摘要:

J. CHEM. soc. DALTON TRANS. 1988 297 1Structures of Borate-Aldohexose and Borate-Chetohexose Complexes inAqueous Solution. A Thermodynamic StudyRoberto ArugaDepartment of Analytical Chemistry, University of Turin, via Giuria 5, 70725 Turin, ItalyHeats of complex formation of the borate ion with the carbohydrates D-galactose, D-glucose,D-fructose, D-mannose, and L-sorbose have been determined by direct calorimetry. By means ofthe equilibrium constants the corresponding Gibbs functions and entropies were also obtained. Thepresent data refer to the aqueous medium, at T = 25 "C and I = 0.1 mol dm? Examination of theresults indicates a leading role, in several cases, of solute-solvent external factors (e.g.desolvations of the borate ion) in determining the different stabilities of the complexes.Amongstvarious factors for the ligands, some appear of minor importance, while the contribution of others(such as the strength of the borate-hexose C-0-B bonds) is fairly constant throughout one series.The present data seem also to confirm the presence of the furanose cyclic form of the carbohydratein the complex.The importance of the reaction of boric acid with polyhydroxycompounds in analytical chemistry and in the investigation ofmolecular structures is well known.' Such reactions could havea remarkable interest also in phytobiology, as the absorption ofborate by many plants could take place through polyhydroxycompounds such as polysa~charides.~~~At first, borate-carbohydrate complexes in solution weremainly investigated by means of the determination of stabilityconstants, even at different temperature^.^-" It must be noted,in general, that these quantities are merely conditionalconstants.In fact, when they are defined, the expression of theconcentration (or activity) of uncomplexed carbohydrate is thesum of the concentrations of various forms (i.e. a$, pyranosic,furanosic. ctc.) present in the solution, one only of whichpresumably reacts with borate (see below). Then the magnitudeof the constants is dependent on the experimental conditionschosen. In other words these quantities are useful for calculatingthe actual concentration of the complex under particularconditions, but they may not lead to reliable and generalconclusions about the stability and structural features of thecomplex.Among spectral methods, n.m.r.has recently given someinformation on the borate esters with D-fructose, D-glUCOSe, andother polyhydroxy compounds.' 'As calorimetric values of enthalpy and entropy are notavailable in the literature for these associations,12 it has beenthought of interest, in the present work, to determine the abovequantities by direct calorimetry for the reaction of borate withD-galactose, D-glucose, D-fructose, D-mannose, and L-sorbosein aqueous solution.ExperimentalReagents and Solutions.-Analytical-grade reagents werealways used: boric acid, C. Erba RPE-ACS 99.8%; D-galactose,D-glucose, D-fructose, D-mannose, and L-sorbose, Fluka puriss.Ionic strength and pH were adjusted to the desired values bysodium nitrate, sodium hydroxide, and nitric acid (C.ErbaRPE). To avoid any influence by mutarotation, the carbo-hydrate solutions were used at least 24 h after preparation.Eyuipment.-Calorimetric measurements were made at 25 "Cwith an LKB 8700-2 Precision Calorimetry System (isoperibol,incremental-titration type) and an LKB 8726-1 lOO-cm3titration vessel. Fuller details on the instrument and thecalibration have been reported.' The calorimeter was equippedwith a Radiometer ABU 12b autoburette for the addition oftitrant. pH Measurements were carried out with a Metrohm605 potentiometer. The calorimetric experiments were per-formed in a room kept at a temperature constant to within0.3 "C (Branca Idealair 'Zero' air-conditioning system).Procedure.-Three or four series of measurements, withdifferent concentrations of reagents, were carried out for eachborate-hexose system.Each series was performed in thefollowing manner. Successive amounts of sodium boratesolution (2,505 & 0.002 cm3) were added to a hexose solution(88.00 cm3) in the calorimetric vessel, and the heat for eachaddition measured. The hexose solution was previously broughtto pH 5.2 & 0.2. Sodium borate solutions were obtained fromboric acid solutions brought to pH 3 11.6 by NaOH. The pHvalues and the borate-hexose concentration ratios were similarto those used in the determination of the corresponding stabilityconstant^.