外文原文-不同孔隙分布和表面氧化物质含量的活性炭对低浓度苯和甲苯的吸附特性.pdf

Behaviour of activated carbons with diff erent pore size distributions and surface oxygen groups for benzene and toluene adsorption at low concentrations M A Lillo Ro denas D Cazorla Amoro s A Linares Solano Departamento de Qu mica Inorga nica Universidad de Alicante Apartado 99 E 03080 Alicante Spain Received 28 October 2004 accepted 17 February 2005 Available online 21 March 2005 Abstract This paper deals with the study of the eff ect that the porosity and the surface chemistry of the activated carbons have on the adsorption of two VOC benzene and toluene at low concentration 200 ppmv In this sense activated carbons with very diff erent porosities and contents in oxygen surface groups have been tested Our results regarding the eff ect of the porosity show that the volume of narrow micropores size 0 7 nm seems to govern the adsorption of VOC at low concentration specially for benzene adsorption Regarding the surface chemistry AC with low content in oxygen surface groups have the best adsorption capacities Among the AC tested those prepared by chemical activation with hydroxides exhibit the higher adsorption capacities for VOC The adsorption capacities achieved are higher than those previously shown in the literature for these conditions specially for tol uene Adsorption capacities as high as 34 g benzene 100 g AC or 64 g toluene 100 g AC have been achieved 2005 Elsevier Ltd All rights reserved Keywords Activated carbon Adsorption Porosity Surface properties 1 Introduction VOC are pollutants present in gas and or liquid streams of many industrial applications They are very harmful for both human health and environment even at very low concentrations 1 6 It can be remarked that they are 1 agents that destroy the ozone stratospheric layer 2 precursors of photochemical oxidants 3 agents of the acid rain 4 elements of the climatic change 5 agents that aff ect the nervous system and 6 carcinogenic and mutagenic agents For all these reasons a considerable eff ort has been dedicated in the last years regarding their removal The most important methods used for treating VOC gaseous streams are 3 7 i absorption ii adsorption iii condensation iv thermal oxidation and v cata lytic oxidation All these methods provide good results when using the appropriate conditions of concentration fl ow and temperature Many of the gaseous streams involved in industrial processes contain VOC at very low concentrations Both legislation and human health determine that these streams must be treated to remove the organic com pounds Considering the low concentration of the VOC in the streams adsorption is one of the techniques which provides better results 2 3 7 8 The adsorption of organic pollutants can be carried out employing diff erent adsorbents Among them acti vated carbons are one of the best options due to their hydrophobic properties and their high surface area Carbon 43 2005 1758 1767 0008 6223 see front matter 2005 Elsevier Ltd All rights reserved doi 10 1016 j carbon 2005 02 023 Corresponding author Tel 34 965 909 350 fax 34 965 903 454 E mail address mlillo ua es M A Lillo Ro denas and because they are useful for adsorbing molecules with molecular weights between 45 and 130 2 3 6 In this work the adsorption of two of the most com mon VOC benzene and toluene at very diluted concen tration i e 200 ppmv concentration will be analysed The relative pressure for 200 ppmv concentration is 1 6 10 3and 5 4 10 3for benzene and toluene respectively at 298 K saturation pressures for benzene and toluene at 298 K are 126 9 and 37 9 mbar respec tively calculated using the corresponding Antoine coef fi cients 9 Many studies regard the abatement of benzene and toluene at low concentration 2 8 10 23 These studies try to fi nd adsorbents with high adsorption capacities and fast kinetics Table 1 summarises the results of the papers related to the adsorption of benzene and toluene on AC As it can be seen in this table a large variety of AC have been used granular pellets and activated carbon fi bres Unfortunately in most of these studies we cannot fi nd an appropriate characterization of the porosity of the activated carbons used Usually they only include BET surface area data of the activated carbons very few include data about the micropore volume calculated by nitrogen adsorption and none of them deals with the characterisation of the volume of narrow micropores It mustbeunderlinedthatthevolumeofnarrowmicropores size 0 