrate transfer (e.g., when additional cover- ing membranes decreasing the rate of the substrate transfer from the solution into the enzyme-containing layer are used), etc.16 If the reaction occurs in the diffusion-dependent regime, the sensitivity of inhibitor assays is lower than in the kinetic regime.A decrease in the concentration of active sites of enzymes as a result of their interaction with their inhibitors is partly compensated for by the involvement of free, nonreacted sites in the enzymic reaction which occurs independently of the enzyme ± inhibitor interactions. Thus on going from the native to the immobilised enzyme, the limits of detection of Hg(II) with choline oxidase and glycerol 3-phosphate oxidase increase 2- and 2.5-fold, respec- tively, while the limits of detection of Hg(II) with lactate oxidase and glutamate oxidase, of cadmium with D-amino acid oxidase and of V(V) with alcohol oxidase increase more than 20-fold.44 Moreover, many metal ions inhibit native enzymes without influencing the activities of immobilised preparations.The limit of detection of paraoxon for a luminescent biosensor containing ChE, choline oxidase and peroxidase was 0.75 mg litre71 when ChE was used as a solute and the other two enzymes were immobilised in an acrylate gel and 125 mg litre71 when all the three enzymes were immobilised.79 The diffusion-dependent and kinetic regimes represent extreme functioning regimes for biosensors and are separated by a vast transition region where the biosensor response is deter- mined by both kinetic parameters of the enzymic reaction and the diffusion coefficients of the reagents. A comprehensive mathe- matical description of responses of amperometric and potentio- metric biosensors has been given.3, 80 ± 84 However, these models do not take account of the pH dependence of the rate of the enzymic reaction, which is a rough approximation, since the change in pH value of an enzyme-containing membrane can vary from 1 to 3 units during the measurements of biosensor responses.No simulation of the effect of inhibitor on responses of biosensors including stationary and dynamic responses caused by the addition of substrates (dynamic or transition responses) have been carried out thus far. This may be due to significant complication of the mathematical model in comparison with the description of kinetics of reactions of immobilised enzymes. Besides, in the majority of cases kinetic parameters of individual stages of the inhibition reaction essential for simulation are unknown.In the simplest case, the effect of an inhibitor can be described in terms of a mathematical model of the biosensor response in the absence of the inhibitor. In the case of uniform distribution of the inhibitor throughout the membrane, a decrease in the biosensor response can be simulated using the dependences of the biosensor response on the substrate concentration calculated for different rates of enzymic reaction. A change in the biosensor response as a result of its contact with the inhibitor will be similar to that observed in the case where a smaller concentration of the enzyme is used for the preparation of the membrane. This approach does not take into account the effect of enzyme immobilisation on the mechanism of inhibition.In particular, on going from native to immobilised ChE the mechanism of inhib- ition by N-methyl-4-piperidinyl benzylate or tacrine (1,2,3,4- tetrahydro-9-aminoacridine) 85 is changed, which is associated with steric factors (e.g., limited access of the inhibitor to the membrane or a change in the ratio of substrate/inhibitor concen- trations in the membrane in comparison with that in solution) rather than with changes in the mechanism of enzyme ± inhibitor1050 interactions. The necessity of estimating the contribution of the final rate of the inhibitor transfer into the membrane brings about additional problems. III. Determination of kinetic parameters of inhibition The choice of methods for quantitative evaluation of effects of inhibitors on immobilised enzymes incorporated into biosensors is determined by the type of interactions in `enzyme ¡¾ sub- strate ¡¾ inhibitor' systems and the procedure used for measuring of analytical signals implemented in biosensors.Hereinafter it is assumed that the enzymic reaction occurs in accordance with the classical Michaelis ¡¾ Menten mechanism:4, 5 E+S (1) E, E S P where E is the enzyme, S is the substrate, ES is the enzyme ¡¾ substrate complex and P is the enzymic reaction product. An important feature of the inhibition mechanism is the reversibility of interaction of the inhibitor with the active site of the enzyme. It is necessary to distinguish between reversible and irreversible enzyme inhibition upon its reaction with the inhibitor in the absence of the substrate depending on the type of dissociation of the enzyme ¡¾ inhibitor complex.k1 k2 (2) E I E+I (E I) 0 E, k71 I0 I00 where I is the inhibitor, E7I is the enzyme ¡¾ inhibitor complex and I0 and I00 are the inhibitor decomposition products. If the dissociation of the E7I complex occurs with simultaneous decomposition of the original inhibitor, the inhibition is irrever- sible (k2 44 k71); if the dissociation of the complex into the original molecules prevails (k71 44 k2), the inhibition is rever- sible. The indication to the type of inhibition does not provide any information about the possibility of restoration of enzymic activity after the interaction of the inhibitor with the enzyme.Thus carbamates cause irreversible inhibition of ChE, but spontaneous hydrolysis of carbamoylated ChE occurs very fast and the enzyme activity is completely restored. The reactivation rate constant of carbamoylated acetyl cholinesterase (AChE) from electric eel is 0.012 min71, while that of phosphorylated AChE is 1075 min71 and the corresponding value for the hydrolysis of the acetylated enzyme formed upon enzymic cleavage of the substrate is 66105 min71 (see Ref. 86). Inhib- ition of urease with heavy metal ions is also expected to be reversible; however, the restoration of enzymic activity requires the use of complexones, e.g., EDTA.51, 52 Toxic irreversible inhibitors, e.g., contaminants-xenobiotics, are more dangerous to living organisms than reversible inhibitors because of the low rate of spontaneous reactivation and the lack of efficient detox- ification mechanisms.1. Irreversible inhibition Irreversible inhibition is evaluated quantitatively by the level of enzymic activity before and after the contact of the enzyme with the inhibitor. If the concentration of the inhibitor is much higher than that of the active sites of enzymes, irreversible inhibition will be described by a second-order kinetic equation 87 (3) ln u0 a kIIcIt, ut where u0 and ut are the initial rates of the enzymic reactions before and after the contact of the enzyme with the inhibitor, respec- tively, kII is the inhibition bimolecular rate constant, cI is the G A Evtugyn, H C Budnikov, E B Nikolskaya concentration of the inhibitor and t is the duration of the enzyme ¡¾ inhibitor interaction (incubation step).The value of kII depends on the nature of the enzyme and the inhibitor as well as on the reaction conditions and rate constants in separate steps of the reaction. In particular, at t?0: . kII? k1k2 k¢§1 a k2 The value of kII describes the inhibitory activity of the toxicant better than I50, since it allows comparison of the results obtained under different experimental conditions, particularly, for differ- ent t. If the concentrations of the enzyme active sites and the inhibitor are comparable, the dependence of the relative decrease in the rate of the enzymic reaction on the inhibitor concentration for strictly irreversible inhibition is linear.However, this depend- ence, as applied to biosensors, is not realised, since the specific activity of immobilised enzymes should be rather high in order to provide high accuracy of measurement of the response. In the case of irreversible inhibition (i.e., at low rate of spontaneous reactivation), the inhibiting effect depends on the time of incubation and the calibration curve for the inhibitor assay is linear in the cI7ln(u0/ut) coordinates and passes through the origin of the coordinates. Moreover, the relative decrease in the activity of the enzyme (the biosensor response) as a result of incubation does not depend on the concentration of the substrate, which is especially important for optimisation of conditions of inhibitor assays. 2.Reversible inhibition Akinetic description of reversible inhibition is more complex than that of irreversible one, since the interaction of the enzyme active site with the substrate and the inhibitor may occur in different ways: k1 k2 E+S E S E+P k71 (4) I I Ki aKi k1 bk2 E I + S E I + P E S I ak71 Here E7S7I is the enzyme ¡¾ substrate ¡¾ inhibitor complex, Ki and aKi are the equilibrium constants of the enzyme ¡¾ inhibitor interaction, k1, k71, k2, ak71 and bk2 are the rate constants of the corresponding steps of the enzymic reaction. Usually, for the inhibitors a>1 and b<1, but at b?0 the coefficient a is <1.For enzyme activators, the kinetic mechanism of which is described by the same equation, a<1 and b>1. Knowing the dependences of reaction rates on concentrations of the substrate and the inhibitor, one can determine the type of inhibition and the values of Ki and aKi.88, 89 In actual practice, extreme cases may occur. Thus at a??, complete competitive inhibition takes place. The latter also includes the interaction of the enzyme with an irreversible inhibitor in the presence of the substrate. If a?0, the inhibition is referred to as non-competitive. The corresponding constants for the formation of inhibitory complexes, Ki and aKi, are termed as competitive and non- competitive ones. Mixed type of inhibition is characterised by a combination of competitive and non-competitive inhibition, i.e., in this case the inhibitor binds both to the active centre of the enzyme and the enzyme ¡¾ substrate complex.Incompetitive inhib- ition is a particular case of mixed inhibition, if a=1 and b=0. The competitive (K) and incompetitive (K0) components of mixed inhibition are calculated from the dependence of relative changes in the rate of the enzymic reaction on the concentrations of the substrate and the inhibitor: 0 (5) uui a KMO1 a cI=KU a cSO1 a cI=K0U , KMBiosensors for the determination of environmental inhibitors of enzymes where u0 and ui are the reaction rates in the absence and in the presence of the inhibitor and KM is the Michaelis constant of the enzymic reaction in the absence of the inhibitor.The total inhibiting effect is described by the reduced inhib- ition constant K Ki 7= KiK0i. i a K0i The constants kII and Ki reflect the sensitivity of the enzyme ¡¾ substrate system to the inhibitor. Their values can be used for the identification of inhibitors using a set of enzymes differing in their sensitivities to toxic agents. If the experiments are run under identical conditions (incubation time for irreversible inhibitors, substrate concentration for reversible inhibitors), it is sufficient to determine the relative changes in the rate of the enzymic reaction according to the formulae: irreversible inhibition ; (6) a ln u0 u ln u0 u i i 2 1 OkIIU1 OkIIU2 reversible inhibition 0 0 .a i ui 2 1 uu u OKi U1 OKi U2 The indices 1 and 2 refer to the kinetic data obtained in experiments with enzyme preparations differing in the degree of purity, the method of isolation, the source, etc.90 The relative sensitivity of enzymes to a definite group