LYH-FYP presentation slide on Pebax for CO2 capture -__ 20Dec12 (1)
- 格式:pptx
- 大小:4.68 MB
- 文档页数:31
Carbon Dioxide Capture from Coal-Fired Power Plants in ChinaSummary Report for NZEC Work Package 3September, 2009Prepared by:Doosan Babcock Energy Limited Jacqueline GIBSON and Diemo SCHALLEHNTsinghua University,Department of Chemical EngineeringZHENG Que and CHEN JianTsinghua University,Department of Thermal Engineering, KeyLaboratory for Thermal Science and PowerEngineering of Ministry of EducationWANG ShujuanGreengen Co. Ltd, Beijing CAO JiangImperial College London,Mechanical Engineering Department,Energy Technology for SustainableDevelopment GroupJon GIBBINS and Mathieu LUCQUIAUDNorth China Electric Power University, KeyLab of Condition Monitoring and Controlfor Power Plant Equipment of Ministry ofEducationYANG Yongping, XU Gang and DUAN LiqiangTsinghua University, BP Clean EnergyResearch & Education CentreXU ZhaofengWuhan University HU Jicai and LI JiZhejiang University, State Key Laboratory of Clean Energy Utilization FANG Mengxiang, YAN Shuiping and LUO Zhongyang50100150200250300350400Advanced supercrit 800MW Oxyfiring Post ‐com MEA Post ‐com aqueous ammonia Pre ‐com IGCC Pre ‐com polygen Existing 600MW supercrit Post ‐com MEA Post ‐com aqueous ammonia Existing 300MW subcrit Post ‐com MEA Post ‐com aqueous ammonia2004006008001000Advanced supercrit800MWOxyfiring Post ‐com MEA Post ‐com aqueous ammonia Pre ‐com IGCC Pre ‐com polygen Existing 600MW supercrit Post ‐com MEA Post ‐com aqueous ammonia Existing 300MW subcrit Post ‐com MEA Post ‐comaqueous ammonia 05000100001500020000Advanced supercrit 800MW Oxyfiring Post ‐com MEA Post ‐com aqueous ammonia Pre ‐com IGCC Pre ‐com polygen Existing 600MW supercrit Post ‐com MEA Post ‐com aqueous ammonia Existing 300MW subcrit Post ‐com MEA Post ‐com aqueous ammonia1020304050Advanced supercrit 800MW Oxyfiring Post ‐com MEA Post ‐com aqueous ammonia Pre ‐com IGCC Pre ‐com polygen Existing 600MW supercrit Post ‐com MEA Post ‐com aqueous ammonia Existing 300MW subcrit Post ‐com MEA Post ‐com aqueous ammoniaN e t E f f i c i e n c y (% L H V )C o s t o f E l e c t r i c i t y(R M B /M W h )Power plant net efficiencies with and without CO 2captureCost of electricity wi th and with out CO 2capture, 85% load factor, 10% discount rate, tax not includedCoal costs: 16 RM B/GJ / 32 RM B/GJ Capital expenditureCAPEX = Total Installed Cost + 10% contingency + 7% owner’s costs, financing costs and taxes not includedRelative to new build advanced supercriticalpulverised coal plantRelative to existing 600MW supercritical coa lplantRelative to existing 300MW sub ‐critical coalplantC A P E X (R M B /k W )C o s t o f A b a t e m e n t (R M B /t C O 2)Cost of abatementExecutive SummaryThis report draws on the capture plant case studies by NZEC Work Package 3 partners to present, in a unified and consistent form, the technical and economic performance of power generation with CO 2 capture in China. Complementary work on performance calculations by Chinese and UK partners has been combined where appropriate to give estimated final overall plant efficiencies and CO 2 emission levels. Estimated capital costs have been used to derive levelised costs of electricity and emission abatement costs on a consistent basis.Advanced new build capture technologies that have still to be demonstrated, oxyfiring, post ‐combustion with aqueous ammonia and pre ‐combustion capture on IGCC, are predicted to achieve similar power plant efficiencies of 35.6, 35.7 and 36.8% respectively. CAPEX values for these plants are approximately 9000 to 10000 RMB/kW net, with an estimated +/‐ 30% uncertainty. Levelised costs of electricity for these options are estimated to have a range of approximately 370‐410RMB/MWh for a coal cost of 16 RMB/GJ and approximately 40% higher for a coal cost of 32 RMB/GJ. Costs of abatement are calculated relative to the standard alternative plant that would be built, an advanced supercritical plant. Since the cost of abatement is based on the differences betweenrelatively larger numbers there is more variance in the results, from an estimated 140 RMB/tCO 2 for oxyfiring to 200 RMB/tCO 2 for IGCC+CCS. These values are, however, very sensitive to estimates for capture plant costs, particularly the additional capital costs, and should be regarded as preliminary. Post ‐combustion capture with an MEA ‐based solvent, an older technology, is predicted to have a generally less favourable performance. Polygeneration of electricity and methanol also appears to have a low efficiency and high capital costs, but the methanol production is not taken into account.Table of contents1. Background and supporting case study reports 11.1 Background and scope 11.2 Contributing reports 12. Power generation from new build pulverised coal plants with CCS(DB, DCE, DTE, IMP, NCEPU, ZJU) 2 2.1 Advanced supercritical base plant for new build case studies with and without capture 2 2.2 CO2 capture using oxyfiring (DB, ZJU) 5 2.3 Post-combustion CO2 capture (DCE, DTE, IMP, NCEPU, ZJU) 6 2.3.1 MEA post‐combustion capture system (DCE) 6 2.3.2 Aqueous ammonia post‐combustion capture system (DTE) 72.3.3 Thermal integration between the power plant steam cycle and post‐combustion capture units and calculation of overall plant efficiency and cost with capture (DCE, DTE, IMP, NCEPU, ZJU) 83. Post‐combustion capture from existing Chinese power plants (NCEPU) 11 3.1 Chinese 600MW supercritical unit 11 3.2 Chinese 300MW sub‐critical unit 113.3 Performance without and with CO2 capture 114. Electric power generation with pre‐combustion CO2 capture (GG, THCEC) 12 4.1 Background 12 4.2 GS1‐ 2×400 MW IGCC using TPRI gasifier with CCS (GG) 144.3 GS2‐ 1×400 MW polygeneration system with CCS (THCEC) 155. Summary of CO2 capture performance for different plant types 17Appendix A – Case study basis 21 Appendix B – Tabulated summary of technical and economic performance 23 of coal fired power plants with CO2 capture under Chinese conditionsAppendix C ‐ CO2 capture using membrane gas absorption technology 24 Appendix D ‐ Levelised cost of transporting CO2 as a function of mass flow rate 261. Background and supporting case study reports1.1 Background and scopeThis report draws on the following capture plant case studies by NZEC Work Package 3 partners. Salient features of these studies are presented in this summary report but for detailed information readers should refer to the original source(s). These are indicated in the text by the appropriate partner abbreviations shown in brackets after section headings.The principal purpose of this report is to present in a unified and consistent form the technical and economic performance of the main capture case studies. Complementary work on performance calculations by Chinese and UK partners has been combined where appropriate to give final overall plant efficiencies and CO2 emission levels. Estimated capital costs have been used to derive levelised costs of electricity and emission abatement costs on a consistent basis. These are presented in graphical and tabulated forms.1.2 Contributing NZEC Work Package 3 ReportsPartnerAbbreviationFull Name Report TitleDB Doosan Babcock Energy Limited J R Gibson and D Schallehn, China – UK NearZero Emissions Coal (NZEC) – OxyfiringOptionsDCE Tsinghua University, Department ofChemical Engineering Que Zheng and Jian Chen, Cost estimation for CO2 Capture with a MEA absorption processDTE Tsinghua University, Department ofThermal Engineering, Key Laboratoryfor Thermal Science and PowerEngineering of Ministry of Education Wang Shujuan, Carbon Dioxide Capture Using MDEA and Ammonia SolutionsGG Greengen Co., Ltd, Beijing Case Study for IGCC Power Plant In China(with CCS)IMP Imperial College London, MechanicalEngineering Department, EnergyTechnology for SustainableDevelopment Group Mathieu Lucquiaud, Steam cycle calculations for capture‐ready steam cycle and retrofits with MEA, ammonia and MDEA solventsNCEPU North China Electric Power University,Key Lab of Condition Monitoring andControl for Power Plant Equipment ofMinistry of Education Yongping Yang, Gang Xu and Liqiang Duan, Carbon Dioxide Capture from Existing Coal‐Fired Power Plant in ChinaTHCEC