_.,
JOURNAL OF RESEARCH of the National Bureau of Standa rds- A. Physics a nd Chemistry Vol. 66A, No.3, May- June 1962 Oxidation of Aldoses With Bromine! Horace S. Isbell (February 12, 1962) Rates of oxidation of aldoses with bromine have bee n reappraised and interpreted in th e light of present concepts of conformation and reaction mechanism. It is suggested that differences in t he rates of ox id ation of the a and /3 anomers are largely determined by differences in the free energy req uired by the reactants for passing from the ground state to the complex in the transition state. Structures for t he aldoses in the ground states and in the transition states are postulated, a nd factors affectin g the energy required for reaching the transition states from t he ground states are di scussed. The relative ra tes of ox idation are in accordance with the hypothesis that cach of the aldoses in the ground state has the conformation predicted by Reeves, and, in the transit ion state, has a conformation in which the oxygen atom of the C1-hydroxyl group lies in the plalle formed by t he r in g oxygen atom, C1, C2, and C5. Presumably, this conformation is stabilized by resonance in volving the oxygen atom of the ring. For a ldoses havin g high stability in one cha ir conformation, the rates of oxidat ion of the anomers d iffer widely; in each in stance, t he anomer in which the Cl-hydroxyl group is axial is ox idized more slowly t han the anomer in which this group is equatorial. For aldoses having less stabil ity in a cha ir conformation, the rates of oxidation of t he a nomers difrer lcss widely, but, nevertheless, show a defin ite corrclation with the angular posit ion of the Cl-hydroxyl group relative to the plane of t he ring. For aldoses for which the stabili ty in both chair conformations is so low that they probably exist in a variety of conformations, the rates of ox idat ion of the anomers show little difference and no particular correlation with the angular position of the Cl-hydroxyl group. The presence 0 " absence of an oxygen atom in the rin g is used to account for the large diff erences between the rates of bromin e ox idation of t he aldoses a nd those of deri vati ves of eyclohexanol. Differences in conformation in the t ra nsition statp, associated with t he presence or absence of this oxygen atom, likewi se accoun t for t he fact that t he relative rates of oxidation of t he axial and equatorial isomers in the two classes of compound a re reversed. Because of uncerta in ty as to the anomeri c configurations co mmonl.v assigned to son1(' of t Il e aldoses, the co nfigurations of 22 aldoses were reappra ised. Advantage was taken of t he principle that the anomer p reponderatin g in the equili brium solu tion has l1'ans hydroxyl groups at Cl and C2. Except for crystalli ne D-glyceTo -D-ido-heptose , the assignments of con fi guration based on this p ri nciple agree with t he configurations genera ll y accepted. Classifi ciation of cr ys talline D-glyceTo-D-ido-heptose as an a-D-pyranose necessitates co rrection of ea dier reco rds in which t hi s suga r was co nsid ered to be a /3 -D-py ranose. In accorda nce with the author's earli er formulation, oxid ation of the axial a nomer is beli eved to take place by two co urses: (1 ) di rect ox id ation and (2) conversion to the eq uatori al a nomer by the anomerizat ion reaction and the subsequent oxid ation of this ano mer. The relative importa nce of the two courses is not considered in thi paper . It is poin ted out, however, that the actual difference in t he rates for t he direct ox id ation of the two anomers must be at least as great as that observed For the overall rates of oxidation. 1. Discussion of Prior Work plane or the ring, and Lha.t "a comparison or the reaction rates or alpha fLnd beLa pyranoses should ULilization of bromine-oxidation measurem ents provide inform ation about the con formation of for st u d~T of the ring structure of aldoses began with the pyranose ring" (p . 