of inhibitors is established in a similar way. By illustration, the sensitivity of ChE isolated from different sources can be charac- terised by the relative value: exp , IIn=k0IIU n XlnOki where the index i relates to the enzyme the relative sensitivity of which is being calculated, the index 0 relates to the reference enzyme and n designates the number of inhibitors. The ratio of inhibition constants, kII, is established for each insecticide and is then summated for all the pesticides under study.91 Estimation of inhibition parameters of immobilised enzymes including those incorporated into biosensors can also be used in the case of kinetic regimes.Biosensors in which the enzyme is immobilised on the electrode surface (without any carrier) or in a thin hydrophilic layer which has no substantial effect on the diffusion coefficients of the substrate and the inhibitor are best suited for this purpose. Albumin, polysiloxanes, silicates, hydro- philic polymers of the Eastman AQ 55D type, etc., are used as carriers. The feasibility of formal equations for Michaelis ¡¾ Menten kinetics provides indirect evidence for the lack of diffusional limitations.In particular, for tyrosinase- or peroxidase-based biosensors the current (I) of the reduction of the reaction product (quinone) or of the oxidised form of the mediator can be used as a measure of the rate of the enzymic reaction. The calibration curve for the substrate is linearised in double reciprocal coordinates I717c¢§1 S , whereas the segment intercepted on the abscissa gives the value of K¢§1 M as well as the Lineweaver ¡¾ Burk equation. Other kinds of anamorphoses of kinetic curves known in homogenous enzymic reaction kinetics can also be used for this purpose. Thus a calibration curve for determining reversible competitive or incom- petitive inhibitors at a constant concentration of the substrate is linearised in the coordinates I717c¢§1 I ; the inhibition constant, Ki, is determined from the segment intercepted on the abscissa.24 In some cases, it is more preferable to use the degree of inhibition of the enzyme (Y) instead of the biosensor response.This offers a possibility of using the data obtained for different 1051 experimental conditions for calculation. For example, the value of Y relates to the response of an amperometric biosensor: cI cE¢§I (7) , a a DI I0 cI a Ki Y a cE¢§I a cE where cE7I, cE and cI are the concentrations of the enzyme ¡¾ inhibitor complex, of the free centres of the enzyme and of the inhibitor, respectively, DI is the change in the biosensor response upon inhibition and I0 is the biosensor response before the contact of the enzyme with the inhibitor.29, 31 ¡¾ 34 Derivation of Eqn (7) a priori entails a fully competitive type of inhibition, since it is in this particular case that the decline of the response will be proportional to the concentration of the enzyme ¡¾ inhibitor complex.For practical purposes, the ratio (7) is linearised in the Y/ cI7Y coordinates (the Scatchard equation) to determine Ki:31 (8) I cY a ¢§ YKi a I0 Ki or in the Y/(17Y)7cI coordinates (the Hill equation):92 I cI (9) , a I50 x x cKi Y 1 ¢§ Y a 0 ¢§ I a I I where x is the empirical Hill coefficient; x=1 in the case of the Michaelis ¡¾ Menten mechanism, the value x for organic solvents increases to 3 ¡¾ 4.18, 22, 29 The Hill coefficient serves as a criterion of efficiency of the catalysis I0/K0M (K0M is the experimental value of the Michaelis constant) and the effect of the solvent on the kinetics of the enzymic reaction.Equations (7) ¡¾ (9) were obtained assuming that the response of the biosensor is proportional to the rate of the enzymic reaction. On the whole, the dependence u7I is sigmoidal and is approxi- mated by a linear dependence in a narrow range of substrate or inhibitor concentrations. In the case of irreversible inhibition, the inhibition bimolecu- lar constant is calculated from Eqn (3). By expressing u0/ui through the degree of inhibition Y, we obtain the following ratio: k (10) II a ¢§lnO1 ¢§ YU . cIt The plots are linearised in the ln (17Y)7cI coordinates or, if the degree of inhibition is expressed on a percentage basis, ln(1007Y)7cI.93 If the kinetic parameters kII and Ki are used for comparative characterisation of the enzyme ¡¾ inhibitor interaction, it should be remembered that they do not provide any direct information about the analytical characteristics of inhibitor assays, such as limits of detection or a concentration range.IV. Optimisation of operating conditions for inhibitor assays 1. Concentration of the substrate A crucial role in the choice of experimental conditions belongs to the choice of working concentrations of the substrate. In irrever- sible inhibition, the relative decrease in the rate of enzymic reactions does not depend on the measurement conditions. If a change in the rate of an enzymic reaction by 15%¡¾ 25% is taken as a statistically significant quantity, the limit of detection estimated by Eqn (3) is (0.1370.22)/kIIt In particular, the calculated values of the limit of detection of organophosphorus pesticides (t=2 ¡¾ 30 min, kII= 105¡¾ 107 mol71 litre min71), are in the range of 1079 ¡¾1077 mol litre71, which is consistent with the published data for enzymic assays of pesticides with free ChE.8, 86, 941052The value of u0/ut can be determined by analysis of any characteristics of the system which are proportional to the reagents concentration, e.g., the time needed for reaching a definite (normally, not more than 30%) level of substrate conversion: 0 uu à ti , t0 t where t0 and ti are the times of the enzymic reactions before and after the interaction of the enzyme with the inhibitor, respectively.The choice of small initial concentrations of substrates (cS 55 KM) has a number of advantages. If a substrate partic- ipates in several enzymic reactions (the presence of enzymes with low substrate specificity in the original preparation, significant differences in the kinetic parameters of isoenzymes) or if several substrates interact with the same enzyme, the general view of the equation describing the rate of the target enzymic reaction will be complicated: maxc1S S M á c1S M ià2u1 u10 à , Xiàn (11) c i 1 á Ki Ki max is the maximum rate of the enzymic reaction; the index S is sufficiently small (ciS55K0M), the equation is where u1 1 relates to the target enzymic reaction the rate of which is being determined and the index i relates to reactions between competing substrates/enzymes which occur in parallel.If the initial rate of the substrate ci simplified: K1 u0 1à u1max c1S , M and the effect of simultaneously occurring reactions can be neglected.95 For competitive and mixed ± competitive (a>1) inhibition, the limit of detection of the inhibitor decreases with an increase in the substrate concentration. At cS 55 KM, the limit of detection of environmental pollutants (Ki=1074 ± 1076 mol litre71) by the most common methods is (0.15 ± 0.25)Ki or in the range of 1075 ± 1077 mol litre71. At cS 44 KM, the limit of detection of non-competitive inhibitors (a=1) does not depend on the substrate concentra- tion.For incompetitive and mixed-non-competitive inhibition (a<1), the limit of detection of inhibitors decreases with an increase in the substrate concentration. In other cases, any increase in the substrate concentration will increase the limit of detection of the inhibitor. While estimating limits of detection, one should remember that inhibitory effects are established on the basis of results of two response measurements, viz., before and after the contact of the enzyme with the inhibitor. Therefore, the minimum value of the degree of inhibition is equal to the double measurement accuracy of the biosensor response to the substrate. The metrological characteristics of sensors incorporated into biosensors can be used as a preliminary estimate.Thus the magnitude of steady-state oxidation (or reduction) currents on solid electrodes can be measured with *1%± 2% accuracy, while that of the potential of the ion-selective electrode is *3%. Consequently, the mini- mum significant inhibition corresponds to a decrease in the amperometric and potentiometric responses by not less than 2%± 4% and *6%, respectively. Even in those cases where the calibration curve for the inhibitor is plotted from the absolute value of the response, it is assumed that the initial response of the biosensor (before its contact with the enzyme) is constant. In other words, the measurement error of the biosensor response after its contact with the inhibitor includes the measurement error of its response to the substrate.Therefore, whilst choosing a detection method (for irreversible, non-competitive, incompetitive and mixed ± non-competitive inhibitors) preference should be given G A Evtugyn, H C Budnikov, E B Nikolskaya to the methods which ensure more accurate measurements of the biosensor response rather than the possibility of detection of smaller concentrations of the substrate. In other cases, sensors allowing maximum reduction of working concentrations of the substrate will be the sensors of choice. 2. Enzyme reactivation The ability of enzymes to regain their activity after measurement of the inhibitory effect is a prerequisite for repeated application of biosensors.In the case of reversible inhibitors, this is achieved by washing of biosensors or enzyme-containing membranes with working buffer solutions which may contain stabilisers or enzyme cofactors preventing its inactivation. In some cases, natural reactivation occurs so slowly that under real experimental conditions it is necessary to use special reagents accelerating the dissociation of enzyme ± inhibitor complexes. For example, heavy metal assays make use of complexing or precipitating reagents, such as EDTA, potassium iodide, sulfur-containing organic compounds (cysteine, dithiothreitol, etc.);36, 51, 52, 96, 97 while anti- dotes (i.e., substances that counteract poisons in human beings) are often used as reactivating reagents in the analysis of organic inhibitors.Thus 2-PAM (2-pyridinaldoxime methiodide) or TMB-4 [1,10-trimethylenebis(4-hydroxyiminomethyl)pyridinium dibromide monohydrate] which accelerate hydrolysis of the phosphorylated active site of ChE are used for reactivation of the enzyme after its interaction with organophosphorus com- pounds.98 The efficiency of enzyme reactivation depends on the nature of the reactivating agent, the degree of enzyme inhibition, the nature of the carrier and the time elapsed between inhibition and reactivation. By illustration, phosphorylated ChE which is reac- tivated at *70% inhibition restores up to 95%± 100% of its initial activity, which makes it possible to use the same immobi- lised enzyme for 15 ± 20 assays.8 When heavy metal salts are used as inhibitors, the reactivation amounts to 70% ±80% if the enzyme is immobilised on nitrocellulose 96 and up to 100% if gelatine is used as a carrier.97 The effects of reactivating reagents, similar to those of inhibitors, are strictly selective.Therefore, the former can be used for separation of inhibitory effects of individual toxic agents present in the same solution as well as for selective determination of individual inhibitors. Thus it was proposed to determine Fe(III) ions according to their reactivating effects on peroxidase inhibited with cyanides or salicylic acid,53 and to determine EDTA accord- ing to its reactivating effect on urease inhibited with Cu(II) ions.52 An addition of sodium iodide allows one to separate the inhibitory effects of Ag(I), Hg(II) and other heavy metal