Tsinghua University, BP Clean EnergyResearch & Education Centre XU Zhaofeng, Polygeneration using two‐stage slurry gasifier with CCSWHU Wuhan University Hu Jicai and Li Ji, CO2 TransportZJU Zhejiang University, State KeyLaboratory of Clean Energy UtilizationMengxiang Fang, Shuiping Yan and Zhongyang Luo, Carbon Dioxide Capture from a New‐built Ultra Supercritical PC Power PlantUsing Hollow Fibre Membrane Contactors 2. Power generation from new build pulverised coal plants with CCS (DB, DCE, DTE, IMP, NCEPU, ZJU)2.1 Advanced supercritical base plant for new build case studies with and without capture (DB)New‐build, pulverised coal (PC) capture studies are based on Advanced Supercritical (ASC) Boiler Turbine (BT) technology. Advanced supercritical boilers are already operating in China, with an estimated Total Installed Cost of 5000 RMB/kW based on Chinese partners’ experience. Present state‐of‐the‐art advanced supercritical boiler/turbine technology is used, with steam turbine inlet conditions of 280 bar / final superheat temperature of 600°C / final reheat temperature of 610°C, giving an efficiency for the site and coals specified of 43.9% LHV net. Emissions controls consist of DeNOx, particulate removal and DeSOx plant. A unit with an output of 824.3 MW net is used as the base case for new build pulverised coal options with capture. This size of unit is typical for new advanced supercritical boilers planned in Europe and also gives PC capture options with net MW outputs that are close to the limiting size for IGCC+CCS options. Larger PC base units (e.g. 1000‐1200 MW) are feasible and are likely to be built in China, but the output of the IGCC+CCS unit is limited by the size of the largest gas turbines available.A block flow diagram for the base case ASC PC plant is shown in Figure 2‐1. A brief description of each of the units identified in the block flow diagram is given below, with a sub‐section adding how, if at all, it might need to be modified for use in a PC‐based capture plant. The base case unit burns coal in air and so also do the post‐combustion capture cases; the base plant design remains essentially unmodified for these apart from the Turbine Island. For oxyfiring capture, the coal is burnt using oxygen instead of air, mixed with recycled combustion products at the burners to moderate flame temperatures. For oxyfiring capture certain aspects of some the units in the base plant design will have to be changed while others would be used unaltered, as indicated.Boiler: The advanced supercritical boiler is based on the state‐of‐the‐art Doosan Babcock Two‐Pass single reheat BENSON boiler with Posiflow™ Technology, Balanced Draught, and Gas Biasing for reheat steam temperature control.Coal and ash handling: A conventional system is employed, with the design of the furnace bottom ash and fly ash systems from the boiler and downstream particulate removal systems following modern conventional air fired power generation plant practice which aims to minimise tramp air ingress.Pulverised Fuel Milling Plant: Coal will be milled using conventional pulverisers. In this case, tubemills are selected to pulverise the coal to a suitable fineness (typically >80% passing through 75microns). If the coals are inclined towards sub‐bituminous coals then roller or ball mills could be substituted.Modifications for capture cases: In any pulverised fuel combustion plant with a direct firing system the primary gas supplied to the mill is required to both dry the as‐fired coal and to convey the pulverised coal from the mill to the burners. In the case of air firing, hot combustion air is supplied from the air/gas heater, and in the case of oxyfiring, flue gas is recycled from the FGD outlet and passes through a gas/gas heater. The ability of the primary gas to dry the fuel is driven by two factors, the overall temperature and heat capacity of the gas entering the mill, and its moisture content. In both oxyfiring and air firing the temperature of the primary gas is similar, although the volumetric heat capacity of recycled flue gas is greater than that of air due to the higher density of CO2 compared to nitrogen. When operating in air firing mode the moisture in the combustion air (i.e. its relative humidity) is low, making the drying process easier as the air has a high capacity to absorb evaporating moisture. For oxyfiring operation the raw recycled flue gas typically has a substantially higher moisture content which limits the drying capacity. This reduction in drying capacity can be mitigated by both increasing the mill outlet temperature (although there are practical limits to this) and by reducing the moisture content of the recycled flue gas.Turbine Island: State of the art advanced supercritical turbine technology will be utilized. Modifications for capture cases: This will be used without modification from the air‐firing design for oxyfiring, but with provision for steam extraction after the intermediate pressure (IP) cylinder for post‐combustion capture (see section 2.3.3 below). Rejected heat from the capture equipment is used to minimise the steam extraction for feed water heating for all capture options.DeNOx: New build plant is required to meet NOx emissions regulations (450mg/Nm3 @ 6%O2). Technologies based on primary NOx reduction measures, (low NOx burners and overfire air (OFA)), and secondary measures, (Selective Catalytic Reduction (SCR)), can be used.Modifications for capture cases: DeNOx plant is not a requirement within the oxyfiring boiler island as the NOx is captured in the downstream compression plant. However, DeNOx plant will need to be installed for post‐combustion capture cases with air firing.ESP: Particulate removal plant is essential to both meet dust emission level limits, as defined by the applicable environmental legislation, and to protect downstream flue gas fans and blowers from excessive erosion. Electrostatic Precipitators (ESP) are used in the NZEC base case.Modifications for capture cases: When operating in oxyfiring mode the particulate removal plant ensures that the FGR streams are relatively dust free. The proposed oxyfiring and post‐combustionplants both also utilise an ESP . Previous studies have shown that, as a direct result of the performance of the airheater / gas ‐gas heater module, the operating temperature of the electrostatic precipitator (ESP) is increased during oxyfiring operation to typically between 160°C and 200°C. Normally higher temperatures lead to less gas residence time in the ESP with a resulting loss in particulate collection efficiency. However, oxyfiring flue gas has a higher density than air ‐fired flue gas and this mitigates the loss in efficiency arising from the increased temperature. DeSOx: For air ‐firing an FGD is required to meet emissions legislation.Modifications for capture cases: Particular attention will be given to achieving a very high level of removal (<30 ppm ) for post ‐combustion systems to minimize solvent loss. The FGD plant also cools the flue gas by evaporative cooling. In oxyfiring mode, an FGD plant is used to treat the flue gas to ensure that the primary and secondary flue gas recycle streams lead to sulphur dioxide (SO 2) and sulphur trioxide (SO 3) concentrations in the furnace that are no worse than the equivalent experience for high sulphur coals with air firing 1 and residual SOx is removed by processing as the CO 2 is compressed.Air Firing OnlyAdditional to normal power plant supplyFigure 2-1 Advanced supercritical steam plant, also showing scope of equipment for use asthe base plant for air firing with post-combustion capture and with oxyfiring(flue gas recycle for oxyfiring only)1For very low sulphur coals, with a sulphur content of 0.2% w/w daf or lower, FGD is not expected to be required for oxyfiring operation. Coals with somewhat higher sulphur levels might also be satisfactory for oxyfiring operation without using an FGD but this would depend on the coal calorific value and hence firing rate and also on the details of the flue gas recycle approach adopted.2.2 CO2 capture using oxyfiring (DB, ZJU)Overall, the key features of oxyfiring vs air firing are:•Incorporation of an air separation unit for removal of N2 to supply a nearly pure O2 stream into the recycled flue gas for the combustion process.