Combinatorial Search for Optimal Hydrogen-Storage Nanomaterials Based on Polymers

俯冲带变质过程中的含碳流体刘景波【摘要】俯冲带含碳岩石通过俯冲过程的变质反应生成了含碳水流体、富硅酸盐的超临界流体和含碳熔体.不同类型流体的形成与岩石成分和岩石经历的温压条件相关.岩石中碳酸盐矿物脱碳反应的温压条件取决于岩石起初的流体成分:有水存在时,反应发生在低温条件下.在高压条件下,碳酸盐矿物在水或含盐水流体的溶解是生成含碳流体重要的机制,其导致的碳迁移作用可能超过脱碳变质反应的作用.高温条件下,含碳岩石的部分熔融可以生成含碳的熔体,这在热俯冲环境和俯冲带岩石底辟到上覆地幔的情况下是碳迁移重要载体.富硅酸盐的超临界流体可能是在第二临界端点上形成的超临界流体,目前在超高压岩石中观察到的非花岗质成分的多相固体包裹体被认为是这种流体结晶的产物,然而对其理解尚存在很多问题,需要进一步的实验研究.地表含碳岩石在俯冲带被带到深部,俯冲带地温特征的不同导致了不同类型含碳流体的形成,这些流体运移至上覆地幔引起岩石部分熔融产生含碳的岛弧岩浆,岩浆喷出到地表释放了其中的碳,这构成了俯冲带-岛弧系统的碳循环.【期刊名称】《岩石学报》【年(卷),期】2019(035)001【总页数】10页(P89-98)【关键词】俯冲带;碳循环;含碳流体;多相包裹体;熔体包裹体【作者】刘景波【作者单位】中国科学院地质与地球物理研究所,岩石圈演化国家重点实验室,北京100029;中国科学院大学地球与行星科学学院,北京100049【正文语种】中文【中图分类】P542.5;P588.3俯冲带-岛弧系统的碳循环模式可以概括这样一种图景：含碳岩石通过俯冲过程的变质作用形成含碳流体，含碳流体运移至上覆地幔楔交代其中的岩石导致部分熔融产生含碳的岛弧岩浆，岩浆上升到地表将碳以CO2形式释放到地表的物质圈层中去。

岛弧火山作用释放的CO2在碳同位素组成上证明了这种过程的存在。

岛弧岩浆的δ13 C在-0.1‰～-11.6‰之间，这种成分的碳是俯冲带的碳酸盐（δ13 C＝0‰）、蚀变大洋玄武岩及其下覆地慢岩的碳（δ13 C＝-5‰）和俯冲岩石中的有机碳（δ13 C＝-30‰）混合的结果（Sano and Marty，1995；Shaw et al.，2003；De Leeuw et al.，2007）。

Co-digestion of food waste and sludge for hydrogenproduction by anaerobic mixed cultures:Statistical key factors optimizationChakkrit Sreela-or a ,Pensri Plangklang a ,Tsuyoshi Imai b ,Alissara Reungsang a ,c ,*aDepartment of Biotechnology,Faculty of Technology,Khon Kaen University,A.Muang,Khon Kaen 40002,Thailand b Division of Environmental Science and Engineering,Graduate School of Science and Engineering,Yamaguchi University,Yamaguchi 755-8611,Japan cFermentation Research Center for Value Added Agricultural Products,Khon Kaen University,Khon Kaen 40002,Thailanda r t i c l e i n f oArticle history:Received 20February 2011Received in revised form 22May 2011Accepted 24May 2011Available online 25June 2011Keywords:Bio-hydrogen Co-digestion Food waste SludgeResponse surface methodology Optimizationa b s t r a c tFactors affecting hydrogen production from the co-digestion of food waste and sludge in batch fermentation by anaerobic mixed cultures were optimized using Response Surface Methodology with Central Composite Design.Investigated parameters included C/N ratio,inoculums concentration,Na 2HPO 4concentration and Endo nutrient addition.The exper-iments were conducted in 120mL serum bottles with a working volume of 70mL.Results revealed that the optimum conditions were a C/N ratio of 33.14,inoculums concentration of 2.70g-VSS/L,Na 2HPO 4concentration of 6.27g/L and Endo nutrient addition of 7.51mL/L.Under the optimal conditions,a maximum hydrogen yield (HY)of 102.63mL H 2/g-VS added and specific hydrogen production rate (SHPR)of 59.62mL H 2/g-VSS h were obtained.C/N ratio and inoculums concentration showed the greatest individual and interactive effects on HY and SHPR (P <0.05).Endo nutrient addition also had an individual effect on SHPR (P ¼0.0124).The confirmation experiment under optimal condition showed an HY and SHPR of 101.14mL H 2/g-VS added and 59.43mL H 2/g-VSS h,respectively.This was only 1.01%and 1.00%,respectively,different from the predicted values.Copyright ª2011,Hydrogen Energy Publications,LLC.Published by Elsevier Ltd.All rightsreserved.1.IntroductionAs a sustainable energy,hydrogen is a promising alternative renewable energy and is considered as a clean and environ-mentally friendly energy.When hydrogen is combusted with oxygen,water is obtained as a by-product [1].Hydrogen has a high energy yield of 122kJ/g,which is 2.75times greater than that of hydrocarbon fuel [2].Hydrogen production from bio-logical processes can be divided into two types i.e.photo-fermentation by cyanobacteria,algae,photosynthetic and chemosynthetic e fermentative bacteria and dark fermenta-tion by anaerobic bacteria [3].Hydrogen production from the dark fermentation process has advantages over the photo-fermentation process which is a low operating cost because light is not required and the rate of hydrogen production is greater [4].In Thailand,the generation of food waste is about 20,041tons per day,which accounts for 50%of municipal solid waste*Corresponding author .Fermentation Research Center for Value Added Agricultural Products,Khon Kaen University,Khon Kaen 40002,Thailand.Tel./fax:þ6643362121.E-mail address:alissara@kku.ac.th (A.Reungsang).A v a i l a b l e a t w w w.s c i e n c e d i r e c t.c o mj o u r n a l h o m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /h ei n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 36(2011)14227e 142370360-3199/$e see front matter Copyright ª2011,Hydrogen Energy Publications,LLC.Published by Elsevier Ltd.All rights reserved.doi:10.1016/j.ijhydene.2011.05.145[5].Food waste consists mainly of starch,protein,and