ions on urease 51 and ChE.98 An excess of reactivating reagents can decrease the enzyme activity down to complete irreversible inhibition as a result of elimination of cofactors from the active site or the formation of more potent inhibitors, viz., the products formed in the reaction of the reactivating reagent with the toxic compound.99 However, in some cases, reactivating reagents act as allosteric activators of enzymes.3. Acidity of the working solution The activities of enzymes and their sensitivities with respect to inhibitors are strongly pH-dependent; therefore, a crucial role in inhibitor assays is played by the correct choice of buffer solutions and pH.In the overwhelming majority of studies, the optimisation of the composition of reaction media is aimed at eliciting maximum response of the biosensor with respect to the substrate, but not to the inhibitor. Indeed, in many cases, especially in the case of irreversible and non-competitive reversible inhibition, the pH of the medium corresponding to the maximum inhibitory effect is close to that providing optimum functioning of the enzyme (e.g., in the cyanide assay with cytochrome oxidase,18 in dithiocarba- mate and cyanide assays with tyrosinase and peroxidase 27, 59 andBiosensors for the determination of environmental inhibitors of enzymes in organophosphorus pesticide assays with ChE and tyrosi- nase).8, 30 In all those cases where the inhibitors had competitive or irreversible inhibiting action, the maximum sensitivity of the response of amperometric biosensors with respect to inhibitors was determined by the I0/Ki ratio and the magnitude of the biosensor response to the substrates.In many cases, the pH of the buffer solution can strongly influence the extent of inhibition. In particular, such effects are observed if the inhibitor is involved in acid ± base reactions; it is noteworthy that only one of the equilibrium forms interacts with the enzyme. For example, the pH dependence of the inhibitory effect of metal ions on ChE is due to accumulation in solutions of metal hydroxo complexesMOH+, which competitively inhibit the enzyme.100 As a result, the effects of Cu(II), Ni and Fe(II) ions are manifested in a very narrow range of pH which is limited from the alkaline region by the commencement of precipitation of the corresponding hydroxides. Apparently, the optimum value of pH (5.0) for Hg(II) determination with peroxidase in the presence of thiourea is associated with the hydrolytic stability of the inhibitor.53 Colloidal solutions of heavy metal hydroxides can also inhibit enzymes, at least after long-term incubation.Thus the inhibitory effect of Hg(II) salts on urease was studied at pH 8.9;101 the maximum inhibitory effect of beryllium ions on alkaline phosphatase is manifested at pH 9.8.102 Instability of inhibitors is yet another reason for the deviation of the working pH from the optimum values.Some organo- phosphorus pesticides are spontaneously hydrolysed in alkaline media, which markedly decreases the reproducibility of inhibition of free and immobilised ChE.8 Immobilisation of enzymes can significantly alter both the value of pH corresponding to the optimum enzymic activity and the sensitivity of the enzyme to the inhibitor. This is due to a change in ionic equilibrium constants caused by the differences in dielectric conductivity and ionic strength of the microenviron- ment of the enzyme active site and the solution. Enzyme carriers and ballast proteins being electrolytes can represent ion-exchange materials thereby increasing buffer capacity of the enzyme micro- environment and decreasing the rate of removal of hydrogen ions formed upon enzymic reactions into the solution.78, 103 Therefore, the results of investigations of free enzymes demand verification when passing to the study of immobilised enzymes. For example, if ChE is immobilised by microencapsulation into cellulose trinitrate, the maxima of the enzymic activity and sensitivity towards organophosphorus pesticides are shifted to the alkaline region (to pH 9 ± 10).104 The pH maximum is not changed upon immobilisation of ChE on paper (7.5 ± 8), but the sensitivity of the biosensor response to pH decreases in comparison with the corresponding dependence of the free enzyme.8 In some cases, a change in the pH dependence of inhibition due to immobilisation extends the range of compounds that can be determined.Thus hydrazinium dialkyl dithiophosphates inhibit microbial carboxyesterase at pH 3.5 ± 4.5 but have no effect on free ChE. On the other hand, these compounds cause reversible inhibition of ChE immobilised on paper with a gently sloping maximum at pH 6.0.105 When immobilised enzymes are incorporated into biosensors, the working range of pH can be narrowed due to the instability of the material used for the preparation of biosensors. For example, destruction of the electrode material limits the pH value for the operation of a cholinesterase biosensor based on a carbon ± epoxide composition:106 the assays of organophosphorus pesti- cides were carried out at pH 7.0, although the maximum enzymic activity was observed in the pH range 8.0 ± 9.0.In the case of electron transfer mediators, the response of oxidoreductase biosensors was also pH-dependent, which can be ascribed to the change in the redox potential of the mediator.35, 107, 108 1053 4. The presence of effectors The effect of compounds present in solution and affecting the enzymic activity is non-additive. Irreversible inhibitors used in low concentrations represent the only exception. The presence of both a reversible effector and a substrate reduces the irreversible inhibition owing to the so-called protective effect.109 In the course of incubation, the active site of the enzyme binds to the reversible effector to form a labile complex, which is inaccessible to an irreversible inhibitor. Following addition of the substrate, the complex dissociates, resulting in partial restoration of enzymic activity.In the majority of cases, the protective effect impairs the analytical characteristics of irreversible inhibitor assays, viz., any effectors present in the sample under study including such widespread compounds as alkali and alkaline- earth metal salts, can increase at definite concentrations the limit of detection of the irreversible inhibitor. Therefore, the maximally permissible concentration of salts in the sample is specified to meet the requirements inherent in the methods of enzymic analysis of environmental pollutants irrespective of whether a free enzyme or a biosensor is used in the assay. In some cases, enzyme effectors produce a synergistic effect, resulting in a considerable decrease in the limit of detection of the inhibitor or in an increase in the selectivity of measurement.Thus peroxidase ensures a highly sensitive determination of heavy metal salts; the inhibitory effects of Hg(II), Pb, Cd and Bi(III) increase following preincubation of the enzyme with thiourea or with 1,4- dithiothreitol for Bi(III) 54, 55 bringing about nonspecific inactiva- tion of enzymes. With o-dianizidine as a substrate, the limit of detection of Hg(II) is 161075 mg litre71, that with 3,30,5,50- tetramethylbenzidine is 361076 mg litre71. Cadmium and bis- muth interfere with the assay when present in a 1000- or 10000- fold excess with respect to mercury.In contrast, organomercurials suppress the inhibitory effects of sulfur-containing compounds (phenylthiourea, 1,4-dithiothreitol). Thus, the selectivity and sensitivity of the method can be varied by changing the duration of incubation, the nature of the substrate and the second effector. In the presence of magnesium ions, the sensitivity of zinc 110 and beryllium 102 ion assays with alkaline phosphatase increases, but the reproducibility of the signal decreases, as a rule. A sharp increase in the sensitivity of ChE to pesticides (chlorofos, 2,4-D, DDT) and heavy metals was observed 104 in alkaline borate buffer in the presence of calcium ions. This effect is manifested within rather narrow ranges of pH and effector concentrations. Interest- ingly, 2,4-D and DDT do not inhibit free ChE; therefore, in this case either a cooperative effect of calcium ions and pesticide molecules on the active site of the enzyme is manifested or indirect indication of toxic agents, e.g., owing to elimination of the activating effect of calcium ions on ChE upon its incorpora- tion into the enzyme ± inhibitor complex or to spatial blockade of the access of hydrophobic pesticide molecules to the enzyme- containing membrane.V. Cholinesterase-based biosensors Cholinesterase is being especially widely used and thus deserves special consideration. This enzyme is the subject of up to 60% of publications devoted to biosensors. The interest in ChE of practitioners in environmental and analytical monitoring and toxicology is due to the broad range of inhibitors which involve such widespread toxicants as organophosphorus compounds, heavy metal salts, surfactants, etc.8, 86, 94 Acetylcholinesterase provides hydrolytic cleavage of a natural neurotransmitter, viz., acetylcholine, i.e., it is responsible for the discreteness of transmission of nerve pulses.111 ChE H2O MeC(O) ChE MeCOOH+ChE Me3N+CH2CH2C(O)Me Acetylcholine Acetylated ChE (12) Me3N+CH2CH2OH Choline (alcohol)1054Cholinesterases differ in their substrate specificities.Thus acetylcholinesterase (AChE) (EC 3.1.1.7) hydrolyses acetylcho- line at the highest rate and does not hydrolyse butyrylcholine. Its activity is inhibited by an excess of a substrate, since deacylation is inhibited by the binding of the second acetylcholine molecule to the anionic site of the acyl-enzyme.112 Butyrylcholinesterase (BChE, EC 3.1.1.8) catalyses hydrolysis of butyrylcholine with the highest efficiency but can also react with acetylcholine.Therefore, this was termed pseudo- or nonspecific cholinester- ase, in contrast with the true, specific AChE. At present, various ChE preparations for biochemical assays including the most popular AChE reagents from electric eel (`Sigma', USA, `Boehringer', Germany) and BChE from equine blood serum (Joint-Stock Company `Biomed', Perm, Russia; `Sigma', USA) are manufactured on an industrial scale. These preparations are available as freeze-dried powders, AChE is also purchased as frozen aqueous solutions containing ammonium sulfate.Numerous esters of phosphoric, phosphonic and pyrophos- phoric acids, carbamates and sulfonic acid esters are irreversible and specific inhibitors of ChE. They bind covalently to the active site of the enzyme to form phosphorylated (carbamoylated, sulfonylated) ChE. Organothionophosphorus compounds (para- thion, diazinon, malathion, etc.) are very weak reversible inhib- itors of ChE in vitro. This is due to the inability of the sulfur atoms to form hydrogen bonds with the hydroxy group of serine-200. In the organism of insects, thionophosphorus pesticides are oxidised by nonspecific oxidases to oxygen analogues. Thus parathion gives paraoxon, malathion gives malaoxon, etc.113 As a result, the inhibitory effect of pesticides increases 10 000-fold and even more.Metabolic activation of the insecticide as a result of the so-called lethal synthesis decreases the hazard of acute poisoning of humans with simultaneous preservation of high insectoacaricidal activity of the preparations. Prior to determination of thionophosphorus pesticides with ChE, they were oxidised to the corresponding oxygen analogues with bromine water 114 or 50% nitric acid.115 An effective procedure has been also developed for the oxidation of