Stable Sulfur Isotope Fractionation by Anion Exchange Chromatography . Production of Compounds Enriched in 34 S

A separação do isótopo S foi obtida pela reação de troca isotópica en tre solução de ácido sulfuroso e ânions bissulfito adsorvidos em resinas aniônicas tipo amônio quaternário (Dowex 1X8 e Dowex 2X8, malha 100-200) acondicionadas em seis colunas de resina de troca aniônica. Cada coluna de resina possuia 130 cm de comprimento com diâmetro de 2,2 cm. As colunas foram conectadas em série du rante o deslocamento da banda de bissulfito. Para a experimentação, a banda de bissulfito fixada na resina aniônica, inicialmente na forma OH, foi eluída subsequentemente com 0,2, 0,3, 0,4 ou 0,6 mol L HCl em solução. Ácido clorídrico em solução foi introduzido juntamente com nitrogênio sob pressão de 245 kPa, para impedir du rante o deslocamento, a liberação de gases e a oxidação da banda de bissulfito. Os experimentos utilizando a resina Dowex 1X8 mostraram que os melhores resultados em termos de máximo enriquecimento (18,66 % em átomos de S) e da massa to tal acumulada de S (161,7 mg, excesso) foram obtidos eluindo a banda de bisssulfito por 50 m com HCl 0,2 e 0,3 mol L, respectivamente. A substituição da parte empobrecida da banda com bissulfito nat u ral foi efetuada a cada deslocamento de 10 m.

ab sorp tion, carboxylase ac tiv ity, ribulose-2P, and the phosphorylation re ac tion, and it is es sen tial in the N fix ation pro cess in le gumes 1 .
More stud ies are needed to more pre cisely de fine the im por tance of sul fur in plant nu tri tion, since lit tle at ten tion has been given to this mat ter up to now.Al though re ports about sul fur de fi ciency in cul ti vated plants around the world, and es pe cially in Brazil are com mon, few stud ies on this nu tri ent have been car ried out .There is an ur gent need for re search in this field 2 .
The iso to pic tech nique has proven to be very valu able for elu ci dat ing as pects re lated to sul fur as a plant nu tri ent, and also for eval u at ing its uti li za tion by cul ti vated plants.The ra dio iso tope 35 S has been used as a tracer in stud ies of the soil-plant sys tem, due to its avail abil ity, rel a tively low cost, and non-sophisticated an a lyt i cal tech nique. 35S has a de cay of β -0.167 Mev and a half-life of 88 days 3 .Currently, there is a ten dency, at the in ter na tional level, to encour age the use of non-radioactive tech niques in iso to pic tracer re search, es pe cially in field ex per i ments.
The use of 34 S as a sulfur tracer al lows for the devel op ment of ex per i ments with out time re straints.Be cause the ex per i men tal ma te rial is not ex posed to ra di a tion, special safety pre cau tions are not needed.Hence, the use of the tracer tech nique with the 34 S sta ble iso tope is an al ter na tive for eval u a tions of sul fur dy nam ics in the soil-plant sys tem.Re cently, lab o ra to ries in the USA have pro duced compounds enriched with the 34 S stable isotope.Because of this and also due to prog ress in the an a lyt i cal tech nique us ing mass spec trom e try, the first stud ies on the 34 S sta ble iso tope us ing marked com pounds have been devel oped 4,5 .
Al though the stud ies us ing com pounds en riched with 34 S are re cent, the re sults ob tained show tre men dous po tential and are a good al ter na tive in stud ies us ing this im portant nu tri ent for plants, es pe cially in long term pro jects.
It is fun da men tal when us ing the 34 S tracer in re search pro jects that the method of iso to pic sep a ra tion and the produc tion of la beled com pounds in the iso tope be dom i nated.It is fea si ble to sup ply Bra zil ian re search ers in soil, en viron men tal and bi o log i cal sci ences in gen eral with this prod uct at a rea son able price, one which is lower than the for eign prod uct, and sim i lar to that of the 35 S ra dio iso tope.
The goal of this work is to study the method of 34 S enrich ment by an ion ex change chro ma tog ra phy in resin columns.There is lit tle in for ma tion in the lit er a ture on the tech ni cal and op er a tional de tails of the resin col umn system for the en rich ment of 34 S, nor much re lated to the simul ta neous sep a ra tion of the other sta ble sul fur iso topes and iso to pic anal y sis.Hence, dif fer ent ex per i men tal con ditions have been eval u ated (e.g.resin type, eluent, chem i cal con cen tra tion, and band dis place ment) in or der to ob tain 34 S en rich ments up to of 20 atom %.Studies have also been car ried out to ob tain si mul ta neous sep a ra tion of the other less abun dant sta ble sul fur iso topes ( 33 S and 36 S).Af ter obtain ing 34 SO 2(aq) in the iso to pic en rich ment col umns, it is pos si ble to pre pare other in or ganic com pounds en riched with the iso tope (NaH 34 SO 3 , Na 2 34 SO 3, Na 2 34 SO 4 , (NH 4 ) 2 34 SO 4 ).