524 of [6]) . Subsequently, the observation of Isbell and Hudson [1,2] 2 that, through the brillian t work of R eeves [7 ,8], H asscl in slightly acid solution, pyranoses are oxidized to and Ottar [9], Beckett fLnd co-workers [10], fLnd 1,5-la,ctones without cleavage of the ring. Isbell others, factors that affect tho conrormation or thc and Pigman [3,4,5 (pp. 435 to 456)] showed that, pyrfLnoses have become known. Theoretical con for many aldoses, the a and (3 anom ers are oxidized siderations fLS well as extensive experimental (hta at widely different rates, and Isbell [6] sought to have established that Isbell fLnd Pigman's slowly classify the anomers of aldoses according to the oxidizable "alpha" anomers have fL chair con form a angular position of the Cl-hydroxyl group relaLive Lion wiLh an ftxi al C I-hydroxyl group and their to Lhe phne of a strainless, pyranoid ring. \iVhen rapidly oxidizable "beta" anomers have a chair those studies were made, the concept of conform ation conform ation with ftU equatori.al Cl-hydroxyl group WflS just beginning to emerge. Nevertheless , it [11]. was concluded that the aldopyrfLnoses exist in a Kinetic studies have shown that oxidfttion by defi. nite conformfLtion in which the a and (3 positions bromin e is a complex process. Bunzcl and Mftthews are not s:rmmetricfLlly located with respect to the [12] found that the reaction is rapid in neutral solu tion and slow in fLcid. Perlmutter-lIaym fLn and
1 Presen ted, in part, before the Division of Carbohyd rate Chemistry at the P ersky [13] recently eonfu'm ed th e work' of Bunzel l32d M ee ting of the American Chem ica l Society, at New York, N. Y., 0 11 Se p tem and Mathews and presented evidence L]mt the ber 8, 1957. ' Figures in brackets indicate the literature references at the end ofthi5 paper. anionic form of D-glu cose is oxidiz ed much faster 233 than D-glucose itself. Bunzel and Mathews also C4 established that the rate of oxidation of pre-equili BH ___ _ ~6' C 2 0 /\ C5 brated D-glucose is first order with respect to the I ~ aldose and fr ee bromine. Several investigators have /' I C3 advanced hypotheses to explain the difference in the ~ , J~ rates of oxidation for the a and {3 anomers. Isbell 'Br ' and Pigman [4] showed that oxidation of the sepa rate anome1"S is complicated by the simultaneous 2 C4 3 anomerization r eaction. In the oxidation of a-D BH -----O-::'\ ~. C5 glucose, form ation of the more reactive {3 anom er and ICI oxidation thereof raises the overall rate of oxidation; :21 C3 in the oxidation of {3-D-glucose, formation of the less H___ f'. - Br----·Br reactive a anomer lowers the overall rate. Thus, th e actual difference in the rates for the direct oxida tion of the two anomers must be at least as great as C4 the observed difference in the overall rates. Barker, ,.0 C~ ° /\ C5 Overend, and R ees [14] consider the values for the / ~ H/ : C3 differences in the overall rates to be of little sig ,,_____ (H nificance because of the anom eriz ation of the slowly 'Br-l__ Br ) ~ oxidizable an omers, which, the~T believe, takes plaee nearly co mpletely b efore oxidation. Bentley [15] FIGU RE 1. H ypothetical t?"ansilion slales. suggested that the higher activity of the anom e1'S in (a) Intramolecular, base-catalyzed mechanism. whieh the C1-hydroxyl group is in an equatorial (b) Intermoleenlar, base-catalyzed llleehanism. (c) Intra molec ular, concerted mechaniSlll. position arises from the release of strain in the rate 1 determining step. H e also noted that the ratios of rates of oxidation for the anom eric aldoses appear to experimen tal data. Structures (a) and (b) arc b e related to th e conform ational stability of the symbolic of either a fast removal of a proton from 1 aldoses. The subj ect was discussed in an excellent the C1-hydroxyl group followed by a rate-determin review by Shafiz adeh [16] . ing hydride transfer from C1, or a partly concerted Various luechanisL11s have been proposed for the process in which the proton removal is not a com oxidation of aldoses with bromine. Isbell and pletely separate step. Structure (b) is similar to Pigman [4] suggested rapid addition of free bromine one proposed by Swain, Wiles, and Bader [21] for to the ring oxygen atom, followed by slow elim.ination the transition state in the oxidation of 2-propanol 01' hydrogen bromide at a rate depending on the stereo with bromine. Structur e (c) is