thionophosphates with chlorine or bromine generated in situ by galvanostatic electrolysis.116 Quaternary ammonium salts, acridine derivatives and metal ions are also known as reversible inhibitors of ChE. The activity of ChE decreases also in the presence of polar organic solvents miscible with water.117 Cholinesterase is widely used for the determination of residual organophosphorus and carbamate-type pesticides in water, aqu- eous and aqueous-organic extracts from plants, soil, food prod- ucts, etc.The published data (1993 ± 1998) concerning cholinesterase biosensors designed for pesticide assays are listed in Table 2. Enzyme ± inhibitor specificity. The analytical parameters of determination of ChE inhibitors strongly depend on the source of the enzyme. Cholinesterases from insects manifest the highest sensitivity with respect to organophosphorus pesticides.91, 150, 151 Thus the inhibition constant, kII, of dimethyl dichlorovinyl phosphate (DDVP) is 56105 mol71 litre min71 for AChE from Drosophila melanogaster, while that for AChE from bovine erythrocytes is <103 mol71 litre min71 for (see Ref. 91).Chol- inesterase from cockroach allows one to determine DDVP at a concentration of 861079 mol litre71, whereas in the case of mammalian enzymes the limits of detection are 10 ± 100 times as high.150 Insect ChEs are also sensitive to organochloric pesticides (2,3-D, DDT, etc.), which seem to be reversible inhibitors. These enzymes are difficult to isolate in sufficiently large amounts, which restricts their practical application. In future, the situation may change owing to the progress in gene engineer- ing. The preparation of AChE from Drosophila melanogaster by incorporation of a mutant gene responsible for AChE synthesis into the genetic apparatus of microorganisms has been described.The yield of AChE was several milligrams per litre of the cultural fluid. This preparation is characterised by high sensitivity to G A Evtugyn, H C Budnikov, E B Nikolskaya organosphosphorus and carbamate-type pesticides despite its low activity in comparison with the natural AChE from the same source.91 Cholinesterases from blood serum 152 and brain 153 of various fish and avian 154 species as well as from the squid optical ganglion 155 are promising for analytical determination of inhib- itors. The problems of low stability can be partly solved by immobilisation of the enzymes in gelatine or agar. The majority of cholinesterase biosensors utilise only three commercial preparations of ChE, viz., AChE from the electric organ of eel Electrophorus electricus, bovine erythrocytes and BChE from equine blood serum.Comparison of analytical characteristics (see Table 2) demonstrates difficulties in establish- ing enzyme ± inhibitor specificities for different biosensors. Thus the reported limits of detection of paraoxone using inhibition of AChE from electric eel vary from 1077 (Refs 47 and 134) to 4610711 mol litre71 (Ref. 136); for chlorofos (trichlorofon) and its active principle (DDVP), the scatter in the limits of detection amounts to five orders of magnitude, viz., from 10711 (Ref. 104) to 1076 mol litre71 (Ref. 134). The experimental value of kII for AChE from electric eel (<103 ±46104 mol litre71 min71) is smaller than that for BChE from equine blood serum (86104± 26105 mol litre71 min71) according to different authors (Refs 86, 91 and 156).It should be mentioned that the sensitivity of chlorofos detection with AChE- based biosensors is higher and the limit of detection is lower than the corresponding analytical characteristics of butyrylcholinester- ase-based biosensors. The reason, as in the case of many other similar discrepancies, may be due to diffusion-dependent inhib- ition of the biosensor response. Commercial preparations of AChE from electric eel normally have high specific activity, which exceeds manyfold (sometimes, hundreds times) that of BChE from equine blood serum. The latter enzyme forms denser films upon immobilisation than the immobilised AChE with high content of the ballast protein.As a result, the biosensors operate in a diffusional-kinetic regime, and the inhibitory effect of the pesticide is reduced. Special mention should be made of early studies aimed at determining thionophosphorus pesticides without preliminary oxidation. Thus the limits of detection of malathion (carbofos) without preliminary oxidation were 0.1 (Ref. 122) and even 1.6 ± 4.6 mg litre71 (Refs 157 and 158), whereas that after oxida- tion with bromine water was of the order of 1074 mg litre71 (Ref. 159). From the standpoint of toxicology, it is the inhibition of AChE that presents special interest. The calculated values of inhibition constants of AChE (kII) correlate with the mean lethal concentrations (LC50) for mice and rats as well as with hygiene and sanitaryMPCstandards for natural water reservoirs.86, 113, 134 Nevertheless, the use of BChE incorporated into biosensors gives in general comparable limits of detection of pesticides which differ not more than 100-fold in both directions (see Table 2).In order to extend the range of compounds that can be determined by this method, some investigators deliberately decrease the selectivity of the biosensor response through simultaneous use of several ChE isolated from different sources.115, 160 Tissue homogenates or biological fluids possessing esterase activity (blood, liver homo- genates, etc.) are used for the same purpose.161 ± 163 By selecting substrates of different specificity, one can vary the set of esterases the activity of which is to be monitored.Thus the use of indophenyl acetate makes it possible to record the overall response of BChE and carboxylesterases. Acetyl(thio)choline is used to detect the responses of AChE and BChE, while butyryl(thio)choline is used for the BChE signal. This extends the potentialities of identification of toxic agents. Immobilisation (stabilisation) of ChE sometimes plays a crucial role. Immobilisation decreases (or rarely increases) the sensitivity of biosensors with respect to inhibitors in comparison with the free enzyme. Covalent immobilisation by cross-linking of ChE on the electrode surface or an appropriate carrier using glutar-Table 2.Some characteristics of methods used for the determination of pesticides with ChE-based biosensors and experimental parameters, viz., concentration range (c/mol litre71) and limits of detection (LD/mol litre71). Enzyme source, substrate Potentiometric biosensors AChE from electric eel, BChE from equine blood serum, acetyl- and butyrylcholine AChE from electric eel, acetylthiocholine AChE from electric eel, acetylcholine BChE from equine blood serum, butyrylcholine Amperometric biosensors AChE from electric eel, acetylthiocholine Immobilisation / stabilisation cross-linking with albumin by vapour of glutaraldehyde on the electrode surface, poly- ethyleneimine and glutaraldehyde sorption on activated nylon (biodin and immunodin) cross-linking with peroxidase and choline oxidase by poly- ethyleneimine on carbon electrodes cross-linking by glutaraldehyde on paper, in gelatine and nitrocellulose (NC) cross-linking by glutaraldehyde on paper, nylon and nitrocellulose cross-linking with albumin by glutaraldehyde on cellulose acetate BChE in a working buffer solution incorporation into a carbon paste with an epoxide resin Method pH-ISFT; conductometric (Pt electrodes) glassy carbon electrode, by a redox mediator potential pH-ISFT glassy carbon electrode glass pH-ISE the same pH-ISFT choline oxidase sensor based on a Pt ±H2O2 sensor carbon-paste electrode Conditions phosphate buffer, 2.561073 mol litre71, pH 7.4, cS=861073 mol litre71, incubation for 10 min citrate buffer, 0.1 mol litre71+ +MgCl2, pH 7.2, cS=161073 mol litre71, incubation for 10 min phosphate buffer, pH 7.0 and 9.0, cS=261072 mol litre71 phosphate buffer, pH 7.5, cS=161072 mol litre71, incubation for 10 min phosphate buffer, pH 7.9, cS=261072 mol litre71, incubation for 10 min the same phosphate buffer, pH 7.4, incubation for 10 min phosphate buffer, pH 7.0, BChE, 0.4 U ml71, 30 8C, cS=361074 mol litre71 phosphate buffer, pH 7.0, cS=161074 mol litre71 Pesticides trichlorfon paraoxon- methyl DDVP malathion trichlorfon diazinon zolone DDVP diazinon in the presence of surfactants DDVP zolone paraoxon carbofuran carbaryl paraoxon DDVP Parameters of inhibitor assay Note LD c 1077 1076 1077 1076 77 261077 7 261076 7 0.1 a 71079± 1075 72610713 70.003 ± 0.3 a 7 0.002 ± 0.3 a 7 0.001 ± 0.4 a 7 0.004 ± 0.4 a 7 0.004 ± 0.7 a 7 0.5 ± 5.0 a 7 0.2 ± 1.0 a 7 0.1 ± 2.0 a 7 7 around 1079 6.7 ± 3.5 b 7 7.5 ± 5 b 70.5 c 7 0.2 c 7 2.2 c 7 20 c 7 2.7 c 7 22 c 7 Ref.118 ± AChE 120 BChE AChE BChE 121 flow-injection regime 122 incubation for 1 h 123 native BChE immobilised BChE 93, paper, gelatine 124 NC paper gelatine NC paper gelatine NC 125 40 126 incubation for 30 min incubation for 1 h 106 without incubationTable 2 (continued). Enzyme source, substrate AChE from bovine erythro- cross-linking by carbodiimide cytes, acetylthiocholine AChE from electric eel, 4-aminophenyl acetate AChE from electric eel, acetylcholine AChE from electric eel, acetylthiocholine AChE from bovine erythro- cross-linking by glutaraldehyde cytes, acetylthiocholine AChE from electric eel, BChE from equine blood serum, acetylthiocholine AChE from electric eel, acetylthiocholine AChE from bovine (be), human (he) and electric eel (ee) erythrocytes, BChE from equine blood serum (eu), acetyl- and butyrylthiocholine Immobilisation / stabilisation on glassy carbon electrodes cross-linking with albumin by glutaraldehyde on nylon cross-linking by glutaraldehyde incorporation into a photo- polymerised gel PVA-SbQ (8%) on a magnetic carrier cross-linking of two enzymes with albumin by glutaraldehyde cross-linking by carbodiimide on the electrode surface the same cross-linking with albumin by glutaraldehyde Method rotating disk electrode Pt electrode choline oxidase sensor based on a Pt ±H2O2 electrode Pt electrode the same carbon-paste electrode with cobalt phthalocyanine the same "Pt microplanar electrode Pesticides Conditions trichlorfon universal buffer, pH 8.0 carbaryl phosphate buffer, pH 8.0, cS=1.261072 mol litre71, paraoxon incubation for 5 min paraoxon phosphate buffer+0.1% NaN3, pH 7.0, cS=561074 mol litre71, incubation for 30 min paraoxon phosphate buffer, pH 8.0, cS=1.261073 mol litre71, 30 8C, incubation for 30 min carbofuran phosphate buffer, paraoxon- pH 8.0 ethyl paraoxon- methyl malaoxon carbaryl phosphate buffer, pH 7.0, cS=561074 mol litre71, DDVP 30 8C paraoxon paraoxon phosphate buffer, pH 7.0, cS=161073 mol litre71, 37 8C, incubation for 5 min DDVP phosphate buffer, pH 7.0, cS=161073 mol litre71, 37 8C, paraoxon incubation for 20 min aldicarb phosphate buffer, pH 7.0, cS=161073 mol litre71, incubation for 10 min carbaryl carbofuran methomil propoxur Parameters of inhibitor assay Note LD c K 7 7 i=5.761077 (see e) 127, 561077± 561075 161077 561075± 161075 161077 1.3 b 1.5 ± 15 b 10710 ±1075 73 b 7 3 b 7 7 b 7 7 b 7561076 161076 161077 361078 7 661079 7 4610711 7 430,c 190,c 7 3200,c 630 c AChE (ee), BChE 7 30,c 0.092,c 7 1300,> >104 (see c) 3.1,c 0.08,c 7 150,c 360 c 39,c 0.19,c 7 0.88,c 1.9 c 1.5,c 0.68,c 7 >104, 11000 c Ref.pI50=4.37e 128 129, 130 131 132 flow-injection regime kII =2.661075 (see d) 133 kII=2.161075 (see d) kII =1.861075 (see d) kII =1.461075 (see d) 134 without incubation 135 extraction with heptane 136 137, for AChE (be), 138 (he), BChE (eu), respectively the same """Table 2 (continued). Enzyme source, substrate AChE from electric eel, acetylcholine AChE