Modelling and optimization of continuousbiosorption of heavy metal ions from water

Abstract

<p>U disertaciji je ispitana mogucnost kori&scaron;cenja<br />izluženih rezanaca &scaron;ecerne repe za adsorpciju<br />jona bakra u koloni sa nepokretnim<br />slojem. Eksperimenti su izvedeni primenom<br />Boks-Benkenovog eksperimentalnog dizajna<br />na tri nivoa sa tri promenljive: koncentracija<br />napojnog rastvora (50-150 mgL^-1), masa<br />adsorbenta (8-12 g) i pH napojnog rastvora<br />(4,0-5,0). Dobijene prevojne krive su<br />fitovane uobicajenim empirijskim modelima,<br />Boart-Adamsovim i modelom doza-odziv.<br />Na osnovu uocene asimetrije prevojne<br />krive, predložen je novi matematicki model.<br />Novi model predlaže prevojnu krivu koja<br />se sastoji od dva dela, koji u zbiru daju<br />kumulativnu krivu koja dosledno pokriva<br />uocenu asimetriju. Uzimajuci u obzir najnižu<br />prosecnu sumu kvadrata odstupanja, SSer<br />(3,610^-3) i najvi&scaron;u prosecnu vrednost<br />koeficijenta determinacije, R² (0,9989),<br />novi model je pokazao najbolje fitovanje<br />eksperimentalnih podataka, u poreenju sa<br />uobicajeno kori&scaron;cenim modelima adsorpcije<br />u koloni. RSM i ANN modelovanje je dalje<br />kori&scaron;ceno za evaluaciju procesnih parametara<br />i optimizaciju procesa. Najznacajnija<br />nezavisno promenljiva velicina, koja ispoljava<br />negativan uticaj na odabrani odziv (kriticno<br />vreme), bila je koncentracija napojnog<br />rastvora, dok je masa adsorbenta ispoljila<br />slabiji i pozitivan uticaj na odziv. Optimizacijom procesa metodom ANN je<br />utvreno da se najveca vrednost kriticnog<br />vremena (341,4 min) postiže pri sledecim<br />uslovima: C0=50 mgL^-1, ma=12 g and pH<br />4,53, dok metoda RSM smatra pH vrednost<br />faktorom koji nema znacaja po odziv u<br />ispitivanom opsegu, i postiže maksimalno<br />kriticno vreme od 314,8 min, pri istim ostalim<br />uslovima kao i ANN.<br />Iskori&scaron;ceni adsorbent, zasicen jonima bakra,<br />je zatim spaljen i dobijeni pepeo je iskori&scaron;cen<br />kao punilo u proizvodnji gume. Proizvedena<br />su cetiri uzorka gume sa razlicitim sadržajem<br />pepela. Pracen je proces vulkanizacije<br />gume, odreena su mehanicka svojstva<br />gotovog proizvoda i izvreno je poreenje<br />dobijenih rezultata izmeu uzoraka sa i<br />bez dodatka pepela. Rezultati su pokazali<br />da dodatak pepela u strukturu gume ne<br />utice znacajno na njene mehanicke osobine.<br />Meutim, povecanje sadržaja pepela u gumi<br />dovodi do povecanja energije aktivacije<br />procesa vulkanizacije, &scaron;to se može smatrati<br />unapreenjem procesa proizvodnje gume u<br />slucaju proizvoda velikih dimenzija. Na<br />kraju, ispitano je izluživanje bakra iz gotovih<br />proizvoda od gume. Rezultti su pokazali da je<br />kolicina bakra izluženog u vodu veoma niska,<br />reda velicine 0,001%, &scaron;to se može smatrati<br />zanemarljivim.</p>

<p>The potential use of sugar beet shreds for<br />copper ions removal from aqueous solution<br />in a fixed-bed column was investigated.<br />Experiments were performed using<br />Box-Behnken experimental design on three<br />levels and three variables: concentration of<br />the inlet solution (50-150 mgL^-1), adsorbent<br />dosage (8-12 g) and pH of the inlet solution<br />(4.0-5.0). The obtained breakthrough curves<br />were fitted with two common empirical<br />models, Bohart-Adams and dose-response.<br />Observing the asymmetric shape of the<br />breakthrough curves, the new mathematical<br />model was proposed. The new model<br />proposes the breakthrough curve composed<br />of two parts, sum of which gives the<br />asymmetrical S-shaped curve, accurately<br />matching experimental data. Regarding the<br />lowest average SSer (3.610^-3) and highest<br />average R² (0.9989), new model exhibited<br />the best fit comparing to the commonly used<br />models. RSM and ANN modelling were<br />employed for process variables evaluation and<br />optimization. The most influential parameter<br />exhibiting negative influence on target<br />response (critical time) was concentration of<br />the inlet solution, while the adsorbent dosage<br />exhibited positive influence. Optimization procedure revealed that the<br />highest critical time (341.4 min) was achieved<br />at following conditions: C0=50 mgL^-1,<br />ma=12 g and pH 4.53 by ANN, while<br />RSM considered pH as insignificant factor<br />and obtained 314.8 min as the highest<br />response. Discarded adsorbent, saturated<br />with copper ions, was then incinerated and<br />the obtained ash was used as a rubber filler.<br />Four rubber samples containing different<br />amounts of added ash were produced. The<br />vulcanization process was monitored and<br />mechanical properties was measured and<br />compared with the samples with no ash<br />added. The results revealed that addition of<br />ash has no significant effect on the rubber<br />mechanical properties. However, as the<br />ash content increases, the activation energy<br />of the vulcanization process increases too,<br />which can be considered as an improvement<br />in the cases of large-dimensioned rubber<br />products production. Finally, the leaching<br />of the copper from rubber products was<br />investigated. Results showed that very small<br />amount of copper, order of magnitude 0,001%,<br />were leached from the rubber into the water,<br />which can be considered as insignificant.</p>