similar to an inter merie position of the C1-hydrogen atom. Lichten mediate proposed by Bark er , Overend, and Rees [22] and Saxe [171 postulated rapid addition of bromine for the oxidation of tert-butyl cycloh exanol, but differs to the anomeric oxygen atom; this is followed by in certain res pects to be discussed in section 3. elimination of the hydrogen atoms of C1 by base In a preliminary communication, Isbell [23] catalysis, and formation of two bromide ions. Fried advanced the hypothesis that the relative rates of berg and K aplan [18] proposed elimination of the oxidation of aldoses depend 011 variation in the C1-hydrogen atom as a hydride ion by direct attack difference in free energy for the reactants in the of bromine, a concep t advanced by Kaplan [19] ground state and in a h'ypothetical transition state from study of the oxidation of ethanol and acetalde in which the Cl-hydroxyl group is restricted by hyde. Bentley [15] suggested elimination of the resonance to the plane formed by the ring-oxygen C l-hydrogen atom as Br2H - from a bromine-con atom, C1, C2, and C5. This concep t will be de taining in termediate. Isbell [20] suggested a bicyclic veloped in the next section of this paper. intermediate that decomposes intramolecularly with elimination of hydrogen bromide and a bromide ion (see p. 17 of [16]) . Although the precise structure 2 . Rationalization of Reaction Rates in of the bromine-aldose complex in the transition Terms of Free-Energy Differences state seems obscure, the structures given in figure 1 appear to account satisfactorily for many of the In accordance with transition-state t heory [24, 25, 26, 27], the rate of reaction depends on the difference 3 T he relative rates of anomcl'ization and oxidation depend on pII, concentra in free energy between the reactants in the ground tion of free bromino, and other factors. Whon tho rate of anomerization is high in comparison with the rate of oxidation , there is little difference in the overall rates state and an activated complex in the transition of oxidation for the anomers. Isbell and Pigman [4] represented the rate eonstant sta,te. 4 The difference in free energy depends solely for the overall oxidation of a-D-glucose as aka+kt , wllere a is the concentration of free bromine; ka, the rate constant for the direct oxidat.ion of a-D-glueose and kl on the nature of the two states and is independent of the rate of con version of the a anomer to tho {3 anomer and the oxidation thereof. I t was assumed, as a rough a pproxilnation (when tho bromine concentration is the manner in which the reaction proceeds. In th e high), that kJ corres ponds to the rate of conversion of a-i)-glucose to {3-D·glucose (calculated from the rate of mutarotation of a-D-glucoso at an acidity and tern pera present study, the structure of the aldose was varied ture co mparable to the conditions prevalen t duri.ng the measurements of oxida while other factors were held constan t in so far as tion). Undcr the conditions selected to minimize anomerization, it was esti m ated that about 80 percent of the overall oxidation of a-i)-glucos] takes place directly, and 20 percent indirectly by co nversion of a-IJ-gl ucose to ~- D -g l uco se 4 1"l1e free energyofactiyation LlF::I:: = - RTI n K =, where K "'= is theeQuilibriun and oxidation thereof. E valuation of the relative importance of thc direct constant fo r the aldos9-bJ'omine complex in the tran sition state. 'rhe frce-ell en oxidation and thc indirect oxidation was not attempted for other aldosos because, change in passing from the ground state to the transition state in yolvcs chan in some instances, the m utarotation reaction involves several molecular sp:::lc ies in entropy as well as energy, ancl no attempt has been made to assess the rclal and the value of kJ is highly uncertain (see p. 147 of [3]). importance of the two factors. 