from bovine and electric eel erythrocytes, BChE from equine blood serum, acetyl- and butyrylthiocholine AChE from electric eel, acetylthiocholine AChE from electric eel, acetylcholine BChE from equine blood serum, butyrylcholine Immobilisation / stabilisation covalent immobilisation by carbodiimide on an affinity membrane UltraBind incorporation of enzymes immobilised on aminated silica gel into a carbon paste incorporation into photo- polymerised gel PVA-SbQ (8%) incorporation with choline oxidase Pt microelectrode into a Langmuir ± Blodgett film cross-linking by carbodiimide in an agarose gel in the pores of a carbon electrode incorporation into a carrageenan gel with choline oxidase cross-linking with albumin by glutaraldehyde on nylon, nitrocellulose and paper Method choline oxidase sensor based on a Pt Clark's electrode Pt electrode combined with an enzyme column Pt electrode porous carbon electrode Clark's electrode carbon-paste electrode Pesticides Conditions phosphate buffer, 0.01 mol litre71, paraoxon pH 7.0, cS=161072 mol litre71, 30 8C, incubation for 10 min carbofuran phosphate and Tris-, borate buffers, pH 6.0 ± 10.0, cS=161074 mol litre71 carbaryl paraoxon chlorofe- vinphos dipterex phosphate buffer, pH 8.0, cS=1.261073 mol litre71, 30 8C trichlorfon phosphate buffer, pH 7.5, cS=161073 mol litre71 paraoxon phosphate buffer, pH 7.0, Meldola Blue mediator, cS=461074 mol litre71, flow-injection regime paraoxon hexane : chloroform mixture (1 : 1) malathion saturated with water, cS=261073 mol litre71, ethylparathion 1.0 ± 10.0 aldicarb incubation for 5 ± 20 min carbaryl diazinon phosphate buffer, pH 7.9, cS=661075 mol litre71, flow-injection regime Parameters of inhibitor assay Note LD c 161078 7 161078 7 161076 7561077± 161075 561077 2.2 c 7 22.1 c 2.0 c 7 2.8 c 7 3.6 c 7 5.661076 77 around 1076 1.4 c 4.5 c 9.4 ± 56.4 4.5 c 10.0 ± 40.0 0.5 c 4.5 c 10.0 ± 60.0 0.5 c 1.0 ± 10.0 8.2 ± 6.3 b 7 Ref.139 native AChE, buffer native AChE, 0.05% acetone or 5% cyclohexane immobilised AChE, buffer or 1% acetone immobilised AChE, 1% cyclohexane 140 AChE BChE AChE AChE BChE 141 142 143 144 145Table 2 (continued). Enzyme source, substrate Miscellaneous methods AChE from electric eel, BChE from equine blood by glutaraldehyde serum, acetyl- and butyrylcholine ChE, indoxyl acetate AChE from electric eel, acetylcholine AChE from electric eel, indolyl-3-acetate Note. The following abbreviation were used: pH-ISE is pH-sensitive ion-selective electrode; DDVP is dimethyl (2,2-dichlorovinyl) phosphate; a mg litre71; b the values of pc=7log c/mol litre71 are listed; c mg litre71; mol71 litre min71; e mol litre71.d Immobilisation / stabilisation cross-linking with albumin sol-gel immobilisation in tetramethyl orthosilicate covalent immobilisation by isothiocyanate on glass cross-linking by vapour of glutaraldehyde Method conductometric, microplanar Pt electrodes fuorescent optode fibre-optic sensor, change in the colouring of Bromothymol Blue piezocrystal sensor Conditions phosphate buffer, pH 7.5, cS=861073 mol litre71, incubation for 10 ± 60 min phosphate buffer, pH 7.5, cS=1.6361074 mol litre71 Tris-buffer, pH 7.5, cS=161071 mol litre71 phosphate buffer, incubation for 5 min Pesticides paraoxon- methyl paraoxon trichlorfon phenitrothion 7 azinphos-ethyl 7 methidathion naled mecarbam paraoxon carbofuran paraoxon carbaryl Parameters of inhibitor assay Note LD c 561077 7 161078 7 561077 7 5610711 717.8 b 17.9 b 37.6 b 7 0.36 b 7 13.9 b 7561077± 561076 1.161077 561078± 561077 1.561078 561078 7 161077 7 Ref.146 lower assay limit incubation for 10 min 147 148 149Biosensors for the determination of environmental inhibitors of enzymes as vapour (in solu- or aldehyde anor tion) 118 ± 120, 129, 130, 133, 134, 137, 138, 149 aqueous carbodiim- ide 127, 128, 135, 136, 140 are conventionally used.A decrease in enzymic activity as a result of covalent immobilisation depends on the concentration of the cross-linking reagent, the duration of treatment, the initial activity of the enzyme and the nature of the carrier. The losses are minimal (20% ± 40% of the initial enzyme activity) when protein films are prepared from albumin and gelatine, which are also cross-linked, and in the immobilisation of ChE on aminated carriers of the Immobilon or Biodin type. Highly reactive enzyme-containing membranes are prepared by incorporating enzymes into hydrophilic gels based on acryl- amide,79, 159 a N-vinylpyridinium ± vinyl chloride copoly- mer,132, 142 polyethyleneimine,121, 123 tetramethyl orthosilicate 147 and polyurethane.164 Biosensors prepared with such membranes are characterised by a rapid response to the substrate; however, the lifetime of immobilised enzymes is much less than that of enzymes obtained by covalent immobilisation on inert carriers. Under field conditions, it is especially convenient to use biosensors with disposable membranes which can easily be replaced in the case of enzyme inactivation.Disposable mem- branes are prepared on the basis of inert carriers, such as nylon,122, 125, 145, 165, 166 paper and various cellulose deriva- tives,40, 93, 96, 104, 105, 124, 167, 168 aluminium hydroxide,164, 169 indus- trially manufactured membranes (for dialysis and immunoblotting) or carriers for affinity chromato- graphy.122, 140, 141 The largest losses (up to 99% of the original ChE activity) were observed upon cross-linking of enzyme globules with aqueous solutions of glutaraldehyde on inert carriers, which do not react with the cross-linking reagent,124 such as cellulose trinitrate or paper.Such considerable losses are compensated for by high storage stability of the membranes (up to 6 months at room temperature) and operation lifetime (for over two weeks in a buffer solution within the biosensor). As a rule, immobilisation of ChE does not envisage the use of additional water-soluble stabilisers.However, in some cases 0.1% ± 3% glycerol, glycine, non-ionogenic synthetic surfac- tants, 0.01% ± 0.05% gelatine, magnesium and calcium salts (in the case of AChE) are used to stabilise the activity of the free enzyme. Stabilisers are necessary when ChE is used in aqueous solutions for analytical purposes, since the stability of even freshly prepared solutions is low. Thus ChEs from insects are completely inactivated within several hours,170 AChE from electric eel is stable in a concentrated ammonium sulfate solu- tion. Stabilisers increase the storage stability of enzymes. They form films in which the enzyme is in a solid solution. Such preparations are readily soluble and yield stable solutions the activity of which is not changed during the measurement of the inhibitory effect.In other cases, stabilisers are added directly to the enzyme solution for the same purpose. The addition of stabilisers also decreases the losses of enzymic activity in the course of immobilisation.136 Cellulose and dextran derivatives, other polyalcohols and some non-ionogenic surfactants are used as ChE stabilisers. Probably, they inhibit nonspecific sorption of the protein on glass or a plastic, integrate inhibitors (especially, transition metal ions) into inert complexes or stabilise the tertiary structure of the protein. Incorporation of ChE into tregalose 169 and N-phthalylchitosane 73, 116, 155, 171 films for the development of colorimetric indicator devices designed for the determination of ChE inhibitors has been described.Stabilisers suppress the interfering effects of metal ions but do not change the kinetic characteristics of irreversible inhibition of ChE by organophos- phorus pesticides. With a decrease in the diffusion-dependent limitations of biosensor responses, i.e., in the case of thinner membranes and lower specific activity of enzymes, the results of inhibitor assays obtained with free and immobilised enzymes are close. Therefore, amperometric biosensors based on membranes with lower enzy- 1059 mic activities than those used in potentiometric biosensors acquire certain advantage due to higher sensitivity of substrate assays. On the other hand, comparison of efficiencies of potentio- metric and amperometric biosensors under identical conditions shows that the advantages of amperometric detection are not so obvious.94 Thus in experiments with disposable membranes with ChE immobilised on nylon, nitrocellulose or in gelatine, potentio- metric biosensors manifested lower limits of detection of para- oxon and DDVP.93, 172 This may be due to the interfering Faraday effects, particularly, to additional generation of hydrogen ions upon thiocholine oxidation on the electrode.ChE Me3N+(CH2)2SC(O)Me +H2O Me3N+(CH2)2SH+MeCO2H(13) Me3N+(CH2)2S S(CH2)2N+Me+2H+ 2Me3N+(CH2)2SH 7e Calculation of inhibition bimolecular constants from calibra- tion curves of pesticide assays utilising a potentiometric biosensor with disposable membranes according to Eqn (10) gave over- estimated values in comparison with inhibition of the free enzyme under identical conditions.One of possible reasons for increased sensitivity of pesticide assays in the case of immobilised ChE is adsorptive accumulation of pesticides on inert carriers. In partic- ular, this can be confirmed by the fact that non-ionogenic surfactants used in concentrations which do not influence the activity of native ChE, produce differentiating effects on the response of cholinesterase-based biosensors to various inhibi- tors, viz., the irreversible inhibition effect of pesticides is increased, while the reversible inhibition effect of fluorides or heavy metal ions is decreased. Small concentrations of surfactants also suppress the protective effects of reversible inhibitors in organophosphoric pesticide assays.125 This effect is due to sorption of surfactants on the membrane surface, which facili- tates the transfer of hydrophobic organic pesticide molecules to the membrane, while the transfer of small hydrophilic ions (reversible ChE inhibitors) is impeded.The effect of surfactants is a maximum for hydrophilic carriers (paper, nylon) and a minimum for nitrocellulose membranes. The difference between inhibition bimolecular constants for free and immobilised enzymes changes in the same series. For hydrophobic nitro- cellulose, the difference is a maximum, while that for hydrophilic paper and gelatine is virtually absent.173, 174 Increased sensitivity of inhibitor assays upon enzyme immo- bilisation is observed not only for ChE.The limit of detection of heavy metals with oxidoreductases immobilised in albumin films decreases in the following order (mg litre71): from 0.25 to 0.05 for Hg(II), from 1.00 to 0.05 for Cu(II) (alcohol oxidase), from 1.00 to 0.10 for Ni (sarcosine oxidase) and from 0.50 to 0.015 for Se(IV) (glutathione oxidase).44 Obviously, electrostatic interactions of metal ions with charged carriers play an important role in these processes along with nonspecific sorption. Monitoring of environmental objects. Virtually, all the data listed in Table 2 relate to model solutions containing only one ChE inhibitor. The use of cholinesterase-based biosensors for the analysis of real objects is complicated, firstly, because of the presence of considerable amounts of natural enzyme effectors (alkaline-earth metal ions, biogenic amines, etc.) in analysed samples and secondly, due to simultaneous presence of several inhibitors with different mechanisms of action.This fact signifi- cantly complicates the interpretation of experimental results: one has either to confine oneself to qualitative detection of anti- acetylcholinesterase compounds in analysed samples or to have recourse to their preliminary