fat,witha small amount of cellulose and hemi-cellulose which are possible sources for bioenergy production[6].Due to its high organic content and its easily hydrolyzable nature,food waste is a good candidate to be used as the substrate for producing hydrogen by dark fermentation.However,food waste may lack of a nitrogen source which is a nutrient essential for hydrogen production[7].Therefore,a search for a nitrogen source to co-digest with food waste in order to achieve a maximum hydrogen production is needed.Recent research has been investigating the use of sludge as a substrate to produce or co-produce renewable energy i.e.hydrogen and methane by an anaerobic digestion process because of its high carbon and nitrogen content[8].Approximately160tons of sludge is generated daily in Thailand[9].Therefore,a utiliza-tion of sludge for production of hydrogen is one of the alter-nate approaches to reduce or get rid of this abundant amount of waste.Co-digestion of food waste and sludge to produce hydrogen has been reported and it was found that the addition of sludge to food waste supplied a more balanced carbon to nitrogen(C/N)ratio[10].The efficiency of hydrogen production is greatly influ-enced by environmental factors such as temperature,pH, nutrient,ferrous iron and substrate concentration[11].Co-digestion of food waste and sludge to produce hydrogen may encounter the problem of the inoculums being outgrown by normalflora in food waste and sludge.Therefore,the inoculums concentration needs to be optimized.The C/N ratio is important in a biological process.Microbes require a proper nitrogen supplement for metabolism during fermentation[12].A proper C/N ratio could enhance the bacterial productivity of hydrogen suggesting that nitrogen should be supplied at the optimal amount[13].The pH is also one of the factors controlling anaerobic biological processes.A buffer is required to reduce thefluctuation of pH during hydrogen fermentation because a formation of hydrogen is always accompanied by volatile fatty acids(VFAs)or solvents.A failure in pH control from the imbalances of alkalinity,pH and VFAs concentration might result in an interruption in hydrogen production[13]and inhibit the growth of hydrogen producers[14].Therefore,the addition of buffer at a suitable concentration to counteract a decrease in pH would remove this limitation.Another important envi-ronmental factor affecting hydrogen production is nutrients. Hydrogen producing bacteria need nutrients for cell biomass and metabolites production[13].In addition,nutrient can increase microbial activity so that they can effectively convert soluble organic matters into hydrogen[15].From the aforementioned reasons,it can be seen that in order to efficiently produce hydrogen there is a need to optimize these environmental factors.However,it is labo-rious and time consuming to perform the optimization by a conventional technique or known as“a one factor at a time”method[16].A statistical experimental design response surface methodology(RSM)can eliminate this limitation[17].It is not only a time saving method but also can minimize the error in determining the effects of parameters as well as be able to demonstrate the interactive effects among the tested variables[18].In this study,RSM with central composite design(CCD)was used to study the effects of C/N ratio,inoculums concentration,Na2HPO4 concentration and Endo nutrient addition on hydrogen yield (HY)and specific hydrogen production rate(SHPR).CCD has the advantages in term of rotability and the ability to analyze all the quadratic and interaction effects[19].In addition,CCD offers5level of factors(i.e.Àa,À1,0,1,þa) which facilitating thefindings of the optimal point wider than the other designs[19].Statistical optimization design on bio-hydrogen produc-tion has recently been reported in literatures[20e24].For instance,Saraphirom and Reungsang[11]studied the factors affecting bio-hydrogen production from sweet sorghum syrup by anaerobic mixed cultures using CCD.Lee et al.[21]inves-tigated the effects of temperature,pH and starch concentra-tion on fermentative hydrogen production from starch by mixed anaerobic microflora using RSM.Pan et al.[22]inves-tigated process parameters on bio-hydrogen production from glucose by Clostridium sp.Fanp2using Plackett e Burman design and Box e Behnken design.O-Thong et al.