•Recirculation of flue gas back to the boiler (via an FGD for the NZEC coals) providing a transport medium for the pf and to maintain the radiative and convective heat transfercharacteristics of the furnace and boiler.•Incorporation of gas‐gas heaters instead of conventional air‐preheating arrangements. •Incorporation of a flue gas cooler and condenser to recover heat into the steam cycle condensate and feedwater preheating systems.•CO2 compression and inerts separation plant, incorporating additional sulphur removal This is considered a low risk approach for an operator needing to be able to maintain electrical output because the plant designs retain full air firing capability, minimising commercial and technical risk should the oxyfiring components be unavailable, by allowing continued generation if the oxyfiring plant is out of service. It is envisaged that an air firing capability will only be applied to the initial demonstration plants, however, as there are significant costs associated with retaining this on an oxyfiring plant.This configuration thus offers an evolutionary approach from well known and understood plant:•T he plant is designed with operational and practical experience of air‐firing. •Similarities in design and operation to conventional air fired, pulverised coal power plants for each of the major plant systems and components.•New plant systems (ASU, compressors) and components are well proven designs and commercially available at the required capacities.The oxyfiring case study by Doosan Babcock has shown that for an advanced supercritical power plant designed to capture CO2 the penalty in terms of cycle efficiency is 8.3 percentage points compared with an air fired case with no CO2 capture, giving a final efficiency of 35.6 % LHV net. This penalty has been mitigated by recovering heat within the system to offset the additional power consumption of the CO2 compression plant and air separation unit. An approximate estimate is that capital costs for such an oxyfiring plant will be 48% higher on a RMB/kW basis than for a conventional air‐fired plant (but based on UK experience, rather than direct Chinese market costing), giving a Total Installed Cost of 7390 RMB/kW for an assumed base plant TIC (based on estimates from Chinese partners) of 5000 RMB/kW. Unlike post‐combustion CO2 capture systems, oxyfiring plants do not need to make up any CO2 capture solvent losses, so this fairly significant contribution to running costs is avoided.2.3 Post‐combustion CO2 capture (DCE, DTE, IMP, NCEPU, ZJU)Post‐combustion CO2 capture based on the use of monethanolamine (MEA) solutions was studied in conventional packed absorber columns (DCE) and in membrane contactors (ZJU – also see Appendix C). While the latter offer potential advantages for gas and solids handling, the selected membrane materials were limited to more dilute MEA solutions (20% w/w) than the packed columns (30% w/w). This meant that the performance of the membrane system was worse and therefore no results are reported in the detailed cost comparison; additional work is needed on this novel approach if it is to have the potential of becoming competitive with other options. Similarly, the use of methyldiethanolamine (MDEA) in a conventional packed column, examined by DTE, is also not reported in this summary since it did not appear to offer any significant advantage over MEA.The use of aqueous ammonia as a CO2 solvent was also studied by DTE. While still requiring further development work this approach appears to have the potential to offer significant advantages in terms of reduced energy penalty and capital costs. It has therefore been included in the comparative analysis as a possible example of the benefits that improved post‐combustion capture, using this or other ‘second generation’ solvents, may be able to deliver, compared with plants using designs based on the ‘industry standard’ MEA solvent.2.3.1 MEA post‐combustion capture system (DCE)A schematic flow diagram for the MEA capture process is shown in Figure 2‐2.Figure 2‐2 Process flow diagram for MEA captureThe conventional MEA process is a temperature swing absorption process. The removal of CO2 from flue gas is achieved by contacting the feed flue gas in an amine absorber with an aqueous solution of an alkanolamine at a low temperature, where the carrier solution, an aqueous amine solution absorbs CO2 to form carbamate or bicarbamate ions and become a CO2‐rich solution. Meanwhile, CO2 is removed from the flue gas. In a solvent regenerator, the CO2‐rich solution then liberates the dissolved CO2 at an elevated temperature to reverse the absorption reaction and becomes a lean solution. A high purity CO2 stream is released from the solvent regenerator and is then compressed to 110 atmospheres. The regenerated CO2‐lean amine solution is then cooled and recycled to the amine absorber from the solvent regenerator for further CO2 removal. Both thermal energy and electrical energy, the latter principally for solvent pumping, flue gas blowers and CO2 compression, are required to run the MEA capture process. In this study a standard value of 150 kWh per tonne of CO2 captured is estimated to be required for the electrical energy input.2.3.2 Aqueous ammonia post‐combustion capture system (DTE)As shown in Figure 2‐3 a similar flow diagram is used for post‐combustion capture using aqueous ammonia solutions but, because of the volatility of the ammonia, a condenser must be provided inside the stripper. The lean solution from the stripper is cooled by the rich solution and further in an externally cooled heat exchanger in order to reach the defined absorber inlet temperature. In addition, because of the volatility of ammonia, the N2 leaving the absorption column and the CO2 leaving the desorption column will contain some NH3. Additional water washing columns are used to absorb this NH3.The CO2 absorber for ammonia capture is similar to SO2 absorbers (i.e. flue gas desulphurisation plants ‐ FGD) and is designed to operate with a slurry feed. The flue gas flows upwards in counter current to the slurry containing a mix of dissolved and suspended ammonium carbonate and ammonium bicarbonate; 90% of the CO2 from the flue gas is captured in the absorber. The CO2 rich slurry from the absorber contains mainly ammonium bicarbonate as the dispersed solids, which dissolves as the temperature increases in the heat exchanger to about 80°C before it enters the stripper.Figure 2‐3 Process flow diagram for aqueous ammonia capture2.3.3 Thermal integration between the power plant steam cycle and post‐combustion capture units and calculation of overall plant efficiency and cost with capture (DCE, DTE, IMP, NCEPU, ZJU)Effective integration between the base power plant and the post combustion capture equipment is essential to keep the efficiency penalty to a minimum. As shown in Figure 2‐4, low‐pressure steam for solvent regeneration is extracted from the crossover pipe between the intermediate pressure (IP) and low pressure (LP) turbine cylinders and heat recovered from the capture plant (principally from initial CO2 cooling) and from the CO2 compressor intercoolers is recovered and used for condensate heat (avoiding the need to extract steam from the LP turbine).The steam turbines have been designed with an intermediate pressure (IP) turbine capable of operating with a floating outlet pressure, a design that (with slightly different modifications) is suitable for both capture ready steam plants and new build power plants that are capable of flexible operation. A floating IP turbine has the advantage of avoiding the throttling losses of a controlled extraction system with a valve at the LP turbine when the plant is operating with capture while not affecting the plant efficiency without capture. The IP turbine outlet pressure is higher than the operating pressure of the solvent reboiler when the plant does not capture CO2. When CO2 is captured steam is extracted and sent to the solvent reboiler the IP outlet pressure drops to the pressure required to feed the solvent reboiler.LP steam extraction for feed