234 possible. For this reason , the differenees in 1'<1 tes oJ EOUATORIAL ANOMER I:eaction ,C,111 be ~ss,um ed to fl ri e for th e most par t from dliIer ences In iree en ergy needed to conver t th e O ~' C2 0 various aldose Lo th e structures chal'itcteristi c of the 4 transition slates.'l The aldose ill the ground stilte C I (q would h ave t he con [ormation o[ lowest [ree-energy, C3 " H proba bly the on e seleeted by R eeves from con sidera .H tion of instability factor s [sj. In Lhe activated inter ' Br ' mecli itte in the transition state, the lUTitn gemen t of GROUND STATE TRA NSITION STATE the f1lo ms of th e carbohydrate portion would lie be tween that of th e aldose and that of the b ctone. F urthermore, stabili zation by resonan ce would [avor II XIAL ANOMER <~ tr~nsition state in whi ch the anom eric oxygen atom hes 111 the sflm e ph~n e . as .C l , C2, CS , and the r in g ?xygen ato~n. Va~'] atIO n 111 t.lw free energy involved H ~C2 0 C 4 III overeommg th e mtramolecular forces encountered CI in moving the atoms from the ground state to this C3 C5 transition state provides a basis Jor interpretation of 0 - the rates of reaction. Thus, with an aldose havin o' GROU ND STATE TRANS ITION STATE an equatorial Cl-hydroxyl group, a small movement FIGURE 2. Ground stales and t?"ansition states f or the anions of at Cl wou~cl sufiice Lo g!v:e the r equi site arrangem en t the equatorial and axial anomers. of atoms (01' the Lntn ~ l tlOn sLate (see fig. 2); little In tlw transition states, carbon atoms I, 2, ancl5, as well as theoxygcn atoms tend energ,v would be l'eqllll'ed fI,nci the rate of reaction LO lie 111 the same plane. ' would be high. With an aldose having an axial Cl ~ 49. The rates for the fl,xial all Olll ers of group 2 aver I hydroxyl group, JormrtLion o[ the tml1 silion state age 2.7, wh er eas the rates for t he equatorial anomers would require di ~ l~l aee l1l en t or the anomeric oxygen aye!'age ~7. Th.us, less energy is required for acti a.tom from a yOS] t~on perpendlC ular to the pyranose vaLlOn o[ the axwl anoill ers or t he aldoses of group 2 nng to a posltlOn 1I1 the plane of th e 1'1110' in fn,et a . f . , to " than r o~' t il <;> se or group 1, but ).1101'e energy is required CI hmge 1Il con ormatIOn . ['h e rree energy required for such a change depends on the structure o[ the for aetlVatIOll of the eQuatol'wl anomers. I n both groups, the modification having an equaLorial Cl e ~l tjre . molecule; . itnd hen ce would be expecLed to dlfi er lor the varIOU S aldoses. hydrox.vl group in tbe ground state is oxidized Ill ore Ta ble 1 gives It summary or the ntLes o[ oxici alion ~' apid l y t ~l~n .i ts .anomer . Thu , t he energy barrier (relitLive to the rate [or a-D-glucose ) for t he nJdoses III the .0xldatlOll ]S lower for the equatorial t han [or sLud iecl by I sbell and Pigman, as well as herelofore t he aXial f1no mers. The observations are in accord unpublished rate m e
1'fol'C reactive anomer Less reactive anomer
Instability Instability Postulated Relative factors for Postulatecl Relative factors for R a tios o[ Aldose b conformation rate of conforma- conformation rate of confonna- overall ra te in ground reactioll d Lion with in groulld reaction d tion with constants (or state c eq un Loria I " stat,e c eq uatorial ' the anomers Cl-O II Cl- On
Oroup I
D-X ylose. ___ ._. ______. ___ .______f3-C A 52 none a-C A 2.8 Ll2. 3, 4 19 D-G Jucosc. __.______fJ -C A 39 nono (t'-C A 1. 0 II , .6.2. 3.4.5 39 D-glycao-L-ylttro- lleptose ______f3 -CA 53 none a-CA 0.8 It. 1'>2.3.4_5 66 M a ltose (4-0-a-D-glucopyranos yl-D-glucose) ______f3-C A 48 none a -C' A .8 1-1,1'>2.3.4.5 60 Lactose (4-0-,B-D-galactopyranosyl-D-g lucose) ______fj -C A ~o none a-CA . 9 II , Li2.3.4. 5 33 L-Arabinose______a-CE 52 4 f3-CE ~. 0 1'>2.3 l i D-081actose ______{j-CA 50 4 a-CA 1. 3 IT , 1'>2. 3.5 ~8 D-glycerO-D-(!alacl o- I-l eplose ______f3-CA 58 4 a-CA t. 3 11, 1'>2.3.5 45 D-(!lycero-I.-(!alaclo-H eptose ______f3-CA 56 4 a-CA 1.1 It, 1'>2.3.5 51 ------1-----1------1-----1------1------A vcrage______49 ______1. 4
Oroup 2
D-Lyxose______f3- C A 14 1'>2 a-CA 4. 9 3.4 2.9 D-Mannose ______f3-CA 24 L\2 a -CA t. a II. 3. 4. 5 15. 0 6- DcoxY- I.-ma nnose ______fj-CA 24 1'>2 a -CA 2. 8 II. 3. 4_ 5 8.6 D-(/lycero-I.-mal1no-Heptose______fj- C:\ 56 1'>2 a -CA 2. 3 II. 3. 4. 5 24.3 ])-'1'alose ______f3-C A 22 1'>2.4 a -CA 2. 4 II. 3 ..> 9.2 D-(!lycerG-D-lalo- ll e ptose , ______f3-CA :13 1'>2.4 a-CA 3.5 I I. 3. 5 9.4 D-yl ycerO-L-lalo- 1[ e pLose ,______f3-CA 41 1'>2.4 a-CA 2. 9 II. 3. 5 14 . 1 D-OUlo se ______f3-CA J3 3, 4 a-CA 2.2 Li2. 5 5.9 D-gl'Yf'ero-D-yulo-li eptosc______f3-CA .1 2 3, 4 t.4 1'>2.5 8_ 6