separation at the expense of simplicity and rapidity, which are the main advantages of enzymic assays. Commercial production of test systems based on ChE and designed for monitoring residual organophosphorus and carba- mate-type pesticides in environmental objects and foodstuff has already been set up.These systems, which represent indicator kits with colorimetric indication are manufactured by `Boehringer1060 Mannheim' (Germany) 175 and `Ohmicron Environmental Diag- nostic Inc.' (USA) 176 and ensure qualitative detection of organo- phosphorus pesticides with limits of detection ranging from 0.05 to 30 mg litre71. In the 1970's, the manufacturing company `Midwest Research Institute' (USA) produced a family of automated indicator systems CAM-1,2 (Continuous Alarm Monitor), which represent alarm devices for detecting toxic war gases of neuroparalytic action and are adapted to civil needs.164 These systems are supplied with pumps and aspirators and operate in a semi- automatic regime.The minimum concentrations of pesticides detectable in air and water are as follows (mg litre71): paraoxon, 0.1; DDVP, 0.2; diazinon, 0.2; dursban, 4.3; difonat, 10; mala- thion, 17 ± 70 (depending on the brand); temic, 0.5; furadan, 0.75; mesuron, 12 and sevin, 20. Their service cycle includes automated substitution of enzyme-containing polyurethane reactors and electrochemical purification of electrodes, they are supplied with an acoustic signal device of contamination. A potentiometric biosensor containing AChE immobilised in gelatine on the surface of a pH-metric glass electrode was used to control the quality of waters in the river Seine (France).177, 178 The limits of detection of pesticides with this biosensor were as follows (mg litre71): 0.003 (ethylparaoxon), 0.005 (methylparaoxon) and 0.014 (malathion, ethyl- and methylparathion). Some countries including Russia have begun the commercial production of indicator tubes which contain ChE and chromogenic substrates and are designed for standard aspirators used in air sampling.179 At present, the possibility to use ChE for environmental and analytical monitoring is chiefly associated with the creation of a market of cheap electrochemical sensors which can form the basis for mass-scale production of disposable cholinesterase-based biosensors that are compatible with standard measuring devices (voltamperographs, ionometers, etc.).Such sensors can be manu- factured by photolithographic methods.Because of High-Tech production, small size (several square millimetres) and low enzyme expenditure (*0.1 unit per electrode), the cost of such biosensors does not exceed US $ 1 ± 5. The production of pH- sensitive field transistors and microplanar conductometric devices is currently under way.118 ± 120 A technology for the production of screen-printed metal-coated and carbon-paste electrodes with plastic mounts has been developed.180 The cost of equipment needed for their production amounts to US $ 30 000, which is comparable with the cost of a general-purpose voltamperograph. At present, screen-printed electrodes are manufactured in the Czech Republic,133 Italy 180, 181 and Germany.182 All of them represent a combination of a working, auxillary electrode and a reference electrode (Ag/AgCl), which have various geometries, are arranged into a single block (sometimes, in pairs) and operate in a differential regime.The progress in the design of biosensor elements demands adequate solution of methodological problems. As expected, the effect of the matrix of the analysed subject does not seem to be of crucial importance.115 Biosensors are used for the determination of residual organophosphorus and carbamate-type pesticides in vegetables, fruits, juices,137, 138, 183 soils and silt sediments.134, 184 The only limitation in the analysis of surface waters is their mineralisation, which should not exceed 0.1 mol litre71 (Refs 122 and 177). It has to be specified, however, that the accuracy of pesticide assays with biosensors was checked using a method of additives rather than independent methods.Chromato- graphic methods were used for the estimation of the correctness of measurements only in the case of carbamate-type pesticide assays in vegetables,138 of organophosphorus and carbamate-type pesti- cide assays in fruits and vegetables 183 and of organophosphorus pesticide assays in natural waters (freezing and subsequent thawing of standard samples).128 The divergence of experimental results was*30%, which was accepted to be satisfactory. A number of simplified sampling techniques have been developed which help elimination of laborious procedures con- nected with the removal of organic solvents. Thus, a 10-minute G A Evtugyn, H C Budnikov, E B Nikolskaya extraction with acetone followed by 10-fold dilution of extracts allows reliable determination of up to 0.1 mg kg71 of pesticides (diazinon, chlorofos, tetraethyl monothiopyrophosphate) in wheat grains and combined fodders.172 However, in this method the level of extraction is 60%± 80%, which means that it is only possible to detect hazardous contamination of the products with pesticides.If the information about the nature of toxic agents is absent, a reduced biosensor response should not necessarily be related to the concentration of the compound under study. The difference in inhibition bimolecular constants does not allow one to regard the degree of inhibition as a measure of net (molar or mass) content of pesticides even in the absence of interfering factors (diffusion- dependent inhibition of biosensor responses, protective action of effectors, etc.).A concept of a standard toxicant has been proposed according to which the net content of pesticides is expressed in units of concentrations of the pesticide used as a standard judging from the inhibitory effects they produce. DDVP,164 paraoxon,132 parathion 115 and carbaryl 180, 181 were recommended as candidates for such standards. The inhibiting effects of other pesticides are calculated with the help of relative toxicity coefficients. The latter are determined in model pesticide solutions of equal concentration by estimating the degree of inhibition of ChE (i.e., a relative decrease in the biosensor response).Undeniably, these coefficients are not equal to the ratio of inhibition constants and depend on experimental param- eters including the choice of pesticide concentrations. Similar limitations are typical of calculations of pesticide concentrations expressed as factors of excess over MPC. Simultaneous use of several biosensors including enzyme- based ones affords additional possibilities for differentiation of biosensor responses. In contrast with the above-described deter- mination of the ratio of inhibition constants (kII)1/(kII)2, in this case the result is obtained instantaneously, which makes it possible to use a set of biosensors 47, 63, 185 (or a multichannel multisensor 40, 41) for on-line or even flow measurements. Hydrolytic enzymes, e.g., urease,40, 41, acid and alkaline phosphatases 47, 185 and glucose oxidase 40, 41 can also be used for this purpose together with ChE.Unfortunately, enzyme ± inhibitor specificity of these enzymes does not allow simultaneous qualitative and quantitative determination of inhib- itors. In the best case, adequate interpretation of experimental results can be achieved through the use of chemometric methods, e.g., by estimating individual contributions of Hg(II) ions and the overall effect of all pesticides.185 However, in this case, too, the mutual effect of solution (sample) components is questionable, since the information about the relative sensitivity of biosensor responses is obtained for single-component solutions.Even with the use of experimental planning methods for the choice of conditions for detecting inhibitor signals, the possibility to separate the contributions of individual inhibitors was estab- lished for sufficiently high concentrations (1 mg litre71), which exceeded the existing hygiene and sanitary standards a thousand- fold.To some extent, the selectivity of effects produced by reac- tivating reagents can assist in interpretation of experimental results. Thus the anti-cholinesterase effect of metal ions can be suppressed by EDTA and sodium iodide, which allows manifesta- tion of the toxic effect of pesticides.1, 97 Contrariwise, ChE added to analysed samples can bind trace amounts of anti-cholinesterase pesticides and then the toxic effects of other pollutants can be determined with other enzymes or indicator organisms.1 In another approach, the very fact of ChE inhibition is considered as a proof of contamination without regard for the qualitative and quantitative composition of the analysed sam- ple.186 This approach is widely used in toxicological studies which make use of a special term, `anti-cholinesterase activity' for a substance or tested subject.187, 188 The activity of AChE from blood or brain serves as a measure of chronic poisoning of living organisms with sublethal concentrations of pesticides.189, 190 TheBiosensors for the determination of environmental inhibitors of enzymes total inhibitory effect of toxic agents (i.e., the anti-cholinesterase activity of the medium) can also be used for detection of hazard- ous poisoning levels and for outlining the main tendencies in their variability.Studies of inhibitory effects of silt sediments in the rivers of the Morava Basin which employed biosensors based on AChE from electric eel revealed a relationship between biosensor signals (the inhibition was studied by following relative changes in the dynamic responses) and agricultural impact on the given territory.134 The biotesting conducted by the methods accepted in biological monitoring of surface and sewage waters established a correlation between the degree of inhibition of BChE immobi- lised on paper or nitrocellulose and the death rate or retardation of growth of hydrobionts of the Cladocera family (crayfish Daphnia magna Straus) and the protococcal alga (Chlorella vulgaris Beyer).191, 192 In the first place, this correlation was established for specific inhibitors of ChE.An analysis of real environmental objects revealed a certain divergence of results obtained by different methods. For example, studies of more than 200 samples of sewage waters from industrial and agricultural enter- prises with the aid of potentiometric cholinesterase-based bio- sensors and acute toxicity tests of waters for the presence of Infusoria (Paramecium caudatum) have made it possible to classify all sewage waters into three groups according to the contamina- tion source, viz., (1) sewage waters from machine-building plants which are predominantly contaminated with heavy metals (high death rate among Paramecium caudatum with moderate levels of ChE inhibition); (2) sewage waters from petrochemical plants and enterprises of food industry which are contaminated with petro- leum products, fats and other high-molecular-weight organic compounds (high death rate among Paramecium caudatum in the absence of ChE inhibition) and (3) sewage waters from agricul- tural enterprises and surface run-off wastes from rural areas (lack of toxicity for Infusoria, low and medium levels of ChE inhib- ition).193 The methods which combine biological and biochemical parameters of water quality (and, in prospect, other generalised hydrochemical parameters) seem to hold especially great promise, primarily due to the possibility to extend the range of potential toxic agents present in environmental objects.VI. Conclusion The explosive development of enzymic