water heating is reduced to 10% of the non‐capture value by heating boiler condensate using heat recovered from cooling the CO2 after the stripper and compressor stages. Calculated power plant performance without and with CO2 capture is shown in Table 2‐1 below.To calculate the performance of the new build post‐combustion capture cases the base case advanced supercritical plant was assumed to be modified slightly to operate with the steam cycle shown in Figure 2‐4. When the steam required for operating the CO2 capture plants, determined by modelling the capture systems described in 2.3.1 and 2.3.2 (and in more detail in DCE and DTE case study reports) is extracted from the steam cycle, and also taking into account the heat recovery from the capture and compression processes for condensate heating, the power output (calculated using gPROMS models as described in the Imperial report), falls to the values shown in Table 2‐1 below. Additional electrical power to run the capture plant’s fans and pumps and the CO2 compressors, calculated from the specific values per tonne of CO2 captured found in the case studies, has also tobe subtracted to give the reduced capture plant net outputs and net efficiencies shown in Table 2‐1 below.Table 2‐1 Performance of new build advanced supercritical power plantwithout/with post combustion CO2 captureand principal energy requirement assumptionsSolvent MEA NH3WITHOUT CAPTURENet plant power output (MWe) 824.3 824.3Efficiency (%LHV net) 43.9 43.9IP turbine outlet pressure (bara) 9.34 25.53WITH CAPTUREIP turbine outlet pressure (bara) 4.17 12.57Solvent reboiler steam pressure (bara) 3.67 12.06Heat for solvent regeneration (MJ/kgCO2) 3.54 3.23Turbine output power loss (MWe) 116.0 129.1Capture and compression electricity150 43.4requirements (kWh/tCO2)86.1 25.0Capture/compression plant power consumption(MWe)Net plant power output (MWe) 621.5 670.3Efficiency (%LHV net) 33.1 35.7capture process Floating pressureFigure 2‐4 Steam turbine/capture plant integration with a floating intermediate pressure turbineThe main additional assumptions and estimating procedures for post ‐combustion capture plant capital costs are shown in the summary of technical and economic performance in Appendix B. The assumed additional charges for the capture and compression equipment and the, relatively minor, modifications to the base plant are based on the NCEPU case study results. A conservativeassumption has been made that these capture and compression equipment costs (expressed as RMB/kW thermal) will be the same irrespective of the base plant size (across the range covered in WP3), because of uncertainties regarding the actual size of the individual absorbers and strippers that could be used (currently considered to be limited to 300‐400 MW electrical output equivalent). If larger capture equipment unit sizes prove to be feasible in the future then some consequent economies of scale would be expected, although the effect on overall plant costs would be limited since the capture equipment itself is estimated to be only 20‐25% of the total power plant capital cost with CO 2 capture.。
Journal of Energy Chemistry26(2017)1007–1013/journal-of-energy-chemistry/Contents l i sts a v ai l a ble a t Sc i enceD i rectJournal of Energy Chemistryj ourn a l ho m ep a ge:www.elsev i er.co m/loc a te/j eche mResorcinol–formaldehyde resin-based porous carbon spheres with high CO2capture capacitiesXuan Wang a,Jin Zhou a,∗,Wei Xing b,Boyu Liu a,Jianlin Zhang a,Hongtao Lin a,Hongyou Cui a,Shuping Zhuo a,∗a School of Chemistry and Chemical Engineering,Shandong University of Technology,Zibo255049,Shandong,Chinab School of Science,State Key Laboratory of Heavy Oil Processing,China University of Petroleum,Qingdao266580,Shandong,Chinaa r t i c l e i n f oArticle history:Received4May2017Revised10July2017Accepted13July2017Available online5October2017Keywords:CO2capturePorous carbonCarbon sphereUltra-microporeResorcinol formaldehyde resins a b s t r a c tPorous carbon spheres are prepared by direct carbonization of potassium salt of resorcinol–formaldehyde resin spheres,and are investigated as CO2adsorbents.It is found that the prepared carbon materials still maintain the typical spherical shapes after the activation,and have highly developed ultra-microporosity with uniform pore size,indicating that almost the activation takes place in the interior of the polymer spheres.The narrow-distributed ultra-micropores are attributed to the“in-situ homogeneous activation”effect produced by the mono-dispersed potassium ions as a form of–OK groups in the bulk of polymer spheres.The CS-1sample prepared under a KOH/resins weight ratio of1shows a very high CO2capture capacity of4.83mmol/g and good CO2/N2selectivity of∼17–45.We believe that the presence of a well-developed ultra-microporosity is responsible for excellent CO2sorption performance at room temperature and ambient pressure.©2017Science Press and Dalian Institute of Chemical Physics,Chinese Academy of Sciences.Publishedby Elsevier B.V.and Science Press.All rights reserved.1.IntroductionCarbon dioxide(CO2)has been recognized to be the biggest driver of global warming which is one of most serious prob-lems that our world are being faced with[1].Carbon capture and storage(CCS)is the technique of capturing CO2from large point sources,and is considered to be a promising means of mitigat-ing the contribution of CO2to global warming and ocean acidifi-cation[2,3].Furthermore,CO2is an important source of C1chemi-cal engineering,and could be converted into high-valued chemical products via catalyzed polymerization[4,5],photochemical[6,7]or electrochemical reactions[8,9].However,the efficient capture of CO2is primarily essential for the above processes.Currently,CO2 are mainly adsorbed by aqueous ammonia or organic amines via acid-base reactions between CO2and the ammonia/amines.This process is disadvantage of energy-costly regeneration and harsh corrosion.Therefore,physical adsorption process based on porous solid adsorbents has holden more and more attentions[10].Considering main contributions of Von der Waal forces on the physical adsorption,the materials those possess developed micro-porous texture are preferred for CO2capture.Nowadays,several kinds of microporous materials,such as microporous carbon mate-rials[11,12],zeolitic imidazolate frameworks(ZIFs)[13],metal or-∗Corresponding authors.E-mail addresses:zhoujin@(J.Zhou),zhuosp@(S.Zhuo).ganic frameworks(MOFs)[14],covalent organic frameworks(COFs) [15],porous coordination polymers(PCPs)[16],have been inves-tigated,and show excellent performances of CO2capture.Among these materials,porous carbon materials are unique due to their comprehensive advantage of easy preparation,moisture stability, low cost,tailored porosity,and so on.It has been widely accepted that the small micropores in carbon frameworks,especially the ultra-micropores smaller than0.8nm,are most efficient in CO2 capture[17–19].Chemical activation methods are still widely used in preparation of microporous carbon materials.In order to get a high capture capacity of CO2,excess activation agents,like high dosage of KOH,are always used to improve the microporosity via a deep activation.In this way,an inhomogeneous activation is in-evitable which results in wide pore size distributions(PSDs),low carbon yield and enhanced cost.Recently,we have reported a strategy of the direct carboniza-tion of phenolic resin salts in which the mono-dispersed–OK or –COOK groups are introduced via an acid-base reaction between KOH and the hydroxyl(–OH)or carboxyl(–COOH)groups on the phenyl ring of resin[20,21].The existing–OK or–COOK groups could drive an“in-situ homogeneous activation”effect,resulting in highly developed and narrow-distributed ultra-micropores.In this work,we applied the above strategy on the synthesis of mi-croporous carbon spheres due to the reasons of:(i)high ultra-microporosity in the bulk of carbon spheres could achieve excellenthttps:///10.1016/j.jechem.2017.07.0102095-4956/©2017Science Press and Dalian Institute of Chemical Physics,Chinese Academy of Sciences.Published by Elsevier B.V.and Science Press.All rights reserved.1008X.Wang et al./Journal of Energy