~t\. vera-ge ______2i ______2.7
Group 3 (Oxidized at intermed iate ratcs)
J.-Hi bose______f3-CA a. 1 3 D-A lIose g______f3- C A 5. 0 3 D-Altrose g ______a-C A 5.4 4.5 N eolactose (4-0-B-D-galactopyranos yl-D-altl'ose) ______a -CA 5. a 4.5 0-Idose g __ •• ______a-C g a.5 D-(!lycero-o-ido-Il c ptose______a-C E II
a Except where noted otherwise, tbe d ata are [rom t he work o[ Is bell a nd Pig c Postulated conformation based on the instability factors of Rce ves [8Jex pressed man repor ted on pages 436 and 455 of referenee [5J_ ill t he symbols of Isbell a nd 'l'ipsofl [28J. b Since the origilJ al publications of Isbell and Pigman, the following changes d Based on the rate constant for a-o·gluco pyrano'3c as u nity. in nomenclature fl a ve been m ade: e 'I'he instability factors urc cited for t he chnir conformation h aving an equa a-L-arabinose to ,B-L-arabinose, torial C I-O H group regardless of the a,ctual con/ormation of the substance. For ,B-L-ara binose to a-T~- arabino se, the wore reactive anomers t.his con formatioll is presumably that of the grou nd ,B-L-ribosc to a-L-rihosc, stU I.P; for the less reactivo a nomers it is supposedly the less stable co nformation. o-f3-~alahepto sc to D-(!lycero-L-gZuco-heptosc, The nUlllbers designate the individ ual carbon atoms bea ring axial groups (olhor D-a-l il annohcptose to D-(/lycero-D-galaclo-he p toSC' , than hyciro!(en) . II. refers to the enhanced e fTcct (Ilassc\ a nd Otlar) o[ 2 axia l D-,B-guloheptose to D-glycerO-L-(/ulacl o-hepto5e , groups on the same side of the ring, and Li2 refers to a highly unstable arrange D-a-galaheptose to D-(jlycerO-L-17la.nno·hepwso, mcnt of the carbon-oxygen bonds at carbon atom s 1 and 2. D-a-gu!oheptosc to D-glycero-J.-lalo-hc ptosc, f Values calculated from measure ments reported in J . Hescarch ~ BS 20, 97 D-,B-Illannoheptose to D-glycero-D-talo-heptosc, (J9:38) RPI069. D-a-glucoheptose to D-(Jlycero-D-yulo-he pLo H OH as with the axial anomer of glucose. However , the unfavorable effect of this eclipsing would be counter acted, in some measure, by the release of strain arising from separation o[ the h y dl'ox ~d groups of C1 and C3, and hence the change in conformation, as ,,-ith the manna and tala configuratiops, would require less energy than that required in the glucose series. The rate or oxidation for /'l-D-lyxose was found to be low '---- OH '-- OH in compa,rison with the rates [or the configurationally AX IAL ANOMER GROUP I AXIAL ANOMER GROUP 2 r elated aldoses, /'l-D-mannose and D-glycero-/'l-L manno-heptose, but it is close to the rates for /'l-D FIGU RE 3. Relative positions of atoms attached to carbon atoms 1 and 2. gulose and D-glycero-/'l-D-gnlo-heptose_ The relatively low rate of o:-..idation [or /'l-D-Iyxose, and the high (When C 1 is t urned on the CI- C2 axis, high-energy states a rise from ecli Psin g o[ the groups of C I an d C2.) rate for a-D-l.vxose, arises at least in part, from a fas t 236 anomerization r eaction. (See p. 147 of [3]. ) The T ABLE 2. Equilibriwn pro portions of the anomeric aldop ym unusually mall ratio for the rates could be explained noses· equally well by a non-chair conformation for one or I both. of the anomers. It is noteworthy tha t R eeves Composition 01 the equilibrium solu tion I predIct ed for ,B-D-Iyxose a mixture of the chair con Aldose Calculated, Irom formation s or, possibly, a conformation in the boat E sti mated, rrom optical rotation , i bromine oxidation assum ing only two r' skew cycle. constituents . The aldoses of group 3 behave somewhat differ ently from those of groups 1 and 2. The freshly dis cis trans cis trans solved , crys talline aldoses are oxidized at inter anomf' r b anomor b anomer b ano mer h mediate r atcs, and the oxidations o[ Ghe aldoses present in equilibrium mixtures do not show abrupt ch ang ~s m rate. .Small changes in the rates during the OXidatIOn s lllchcate that the equilibrium solutions o.f both D-!:tllose and D-ribose contain a small propor tIon (pOSSIbly 10 %) of a modification oxidized more iIIJ~w"! , ,, . ~!" ,) ii.i::iiii •••• !1 •••• 11 •••••~.! •••••• ~! rapidly than the rest of th e aldose. Both aldoses L- RhamDose ______c ______31. 