Col umn sys tem and ap pa ra tus
The enrichment system consisted of six columns of acrylic, 130 cm in length and having a 2.2 cm internal diameter, filled with 100-200 mesh an ionic resin (Dowex 1X8 or Dowex 2X8), up to a height of 115 cm, in the form of Cl -.Each column had in lower section an acrylic cy lin dri cal ap pa ra tus con nected to the tube by a screw.On the base of this acrylic part two stain less steel screens, 50-and 300-mesh, and one 300-mesh pro pyl ene screen were at tached in se quence.Above them, a neo prene O-ring, made of C 3002-7B, was placed to en sure seal ing.In su pe rior sec tion of the col umn, where the chem i cal so lutions were poured, a PVC pis ton with two neo prene rings was adapted.When moved up wards or down wards this PVC pis ton main tained the level of the so lu tion in the column con stant, at a few mm above the sur face of the resin bed.The col umns were con nected in se ries with the aid of 1 mm in ter nal-diameter poly vi nyl chlo ride tub ing.
The con tain ers with so dium hy drox ide, so dium bisulfite, and deionized wa ter were placed 4 m above the col umns, and the con tainer with hy dro chlo ric acid eluent so lu tion was lev eled with the col umns.The flux of hy drochlo ric acid through the resin lay ers in the col umns was effected with the aid of a sys tem work ing un der ni tro gen pres sure, at ap prox i mately 245 kPa.The en tire sys tem can be ob served in Fig. 1.
For the iso to pic anal y sis, a mass spec trom e ter model AT LAS-MAT CH-4 was used, hav ing the fol low ing features: an a lyzer tube cur va ture ra dius of 120°; ion iza tion by elec tronic im pact; ad mis sion sys tem with mo lec u lar flux; sim ple ion col lec tor, Far a day-type cup; and scan ning analy sis sys tem.

Ex per i men tal de vel op ment
During the development of this study the isotopic separation of 34 S was in ves ti gated us ing two types of resins and four con cen tra tions of hy dro chlo ric acid (0.2, 0.3, 0.4, and 0.6 mol L -1 ) in five ex per i ments (I to V), as shown in Ta ble 1.
At the be gin ning of the ex per i ments the res ins were regen er ated to the OH -form with a so lu tion of 2 mol L -1 sodium hy drox ide, at a flow rate of 3.8 cm 3 min -1 , fol lowed by rins ing the resin lay ers in each col umn with deionized water, and fix ing a band of so dium bisulfite so lu tion in one of the col umns (4 L of so lu tion 0.15 mol L -1 col umn -1 ).This band was then eluted with an HCl so lu tion, at a rate of 3.8 cm 3 min -1 , hav ing a sec ond col umn con nected in se ries.While the band dis place ment oc curred, it was con fined between two well-defined lim its.These lim its were made at front of the band by the equi lib rium con stants of the ionic ex change re ac tions be tween HSO3 -in so lu tion and ad -sorbed hydroxyl an ions, and at the rear of the band by the re ac tion be tween Cl -in so lu tion and ad sorbed HSO 3 -.Dur ing dis place ment (elu tion), sam ples were col lected along the band in or der to eval u ate the 34 S con cen tra tions.The sam pling pro ce dure con sisted of col lect ing small fractions of sul fu rous acid at dif fer ent po si tions along the band length, in one dis place ment col umn.The sam ples taken had equiv a lent amounts or twice as much so dium hy droxide as de sired, mak ing it pos si ble to ob tain so dium bisulfite or so dium sul fite.For the iso to pic anal y sis by mass spectrom e try, the sam ples were pre pared by trans form ing the so dium bisulfite or so dium sul fite into sul fur di ox ide un der vac uum 6 .The re gion of the band with a nor mal iso to pic con cen tra tion of 34 S (nat u ral con cen tra tion thresh old) was main tained at a min i mum of 20% of the band.This con trol was done by the ad di tion, when nec es sary, of a so lu tion of 0.15 mol L -1 so dium bisulfite with 34 S nat u ral con cen tration, fol low ing the dis card of the same amount of band solu tion de pleted of the iso tope (with a con cen tra tion of less than 4.2 atom % 34 S).
In this pro cess, the equi lib rium be tween the sul fu rous acid so lu tion and HSO 3 -ions ad sorbed to the an ion exchange resin (HSO 3 --H 2 SO 3 sys tem), for the frac tion ation of 33 S and 32 S, 34 S and 32 S, or 36 S and 32 S by ionic ex change in resin col umns, is rep re sented by Eq. 1 7 .
where * S rep re sents the iso topes 33 S , 34 S, or 36 S, and (r) and (s) stand for resin and so lu tion, re spec tively.