Chemistry26(2017)1007–1013CO2capture capacities;(ii)the small size of carbon spheres en-sures a short diffusion channels for CO2molecule,(iii)the gaps be-tween carbon spheres allow a rapidflow offlue gas.Herein,potas-sium salts of resorcinol–formaldehyde(RF)resin spheres were pre-pared by the reaction of KOH and RF resin spheres,and were car-bonized into microporous carbon materials.The prepared carbon materials are found to maintain the monodisperse spherical shapes and possess highly developed ultra-microporosity,thus show ex-cellent CO2capture performance,including very high CO2capacity up to4.83mmol/g and good selectivity of CO2-over-N2up to17.2.Experimental2.1.Preparation of resorcinol–formaldehyde resin mono-spheresResorcinol–formaldehyde resin mono-spheres are prepared ac-cording to an extension of Stöber method previous reported[22]. In a typical synthesis,5.0g of resorcinol and2.5mL of ammonia aqueous solution(NH3•H2O,25wt%)were dissolved to a mixed solvent containing500mL of deionized water and200mL of ab-solute ethanol.After stirring at30°C for1h,7.0mL of aque-ous formaldehyde solution(37wt%,equal to2of the formalde-hyde/resorcinol molar ratio)was added into the above solution and further reacted for24h,and subsequently aged for24h at 80°C under a static condition.The brown RF resin mono-spheres were separated from the reaction mixture by centrifugation and air-dried at100°C for24h.2.2.Preparation of porous carbon spheres1g of as-prepared RF resins was mixed with different amounts of KOH(0.5,0.75, 1.0, 1.5and 2.0g)and 1.5mL of H2O,and were stirred for2h to ensure a sufficient reaction.The solid prod-ucts were separated by centrifugation,washed by a little absolute ethanol,and vacuum-dried at50°C.For carbonization,the KOH-treated RF resins were heated under N2atmosphere at200°C for 1h with a heating rate of2°C/min,and further heated at700°C for2h with a heating rate of5°C/min.After cooling to room tem-perature,the porous carbon mono-spheres were liberated by wash-ing with10wt%HCl solution and deionized H2O to neutrality.For convenience,the prepared carbon samples are denoted as CS-x,in which CS and x represent carbon sphere and weight ratio of KOH to RF resins,respectively.2.3.Characterization of CS-x materialsMicroscopic morphology of the prepared CS-x samples was ob-served by a scanning electron microscope(SEM,Sirion200FEI Netherlands)and a transmission electron microscope(JEM2100, JEOL,Japan).Surface chemical properties were determined by en-ergy dispersive spectroscopy(EDS,INCA Energy spectrometer)and X-ray photoelectron spectroscopy(XPS,Escalab250,USA).All gas sorption measurements were conducted on a Micromeritics ASAP 2020static volumetric analyzer.The investigated samples were firstly degassed at350°C for6h under turbomolecular vacuum before all the gas sorption measurements.The porosity proper-ties of CS-x are analyzed by means of N2sorption at−196°C and CO2adsorption at0°C.Brunauer–Emmett–Teller(BET)sur-face area(S BET)was calculated from the N2adsorption branch in a relative pressure(p/p0)range of0.05–0.25.Total pore volume (V Total)was obtained at a relative pressure of0.995.Micropore sur-face area and volume are calculated by the t-plot method.For the calculation of pore size distributions(PSDs),nonlocal density func-tional theory(NLDFT)models are widely applied,and featured in a standard by ISO-15901-3[23].Standard NLDFT models assume an idealizedflat-graphitic-like surface.In contrast to NLDFT models, quenched solid density functional theory(QSDFT)models take into account the roughness features of carbon surface.Several reports have suggest that QSDFT models could provide a better represen-tation of the roughness features of real carbon surface,while the NLDFT models may not achieve[24–26].Therefore,in this work, the PSDs of the prepared carbons were determined by applying QSDFT method and a slit pore model on N2adsorption data at −196°C.The calculation of QSDFT models were conducted on a Quantachrome Autosorb iQ2.02software.In the case of CO2sorp-tion measurement at0°C,NLDFT methods for carbon with slit-like pore model were still adopted since that QSDFT models are not yet available for this measurement.The CO2sorption capacities of the prepared porous carbons were determined by a static adsorption method carried on the ASAP2020analyzer.The CO2adsorption ex-periments were performed at0°C and25°C in the pressure range of0–1.0bar.3.Results and discussion3.1.Microscopic morphology and pore texture of the porous carbon spheresFig.1shows the SEM and TEM observation results of RF resins and the prepared carbon materials.It is clearly seen that the RF resin polymers have regular sphere shape with smooth surface (Fig.1(a)and(b)).After activation,the obtained carbon particles of CS-0.5and CS-0.75remain regular sphere shape,and the surface of the carbons is smooth and no roughness is observed,indicat-ing that the activation occurs in the interior of the carbon mono-spheres.In comparison,some defects and roughness are observed on the surface of carbon particles of CS-1,CS-1.5and CS-2,and the broken spheres appear to be towards more as the weight ratio of KOH to RF resins increases,indicating that the activation took place not only in the bulk but also on the surface of resin spheres. However,the majority of carbon particles still roughly remains the spherical shape even when the weight ratio of KOH to RF resins is up to2.The sizes of RF resin spheres and the carbon spheres are about950and850nm(average values of50spheres),respectively, indicating a shrink after the high temperature carbonization treat-ment.TEM images further confirm the typical mono-dispersion of the prepared carbon spheres(Fig.1(i)–(k)).As shown in Fig.1(k), the breakage of CS-1.5could be clearly observed which is coincided with the observation of SEM.Under high resolution TEM observa-tion,worm-like ultra-micropores are found in the bulk of the pre-pared carbon spheres.Gas sorption measurements could give systematical information of pore texture.The pore texture of the prepared porous carbon spheres wasfirstly analyzed by N2sorption at−196°C.As seen in Fig.2(a)and(b),all the isotherms exhibit common characteris-tics of narrow knees and high N2sorption capacities at very low relative pressures(p/p0<0.01).These characteristics demonstrate that all the CS-x carbon materials have highly developed ultra-microporosity with narrow pore size distribution.The samples of CS-0.5,CS-0.75and CS-1exhibit a type I isotherm,which is typi-cal characteristic of microporous materials(Fig.2(a)).In compari-son,hysteresis loops at the moderate relative pressure of0.45–0.8 are observed on the sorption isotherms of CS-1.5and CS-2,indi-cating the existence of some mesopores in size of several nanome-ters(Fig.2(b)).PSDs calculated by QSDFT model are plotted in Fig.2(c)and(d).As shown in Fig.2(c),the pore texture of the CS-0.5,CS-0.75and CS-1is almost exclusively made up of uni-form ultra-micropores smaller than0.8nm,indicating these car-bons are obtained from a homogeneous activation process.Further-more,as the KOH dosage increase,a slight enlargement of micro-pores takes place,and a few supermicropores in the1.0–1.5nmX.Wang et al./Journal of Energy Chemistry 26(2017)1007–10131009Fig.1.Microscopic morphology of CS-x .SEM images:(a,b)RF resin spheres,(c,d)CS −0.5,(e)CS-0.75,(f)CS-1,(g)CS-1.5and (h)CS-2;TEM images:(i)CS-0.5,(j)CS-1,(k)CS-1.5and (l)CS-0.5.Fig.2.N 2sorption measurement of CS-x materials:(a,b)N 2sorption isotherms,(c,d)PSDs plots calculated by QSDFT model.range are found for CS-2.Fig.2(d)clearly shows the existence of small mesopores in the size of 3–6nm for CS-1.5and CS-2.These results agree well with the analysis of N 2sorption isotherm.The pore textural parameters calculated by N 2sorption data,including specific surface area,micropore surface area,total and micropore volumes,are present in Table 1.Due to the existence of mesopores,the total pore volumes of CS-1.5and CS-2samples are larger than those of other samples.As the KOH dosage increases,the surface area and volume of micropores first increase and then decrease,and CS-1possesses the largest specific surface area,micropore sur-face area and volume,up to 1235m 2/g,1084m 2/g and 0.57cm 3/g,respectively.It is commonly known that N 2sorption measurement always underestimate the real microporosity since that the extremely nar-row micropores (<0.7nm)could not be totally accessible for N 2molecule due to its low diffusion kinetic energy at −196°C.To more accurately detect the microporosity of the CS-x samples,we further carried out the CO 2sorption at 0°C.Fig.3(a)shows the NLDFT micropore size distributions of the prepared carbons,and confirms that most of the micropores are narrow-distributed in the range of 0.5–0.75nm.Dubinin −Radushkevich (D–R)plots could provide in-depth information about ultra-microporosity.In this work,the CO 2adsorption data obtained at 0°C were further analyzed by Dubinin–Radushkevich (D–R)equation:V =V 0exp−(A /βE 0)2=V 0exp =k (RT /β)2[−2.3031g (p /p 0)]2(1)1010X.Wang et al./Journal of Energy Chemistry 26(2017)1007–1013Fig.3.