0 69.0 26.9 73.1 D-gl-ycero- D-galaclo-U eptose______32. 8 67.2 ______~xhibi ~ complex mutarotation r eactions, presumably D-gl-ycero-L-galaclo-H eptose______37.0 63.0 ______lllvolvmg both pyranose and furanose modifications. D-glycero- L-!ll-uco-H eptose______37.2 62.8 D-Altrose, in an equilibrium solu tion, was found to ~:glt~~~~g : f%!'ii~~ ~;~;eo-~ ~::::::: :::: _____ ~ ~ ~ ~ ______8~ ~ 2 :: ::: ::::: : : :::::: :: be oxidized at essentially the same rate as the D-glycero-L-manno- Heptose ______._ 20.6 79.4- __ : : :::::: : : ::::: : : : 1 D·glycero-D-lalo-H eptose d ____ • ______._ 38 62 ___ ._. ______. ___ fr esh~y di ss ol~ ed flldo se, and a carefully purified D-gl ycero-L-lalo-H eptose d __ .______33 67 ______36.8 63.2 solutIOn of D-Idose was found to be oxidized at a Kf~A~~~~~: :::::::-:::-:·:::::::-:::: ~~ : ~ ~~: 3 36. 0 64.0 uni.form rate, abouL six Limes that of a-D-glucose. • Exce pt where mdlcated, these data were com pi led from the work of Isbell NeIther D-altrose nor D-idose gave evidence for a and Pigman given in chapter XXI X or reference [5]. j substan tial proportion of a second modifi CiL tion in . b In anomers <:lesignated cis, t he hydroxyl groups at C1 a nd C2lie on the same 7 Slele ~r the n ng: lD anomers deSigna ted trans, . they lie OU Ol)posi te sides or the ring. th e equilibrium mixture. D-glycero-n-ido-HepLo o knowledge of con fi guration aod prope rtIes IS ID suffiCie nt for classification 01 ~ PJT a ~lO Se IS ~1 e only known crystalline aldose havin g Lhc ano mcl'S. R ~, ~~~~ es calcul ated Irom data reported in J. Research NBS 20, 97 (1938) the ~do conngw·atIOn. When the aldosc was first 1 crystallized [31], an unusual conformation was suspect ed (see p. 530 of [6]); this is now con sidered constituents in the equilibrium solu tions o[ the to b e eE, the conformation in which th e a anomcr above aldoses are respec tively, ,B-D-ribopYl'anose, has an equa Lorial C1-hydroxyl group. Oxidation ,B-D-~llopyrano se, a-D-altropYl'anose, and D-giycel'o of freshly dissolved crystalline D-glyceTo-D-ido-h ep tose a-D-ulo-heptopyranose. Excep t for the last-n amed takes place about 11 times as rapidly as Lhat of aldo ~e, the e assignments agree with Lh e stl'uctm es a-n-glucose. Oxidation of an equilibrium solution preVIOu sly .accep ted. Classification of crystalline of D-glyce ro-D-ido-heptose gave eviden ce 1'01' th e D-glycero-D-ulo-heptose as an a-D-pyranose in the pres~nce ?f a small proportion of a slowly oxidizable CE conformation constitutes an iln portant chanO'o modIficatIOn ; the character of this modification and all work on this compollnd that has been basbed remains to be determined . on the assumption that it was the ,B anomer will In tel'pretation of the oxidation of D-glycero-D -ido need to be revised. On the basis oJ the trans anomer rule, it is postulated that the a PYTanose modi fi cation heptos.e and certain. oth.el' ~ u g ar s is hampered by lack 01 kno\,,rledge 01 therr nng structures and ano of D-idose, also, preponderates in aqueous solutions. meric configurations. When both anomers are known Inasmuch as the rates of oxidation depend Oll the the anomeric configuration of pyranoses can b ~ fr ee energy of the reactants in the ground state as ass i~ n e d on the basis that the more dextrorotatory well as in the transition state, it is of interes t to (or less levorotatory) anomer of an alpha-beta pair c om~al' e the free. ~n ergies . of the aldoses in aqueous has the D-g!ycero config.uration at Cl. When only s olutlOl~ . The drf:l erence .m free energy for a pair of one crystallme anomer IS known and the mutarota anomen c aldoses In solutIOn can be es timated from tion r eaction is complex, the configuration cannot be the proportions 01' the anomers at equilibrium. The assIgned from the optical rotation ' this is the case rates of oxidation clearly show, [01' aldo es in the with crystalline D-ribose, D-allo s~, D-altrose, and gluco series, that the anomers having equatorial D-gl:z;cero-D-ido-heptose . In re-examining the classi C ~-hydroxyl groups preponderate in the equilibrium fi catIOn of these