^.' In this way complex species ofthe type [B,L]'- areabsent (B- = borate ion; L = carbohydrate).At the same timethe above constants (which are conditional constants, seeIntroduction) may be used for a reliable evaluation of theamount of the complex species formed under the presentconditions. The ionic strength of all solutions was adjusted to0.1 rnol dm-3 by NaNO,. The corresponding heat of dilutionwas measured by adding the same amounts of the sodiumborate solution to 0.1 rnol dm-, NaNO, (88.0 cm3) (pH 3 11.6),without any carbohydrate.The following results were obtained under these conditions.(a) A negligible heat of neutralization between H+ and OH-.(b) A negligible heat of reaction between the [B(OH),] - ion andthe proton. By assuming a pK, value of 9.1 for boric acid at25 "C and I = 0.1 rnol dm-, (ref.7) and AH = 13.8 kJ mol-'for its diss~ciation,'~ a heat of smaller than 0.02 J was obtainedfor this reaction. (c) A negligible contribution to the measuredheats from dissociation of polyborate species: since the borateconcentration was lower than 0.025 mol dm-3 at the end of themixing process (Table l), it must have reacted in the monomericformI5 and, moreover, the heat of dissociation of anypolyborates which may be present in the initial solution iscounterbalanced by an equal heat effect during the dilutionexperiment, so that it is eliminated in the calculation of thecorrected heat (see below). The heat for the partialdeprotonation of the hexoses was also calculated. It was foundto be not negligible in some cases.The experimental heats werethen corrected for this process (see treatment of theexperimental data). The pH values in the cell were measuredpotentiometrically, by means of a titration performed in thesame way as the calorimetric one. Finally, it must be noted tha2972 J. CHEM. SOC. DALTON TRANS. 1988Table 1. Experimental data for the mixing of aqueous solutions of sodium borate and hexoses (L) at 25 OC*1 03[BL-]/ 1 03[BL, -I/ 1 O[L]/L lOc,/mol dm-3 10cJmol dm-3 pH, mol dm mol dm-3 mol dm-3 CQJJ CQCJJD-Galact ose 1.30 4.00 9.039.239.359.44D-Glucose 1 .oo 2.99 9.579.9010.0810.20D-Fructose 1.30 0.90 8.809.209.439.57D-Mannose 1.30 2.00 9.109.339.479.56L-Sorbose 1.30 2.00 7.908.708.949.062.254.446.588.661.462.894.305.681.252.604.505.601.312.483.534.471.282.443.494.432.344.396.167.673.596.9910.213.33.607.0010.213.33.843.693.553.412.872.752.642.540.8 10.740.660.601.871.751.641.531.871.751.641.532.435.068.123.256.489.888.0511.213.215.422.528.82.635.027.489.788.3916.624.632.82.224.747.702.545.007.678.0010.810.415.322.328.62.264.4 16.688.848.3816.524.532.6* Sodium borate was added to 88.00 cm3 of L in the calorimetric cell in four successive amounts.Cumulative volumes added: 2.505, 5.010, 7.515, and10.020 cm3.Table 2. Molar thermodynamic quantities for complex formation reactions of the borate ion, [B(OH),] -, with hexoses in aqueous solution at 25 "Cand 1 = 0.1 mol dm-3- AGj*/ AHj*/ ASj*/Hexose j " log Kjb kJ mol-' kJ mol-' J K-' mol-'D-Galactose 121 + 2D-Glucose 121 + 2D-Fructose 121 + 2D-Mannose I + 2L-Sorbose 1 + 22.21 k 0.03'0.18 & 0.172.39 k 0.172.27 k 0.050.49 & 0.052.76 f 0.073.58 k 0.051.36 f 0.054.94 & 0.074.42 & 0.055.80 k 0.0512.6 f 0.21.0 & 0.813.6 k 0.812.9 k 0.32.8 & 0.315.7 & 0.420.4 k 0.37.7 & 0.328.1 f 0.425.2 f 0.333.1 k 0.3 --24.7 f 1.748.5 & 424 f 4-17& 115 & 3-2 k 3-3 & 1-33 + 2-36 f 2-6.81 & 0.0225.15 k 0.08-42 & 4167 f 12125 & 1258 k 846 k 959 + 4-84 & 4-25 & 6- 12.5 & 461.5 k 0.826.8 f 0.8j = 1 or 2 for stepwise reactions: [BL,,]- + L - [BLj]- (B- = borate, L = hexose);j = 1 + 2 for overall reactions: B- + 2L - [BLJ.See refs.5 and 7. The uncertainty given in each case is the estimated standard deviation.the potentiometric and calorimetric experiments showed thatthe achievement of the equilibrium in the cell, after eachaddition of titrant, was rapid enough for carrying out correctcalorimetric measurements with the LKB 870@-2 system.Treatment qf'the E,xperimentul Data.-The heat of deproton-ation of the hexoses was calculated by using the correspondingpK and AH values of acid dissociation reported previously.'Some of the experimental data are collected, as an example, inTable 1, where cB is the initial total concentration of borate inthe titrant solution, cL is the initial concentration of hexose inthe calorimetric vessel, pH,, [BL-I, [BL,-], and [L] are the pHand concentrations of the various species after each addition oftitrant, and where CQ, and CQ,, are the cumulative heats,corrected for dilution and for dilution and deprotonationrespectively.Molar enthalpies of association were determined from theexperimental heats (CQ,,) and the concentrations of thecomplex species: the latter were calculated from values of thecorresponding stability constants determined potentio-metrically 5 , 7 under the conditions of temperature and ionicstrength used in the present study.No contraction in volumewas found (in the limits of the instrumentation used) on mixingthe reagents.The AH values and the corresponding standarddeviations were calculated by means of the numerical methodof minimization of the error square sum for each measurement.' 