Re sults and Dis cus sion
A sum mary of the main re sults is given in Ta ble 2, show ing that Ex per i ments II and III had the best re sults.The data shown in Ta ble 2 in di cate that the best re sults for max i mum en rich ment (18.66 atm% 34 S) and to tal 34 S ac cumu lated mass (161.7 mg, ex cess) were ob tained by 50 m of elu tion of the bisulfite band with 0.2 and 0.3 mol L -1 acid, re spec tively.The ex per i men tal data proved that in ex per iments II and III the en riched band was es tab lished in the last 20 and 35 cm, re spec tively.These re sults, which can be ob served in Fig. 2 and Ta ble 2, show that a higher ac cu mula tion of 34 S mass oc curs in Ex per i ment III (161.7 mg), when com pared to Ex per i ment II.One of the most rel e vant pa ram e ters is the ac cu mu la tion rate (ex cess) of the iso tope 34 S (mg h -1 ).The data in Ta ble 2 in di cate that this rate is approx i mately 20% higher in Ex per i ment III (HCl 0.3 mol L -1 ) when com pared to Ex per i ment II (HCl 0.2 mol L -1 ).
Un der the con di tions used for ex per i ments II and III, the en rich ment curves for 34 S (Fig. 2) were ob tained af ter 50 m of bisulfite band elu tion.It is pos si ble to ob tain from Ex per i ments II and III, 6.46 and 7.37 mmol sul fu rous acid re spec tively as a col umn ef flu ent, which cor re sponds to 1 Ta ble 1. Ex per i men tal con di tions for the en rich ment of 34 S in columns of an ion resin (pres sure 245 KPa).