(a)PSDs plots calculated by NLDFT model,(b–f)D–R plots of the prepared carbons derived from CO 2adsorption data.Table 1.Pore textural parameters of the CS-x samples.N 2sorptionCO 2sorptionSample S BET a S micro b V Total c V micro d V 0eL f(m 2/g)(m 2/g)(cm 3/g)(cm 3/g)(cm 3/g)(nm)CS-0.511719910.660.520.630.75CS-0.75114010490.600.550.650.74CS-1123510840.670.570.670.76CS-1.511879120.880.480.580.74CS-211138540.820.450.500.71a Brunauer–Emmett–Teller surface area;b Micropore surface area calculated by the t -plot method;c Total pore volume;d Micropore volume calculated by the t -plot method;e Micropore volume obtained from CO 2adsorption;fAverage micropore size calculated by D–R plots.in which V (cm 3/g)is the volume filled at a certain temperature (T )and a relative pressure (p /p 0),V 0(cm 3/g)is the micropore volume,E 0(kJ/mol)is the characteristic energy,k is a constant about the pore texture,and βis the affinity coefficient (β=0.35for CO 2),respectively.As shown in Table 1,the micropore volumes obtained from the CO 2sorption (D–R plot,V 0)are higher than those ob-tained from the N 2sorption (t -plot method,V micro ),revealing the presence of extremely small micropores that may be hard to be detected by N 2molecules [27].D–R plots of CO 2sorption on CS-x exhibit well-defined linear curves with a high correlation coeffi-cients (R 2)about 0.999(Fig.3(b)–(f)),reflecting that the CS-x sam-ples possess uniform ultra-microporosity [19,28].According to the empirical equation proposed by Stoecki (Eq.(2)),we could calcu-late the average width of micropores [29].L =10.8/(E 0−11.4)(2)As shown in Table 1,the average widths of micropores are 0.75,0.74,0.76,0.74and 0.71nm for CS-0.5,CS-0.75,CS-1,CS-1.5and CS-2,respectively.It should be noted that the average size of CS-2is lower than those of other samples although this carbon is pre-pared by using excess KOH dosage.The above results indicate that the prepared carbon materials still remain a typical spherical shape after a direct activation pro-cess,and possess narrow-distributed ultra-microporosity closely related to KOH dosage.Therefore,we further investigated the ac-tivation process of the CS-x samples which is illustrated in Fig.4.Firstly,the RF resin mono-spheres are prepared by an extension of Stöber method.When the prepared resins mixed with KOH,the –OH groups of RF resins will immediately react with KOH via fast acid–base reaction,and are change into –OK groups.As the KOH dosage increases,more and more –OH groups will be converted into –OK groups.Based on a simplified molecule structure of RF resins shown in Fig.4,the KOH/RF resins weight ratios of 0.50,0.75,1,1.5and 2are roughly equivalent to the KOH/–OH molar ratios of 0.50,0.75,1,1.5and 2.As shown in Fig.4,in the case of CS-0.5and CS-0.75,only part of –OH groups are converted into –OK groups,in the case of CS-1,all the –OH groups are exactly converted into –OK groups,and in the case of CS-1.5and CS-2,the KOH is excess.Thus,it could be deduced that the actual activated agent during the carbonization process is the produced –OK groups for CS-0.5,CS-0.75and CS-1,while the –OK groups and excess KOH adsorbed on the surface of RF resin spheres for CS-1.5and CS-2.The activation mechanism has been investigated by in-situ XRD measurement in a previous literature [30].In the high tem-perature carbonization process,the phenolic salts could decompose into K 2O or K 2CO 3.These species are further reduced by carbon to metallic potassium via the reactions of K 2CO 3+C →K 2O +2CO and K 2O +C →2K +CO in which partial carbon atoms were etched into CO to give rise to the porosity.The produced K vapors can intercalate into the carbon layers,thus generates even more ultra-microporosity.Therefore,it could be found that the potas-sium ions (K +)mono-dispersed as a form of –OK drive a high efficient chemical activation in the bulk of RF resin spheres,thus develop uniform ultra-micropores while maintaining the spherical shapes.When the KOH is excess,inhomogeneous activation will be occur,thus results in a wide pore size distribution and even break-age of carbon spheres.When all the –OH groups are exactly con-verted into –OK groups (CS-1sample),the activation efficiency is the highest among all the prepared samples,thus the CS-1sampleX.Wang et al./Journal of Energy Chemistry 26(2017)1007–10131011Fig.4.Illustration for preparation of the porous carbonspheres.Fig.5.CO 2adsorption measurements (a)at 0°C and 1bar,(b)at 25°C and 1bar.possesses the largest specific surface area,micropore surface area and volume.3.2.CO 2capture performance of the porous carbon spheresIt has been reported that the CO 2capacities at ambient pres-sure are mainly determined by the micropore lower than 1nm,while the CO 2capacities at high pressure are closely related the micropores and small mesopores.Considering their main ultra-microporosity and lower total pore volumes (<0.9cm 3/g),the CS-x samples prepared here are more suitable to CO 2capture at ambient pressure but not high pressures.Therefore,we investi-gate the CO 2sorption performance of the CS-x samples at ambi-ent pressure.As seen in Fig.5and Table 2,the as-prepared carbon spheres are found to give excellent CO 2capture capacities range from 6.55–7.34mmol/g at 0°C and 4.27–4.83mmol/g at 25°C under 1bar.Among all the samples,CS-1gives the highest CO 2uptake of 7.34mmol/g (32.3wt%)and 4.83mmol/g (21.3wt%)at 0and 25°C,which is attributed to its largest micropore vol-ume.The value of 4.83mmol/g is higher than lots of phenolic resin based porous carbon materials at the same conditions (Ta-ble S1).It is worthy to pointing that the CO 2capacity of CS-1isTable 2.CO 2sorption capacities at different conditions.CO 2sorption capacity (mmol/g)Sample 0°C,1bar 25°C,1bar 25°C,0.15bar CS-0.5 6.69 4.37 1.18CS-0.75 6.92 4.51 1.22CS-17.34 4.83 1.33CS-1.57.26 4.58 1.25CS-26.554.271.24obviously higher than the carbon spheres ever reported,such as CO 2-activated carbon spheres [19,31],N-rich microporous carbon spheres [32–34],porous carbon nitride spheres [35],KOH-activated carbon spheres [11],hollow carbon spheres [36].Based on the ex-isting research results,the excellent CO 2capacities of CS-x should be mainly due to their high developed ultra-micropores with uni-form pore size [17–19].Besides high specific capture capacities,a practical CO 2adsor-bent should have high capture capacity at low CO 2pressure,high selectivity of CO 2-over-N 2,low regeneration energy and good re-cyclability.As shown in Table 2,the CS-1shows a good CO 2cap-1012X.Wang et al./Journal of Energy Chemistry 26(2017)1007–1013Fig.6.