methods for monitoring of environmental objects has become possible because of a combi- nation of fundamental and applied investigations at the bounda- ries between biochemistry, biotechnology, toxicology, analytical chemistry and environmental monitoring. With the advent of biosensors, purely laboratorial complex biochemical methods have been transformed into easy-to-handle measuring tools, which are comprehensible for the untrained personnel and are suitable for field use. This does not mean, however, that the design of biosensors is exclusively confined to the selection of appropriate detectors and immobilisation of enzymes. Investigators are faced with the necessity to solve complicated problems connected with func- tional features of enzymes to be incorporated into biosensor membranes as well as with optimisation of measurement con- ditions and effects of template components.A transition from a free enzyme to a biosensor results in the narrowing of the range of experimental facilities used in the study of biologically active substances. Thus the minimum concentrations of substrates and the specific activity of enzymes are limited, while biosensors sometimes require a priori non-optimal conditions for their functioning, which are restricted by detectability parameters or specific applications. Optimisation of the biosensor design and the choice of service conditions always involve a compromise between the requirements for the accuracy of response detection, the sensitivity of inhibitor (substrate) assays and the limitations imposed by actual conditions of field tests.1061 Realisation of specific functioning of biosensors designed for evaluation of toxic effects of environmental pollutants has stimulated the research carried out for commercial purposes. In the first place, this relates to disposable biosensors based on screen-printed and solid-state sensors as well as to the simplest indicating devices with visual indication of signals. Their practical application is still limited due to complexity of interpretation of experimental results and inadequacy of the standardising basis, which bans the use of the data obtained by biochemical monitor- ing methods for ecological expertise. For this reason, only cholinesterase-based biosensors have found application in envi- ronmental and analytical monitoring, although biosensors allow the detection of virtually all priority-driven toxic agents with the exception of dioxins and radioactive compounds.Nevertheless, the development of biosensors for the determi- nation of environmental pollutants is a very promising trend, primarily because they meet demand. The methods based on the use of biosensors compete favourably with modern methods of toxicological and biological monitoring. To their obvious advan- tages one may relate small dimensions, ease of signal detection, possibility of integration of biosensors into automated control systems and, last but not least, possibility of quantification of toxicological effects.In our opinion, further developments in this field will proceed along three main lines. 1. Further development of the principal elements aimed at improving the metrological and operational characteristics of transducers and transition to polyenzyme sensors (e.g., multi- sensors, sensor arrays), which help establish not only the presence of inhibitors in analysed subjects but also their nature. 2. Further extension of the range of tested compounds, mostly through optimisation of service conditions for measurement and sampling rather than by increasing of the number of enzymes; further improvement and application of biosensors designed for use in organic solvents, particularly those used as detectors in flow-injection analyses and liquid chromatography.3. Further development of methodical and methodological aspects of applicability of biosensors in environmental and analytical monitoring; design of expert systems for estimation of biological (physiological) activity of chemical compounds, toxic effects of solid industrial and vital activity wastes; combination of biosensors with conventional methods of biological monitoring and, as a consequence, increasing the information content of biochemical monitoring; large-scale application of methods for chemometric treatment of experimental results.The integration of different methods and approaches to environmental and toxicological monitoring of environmental objects is still in progress. Recent developments in the field of biosensors are not confined to routine application of enzymes as analytical reagents. Biosensors are used as biochemical models of living organisms and the approaches used therein are related to toxicology and biotesting.193 First attempts have been undertaken to direct alteration of the enzyme structure with a view to changing its inhibiting specificity.91 This approach makes use of `artificial enzymes' (i.e., oligonucleotides or polymers); their synthesis and selection were performed using molecular recogni- tion of specific toxic agents.194 And, finally, the vast experience gained in the study of biosensors has served as a basis for the design of entirely new devices, viz., DNA-sensors allowing direct control over mutagenic and genotoxic properties of environmental pollutants.195 The specificity of their functioning goes far beyond the scope of this review, although their role in the development of new methods of monitoring of environmental objects is now apparent.References 1. A A Tumanov, I A Kitaeva, O V Barinova Zh. Anal. Khim. 48 6 (1993) a1062 2. A P F Turner, I Karube, G S Wilson (Eds) Biosensors. Fundamentals and Applications (Oxford: Oxford University Press, 1987) 3. Yu Yu Kulis Analiticheskie Sistemy na Osnove Immobilizovannykh Fermentov (Analytical Systems Based on Immobilised Enzymes) (Vil'nyus: Mokslas, 1981) 4.Yu A Ovchinnikov Osnovy Bioorganicheskoi Khimii (Foundations of Bioorganic Chemistry) (Moscow: Prosveshchenie, 1987) 5. A Fersht Enzyme Structure and Mechanism (Reading, San Francisco: WH Freeman and Co., 1977) 6. I V Berezin, V K Antonov, K Martinek (Eds) Immobilizovannye Fermenty. Sovremennoe Sostoyanie i Perspektivy T. 2 (Immobilised Enzymes. Current State and Prospects) (Moscow: Moscow State University, 1976) 7. A P F Turner New Trends in Biosensor Development. NATO Advanced Research Workshop, Kiev, 1998 8. E B Nikolskaya, G A Evtugyn Zh. Anal. Khim. 47 1358 (1992) a 9. E B Nikolskaya, G A Evtugyn, T N Shekhovtsova Zh. Anal. Khim. 49 452 (1994) a 10. V N Maistrenko, R Z Khamitov, H C Budnikov Ekologo-Analiti- cheskii Monitoring Supertoksikantov (Ecological and Analytical Monitoring of Supertoxicants) (Moscow: Khimiya, 1996) 11.K R Rogers, C L Gerlach Environ. Sci. Technol. 30 486A (1996) 12. I Karube Biotechnology 7a 685 (1987) 13. Kh Z Brainina, A M Bond Anal. Chem. 67 2586 (1995) 14. S Wring, J P Hart Analyst 117 1281 (1992) 15. A E G Cass, G Davis, G D Francis, H A O Hill,W J Aston, I J Higgins, E V Plotkin, L D L Scott, A P F Turner Anal. Chem. 56 667 (1984) 16. A S Jdanova, S Poyard, A P Soldatkin, N Jaffrezic-Renault, C Martelet Anal. Chim. Acta 321 35 (1996) 17. K R Rogers, L R Williams Trend. Anal. Chem. 14 289 (1995) 18. A Amine, M Alafandy, J-M Kauffmann, M N Peki Anal. Chem. 67 2822 (1995) 19.F A McArdle, K C Persaud Analyst 118 419 (1993) 20. M H Smit, G A Rechnitz Electroanalysis 5 747 (1993) 21. D J Blyth, J W Aylott, D J Richardson, D A Russell Analyst 120 2725 (1995) 22. L Stancik, L Macholan, F Scheller Electroanalysis 7 649 (1995) 23. A J Reviejo, C Fernandez, F Liu, J M Pingarron, J Wang Anal. Chim. Acta 315 93 (1995) 24. J L Besombes, S Cosnier, P Labbe, G Reverdy Anal. Chim. Acta 311 255 (1995) 25. J Wang, L Chen Anal. Chem. 67 3824 (1995) 26. M H Smit, G A Rechnitz Anal. Chem. 65 380 (1993) 27. J Wang, V B Nascimentl, S A Kane, K Rogers,M R Smyth, L Angnes Talanta 43 1903 (1996) 28. J Wang, S A Kane, J Liu, M R Smyth, K Rogers Food Technol. Biotechnol. 34 51 (1996) 29. Q Deng, S J Dong Analyst 121 1979 (1996) 30.W R Everett, G A Rechnitz Anal. Chem. 70 807 (1998) 31. O Adeyoju, E I Iwuoha, M R Smyth Talanta 41 1603 (1994) 32. O Adeyoju, E I Iwuoha, M R Smyth Anal. Lett. 27 2071 (1994) 33. O Adeyoju, E I Iwuoha, M R Smyth Electroanalysis 7 924 (1995) 34. O Adeyoju, E I Iwuoha,M R Smyth Anal. Chim. Acta 305 57 (1995) 35. T-M Park, E I Iwuoha, M R Smyth Electroanalysis 9 1120 (1997) 36. A P Soldatkin, Y I Korpan, G A Zhylyak, C Martelet, A V El'skaya, in Biosensors for Direct Monitoring of Environmental Pollutants in Field (NATO ASI Ser.) Vol. 38 (Eds D P Nikolelis, U J Krull, J Wang,MMascini) (Dordrecht: Kluwer Academic, 1997) p. 281 37. K Stein, J-U Hain Microchim. Acta 93 (1995) 38. J-L Marty, T Noguer Analysis 21 231 (1993) 39. T Noguer, J-L Marty Anal.Chim. Acta 347 63 (1997) 40. N F Starodub, Yu M Shirshov, A L Kukla, N I Kanjuk, A V Prokhorovich, R Merker Electrochem. Soc. Proc. 97 799 (1997) 41. F Starodub, N I Karyuk, A L Kukla, Yu M Shirshov Anal. Chim. Acta. 385 461 (1999) 42. V Volotovsky, N Kim Electroanalysis 10 512 (1998) 43. J-C Gaeyt, A Havuz, A Geloso-Meyer, C Burstein Biosens. Bioelectron. 8 177 (1993) 44. D Compagnone, M Bugli, P Imperiali, G Varallo, G Palleschi, in Biosensors for Direct Monitoring of Environmental Pollutants in Field (NATO ASI Ser.) Vol. 38 (Eds D P Nikolelis, U J Krull, J Wang, MMascini) (Dordrecht: Kluwer Academic, 1997) p. 220 45. J Marcos, A Townshend Anal. Chim. Acta 310 173 (1995) G A Evtugyn, H C Budnikov, E B Nikolskaya 46.F Mazzei, F Botre, C Botre Anal. Chim. Acta 336 67 (1996) 47. T Danzer, G Schwedt Anal. Chim. Acta 318 275 (1996) 48. Y Su, A Cagnini, M Mascini Chem. Anal. 40 579 (1994) 49. D W Bryce, J M Fernandez-Romero, M D L Decastro Anal. Lett. 27 867 (1994) 50. G A Zhylyak, S V Dzyadevich, Y I Korpan, A P Soldatkin, A V El'skaya Sens. Actuators B 24/25 145 (1995) 51. V Volotovsky, Y Nam, N Kim Sens. Actuators B 42 233 (1997) 52. V Volotovsky, N Kim Electroanalysis 10 61 (1998) 53. T N Shekhovtsova, S V Chernetskaya, I F Dolmanova Zh. Anal. Khim. 48 129 (1993) a 54. T N Shekhovtsova, S V Chernetskaya, E B Nikolskaya, I F Dolmanova Zh. Anal. Khim. 49 862 (1994) a 55. T N Shekhovtsova, S V Chernetskaya Anal. Lett. 27 2883 (1994) 56. J Wang Talanta 40 1905 (1993) 57.J Wang, Y Lin Anal. Chim. Acta 271 51 (1993) 58. E I Iwuoha, O Adeyoju, E Dempsey,M R Smyth, J Liu, J Wang Biosens. Bioelectron. 10 661 (1995) 59. P M T Pita, A J Reviejo, M F De Villena, J M Pingarron Anal. Chim. Acta 340 89 (1997) 60. E I Iwuoha, M R Smyth, in Biosensors for Direct Monitoring of Environmental Pollutants in Field (NATO ASI Ser.) Vol. 38 (Eds D P Nikolelis, U J Krull, J Wang,M Mascini) (Dordrecht: Kluwer Academic, 1997) p. 289 61. OAdeyoju, E I Iwuoha,MRSmyth,DLeech Analyst 121 1885 (1996) 62. R Z Kazandjian, J S Dordick, A M Klibanov Biotechnol. Bioeng. 28 417 (1986) 63. G M Christensen, B L Riedel Inter. J. Environ. Anal. Chem. 8 277 (1980) 64. T A Cherkasova, Yu A Leikin, V E Vonskii,M V Deeva, in Analiticheskaya Khimiya Ob'ektov Okruzhayushchei Sredy (Tez.Dokl.), S.-Peterburg, Sochi, 1991 [Analytical Chemistry of Environmental Objects (Abstracts of Reports), St.-Petersburg, Sochi, 1991] Vol. 3, p. 200 65. DMRawson, A J Willmer,MF Cardosi Tox. Assesment 2 325 (1987) 66. D M Rawson, A J Willmer, A P F Turner Biosensors 4 299 (1989) 67. E A H Hall,M Preuss, J J Gooding,M HaÈ mmerle, in Biosensors for Direct Monitoring of Environmental Pollutants in Field (NATO ASI Ser.) Vol. 38 (Eds D P Nikolelis, U J Krull, J Wang, MMascini) (Dordrecht: Kluwer Academic, 1997) p. 227 68. M Purcell, R Carpenter Water Poll. Res. J. Can. 25 175 (1990) 69. I Karube, T Matsunaga, S Mitsuda, S Suzuki Biotechnol. Bioeng. 