Ex per i ment (num ber)
Resin HCl Con cen tra tion (mol L -1 ) Fig ure 1.The col umn sys tem for 34 S en rich ment. 1.So lu tion con tainers; 2. Tap con nec tion be tween the con tainer and the col umn; 3a.Tap con nec tion with waste sys tem or counterflux tap con nec tion with deionized wa ter; 3b.Tap con nec tion be tween col umns; 4. Resin column; 5. Pres sure pis ton; 6. Con nec tion tubes; 7. Pre-purified ni tro gen tank (high pres sure); 8. Hydrochloridic acid con tainer.
lin ear cen ti me ter of the band (at the rear of the band).Ta ble 2 also shows that Ex per i ment V had the least fa vor able results in terms of 34 S iso tope en rich ment.An a lyzing Ta ble 2, it can also be ob served that Ex per iment I pre sented re sults com pa ra ble to Ex per i ments II and III.How ever, the use of the resin Dowex 2X8 in this ex per iment was in ad e quate un der these ex per i men tal con di tions, es pe cially due to dif fi cul ties in vi su al iz ing the band lim its, caus ing iso to pic mix tures.This made it im pos si ble to proceed with the ex per i ment with the dis tance of the band elution su pe rior to 30 m.In Ex per i ments IV and V the dis tance of the band elu tion was achieved only up to 10 m (Ta ble 2), due to the con cen tra tions of hy dro chlo ric and sul fu rous acid, which con stantly caused cor ro sion of the stain less steel at the col umn bases.This caused many dif fi cul ties for the en rich ment pro cess in clud ing resin loss, dis con tinuance of the dis place ment band, iso to pic mix tures, and chance of en rich ment band loss, among oth ers.
In Ex per i ment II, the pro cess had con ti nu ity (rou tine) af ter 50 m of band elu tion.It was pos si ble to ob tain for every 14 m from the last cen ti me ter of band, an av er age of 6.79 mmol of sul fu rous acid en riched at 16.76 at oms % in 34 S.This amount of sul fu rous acid can be con verted to 706, 855, or 964 mg of so dium bisulfite, so dium sul fite, or sodium sul fate, re spec tively.So dium sul fate can be ob tained by the ox i da tion of the so lu tion of so dium sul fite with ox ygen.
Af ter 50 me ters of band elu tion the sul fu rous acid so lution was sam pled at about 12 cm from the rear limit.This had the ob jec tive of eval u at ing the pres ence of sul fate which could orig i nate from sul fite ox i da tion.The con centra tions of sul fate and sul fite in so lu tion were 0.43 and 35.88 mg cm -3 , re spec tively.These re sults in di cate low oxi da tion, since about 1.2% of the ini tial sul fite was ox i dized.There is also an in di ca tion that the sys tem un der ni tro gen pres sure in the res er voir of the HCl eluent so lu tion was ef ficient in terms of pre vent ing the dis solved air in so lu tion (es pe cially O2) from fa vor ing the ox i da tion re ac tion.
In Ex per i ment II the iso to pic con cen tra tion of the sul fur iso topes was also de ter mined.The ra tio of 33 S, 34 S and 36 S iso tope con cen tra tions at 1 cen ti me ter from the back of the band limit to the nat u ral abun dance val ues were 2.0, 4.0, and 5.3, re spec tively.These val ues al low for the in fer ence that the frac tion ation val ues (α) for the iso topes 33 S, 34 S, and 36 S in equi lib rium with the most abun dant sul fur isotope ( 3 2 S) should obey the fol low ing or der: α 36 S > α 34 S > α 33 S.

Con clu sions
Based on the re sults ob tained it can be con cluded that: a) the use of Dowex 2X8 resin for 34 S iso to pic en rich ment is not ad e quate un der the ex per i men tal con di tions used, due to the prob lem of vi su al iza tion of the rear of the band; b) it is pos si ble to ob tain sul fu rous acid with 34 S en rich ment of 20 atom % in an ionic resin col umns of Dowex 1X8, by dis plac ing the bisulfite band with HCl 0.2 or 0,3 mol L -1 eluent so lu tion; c) by us ing the pro posed sys tem of resin col umns, af ter 50 m of elu tion, 6.46 mmol (17.33 atom % of 34 S) and 7.37 mmol (16.93 atom % of 34 S) of sul fu rous acid are ob tained in the last lin ear cen ti me ter of the band  with the use of 0.2 and 0.3 mol L -1 HCl, re spec tively.The av er age rate of ex cess ac cu mu la tion of 34 S is 0.16 and 0.19 mg per hour of elu tion with 0.2 and 0.3 mol L -1 acid, re spectively; d) in rou tine, af ter the first 50 m of band elu tion, under ex per i men tal con di tions, it is pos si ble to ob tain 6.79 mmol of sul fu rous acid for ev ery 14 m of band elu tion with ap prox i mately 17 atom % in 34 S, in the last lin ear cen ti meter at the rear of the band; e) the resin col umn sys tem, working at a pres sure of 245 kPa un der ni tro gen at mo sphere was able to con trol the re lease of gases and es pe cially sul fite oxi da tion.

Ac knowl edg ments
The au thors thank the Fundação de Amparo a Pesquisa do Estado de São Paulo (FAPESP) for the valu able fi nancial sup port for this re search (Pro cess: 90/4365-0).

Fig ure 2 .
Fig ure 2. Pro files of an ad sorbed bisulfite band (80 cm) af ter 50 m of elu tion.
Ta ble 2.Pa ram e ters re lated to 34 S En rich ment Ex per i ments by Ionic Ex change Chro ma tog ra phy (ex per i ments I to V)