7 nm in which primary adsorption takes place has shown to be very important in many gas adsorption applications such as SO2adsorption and methane stor age 24 26 This narrow microporosity can be easily measured by CO2adsorption at 273 K 27 30 Selectingtheconcentrationusedinthispaper 200 ppmv we can observe from Table 1 that the best activated carbon for adsorption of benzene adsorbs 29 g benzene 100 g AC 11 Regarding toluene there are less studies related and few data were obtained in conditions similar to our study At 151 ppmv concentration and 298 K Yun et al 22 achieved adsorption capacities of 30 g toluene 100 g AC No higher adsorption capacities have been obtained for toluene at 200 ppmv in the literature Thus the purpose of this work is to make a system atic study to determine what range of microporosity has higher infl uence in the VOC adsorption at low con centration For that purpose the porosities of the AC in this paper will be characterized in detail Also the eff ect of the content in surface oxygen groups in the VOC adsorption will be studied to optimise the acti vated carbon properties for the adsorption of VOC at low concentration The fi nal aim of this work is to know the activated carbon parameters that control the adsorption of VOC at low concentration It will allow us to improve the preparation step to develop the AC with the best characteristics for low concentra tion VOC adsorption 2 Experimental Ten activated carbons have been selected for this study They have been prepared by diff erent activation techniques and using diff erent precursors two commer cial activated carbons provided by Westvaco an AC prepared by physical activation with steam and chemi cally activated carbons prepared by either NaOH or KOH activation Further details of the chemical activa tion method employed can be obtained from previous works 31 32 Table 2 summarises the nomenclature of the samples employed including the carbon precursor used the method of preparation and the chemical used in the activation The characterisation of the porous texture of the acti vated carbons has been done using physical adsorption of N2and CO2at 77 and 273 K respectively in an Autosorb 6 apparatus Nitrogen adsorption has been used for determining the total volume of micropores pore size smaller than 2 nm whereas the adsorption of CO2at 273 K allows us to assess the narrowest micropores pore size smaller than 0 7 nm 27 30 Table 3 summarises the porosity characterisation of the activated carbons The mean pore size of the AC used applying the Dubinin equation to the CO2adsorp tion isotherms at 273 K 29 is in the range 0 6 0 9 nm The percentage of microporosity has been calculated by means of the following expression microporosity V DR N2 V N2 P P0 0 95 100 1 To study the eff ect of the surface oxygen groups on the VOC adsorption some of these AC have been ther mally treated in inert atmosphere An additional T is added in the nomenclature of the samples thermally treated The thermal treatment consists on heating the AC samples in a furnace with a helium fl ow of 100 ml min using a heating rate of 20 K min up to 1173 K and then cooling in helium This treatment removes most of the surface oxygen groups of the AC whereas the porosity remains nearly unchanged Thus the per formance of the pristine AC and of those whose content in surface oxygen groups has been reduced have been analysed The characterisation of surface oxygen groups present in AC has been done by TPD experiments Dur ing the thermal decomposition of the surface complexes CO2evolves from carboxylic groups and their deriva tives such as lactones and anhydrides while CO is mainly a decomposition product of quinones hydroqui nones and phenols 33 36 The experimental system for TPD experiments con sists of a furnace coupled to a mass spectrometer VG Quadrupoles A helium fl ow of 60 ml min is used and the heating has been done at 20 K min up to 1173 K M A Lillo Ro denas et al Carbon 43 2005 1758 17671759 The gases evolved as CO and CO2are analysed in the mass spectrometer The quantifi cation of CO and CO2 values of the untreated and treated samples are included in Table 4 The quantifi cation of these CO and CO2 Table 1 Adsorption data for benzene and toluene at low concentrations compiled from the literature ReferenceType of AC and weight g Nomenclature and BET S A m2 g V DR N2 cm3 g or m2 g Temperature K Concentration ppmv Benzene ads capacity g 100 g AC Toluene ads capacity g 100 g AC Foster et al 2 