'In order to check the enthalpy values, the experimental heatswere recalculated by using the AH " values obtained. Entropieswere calculated by means of the equation: AG*' = AH" -TAS r). The enthalpy values calculated by taking into accountonly some additions of titrant were equal to the values obtainedfrom the entire titration process. This shows that the conditionsof increasing pH in the calorimetric cell in the course of thetitration (Table 1) do not alter the hexose molecule in the timerequired for a measurement.Results and DiscussionThe molar thermodynamic quantities of complex formation arecollected in Table 2.No stepwise quantities are listed for D-mannose and r>-sorbose, as only [BLJ- was found for theseligands in the citedNo previous calorimetric data are available on the presentequilibria. On the other hand comparisons of the present resultJ . CHEM. SOC. DALTON TRANS. 1988 2973with those obtained from equilibrium constants at varioustemperatures do not seem reliable. The latter, in fact, may be soinconsistent that sometimes they differ even in sign [thefollowing entropy values, for instance, have been obtained forthe successive steps of the borate-D-glucose complex formation:ASl = 20, ASz = -34 (ref. 9); ASl = -27 and ASz = 31 JK-' mol-' (ref.lo)].Examination of the data in Table 2 shows that the factorswhich favour the formation of the present complexes are notattributable in a simple way to enthalpy or entropy alone. Thesequantities appear to be quite different, both for aldoses incomparison with chetoses and for the two steps of reaction. ForD-galactose and D-glucose, for instance, the first step is favouredby enthalpy and opposed by entropy, while an oppositebehaviour is shown in the second step. In the case of D-fructosethe general trend appears reversed in comparison with that ofthe two aldoses.The following factors should be taken into consideration for acorrect examination of the problem of complex stability in thepresent case. ( a ) The stability of the C-0-B ester bonds, whichare the fundamental link between borate and hexose: --C-OH+ HO-B---C-0-B- + H,O. (b) The presence ofhydrogen bonds.(c) The difference in stability between the com-plexed and uncomplexed form of hexose. This contribution mayassume a leading role in certain ring structures. The myo-inositol -borate complex, for instance, is very weak, in spite ofthe presence of three C-0-B bonds. This fact is due to the highenergy content of the 'chair' structure of myo-inositol in thecomplex compared with that of the usual chair.18 (d) Solute-solvent interactions. (e) Hexose-hexose interactions; in the caseof 1 : 2 complex species, mainly of steric nature.In order to evaluate the importance of factor (a) it is necessaryto ascertain in which form the hexose molecule takes part in thecomplex formation.It was found from n.m.r. data" that D-fructose and, probably, D-glucose are present in the 1 : 2 complexwith borate almost entirely in the furanose cyclic structure (seeFigure).In the course of the present calorimetric experiments withD-galactose and I,-sorbose, a small endothermic effect, slightlygreater than the experimental error, was repeatedly observed atthe end of the exothermic main process. Taking into accountthat the present hexoses are prevailingly present in solution inthe pyranose conformation 9 3 2 0 and that some previouscalorimetric measures indicate that the pyranose - furanoseisomerization is endothermic," the above mentioned endo-thermic effect may be considered as confirmation of the n.m.r.results. The fact that no additional thermal effect was observedfor some of the hexoses may be explained by considering thatthe pyranose-furanose isomerization is rather fast under thepresent conditions of pH.I9 Consequently the heat ofisomerization could be entirely superimposed to the heat ofcomplex formation in these cases.Possible exchanges between xand p forms during complex formation should give negligiblethermal effects, as the corresponding molar enthalpies werefound to be very low.2oThe furanose form of