19 1535 (1977) 70.J J Kulys, K V Kadziauskene Biotechnol. Bioeng. 22 221 (1980) 71. S E Strand, D A Carlson J. Water. Pollut. Control. Fed. 56 464 (1984) 72. P A Van Rolleghem, Z Kong, G Rombouts,W VerstraÈ te J. Chem. Technol. Biotechnol. 59 321 (1994) 73. P Solyom Prog. Water Techn. 9 193 (1977) 74. L Campanella, G Vafero,M Tomassetti Sci. Total. Environ. 171 227 (1995) 75. P A Van Rolleghem, W VerstraÈ te, in Proceedings of the 5th European Congress on Biotechnology (Abstracts of Reports), Copen- hagen, 1990 Vol. 1, p. 161 76. A A Bulich Aquat. Toxicol. Hazard Assess. ASTM STP 667 98 (1979) 77. I Karube, T Nakahara, T Matsunaga, S Suzuki Anal. Chem. 54 1725 (1982) 78. A A Klesov Metody Opredeleniya Fermentativnoi Aktivnosti i ikh Unifikatsiya (Methods of Determination of Enzymatic Activity and Their Unification) (Vil'nyus: Mokslas, 1979) p.4 79. A Roda, P Rauch, E Ferri, S Girotti, S Ahini, G Carrea, R Bovara Anal. Chim. Acta 294 35 (1994) 80. S Varanasi, R L Stevens, E Ruckenstein AIChE J. 33 558 (1987) 81. S O Ogundiran, S Varanasi, E Ruckenstein Biotechnol. Bioeng. 37 160 (1991) 82. A N Reshetilov Prikl. Biokhim. Mikrobiol. 32 78 (1996) 83. V V Sorochinskii, B I Kurganov Prikl. Biokhim. Mikrobiol. 33 138 (1997) 84. A L Kukla, Yu M Shirshov Sens. Actuators B 48 461 (1998) 85. L P Kuznetsova, L I Kugusheva, E B Nikolskaya Ukr. Biokhim. Zh. 62 (6) 42 (1990) 86. P Skla'dal Food Technol. Biotechnol., 34 43 (1996) 87. W N Aldridge Biochem. J. 46 451 (1950) 88. T Keleti Basic Enzyme Kinetics (Budapest: Akade'miai Kiado', 1986)Biosensors for the determination of environmental inhibitors of enzymes 89.S D Varfolomeev, V G Gurevich Biokinetika (Biokinetics) (Moscow: Fair, 1998) 90. E B Nikolskaya, S N Moralev Dokl. Akad. Nauk 354 780 (1997) b 91. F Villatte, V Marcel, S Estrada-Mondaca, D Fournier Biosens. Bioelectron. 13 157 (1998) 92. B C Hill, S Marmor Biochem. J. 279 355 (1991) 93. H C Budnikov, G A Evtugyn Electroanalysis 8 817 (1996) 94. J-L Marty, D Garcia, R Rouillon Trends Analyt. Chem. 14 329 (1995) 95. A B Roy Anal. Biochem. 165 13 (1987) 96. H C Budnikov, E P Medyantseva, S S Babkina, A V Volkov Zh. Anal. Khim. 44 2253 (1989) a 97. C Tran-Minh Ion-Selec. Electrode Rev. 7 41 (1985) 98. F Hobbiger,M Pitman, P W Sadler Biochem.J. 75 363 (1960) 99. R I Volkov Dokl. Akad. Nauk SSSR 188 354 (1961) b 100. E B Nikolskaya Zh. Anal. Khim. 38 5 (1983) a 101. W H R Shaw, D N Raval J. Am. Chem. Soc. 83 3184 (1961) 102. A Townshend, A Vaughan Talanta 16 929 (1961) 103. V Tougu, A Pedak, T Kesvaters, A Aaviksaar FEBS Lett. 225 77 (1987) 104. E P Medyantseva, H C Budnikov, S S Babkina Zh. Anal. Khim. 45 1386 (1990) a 105. G A Evtugyn, L V Ryapisova, E E Stoikova, L B Kashevarova, S V Fridland, V Z Latypova Zh. Anal. Khim. 53 988 (1998) a 106. D Martorell, F Cespedes, E Martinez-Fabregas, S Alegret Anal. Chim. Acta 290 343 (1994) 107. G Robinson, D Leech, M R Smyth Electroanalysis 7 952 (1995) 108. M H Smit, A Cass Anal. Chem. 62 2429 (1990) 109.M S Dehlawi, A T Eldefrawi, M E Eldefrawi, N A Anis, J J Valdes J. Biochem. Toxicol. 9 261 (1994) 110. A Townshend, A Vaughan Talanta 17 289 (1970) 111. S N Golikov, V I Rozengart Kholinesterazy i Antikholinesteraznye Veshchestva (Choline Esterases and Anticholine Esterase Substances) (Leningrad: Nauka, 1964) 112. R M Krupka Biochemistry 2 76 (1963) 113. V I Rozengart, O E Sherstobitov Izbiratel'naya Toksichnost' Fosfororganicheskikh Insektoakaritsidov (Selective Toxicity of Organophosphorus Insectoacaricides) (Leningrad: Nauka, 1974) 114. M A Klisenko, M V Pis'mennaya Zh. Anal. Khim. 43 354 (1988) a 115. A P Giang, S A Hall Anal. Chem. 23 1830 (1951) 116. G A Evtugyn, E E Stoikova, R R Iskanderov, E B Nikolskaya, H C Budnikov Zh.Anal. Khim. 52 6 (1997) a 117. N Mionetto, J-L Marty, I Karube Biosens. Bioelectron. 9 463 (1994) 118. A-M Nyamsi Hendji, N Jaffrezic-Renault, C Martelet, P Clechet, A A Schul'ga, V I Strikha, L I Nechiporuk, A P Soldatkin, W B Wlodarski Anal. Chim. Acta 281 3 (1993) 119. S V Dzyadevich, A A Schul'ga, A P Soldatkin, A-M Nyamsi Hendji, N Jaffrezic-Renault, C Martelet Electroanalysis 6 752 1994 120. N Jaffrezic-Renault, C Martelet, P Clechet, A-M Nyamsi Hendji, A A Schul'ga, S V Dzyadevich, L I Netchiporuk, A P Soldatkin Sens. Mater. 8 161 (1996) 121. D M Ivnitskii, J Rishpon Biosens. Bioelectron. 9 569 (1994) 122. C Ristori,M Del Carlo,M Martini, A Barabaro, A Ancarani Anal. Chim. Acta 325 151 (1996) 123. A L Ghindilis, T G Morzunova, A V Barmin, I N Kurochkin Biosens.Bioelectron. 11 873 (1996) 124. E B Nikolskaya, G A Evtugyn, R R Iskanderov, V Z Latypova Zh. Anal. Khim. 51 561 (1996) a 125. G A Evtugyn, H C Budnikov, E B Nikolskaya Analyst 121 1911 (1996) 126. M Bernabei, S Chiavarini, C Cremisini, G Palleschi Biosens. Bioelectron. 8 265 (1993) 127. M Stoycheva Anal. Lett. 27 3065 (1994) 128. M Stoycheva Electroanalysis 7 660 (1995) 129. C La Rosa, F Pariente, L Hernandez, E Lorenzo Anal. Chim. Acta 295 273 (1994) 130. C La Rosa, F Pariente, L Hernandez, E Lorenzo Anal. Chim. Acta 308 129 (1995) 131. C Cremisini, S Di Sario, J Mela, R Pilloton, G Palleschi Anal. Chim. Acta 311 273 (1995) 132. J-L Marty, N Mionetto, S Lacorte, D Barcelo Anal. Chim. Acta 311 265 (1995) 133.A GuÈ nther, U Bilitewski Anal. Chim. Acta 300 117 (1995) 1063 134. P Skla'dal,M Fiala, J KrejeÁ i Inter. J. Environ. Anal. Chem. 65 139 (1996) 135. P Skla'dal, J KrejeÁ i Collect. Czech. Chem. Commun. 61 985 (1996) 136. J J Rippeth, T D Gibson, J P Hart, I C Hartley,G Nelson Analyst 122 1425 (1997) 137. P Skla'dal, G S Nunes, H Yamanaka,M L Riberio Electroanalysis 9 1083 (1997) 138. G S Nunes, P Skla'dal, H Yamanaka, D Barcelo Anal. Chim. Acta 362 59 (1998) 139. S Fennouh, V Casimiri, C Burstein Biosens. Bioelectron. 12 97 (1997) 140. D Martorell, F Cespedes, E Martinez-Fabregas, S Alegret Anal. Chim. Acta 337 305 (1997) 141. G Turdean, I Peter, I C Popescu, L Oniciu Rev. Roum. Chim. 42 879 (1997) 142. M Rehak,M Snejdarkova, T Hianik Electroanalysis 9 1072 (1997) 143.M Khayyami, M T P Pita, N P Garcia, G Johansson, B Danielsson, P-O Larsson Talanta 45 557 (1998) 144. L Campanella, S De Luca, M P Sammrtino, M Tomasetti Anal. Chim. Acta 385 59 (1999) 145. G A Evtugyn, A N Ivanov, E V Gogol, J-L Marty, H C Budnikov Anal. Chim. Acta 385 13 (1999) 146. S V Dzyadevich, A P Soldatkin, A A Shul'ga, V I Strikha, A V El'skaya Zh. Anal. Khim. 49 874 (1994) a 147. A N Diaz, M C R Peinado Sens. Actuators A 38 ± 39 426 (1997) 148. R T Andres, R Narayanaswamy Talanta 44 1335 (1997) 149. J M Abad, F Pariente, L Hernandez, H D Abruna, E Lorenzo Anal. Chem. 70 2848 (1998) 150. M H Sadar, S S Kuan, G G Guilbault Anal. Chem. 42 1770 (1970) 151. G G Guilbault,M H Sadar, S S Kuan, D Casey Anal. Chim. Acta 52 75 (1970) 152. Yu G Zhukovskii, S A Kutsenko, L P Kuznetsova, E E Sochilina, E N Dmitrieva Zh. Evolyuts. Biokhim. Ffiziol. 30 177 (1994) 153. USSR P. 1 359 741; Byull. Izobret. (46) 193 (1989) 154. A P Brestkin, Yu G Zhukovskii, N L Fartseiger Zh. Evolyuts. Biokhim. Ffiziol. 19 138 (1983) 155. E B Nikolskaya, L P Kuznetsova, A A Prokopov Zh. Anal. Khim. 53 749 (1998) a 156. L I Kugusheva, L P Kuznetsova, E B Nikolskaya, O V Yagodina Zh. Anal. Khim. 47 1478 (1992) a 157. P Skla'dal,M Mascini Biosens. Bioelectron. 7 335 (1992) 158. P Skla'dal Anal. Chim. Acta 269 281 (1992) 159. C Tran-Minh, P C Pandey, S Kumaran Biosens. Bioelectron. 5 461 (1990) 160. P Skla'dal,M Pavlik, M Fiala Anal. Lett. 27 29 (1994) 161. O A Malinin Veterinariya 62 (1980) 162. MA Klisenko (Ed.) Metody Opredeleniya Mikrokolichestv Pestitsidov (Methods of Determination of Microquantities of Pesticides) (Moscow: Meditsina, 1984) 163. B.I.Antonov (Ed.) Laboratornye Issledovaniya v Veterinarii. Khimiko-Toksikologicheskie Metody. Spravochnik (Laboratory Investigation in Veterinary Science. Chemical and Toxicological Methods. Handbook) (Moscow: Agropromizdat, 1989) 164. L H Goodson, W B Jacobs Methods Enzymol. 44 647 (1976) 165. M Bernabei, C Cremisini, M Mascini, G Palleschi Anal. Lett. 24 1317 (1991) 166. P Skla'dal Anal. Chim. Acta 252 11 (1991) 167. K Rogers,M Foley, S Alter, P Koga,M E Eldefrawi Anal. Lett. 24 191 (1991) 168. J J Kulys, J D'Costa Biosens. Bioelectron. 6 109 (1991) 169. L H Goodson, W B Jacobs, A W Davis Anal. Biochem. 51 362 (1973) 170. L P Kuznetsova, L I Kugusheva, E B Nikolskaya, O V Yagodina Zh. Anal. Khim. 45 1426 (1990) a 171. V K Nguyen, L Ehret-Sabatier, M Goelder, C Boudier, G Jamet, J M Warter, P Poindron Enzyme Microb. Technol. 20 18 (1997) 172. E B Nikolskaya, G A Evtugyn, A V Svyatkovskii, R R Iskanderov, E V Suntsov, A A Prokopov, S N Moralev, B N Kormilitsyn, V Z Latypova Zh. Anal. Khim. 49 374 (1994) a 173. H C Budnikov, G A Evtugyn, in Biosensors for Direct Monitoring of Environmental Pollutants in Field (NATO ASI Ser.) Vol. 38 (Eds D P Nikolelis, U J Krull, J Wang,M Mascini) (Dordrecht: Kluwer Academic, 1997) p. 239G A Evtugyn, H C Budnikov, E B Nikolskaya 1064 174. H C Budnikov, G A Evtugyn, E P Rizaeva, A N Ivanov, V Z Latypova Zh. Anal. Khim, (1999) (in the press) a 175. Cholinesterase-Hemmtest. Screening Test zur Bestimmung von Cholinesterase Hemmenden Organophosphat- und Carbamat- Pestiziden in Wasser (Mannheim: Boehringer Mannheim, 1994) No. 1293460 176. InQuest OP/Carbamate Screen (Newtown, PA: Ohmicron, 1995) 177. H El Yamani, C Tran-Minh,M Abdul, M Dupont J. Fr. Hydrol. 18 67 (1987) 178. H El Yamani, C Tran-Minh,M Abdul, D Chavanne Sens. Actuators 15 193 (1988) 179. D Tingfa, Z Shiguang, T Mousheng Environ. Pollut. 57 217 (1989) 180. A Cagnini, I Palchetti,M Mascini, A P F Turner Microchim. Acta 121 155 (1995) 181. A Cagnini, I Palchetti, I Lionti, M Mascini, A P F Turner Sens. Actuators B 24/25 85 (1995) 182. S F White, A P F Turner, R D Schmid, U Bilitewski, J Bradley Electroanalysis 6 625 (1994) 183. I Palchetti, A Cagnini, M Del Carlo, C Coppi,M Mascini, A P F Turner Anal. Chim. Acta 337 315 (1997) 184. T Imato, N Ishibashi Biosens. Bioelectron. 10 435 (1995) 185. T Danzer, G Schwedt Anal. Chim. Acta 325 1 (1996) 186. T J Davis, G W Malaney Water Sewage Works 114 272 (1967) 187. V F Kramarenko Toksikologicheskaya Khimiya (Toxicological Chemistry) (Kiev: Vyshcha Shkola, 1989) 188. Fosfororganicheskie Insektitsidy. Obshchee Vvedenie. Gigienicheskie Kriterii Sostoyaniya Okruzhayushchei Sredy (Organophosphorus Insecticides. General Introduction. Hygienic Criteria of the State of the Environment) (Moscow: World Health Organisation, 1990) 189. C H Walker, H M Thompson, in Nondestructive Biomarkers in 190. M E Eldefrawi, K R Rogers, N A Anis, R Thompson, J J Valdes 191. S A Sokolova, G A Evtugyn, R R Iskanderov, A I Startseva, Vol. 63 Vertebrates (Abstracts of Reports), Siena, 1992 p. 9 Am. Chem. Soc. Symp. Ser. 543 114 (1993) V U Smirnova, Yu V Rydvanskii, Yu S Kotov, S A Patin Sb. Nauchn. Tr. GosNIORKh 312 122 (1990) 192. G A Evtugyn, R R Iskanderov, V U Smirnova, V Z Latypova, Yu V Rydvanskii, L N Punegova, G A Marchenko Ekologo- Toksikologicheskaya Kharakteristika g. Kazani i Prigorodnoi Zony (Ecological and Toxicological Characteristics of Kazan and 193. G A Evtugyn, E P Rizaeva, E E Stoikova, H C Budnikov 194. S A Piletsky, H S Andersson, I A Nicholls Macromolecules 32 633 195. L Shenghui, Y Jiannong, H Pingang, F Yuzhi Anal. Chim. Acta 335 Suburban Aria) (Kazan: Kazan University, 1991) p. 126 Electroanalysis 9 1124 (1997) (1999) 239 (1996) a�Russ. J. Anal. Chem. (Engl. Transl.) b�Dokl. Chem. Technol., Dokl. Chem. (Engl. Tran

ISSN:0036-021X    

出版商:RSC    

年代:1999

数据来源: RSC