(a)CO 2and N 2adsorption isotherms on CS-1at 25°C and 1bar,(b)IAST adsorption selectivity of CS-1for binary mixtures of CO 2/N 2=15/85at 25°C,(c)CO 2heats of adsorption on CS-x ,(d)regeneration of CS-1for CO 2sorption.ture capacity of 1.33mmol/g at a typical CO 2pressure of 0.15bar in flue gas.To investigate the selectivity of CO 2-over-N 2,N 2sorp-tion at 25°C is carried out.As shown in Fig.6(a)and S1,CS-1and CS-2show low N 2capture capacities of 0.63and 0.64mmol/g at 25°C and 1.0bar.The large difference in the CO 2and N 2sorp-tion capacities reflects that preferred capture performance of the prepared carbons toward to CO 2.Herein,ideal adsorbed solution theory (IAST)was applied to quantitatively evaluate the selectivity of CO 2-over-N 2[37,38].For binary gas mixtures of CO 2:N 2=15:85(volume ratio),the selectivity plots of CO 2-over-N 2based on IAST are shown in Figs.6(b)and S2(a).CS-1and CS-2show similar adsorption selectivity.High adsorption selectivities (∼25to ∼45)were observed at low pressure;these selectivities drop as the pres-sure increases and reach plateau of ∼17and ∼16for CS-1and CS-2,respectively.The selectivity could be also calculated based on the initial slopes of the CO 2and N 2adsorption isotherms [39–41],and this method is ever used in our previous works [42].The CO 2-over-N 2selectivities of CS-1and CS-2are about 14and 13at 25°C (Fig.S2c and d),which is very close to the IAST values.The preferred adsorption toward CO 2over N 2observed here should be attributed to the higher quadruple moment of CO 2molecules and strong ad-sorption potential of ultra-micropores,leading to its preferred in-teraction with the carbon pore surface.To evaluate the regeneration of energy consumption,the isos-teric heats of adsorption (Q st )were calculated according to Clausius–Clapeyron equation:lnp 1p 2=Q st ×T 2−T 1R ×T 1×T 2(3)in which Q st is the isosteric heats of adsorption,T 1and T 2are the sorption temperatures,p 1and p 2are the pressures at which a spe-cific equilibrium adsorption amount is reached at a certain tem-perature,and R is the gas constant (8.314J/(K mol)).As shown in Fig.6(c),the CS-x samples show a moderate Q st of 20–27kJ/molthat indicates the nature of physical sorption mainly driven by the Van der Waals forces.It should be noted that the Q st values gener-ally decrease with the increasing of average micropore width,and the lowest Q st values (21–23kJ/mol)are given by the CS-1car-bon with the largest average micropore width.This result is due to that the superposition extent of Van der Waals forces between carbon adsorbent and CO 2molecules weakens with the increasing of pore width.Additionally,the surface oxygen-containing groups have been predicted to enhance CO 2adsorption in microporous carbon materials in the absence of water vapor [43,44].As shown in Table S2,all the carbon samples possess high content of oxygen in the range of 8.9at%to 10.2at%.And,three types of oxygen-containing groups,including quinone groups,phenol groups and/or C –O –C ether groups and carboxyl groups are determined according to the binding energy peaks at 531.2,532.3,and 533.5eV (Fig.S2).Finally,we studied the recyclability of CS-1in CO 2sorption appli-cation at 25°C.As shown in Fig.6(d),the CO 2sorption isotherms are totally overlapped,confirming that the as-prepared porous car-bon spheres possess good recyclable stability and could be well re-generated in CO 2capture application.4.ConclusionsWe have prepared porous carbon spheres by direct carboniza-tion of potassium salts of resorcinol–formaldehyde resin spheres.The prepared CS-x materials possess highly developed ultra-microporosity with uniform pore size.It is found that the CS-1sample possess the highest specific surface area,micropore surface area and micropore volume which is prepared under the condition of KOH/resin weight ratio equal to 1.The prepared carbon mate-rials show excellent CO 2capture performance,such as high CO 2capture capacities up to 4.83mmol/g,good selectivity of CO 2-over-N 2,and stable recyclability.We believe that the presence of a well-X.Wang et al./Journal of Energy Chemistry26(2017)1007–10131013developed ultra-microporosity is responsible for excellent CO2ad-sorption performance of the carbons prepared here.AcknowledgmentsWe are grateful for thefinancial supports by the Natural Sci-ence Foundation of China(NSFC21576158,21476132,21576159and 21403130)and Shandong Provincial Natural Science Foundation, China(No.2015ZRB01765).Supplementary materialsSupplementary material associated with this article can be found,in the online version,at doi:10.1016/j.jechem.2017.07.010.References[1] B.Metz,O.Davidson,H.de Coninck,M.Loos,L.Meyer(Switzerland).WorkingGroup III,2005.[2]J.Gibbins,H.Chalmers,J.Energy Policy36(2008)4317–4322.[3]J.Pires, F.Martins,M.Alvim-Ferraz,M.Simões,J.Chem.Eng.Res.Des.89(2011)1446–1460.[4]Y.Wang,Y.Qin,X.Wang,F.Wang,ACS Catal.5(2014)393–396.[5]M.Cheng, D.R.Moore,J.J.Reczek, B.M.Chamberlain, E.B.Lobkovsky,G.W.Coates,J.Am.Chem.Soc.123(2001)8738–8749.[6]Y.Kou,Y.Nabetani,D.Masui,T.Shimada,S.Takagi,H.Tachibana,H.Inoue,J.Am.Chem.Soc.136(2014)6021–6030.[7]W.Tu,Y.Zhou,Z.Zou,J.Adv.Mater.26(2014)4607–4626.[8]M.S.Xie,B.Y.Xia,Y.Li,Y.Yan,Y.Yang,Q.Sun,S.H.Chan,A.Fisher,X.Wang,J.Energy Environ.Sci.9(2016)1687–1695.[9] D.D.Zhu,J.L.Liu,S.Z.Qiao,J.Adv.Mater.28(2016)3423–3452.[10]Q.Wang,J.Luo,Z.Zhong,A.Borgna,J.Energy Environ.Sci.4(2011)42–55.[11]N.P.Wickramaratne,M.Jaroniec,J.Mater.Chem.A1(2013)112–116.[12]J.Wang,I.Senkovska,M.Oschatz,M.R.Lohe,L.Borchardt,A.Heerwig,Q.Liu,S.Kaskel,ACS Appl.Mater.Interfaces5(2013)3160–3167.[13] A.Phan,C.J.Doonan,F.J.Uribe-Romo,C.B.Knobler,M.O’keeffe,O.M.Yaghi,J.Acc.Chem.Res.43(2010)58–67.[14]S.Nandi,S.Collins,D.Chakraborty,D.Banerjee,P.K.Thallapally,T.K.Woo,R,J.Am.Chem.Soc.139(2017)1734–1737.[15]Y.Zeng,R.Zou,Y.Zhao,Adv.Mater.28(2016)2855–2873.[16]L.An,H.Wang, F.Xu,X.-L.Wang, F.Wang,J.Li,J.Inorg.Chem.55(2016)12923–12929.[17]Z.Zhang,J.Zhou,W.Xing,Q.Xue,Z.Yan,S.Zhuo,S.Z.Qiao,J.Phys.Chem.Chem.Phys.15(2013)2523–2529.[18]V.Presser,J.McDonough,S.-H.Yeon,Y.Gogotsi,J.Energy Environ.Sci.4(2011)3059.[19]M.Sevilla,J.B.Parra, A.B.Fuertes,ACS Appl.Mater.Interfaces5(2013)6360–6368.[20]J.Zhou,Z.Li,W.Xing,H.Shen,X.Bi,T.Zhu,Z.Qiu,S.Zhuo,J.Adv.Funct.Mater.26(2016)7955–7964.[21]J.Zhou,H.Shen,Z.Li,S.Zhang,Y.Zhao,X.Bi,Y.Wang,H.Cui,S.Zhuo,J.Electrochim.Acta209(2016)557–564.[22]J.Liu,S.Z.Qiao,H.Liu,J.Chen,A.Orpe,D.Zhao,G.Q.M.Lu,J.Angew.Chem.Int.Ed.50(2011)5947–5951.[23] B.Standard,B.ISO,J.BS ISO.(2005)15901-15901.[24]G.Y.Gor,M.Thommes,K.A.Cychosz, A.V.Neimark,J.Carbon50(2012)1583–1590.[25] A.V.Neimark,Y.Lin,P.I.Ravikovitch,M.Thommes,J.Carbon47(2009)1617–1628.[26]X.Wu,W.Xing,J.Florek,J.Zhou,G.Wang,S.Zhuo,Q.Xue,Z.Yan,F.Kleitz,J.Mater.Chem.A2(2014)18998–19004.[27] A.Wahby,J.M.Ramos-Fernandez,M.Martinez-Escandell, A.Sepulveda-E-scribano,J.Silvestre-Albero,F.Rodriguez-Reinoso,J.ChemSusChem.3(2010) 974–981.[28]M.Dubinin,J.Carbon27(1989)457–467.[29] F.Stoeckli,L.Ballerini,J.Fuel70(1991)557–559.[30]J.Zhou,Z.Li,W.Xing,H.Shen,X.Bi,T.Zhu,Z.Qiu,S.Zhuo,J.Adv.Funct.Mater.26(2016)7955–7964.[31]N.P.Wickramaratne,M.Jaroniec,ACS Appl.Mater.Interfaces5(2013)1849–1855.[32]L.Liu,Z.-H.Xie,Q.-F.Deng,X.-X.Hou,Z.-Y.Yuan,J.Mater.Chem.A5(2017)418–425.[33]N.P.Wickramaratne,J.Xu,M.Wang,L.Zhu,L.Dai,M.Jaroniec,J.Chem.Mater.26(2014)2820–2828.[34]M.Wang,J.Wang,W.Qiao,L.Ling,D.Long,RSC Adv.4(2014)61456–61464.[35]Q.Li,J.Yang,D.Feng,Z.Wu,Q.Wu,S.S.Park,C.-S.Ha,D.Zhao,J.Nano Res.3(2010)632–642.[36]S.Feng,W.Li,Q.Shi,Y.Li,J.Chen,Y.Ling,A.M.Asiri,D.Zhao,-mun.50(2014)329–331.[37]Y.Zhao,L.Zhao,K.X.Yao,Y.Yang,Q.Zhang,Y.Han,J.Mater.Chem.22(2012)19726–19731.[38]S.Gadipelli,Z.X.Guo,J.ChemSusChem.8(2015)2123–2132.[39]G.P.Hao,W.C.Li, D.Qian,G.H.Wang,W.P.Zhang,T.Zhang, A.Q.Wang,F.Schuth,H.J.Bongard,A.H.Lu,J.Am.Chem.Soc.133(2011)11378–11388.[40]R.Banerjee,H.Furukawa,D.Britt,C.Knobler,M.O’Keeffe,O.M.Yaghi,J.Am.Chem.Soc.131(2009)3875–3877.[41]J.An,S.J.Geib,N.L.Rosi,J.Am.Chem.Soc.132(2009)38–39.[42]J.Zhou,Z.Li,W.Xing,T.Zhu,H.Shen,S.Zhuo,mun.51(2015)4591–4594.[43]Y.Liu,J.Wilcox,J.Environ.Sci.Technol.46(2012)1940–1947.[44]W.Xing,C.Liu,Z.Zhou,J.Zhou,G.Wang,S.Zhuo,Q.Xue,L.Song,Z.Yan,J.Nanoscale Res.Lett.9(2014)189.。
葡萄糖在稀硫酸催化下的降解反应动力学1彭新文,吕秀阳浙江大学化工系,杭州(310027)E-mail:luxiuyang@摘要:葡萄糖是纤维素的组成单体。
葡萄糖在酸催化下的降解是从生物质资源出发制备乙酰丙酸过程中重要步骤。
从生物质资源出发制备乙酰丙酸通常是采用1.5%以上的硫酸作为催化剂,既造成严重设备腐蚀,又给环境保护带来很大压力。
为了探索在稀硫酸浓度下水解生物质制备乙酰丙酸工艺的可行性,本文系统地测定了压力5MPa,初始浓度5~20 mg·mL-1、温度160~190℃、硫酸浓度0.05%~0.4wt%范围内葡萄糖的降解反应动力学数据,并以带有平行反应的一阶连串反应动力学模型对数据进行了拟合。
拟合结果表明在实验范围内,葡萄糖降解的主、副反应对葡萄糖均为一级反应;葡萄糖降解的主反应对H+为0.716级,反应的活化能129 kJ·mol-1;副反应对H+为1.06级,反应的活化能为154 kJ·mol-1。
通过对动力学方程进行分析,发现在硫酸浓度到达一定量后(0.3wt%左右),若再增加硫酸浓度,对乙酰丙酸收率影响较少。
因此在综合考虑收率、硫酸用量以及污染等因素的前提下,稀浓度硫酸(0.3%左右)催化降解生物质制备乙酰丙酸工艺是有发展前景的。
关键词:葡萄糖;乙酰丙酸;稀硫酸;降解;动力学中图分类号:TQ 032;O 643.121.引言纤维素含量约占50%的生物质资源,是一种有广阔应用前景的可再生资源,它是由许多D-吡喃葡萄糖彼此以β-1-4糖苷键连接起来的线性高分子化合物,葡萄糖是其组成单体。
乙酰丙酸(levulinic acid,LA)是一种能从纤维素出发,低成本、大规模制备的新平台化合物[1-2]。
从纤维素出发制备LA一般采用1.5~30wt%硫酸做催化剂[3-5]。
但是以高浓度硫酸作催化剂,对反应设备有很大的腐蚀性,且反应后会产生大量的酸性废渣和废液,给环境保护带来严重的问题。