carbohydrate in the present complexescan be justified by examining the possible C-0-B bond pairsformed by the two cyclic isomers of the hexoses. Hexopyranosescould give two C-0-B bonds through two adjacent OH groupsof the ring.Owing to the different orientation of these groups(i.e. axial and equatorial) and the semi-rigid ring structure, thesebonds would be fairly unstable.' Two bonds with 1,3 OHgroups of the ring in 'parallel' position would also be possible,but, in this case too, they are not very stable." Finally, twobonds with the OH groups on the carbon atoms in positions 1,2II I 1 II I 1HOHC5 'IH O H ~ C ~HOHz CIHOHC'CH~OH( 4 )0 CHz OHHOHzCHOHzCIHOHC( 3 1Figure. Structural schemes of aldohexoses (left) and chetohexoses(right) in the furanose form: z-D-galactofuranose (I), z-D-glucofuranose(2), p-u-mannofuranose (3), P-D-fructofuranose (4), and x-L-sorbofur-anose (5). The numbering of the carbon atoms is exemplified in (1) foraldoses, in (4) for chetosesare possible for borate with chetopyranoses.Such bonds arestrong, owing to the ease of orientation of the hydroxy groups ofhexose. They are quite similar to the usual pair of bonds in theborate esters of 1,2-glycols. In the case of the furanose ring astrong binding (quite similar to the preceding one) could beformed through the two OH groups of the open chain inpositions $6. This is possible for aldoses only (Figure).Moreover the difference in orientation (equatorial, axial) of twoadjacent OH groups in cis position is much smaller in thefuranose ring ('envelope' conformation) than in the chairconformation of the pyranose ring. The OH groups of theformer ring are easily oriented as in the case of the 1,2-diol openchain (see ref.21 and Corey-Pauling-Koltun, CPK, models). Atthe same time the corresponding pair of bonds with borateshould lead to an even more stable structure in the case of thefuranose ring than of 1,2-diol, for entropy reasons. Owing to thequasi-rigid ring structure, in fact, the formation of the bondstakes place, in this case, with a smaller loss of conformationalentropy than for the open chain. It can then be concluded,from the preceding experimental and structural observations,that the hexoses are present in the borate complexes in thecyclic furanose form. As concerns the a or p configuration ofhexofuranose, that which possesses a pair of OH groups in cisposition on carbon atoms 1,2 for aldoses (or 2,3 for chetoses)should react with borate.If the structure proposed here is theactual one, then factor (a) above has a fairly constant influenceon the stability of the various complexes.Possible structures of each borate-hexose ester have beeninvestigated by means of CPK models also. As regards the 1 : 1complex of D-galactose (in the form of x-D-galactofuranose),besides two C-0-B bonds on the 1,2 carbon atoms, hydrogenbonds can be present between the OH groups on the 5,6 carbonatoms of the side chain and one OH group of borate (Figure).The consequent loss of conformational freedom by the sidechain, together with the fact that no charge neutralization takesplace in these reactions, could account for the negative A S , J. CHEM. SOC.DALTON TRANS. 1988value ( - 4 2 k 4 J K-’ mol-l). Taking into account the compactarrangement of borate and sugar in the 1 : 1 species, a partialdetachment and a regain of conformational freedom of the firstligand should be caused by the formation of the 1 : 2 borate-D-galactose species. The more positive enthalpy and entropy forthe second than for the first step can be justified, at leastpartially, in this way. It must be noted, in any case, that thestrong positive AS,‘ for D-galactose does not seem to beexplained through this factor alone.The difference of x-D-glucofuranose compared with X-D-galactofuranose lies in the position of the 5,6 side chain, which isnow opposite to borate in the complex species. As no immobi-lization in the first step as well as no detachment in the secondstep is possible for the side chain of the hexose in this case, theless negative A S , and the less positive AS, for D-glucose thanfor D-galactose can be explained in this manner.The corre-sponding enthalpies too are in qualitative accordance with thisexplanation.The main structural features of p-D-fructofuranose in com-parison with aldofuranoses lie in the two CH,OH groupsin positions 2 and 5. From CPK models, the hydroxymethylgroups of D-fructofuranose give two hydrogen bonds with thetwo unbound OH groups of borate. Moreover, in this case thestructure of the ligand being more rigid than in the case