aldoses, advantage was taken of the nnx tu~· e. .Thus! t be !,ree energy of ,B-n -glucose in observation of H aworth and Hirst [32] that the th ~ oXldatlOn l?lxtm e lS less tha.n that of a-D-glucose. I anomer that preponderates in an equilibrium solution WIth aldoses U1 the manna sen es the anomer with of a pyranose has trans hydroxyl groups at C1 and the axial Cl-hydroxyl group preponderates at C2. This principle is applicable 1'01' tentative assign equilibrium, and a-D-mannose in solution must have j a. lower fr e~ - e n e rgy than ,B-D-mannose. Quantita I m ent of configuration, becau e (as shown in table 2) L all aldoses of known configw'ation conform thereto . tIve evaluatIOn of the fr ee energies for the various r" T~u s, if the alcloses are present largely in the pyra sugars in aqueous solutions would be desirable but HOld form, we can conclude that the preponderating it is not feasible at present for lack of precise knowl e ~g e concerning the composition of the equilibrium 7 '[,h e bromine-oxidation lneasurements for D-aUosc, D-altrose and D-idose were conducted by R. SchalIer and B . Y. Foley. ' mIxtures. 630866- 62-4 237 3 . Comparison of the Oxidation of Aldoses 4 . Summary With the Oxidation of Cyclohexanols Early studies of the oxidation of aldoses by the In comparison with the aldopyranoses, cyclo author revealed that the ring forms of the aldoses hexanol and its derivatives are oxidized slowly, and are converted to lactones without cleavage of the the relative rates of oxidation of the axial and equato ring, that free bromine is the oxidant, and that I ~ rial isomers by either chromic acid [33] or bromine there are wide differences in the rates of oxidation I [22] are reversed. According to Barker, Overend, for the different aldoses and for the a and i3 modifi and Rees [22], th e difference in the rates of oxidation cations of a single aldose. The hypothesis was for the axial and equatorial isomers of tert-butylcyclo advanced that the rates of reaction of the anomers hexanol is far less than the difference in rates for a depend on the angular position of the C1-hydroxyl and i3-D-glucose. The authors propose the mechanism ~roup wlth respe~t to the plane of the ring. So depicted in figure 4. Simultaneous removal of the far as known~ thls ,vas the first attempt in any two hydrogen fttoms in the transition state is postu field of chmTllstry to correlate reaction rate and lated to account for a lack of sensitivity of the rate conformation. Recent r eappraisal of the results to a change in acidity. by means of present concepts of reaction-rate theory The oxidation of cycloh exanols differs from that of and of ring conformation leads to the hypothesis aldoses in several inlportant respects. With aldoses that the differences in the rates of oxidation for but not with cyclohexanols, it is postulated by th~ the a and i3 anomers arise from variation in the author that resonance involving the ring-oxygen free energy involved in conversion of the reactants atom aids in the release of th e C1-hydrogen atom. in their ground states to transition states in which r This hypothesis may account for the more rapid the oxygen atom of the C1 -hydroxyl group lies in oxidation of aldoses. In the oxidation of cyclo the plane formed by the ring-oxygen atom C1 ( ~ hexanols, the C1-oxygen atom does not lie in the C2, and C5. Presumably, this conformation r~sult~ same plane as C1, C2, C5, and C6, because th e C1 - from stabilization by resonance. C6 bond (corresponding to the C1- 0 bond of the In the oxidation of derivatives of cyclohexanol, pyranose ring) has no double-bond character, and the oxygen atom, in the transition state, does not there is consequently no resonance stabilization of need to ]~ e in the Salne plane as C1, C2, C5, and C6 such a planar conformation. For this reason, the because, 111 the absence 01' a ring-oxygen atom there C1 -hydroxryl group of the cyclohexanols can fLpproach is no resonance stabilization of such a confol'll~ation. th e transition state from either the axial or the This hypothesis may account for the lower rates equatorial position without a change in the confor of oxidation of the cyclohexanols, for lack of wide mation of