第38卷 第3期2011年3月湖南大学学报(自然科学版)Jour nal of H unan U niver sity(Na tur al Science s)Vo l.38,N o.3M a r.2011文章编号:1674-2974(2011)03-0069-04θ环填料塔中单乙醇胺吸收C O2传质性能研究*那艳清,符开云,梁志武,奚 飞,刘贺磊,李文生,TONIW ACHWUTHIKUL P T (湖南大学化学化工学院,CO2捕获与封存国际合作中心(iCCS),湖南长沙 410082) 摘 要:在乱堆θ环不锈钢填料的自制填料塔中,使用单乙醇胺(M EA)作为吸收剂,研究M EA脱除CO2的去除率和总体积传质系数K G a v,考察了贫液中CO2负载量、吸收剂浓度、液体流量、吸收温度等不同参数对总体积传质系数的影响.实验结果表明,K G a v随贫液中CO2负载量的增大而减小,随吸收剂浓度、液体流量的增大而增大;且进料温度在30~50℃时,K G a v随温度的增大而增大.关键词:二氧化碳;单乙醇胺;填料塔;传质系数中图分类号:TQ028.17 文献标识码:AM ass T ransfer Performance in aθ-ring Packed Tow er fo r CO2 Abso rption Process U sing M onoethanolamine(M EA)NA Yan-qing,FU Kai-yun,LIANG Zhi-w u,XI Fei,LIU He-lei,LI Wen-sheng,TON TIWAC HWU THIK U L P T(Joint I nter na tional Cente r fo r CO2Captur e and Stor age(iCCS),Colleg e of Chemistryand Chemical Eng ineering,Hunan U niv,Chang sha,Hunan 410082,China) Abstract:The mass transfer coefficient(K G a v)and the rem oval efficiency of carbon dioxide for aqueo us MEA-based abso rption pro cess w ere investig ated in a packed absorber filled withθ-ring stainless steel packing.The ex periment results show ed that K G a v decreased w ith the increase of CO2lean-loading,and in-creased w ith the rise o f MEA co ncentratio n and liquid flow rate.In addition,an increase in feed tem pera-ture from30to50℃led to an increase in K G a v.Key words:carbon dioxide;m onoe thano lamine;packed tow er;mass transfer coefficient 随着人类消耗化石燃料的增加,大气中CO2含量逐步上升,温室效应日益严重,致使人类面临着急需解决的环境问题.同时,CO2是一种宝贵的资源,在食品、农业、工业、采油业等很多领域有着广泛的应用.目前,有很多方法可以回收CO2,而化学吸收法因其容量大、吸收效果好、吸收剂可循环使用、成本低并能回收高纯产品等优点,在工业上得到广泛应用.常见的化学吸收剂有氨水、钾碱/钠碱和有机醇胺溶液等[1-3],其中MEA是一种工业上应用最为广泛的有机醇胺吸收剂.填料作为填料塔的重要组成部分,其传质性能的优劣对脱碳效果起着关键作用.θ环填料,又称狄克松(Dix-on)填料,是一种小颗粒高效填料.该填料空隙率大、比重小,表面润湿情况比一般瓷环完全,成膜率高,因而分离效率较高.常彦龙等[4]研究表明θ环填料具有良好的流体力学性能;骆培成等[2]使用θ环填料考察了NaOH对空气中微量CO2的净化情况.但是,目前国内*收稿日期:2010-10-12基金项目:湖南省科学技术厅科技计划重点项目(2010SK2001)作者简介:那艳清(1964-),女,辽宁法库人,湖南大学副教授通讯联系人,E-mail:zw liang@h 湖南大学学报(自然科学版)2011年对θ环填料塔中使用MEA 脱除烟气中CO 2总体积传质系数研究的报道较少,而总体积传质系数K G a v 是研究填料塔传质性能的重要参数,它的确定对于传质速率方程、传质单元高度的计算非常关键.因此对该新型高效散装填料进行传质性能研究有着重要的现实意义.本文在填充θ环不锈钢填料的直径为24mm 的填料塔中,进行MEA 脱除CO 2工艺传质性能研究,考察贫液中CO 2负载量、吸收剂浓度、液体流量、吸收温度等不同参数对传质系数的影响.实验结果可为今后的放大试验及工业上的吸收塔设计提供依据.1 理论分析1.1 MEA 吸收C O 2反应机理目前,公认的MEA 吸收CO 2反应机理是由Ca -plow [5]于1968年提出,并由Danckw erts [6]于1979年补充完善的两性离子机理.首先CO 2与M EA (或DEA )反应会生成一种两性离子的中间产物:CO 2+R 1R 2NH R 1R 2NH -COO -然后,两性离子与溶液中的碱催化剂进行去离子反应,形成一种质子化的产物和氨基甲酸盐离子:R 1R 2NH +COO -+B R 1R 2NCOO -+BH-溶液中的碱催化剂B 包括R 1R 2NH ,OH -和H 2O .1.2 C O 2去除率CO 2去除率η可表示为:η=Y in -Y ou tY in×100%式中Y in ,Y out 为塔底和塔顶气相中CO 2摩尔比浓度.1.3 总体积传质系数的测定微分法是测量填料塔总体积传质系数的方便有效的手段.Aroo nw ilas [7-8]和沈洪士等[9]分别用该法考察了AM P ,M EA ,碳酸钾/哌嗪,AEE +M DEA 等水溶液吸收CO 2的总体积传质系数.根据双膜理论及单相传质速率的方程,在稳态时,用气相浓度差表示的总传质速率方程式如下:N A =K G P (y A -y *A )(1)式中N A 为吸收传质通量,(kmol /(m 2·h ));K G 为气相总传质系数,(km ol /(m 2·h ·kPa ));P 为气体总压,kPa ;y A ,y *A 为气相主体浓度,与液相主体浓度相平衡的气相浓度,无因次.考虑到物料衡算,在逆流连续接触的填料塔中,任意横截面上取d z 的微元高度有:N A a v d z =V d Y A (2)式中a v 为有效相界面积,(m 2/m 3);V 为N 2流量,kmol /(m 2·h );Y A 为气相中CO 2摩尔比浓度,无因次.把式(1)代入式(2)中并整理得:K G a v =V p (y A -y *A )d Y d z (3)式(3)中的d Y /d z 通过测量不同塔高的气相CO 2浓度拟合曲线作Y -z 图,对z 求导即可得到不同高度处的d Y /d z 值.此外,由于M EA -CO 2属于快速反应,在气液相连续逆流接触的吸收过程中,与液相主体中CO 2浓度相平衡的气相CO 2浓度趋于零,即y *A ≈0.在伴随有化学反应的吸收过程中,K G a v 通常沿着填料高度不断变化.实验过程中沿塔底往上到第3个气体取样点处已去除大半甚至全部的CO 2,这里只讨论前3个取样点处的K G a v .在该段填料高度处,y A -y *A ≈y A ,本文忽略y A *的大小.2 实验部分实验装置如图1所示,主体设备填料塔的内径为24mm ,填料高度为1.42m ,塔内乱堆填充天津天大天久科技股份有限公司提供的θ环不锈钢填料(Ф3mm ×3mm ),填料塔外部包有保温材料.N 2,CO 2气体由钢瓶依次经减压阀、质量流量计混合后进入塔底,与经过磁力泵、转子流量计从塔顶进入的吸收剂逆流接触.气、液物料在进入填料塔之前都经过水浴锅加热以便在设定温度进料.吸收过程进行前先调节N 2,CO 2流量,待y CO 2稳定在10%后通入吸收剂.1—CO 2钢瓶;2—N 2钢瓶;3,4—质量流量计;5,10—水浴锅;6—填料塔;7—红外线检测仪;8—富液罐;9—转子流量计;11—磁力泵;12—贫液罐图1 实验装置图Fig .1 Schematic diag ram of ex perimental facility实验中,气相CO 2浓度用便携式红外线CO 2分析仪(NOVA ,Mo del 302K )测量,测量精度为0.1%(体积).由于分析仪工作时所需的气样量较大,不宜直接将取样棒连接塔身测量.每次实验过程中,待吸收装置运行20min 操作状态稳定后,测量点外接气体取样袋.取样过程约为1h ,以保证袋中气体与该塔高处的气体浓度基本一致.贫、富液中的70第3期那艳清等:θ环填料塔中单乙醇胺吸收CO 2传质性能研究CO 2含量采用酸解法测量,即往吸收液中加入过量的H Cl (或H 2SO 4),在强酸作用下,将CO 2分解出来.在封闭系统中,用量气管测量排出的CO 2体积并考虑温度和饱和蒸汽压的影响即可求出吸收液中的CO 2含量.实验中,分别由气、液相计算物料平衡的相对误差均小于5%,误差在可接受范围内.3 结果与讨论3.1 C O 2负载量对η和K G a v 的影响在相同MEA 浓度C MEA 下,贫液中的CO 2负载量α(CO 2/M EA )mo l /mo l 对吸收效果有较大影响.由图2可知,K G a v 随着CO 2负载量的增大而减小.当溶液中CO 2负载量大于0.3m ol /mo l 时,其K G a v 值较新鲜M EA 溶液K G a v 的20%还小.如α分别为0.0027和0.3045mo l /mol 时,在沿塔底往上前3个取样点处的K G a v 值分别为0.345,0.528,1.887和0.018,0.105,0.204kmo l /(m 3·h ·kPa ).这是由于随着CO 2负载量的增大,溶液中的自由MEA 分子减少,与CO 2接触的机会减小,化学反应速率的降低使得化学吸收作用减小,物理吸收作用增大,故在高负载量时,总体积传质系数变化不大.填料高度/m(a )CO 2负载量对η的影响填料高度/m(b )CO 2负载量对K G a v 的影响V N 2—2.50L /min ;y CO 2—10%;C M EA —1.02mol /L ;L —60m L /min ;t —30℃图2 CO 2负载量对η和K G a v 的影响Fig .2 Effect of CO 2lean lo ading on ηand K G a v3.2 吸收剂浓度对η和K G a v 的影响在工业吸收过程中,吸收剂浓度是影响η和K G a v的一个重要参数,选择合适的吸收剂浓度至关重要,因为它影响着整个装置的设计和操作.在30℃时,分别通入1.0,2.0mol /L 的MEA 水溶液.由图3可知,在气、液相进料量一定的条件下,随着MEA 浓度的增大,CO 2去除率和传质速率系数均增大.根据传质机理,吸收剂浓度的提高,使得化学增强因子增大,液相反应速率加快,液相分传质系数大大提高,从而增大了总体积传质系数.值得注意的是,即便增大浓度会不断提高总传质系数,可增大浓度意味着所需吸收剂的费用增加,对设备腐蚀的速率也会增加.因此,在工厂的运行中要整体优化考虑吸收剂浓度参数.填料高度/m(a )M EA 浓度对η的影响填料高度/m(b )M EA 浓度对K G a v 的影响V N2—3.00L /min ;y CO 2—10%;α—0.1152mol /mol ;L —40m L /min ;t —30℃图3 M EA 浓度对η和K G a v 的影响F ig .3 Effect o f M EA co ncentration on ηand K G a v3.3 液体流量对η和K G a v 的影响如图4所示,在C MEA 为1mol /L 时,随着液体流量的增大,总体积传质系数K G a v 和CO 2去除率η增大.产生这种行为的原因可能是增大吸收剂的喷淋密度会增大气液接触的有效面积;增大吸收剂流量会增大液体流动的湍流程度,减小相界面层的液膜厚度,从而增大液相的传质系数;增大吸收剂流量会增大自由MEA 分子,从而增加吸收CO 2的能力.但是,过于增大流量会增大循环和再生吸收剂的花费.故一味增大吸收剂的流量不一定会得到最优化的操作条件.71 湖南大学学报(自然科学版)2011年3.4 进料温度对η和K G a v 的影响如图5所示,对于吸收过程,低温有利于气体分填料高度/m(a )液体流量对η的影响填料高度/m(b )液体流量对K G a v 的影响V N 2—3.00L /m in ;y CO 2—10%;α—0.1431mol /mol ;C MEA —1.02mol /L ;t —30℃图4 液体流量对η和K G a v 的影响Fig .4 Effect of liquid flo w rate on ηand K G a v填料高度/m(a )进料温度对η的影响填料高度/m (b )进料温度对K G a v 的影响V N 2—3.00L /m in ;y CO 2—10%;α—0.1431mol /mol ;C M EA —1.02m ol /L ;L —40m L /min图5 进料温度对η和K G a v 的影响Fig .5 Effect of feeding tempera tur e on ηand K G a v子向溶液中传递;升高温度会降低溶液粘度,有利于分子在溶液中的扩散,且会加快反应速率,两方面的因素都会影响总传质系数.本文对进料温度为30℃,40℃,50℃进行了实验,由图5可知,进料温度升高,吸收效果显著提高.4 结 论本文在θ环填料塔中,对MEA 吸收CO 2过程进行传质性能研究,在考察的参数范围内,得出以下结论:1)在相同条件下,CO 2去除率和总体积传质系数随吸收剂中CO 2负载量的增大而减小.CO 2负载量大的吸收剂化学吸收CO 2作用不显著.2)增大吸收剂的浓度会提高增强因子的作用,加快CO 2与吸收剂之间的传递.3)η和K G a v 均随吸收剂流量的增大而增大,在气液比降低到一定程度时,增大流量对吸收效果影响不大.4)进料温度对吸收效果有较大影响,提高进料温度,去除率和传质系数较大.参考文献[1] 刁永发,郑显玉,陈昌和.氨水洗涤脱除CO 2温室气体的机理研究[J ].环境科学学报,2003,23(6):753-757.DIAO Yong -fa ,ZHENG Xian -yu ,CHEN Chang -he .Study on re -moval mechanism of CO 2greenhouse gas by ammo -nia scrubbing [J ].Acta Scientiae Circumstantiae ,2003,23(6):753-757.(In Chinese )[2] 骆培成,焦真,王志祥,等.填料塔中碱性水溶液对空气中微量CO 2的净化[J ].化工学报,2003,54(6):824-828.LUO Pei -cheng ,JIAO Zhen ,W ANG Zhi -xiang ,et al .CO 2removal from poll uted air using al kal ine s olutions in packed tow er [J ].Journal of Chemical Industry and Engineering ,2003,54(6):824-828.(In Chines e )[3] 孙承贵,曹义鸣,介兴明,等.中空纤维致密膜基吸收CO 2传质机理分析[J ].高等化学工程学报,2007,21(4):556-562.SUN Cheng -gui ,CAO Yi -ming ,JIE Xing -ming ,et al .M ass trans -fer mechanism of CO 2-absorption through a non -porous hollow fiber contactor [J ].Journal of Chemical Engineering of Chinese Univers i -ties ,2007,21(4):556-562.(In Chinese )[4] 常彦龙,冯庆华,张琨,等.θ网环等四种小型高效精密填料流体力学性能测定及比较[J ].过程工程学报,2004,4(6):496-500.CHANG Yan -l ong ,FENG Qing -hua ,ZHANG Kun ,et al .Hydro -mechanical performance measurement and comparison of fou r species of minitype packings [J ].The Chinese Journal of Process Engineer -ing ,2004,4(6):496-500.(In Chinese )[5] CAPLOW M .Kinetics of carbamate formation and break -down [J ].JAm Chem Sci ,1968,90(24):6795-6803.[6] DANCKWERTS P V .The reactions of CO 2with ethanola -mines [J ].Chemical Engineering Science ,1979,34:443-446.[7] AR OONWILAS A ,TONTIWACHWUTHIKUL P .M ass trans ferco -effici ents and correlation for CO 2abs orption into 2-amino -2-meth -yl -1-propanol (AM P )using structured packing [J ].Ind Eng Chem R es ,1998,37:569-575.[8] AROONWILAS A ,VEA W AB A ,TONTIWACHWUTHIKUL P .Behavior of the mass transfer coefficient of structured packings in CO 2abs orbers with chemical reactions [J ].Ind Eng Chem Res ,1999,38(5):2044-2050.[9] 沈洪士,张永春,陈绍云,等.填料塔中混合胺吸收二氧化碳的研究[J ].现代化工,2010,30(2):70-73.SHEN Hong -shi ,ZHANG Yong -chun ,CHEN Shao -yun ,et al .Study on absorption of c arbon dioxi de by blended amines in packed tower [J ].M odern Chemical Industry ,2010,30(2):70-73.(In Chinese )72。