ACFACF 15 900 2985720 ACF 20 1610 2985719 ACF 25 2420 2985714 Cal et al 4 ACF 0 01 0 03 gACF 15 900 29820024 ACF 20 1610 29820026 ACF 25 2420 29820017 Cal et al 10 ACC 0 036 gACC 20 1330 1870 310a50034 Cal et al 11 ACC 0 01 0 03 gACC 15 730 1040 29820028 ACC 20 1330 1870 29820029 ACC 25 1860 2510 29820020 Dimotakis et al 5 ACCACC 1550 0 6129820027 ACC N 1738 0 5929820028 ACC Cl 1523 0 5429820024 ACC O 1105 0 3529820020 Dolidovich et al 6 ACC0 451 b2631 GAC0 3151718 Benkhedda et al 8 Pellets 3 g130029861838 5 Chiang et al 12 GAC 20 gAC 776 0 30430376820 AC Ba 668 0 26830376816 AC Mg 677 0 27630376817 Chiang et al 13 c Very diff erentP1 1470 m2 g 0 522986502326 types and originsP2 1120 m2 g 0 402986502024 P3 1320 m2 g 0 482986502229 P4 880 m2 g 0 312986501614 B1 1120 m2 g 0 402986501726 B2 840 m2 g 0 302986501517 B3 1030 m2 g 0 382986501620 C1 980 m2 g 0 352986502022 Chiang et al 14 GACCarbon A 1472 0 51529840013 Carbon B 1027 0 38029840013 Carbon C 975 0 35329840019 Chiang et al 15 GACAC 783 0 32030361721 AC O3 851 0 34430361719 Huang et al 16 ACFACF6 640 0ACF14 1460 0 32029820021 Noll et al 18 BAC120029821020 29835035 Ryu et al 20 GAC9930 4142981006 29830012 Shin et al 21 GAC 291100 12002974003040 2976003450 Yun et al 22 ACFKF 150029820017 29815130 29834437 Yun et al 23 GAC1100 120030320011 6730 a The experiments seem to be done at this temperature b It does not give details about concentration c The micropore volumes have been calculated by Ar adsorption 77 K using the Horvath Kawazoe method 1760M A Lillo Ro denas et al Carbon 43 2005 1758 1767 groups was done using two calibration cylinders with a concentration of either 10 of CO or CO2in helium Thus the data obtained were corrected for fragmenta tion and mass spectrometer sensitivity The validity of the results obtained using the 10 concentration CO and CO2calibration cylinders was confronted by several calcium oxalate monohydrate calibrations done in the system TG coupled to the mass spectrometer taking into account the CO disproportionation The system used to asses the adsorption of benzene or toluene consists of a BTRS reactor fi xed bed reactor with an internal diameter of 6 mm coupled to a mass spectrometer Balzers Thermocube Most of the AC used are in form of powder except of samples A and C which are granular mean particle diameter of 1 3 and 1 5 mm respectively and B which is in form of pel lets mean particle diameter of 2 2 mm respectively The bed heights are in the range 6 10 mm The samples have been outgassed before the adsorp tion experiments using a helium fl ow of 60 ml min and a temperature of 523 K which has been maintained for 4 h The adsorption experiments in this system have been carried out at 298 1 K The activated carbon bed con tains between 0 07 and 0 11 g of AC The amount of AC has been selected for each sample as the adequate for achieving sharp breakthrough curves The fl ow used during the experiments has been 90 cm3 min Thus the superfi cial velocity calculated as the ratio fl ow section is 318 cm min The gases for the adsorption tests contain 200 ppmv of VOC in helium from cylinders containing these mixtures supplied by Carburos Meta licos S A Thesegaseswereusedtocalibratethemass spectrometer By using this technique the breakthrough curves which represent the concentration of VOC in the outlet gas versus time are obtained From numerical integra tion of these curves the adsorption capacity at 200 ppmv concentration and 298 K has been obtained These values are presented in Table 5 This table also in cludes the ratio between the volume of hydrocarbon ad sorbedintheporesandthevolumeofnarrow micropores calculated by CO2adsorption 273 K 3 Results and discussion 3 1 Samples characterisation Fig 1 presents the nitrogen adsorption isotherms of the samples used in this study From Fig 1 and Table 3 it can be observed that the commercial activated car bons samples A and B have a wide pore size distribu tion They present a wide knee and the slope at relative pressures higher than 0 2 indicates the presence Table 2 Some data about the preparation of the AC samples SamplePrecursorActivation methodActivating agent AWoodCommercialH3PO4 BWoodCommercialH3PO4 CBituminousPhysical activationSteam DAnthraciteChemical activationNaOH EAnthraciteChemical activationNaOH FAnthraciteChemical activationNaOH GAnthraciteChemical activationNaOH HAnthraciteChemical activationKOH ISubbituminousChemical activationKOH JSubbituminousChemical activationNaOH Table 3 Porosity