the molecule. This accounts for the lack differences in the rates of oxidation of the axial and of a large difference in the rates of reaction for the equatorial isomers, and for thc fact that the relative ftxial and equatorial isomers. Because of the intra rates of oxidation of t]IC isomers are reversed from molecular repulsive forces, a cyclohexanol having an those of the axial and equatorial anomel's of the axial hydroxyl group would have a higher free-energy aldopyranoses. in both ground and transition states than a sin1ilar one having an equatorial hydroxyl group, and The author expresses his appreciation to R . U. less energy would be required for the axial isomer to Lemieux of the University of Toronto, and to H. reach the transition state than for the equatorial Eyring of the University of Utah for highly valued isomer. However, the difference in the free energies counsel. He thanks R. S. Tipson and H. L. Frush in the transition states would be less than that in the for assistance in preparing the manuscript, and R . ground states because of modification of the repulsive Schaffer and B. Y. Foley for meaSUTement of the forces by brOlnine. Consequently, less energy would rates of oxidation for D-allose, D-altrose, and D-idose. be required for th e axial isomer to reach th e transition state than for the equatorial isomer, and, therefore 5. References the rate of reaction should be higher for the axial I' isomer, as it, in fact, is. The r elationship is opposite [1] H. S. Isbell and C. S. Hudson, BS J. Research 8, 327 to that for the anomeric aldoses. (1932) RP418. [2] H. S. Isbell, BS J. Research 8, 615 (1932) RP441. [3] H. S. Isbell and W. W. Pigman, J. Research NBS 18, 141 (1937) RP969. [4] H. S. Isbell and W. W. Pigman, BS J . Research 10, 337 (1933) RP534. [5] F. J . Bates and Associates, Polarimetry, Saccharimetry, and the Sugars, NBS Clrc. H O (1942). [6] H . S. Isbell, J. Research NBS 18, 505 (1937) RP990. [7] R. E. Reeves, J. Am. Chern. Soc. 71, 215, 1737, 2116 (1949) ; 72,1499 (1950). FIGU RE 4. Oxidation of axial and equatorial isomers of [8] R. E. Reeves, Advances in Carbohydrate Chern. 6, 107 tert-butulcyclohexanol according to Barker, Overend, and (1951) ; Ann. Rev. Biochem. 27, 15 (1958). Rees [22].) [9] O. Hassel and B. Ottar, Acta Chem. Scand.l, 929 (1947). 238 I ,~ [10] C. W. Beckett, K. S. Pitzer, and R. Spitzer, J. Am. [23] H. S. Isbell, Chern. & Ind. (London) 1961, 593. Cl1em. Soc. 69, 2488 (1947). [24] H. Eyring, Chem. Revs. 17, 65 (1935). [11] R. Bentley, Nature 176, 870 (1955). [25] N.-J. Chil, Conformational analysis of the equilibria and [12] H . H. Bunzel and A. P . Mathews, J. Am. Chem. Soc. 31, rates for the anomerization of acetylated aldo 464 (1909). pyranoses, Ph.D. Thesis (University of Ottawa, [13] B. Perlmutter-Hayman and A. P ersky, J. Am. Chern. Canada, Feb. 1959). Soc. 82, 276 (1960). [26] M. S. Newman, Steric Effects in Organic Chemistry, [14] 1. R. L. Barker, W. G. Overend, and C. W. Rees, Chelll. p. 152 (John Wiley & Sons, Inc., New York, 1956). & Ind. (London), 1960, 1297, 1298. [27] E. S. Gould, Mechanism and Structure in Organic [15] R. Bentley, J. Am. Chern. Soc. 79, 1720 (1957). Chemistry, p. 178 (H enry Holt and Company, New [16] F. Shafizadeh, Advances in Carbohydrate Chern. 13, York,1959). 9 (1958). [28] H . S. Isbell and R. S. Tipsoll, Science 130, 793 (1959); [17] N. N. Li chten and M. H. Saxe, J. Am. Chern. Soc. 77, J. Research NBS 64A, 171 (1960). 1875 (1955). [29] H. S. Isbell and H. L. Frush, J . Research NBS 24, 125 [18] F. Friedberg and L. Kaplan, Abstracts Papers Am. (1940) RP1274. Chem. Soc. 131, 86 0 (1957). [30] H. S. Isbell, J . Research NBS 57, 171 (1956) RP2707. [19] L. Kaplan, J. Am. Chem. Soc. 76, 4645 (1954); 80, [31] H. S. Isbell, J . Am. Chem. Soc. 56, 2789 (1934). 2639 (1958). [32] W. N. Haworth and E. L. Hirst, J. Chern. Soc. 1928, [20] H. S. Isbell, Abstracts Papers Am. Chem. Soc. 132, 1226. ".( 1 D (1957). [33] S. Winstein and N. J. Holness, J. Am. Chem. Soc. 77, [21] C. G. Swain, R. A. Wiles, and R. F. W. Bader, J. Am. 5562 (1955). Chem. Soc. 83, 1945 (1961). [22] 1. R. L. Barker, W. G. Overend, and C. W. Rees, Chem. & I nd. (London) 1961, 558. (Paper 66A3-157) "I 239