ofD-galactofuranose, hydrogen bonds are formed with a smallerloss of conformational freedom.p-D-Fructofuranose can thenbe considered as a structure well enveloping the borate ion andcausing, presumably, strong desolvation phenomena of thelatter. The clearly positive AS1 for D-fructose, together withless exothermic AH,” values than for the two aldoses are inagreement with this possibility. No steric factors, in the secondstep, should oppose the formation of bonds as strong as in thefirst step, while desolvation processes of borate should be nearlyabsent now. The most favourable AH,” together with the mostunfavourable AS, among the stepwise values in Table 2 agreewith this possibility for D-fructose.p-D-Mannofuranose has a structure similar to that of x-D-galactofuranose. Moreover, the 1,2,3 OH groups are all in cisposition in the former hexose. Such a structure, as can beinferred from CPK models, leads to a compact borate-hexosearrangement by means of hydrogen bonds, and to possibledesolvation processes almost as strong as for D-fructose.Thesimilar behaviour of D-mannose compared with D-galactose andD-fructose is well reflected by the overall enthalpy and entropyvalues. In fact, as can be seen in Table 2, the quantities forD-mannose are intermediate between those of the other twohexoses.The two CH,OH groups are oriented differently in WL-sorbofuranose than in p-D-fructofuranose, so that differentsolute-solvent interactions should be present for the twochetoses in the two steps of complexation. In any case, theavailability for L-sorbose of overall values only does not allowone to draw conclusions on this point.In conclusion, the present calorimetric data do not confirmprevious explanations of different stabilities only based oninternal factors.The present results, in particular, do notconfirm the conclusions of a previous study,’ according towhich the higher chelating ability of chetoses in comparisonwith aldoses should be caused by ester bonds with OH groupsoutside the pyranose ring (positions 1 and 2) for the former, andwith OH groups of the pyranose ring for the latter. It can beconcluded, on the other hand, that in several cases solute-solvent interactions are of great importance, among the abovementioned factors, in determining differences in complexstability. Desolvation processes of the borate ion, in particular,seem to play a leading role.Among the other factors cited, the contribution of some isfairly constant throughout the series of ligands such asC-O-B bonds, some (such as hydrogen bonds or hexose-hexose interactions in the 1 : 2 species) seem to have a minorinfluence.As regards the change from the pyranose to thefuranose structure of hexoses, the minor heat attributable to thisprocess in comparison with the total heat should lead to theconclusion that this factor too is of minor importance.References1 J. Boeseken, Red. Truv. Chim. Pays-Bus, 1942, 61, 82.2 M. E. Winfield, A m . J. E.up. Biol. Med. Sci., 1945, 23, 11 1.3 W. J. Evans, V. L. Frampton, and A. D. French, J. Phys. Chem., 1977,4 G. L. Roy, A. L. Laferriere, and J. 0. Edwards, J. I m r g . Nucl. Chem.,5 P. J. Antikainen, Suom. Kemistil. B, 1958, 31, 255.6 J. P. Lorand and J. 0. Edwards, Org. Chem., 1959, 24, 769.7 P. J. Antikainen and K. Tevanen, Suom. Kemistil. B, 1959, 32, 214.8 E. W. Malcom, J. W. Green, and H. A. Swenson, J. Cheni. Soc., 1964,9 P. J. Antikainen and I. P. Pitkanen, Suom. Kemistil. B, 1968,41, 65.10 J. M. Conner, J . Inorg. Nucl. Ciiem., 1970, 32, 3545.1 1 M. Makkee, A. P. G. Kieboom, and H. van Bekkum, Recl. Trav.12 J. J. Christensen and R. M. Izatt, ‘Handbook of Metal Ligand Heats,’13 R. Aruga, Aust. J. Chen?., 1981, 54, 501.14 M. J. L. Tillotson and L. A. K. Staveley, J. Chem. Soc., 1958, 3613.15 T. Pail, Acta Chim. Acad. Sci. Hung., 1976, 91, 393.16 J. J. Christensen, J. H. Rytting, and R. M. Izatt, J. Chem. SOC. B. 1970,17 L. G. Sillen, Acta Chem. Scund., 1962, 16, 159.18 S. J. Angyal and D. J. McHugh, J. Chem. Soc., 1957, 1423.19 W. Pigman and J. Sowden, in ‘The Carbohydrates,’ ed. W. Pigman,Academic Press, New York, 1957.20 S. Ono and K. Takahashi, in ‘Biochemical Microcalorimetry,’ ed.H. D. Brown, Academic Press, New York, 1969.21 L. Hough and A. C. Richardson, in ‘Rodd’s Chemistry of CarbonCompounds,’ ed. S. Coffey, Elsevier, Amsterdam, 1967, vol. IF, p. 11 3.81, 1810.1957, 4, 106.4669.Chiin. Pa?-s-Bas, 1985, 104, 230.Marcel Dekker, New York, 1983.1646.Received 23rd October 1987; Paper 71189

ISSN:1477-9226    

DOI:10.1039/DT9880002971

出版商:RSC    

年代:1988

数据来源: RSC