characterisation of the samples studied SampleBET m2 g V DR N2 cm3 g V DR CO2 cm3 g Microporosity A17570 670 3662 B12970 540 3065 C8830 350 2677 D6560 260 2584 E9320 390 4092 F15940 670 6799 G18720 800 7397 H27460 970 7783 I21230 930 9299 J24780 880 6281 Table 4 Characterisation of the content in surface oxygen groups SampleCO lmol g AC CO2 lmol g AC SampleCO lmol g AC CO2 lmol g AC A1765471AT7120 B2012509 C724759 D1193844DT8641 E1018698 F946312FT15719 G1126290GT6320 H1525297HT3719 I2853718 J1187551 M A Lillo Ro denas et al Carbon 43 2005 1758 17671761 of mesoporosity Sample C prepared by activation with steam has a lower percentage of micro and mesoporos ity The samples prepared by chemical activation from an anthracite i e samples D to H are essentially micro porous These samples present a type I isotherm corre sponding to essentially microporous activated carbons However the isotherms of the activated carbons D to H show that the higher the porosity development the broader the knee of the isotherm due to the widening of the micropore distribution Samples I and J have been prepared by chemical activation from a subbituminous coal by impregnation with KOH for sample I 37 and physical mixing of NaOH in sample J 38 Sample I presents a very narrow microporosity distribution opposite to that of sample J The comparison between micropore volumes calcu lated by nitrogen and carbon dioxide adsorption see Table 3 confi rms the information provided by the shapes of the nitrogen adsorption isotherms Thus the micropore volumes from CO2and N2adsorption are quite diff erent in samples A to C whereas they are very close in samples D E and I corresponding to a narrow microporosity distribution mean pore size close to 0 7 nm 27 Samples G and H present higher micropore volumes calculated by both adsorbates but the micro pore distribution is wider than in samples D E and I Sample J also prepared by chemical activation shows a wide micropore size distribution Table 4 includes the quantifi cation of the surface oxy gen groups of the samples used This table shows that the method used to reduce the oxygen content of these samples thermal treatment is eff ective because the con tent in oxygen groups of those thermally treated samples those with T in the nomenclature is much lower than for the pristine materials Tables 3 and 4 show that we have selected a suitable wide range of AC having diff erent porosity distributions and surface oxygen groups content for VOC removal at low concentration The chemically activated carbons generally present in comparison with the others higher percentage of microporosity 3 2 Eff ect of the activated carbon porosity on the VOC adsorption Fig 2 is an example of the breakthrough curves obtained for benzene and toluene adsorption It includes the adsorption data using around 0 11 g of the AC A Table 5 Quantifi cation of the benzene or toluene adsorption at 200 ppmv concentration on pristine AC Panel A and thermally treated AC Panel B and ratio of the narrow micropore volume occupied by the hydrocarbon SampleBenzene adsorption g 100 g AC Benzene adsorption cm3liq g AC Benzene adsorption cm3liq g V DR CO2 cm3 g Toluene adsorption g 100 g AC Toluene adsorption cm3liq g AC Toluene adsorption cm3liq g V DR CO2 cm3 g Panel A A150 170 48310 361 00 B120 140 46250 290 96 C120 140 53190 220 84 D110 130 50170 200 79 E150 170 43250 290 72 F220 250 38 G230 260 36380 440 60 H270 310 40500 580 75 I340 390 42560 650 70 J290 330 54560 651 04 Panel B AT200 230 64 DT150 170 69270 311 25 FT290 330 50 GT290 330 45560 650 89 HT320 370 48640 740 96 0 5 10 15 20 25 30 35 40 0 00 20 40 60 81 0 P Po n mmol g ABCDEFGHIJ Fig 1 Nitrogen adsorption isotherms 77 K of the samples used in this study 1762M A Lillo Ro denas et al Carbon 43 2005 1758 1767 Fig 2 shows that the adsorption capacity at 298 K is higher for toluene than for benzene This is an agree ment with the higher relative pressure for toluene in rela tion to benzene 5 4 10 3and 1 6 10 3 respectively In both cases a constant 200 ppmv concentration was used Table 5 Panel A collects the adsorption capacities of the AC at 200 ppmv concentration and 298 K for ben zene and toluene for the pristine AC This table shows that the adsorption capacity values for the two studied VOC are very diff erent depending on the AC used Thus the benzene adsorption capacities vary largely from 11 to 34 g benzene 100 g AC while in the case of toluene the adsorption capacities go from 17 to 56 g toluene 100 g AC If we com