[24]investi-gated factors affecting hydrogen production from palm oil mill effluent(POME)under thermophilic condition using an RSM with CCD.Although many studies have been done on the effect of environmental factors on hydrogen production from various kinds of synthetic substrates and wastes but the information on the statistically optimization of environ-mental factors on bio-hydrogen production from a co-diges-tion of food waste and sludge by anaerobic mixed cultures are still lacking.Therefore,this research attempted to optimize C/N ratio, inoculums concentration,Na2HPO4concentration and Endo nutrient addition using the RSM with CCD in order to maxi-mize an HY and the SHPR.The information obtained from this study could pave the way toward a scaling up of the hydrogen production process and/or a continuous hydrogen fermentation process from a co-digestion of food waste and sludge.2.Materials and methods2.1.Preparation of feed stocksFood waste was collected from a cafeteria on the Khon Kaen University campus,Khon Kaen,Thailand.The waste was made up of rice,vegetables,fruits and meats.Bones were removed from the food waste before being mixed with tap water at the volumetric ratio of1:3;it was then ground in a food blender.The pH of the resulting food waste slurry was7.2.Chemical characteristics of the food waste are shown in Table1.The resulting food waste slurry was stored atÀ17 C and thawed in the refrigerator before being used.The sludge was taken from the dissolved airflotation tank of the wastewater treatment plant of a food company in the Northeastern part of Thailand.Dissolved airflotation tank is used as a pretreatment to remove the debris as well as sludge from the factory before the influent is sent to the wastewater treatment plant.The company produces5000tons/day of frozen meat products and1600tons/day of ready meals.The plant handles an average6500m3of wastewater daily andi n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y36(2011)14227e14237 14228generates70e80tons of sludge per day.Chemical character-istics of the sludge are shown in Table1.The collected sludge was stored at4 C prior the usage.2.2.Seed sludge and preparation of inoculumsAnaerobic seed sludge was obtained from the full-scale anaerobic digester of an Upflow Anaerobic Sludge Blanket (UASB)reactor,which belonged to a brewery company in the Northeastern part of Thailand.The UASB is used to produce methane from the wastewater of the beer production process. The collected sludge was pre-heated at105 C for3h in a drying oven(LDO-100E)in order to deactivate methanogens which are hydrogen consumers.The pH and volatile sus-pended solids(VSS)concentration of the sludge were6.8and 7.4g/L,respectively.For inoculums preparation,pre-heated sludge was culti-vated in food waste at20g-COD/L supplemented with0.5mL/ L of Endo nutrient solution[25].The culture was shaken at 150rpm for36h at30Æ2 C before being used as the inocu-lums in the batch experiment.After36h,the seed sludge was centrifuged at12,000rpm for5min,then analyzed for biomass concentration(in terms of VSS)and used as inocu-lums at various concentrations according to the design.Endo nutrient solution used in the bio-hydrogen production experiment was slightly modified in which it did not contain NaHCO3.The composition of the Endo nutrient was as follows (all in mg/L):5240NH4HCO3,125K2HPO4,100MgCl2$6H2O,15 MnSO4$6H2O,25FeSO4$7H2O,5CuSO4$5H2O,and0.125 CoCl2$5H2O.NH4HCO3in the Endo nutrient solution was used as the initial alkalinity.2.3.Bio-hydrogen productionBio-hydrogen production experiment was conducted in 120mL serum bottles with a working volume of70mL.The fermentation broth contained different C/N ratio and concentrations of variables according to the design.Total COD and total N were used to represent the amount of carbon and nitrogen,respectively.In order to obtain substrate ingredi-ents,food waste and sludge were mixed at various ratio so that thefinal C/N ratio were10:1,20:1,30:1,40:1and50:1.The serum bottles wereflushed with nitrogen gas to remove oxygen in the headspace of the bottles in order to create the required anaerobic condition and capped with rubber stop-pers.The bottles were incubated at room temperature (30Æ2 C)and operated in an orbital shaker with a rotation speed of150rpm.At designed time,the total gas volume was measured by releasing the pressure in the bottles using wetted glass syringe[26]and then analyzed for gas content by gas chromatography equipped with a thermal conductivity detector(TCD).Effluent was collected by using a glass syringe and analyzed for VFAs and alcohol by gas chromatography equipped with aflame ionization detector(FID).All treat-ments were conducted in four replications.The hydrogen production was continued until the biogas volume could not be measured.2.4.Experimental design and data analysisCCD was used to study the effects of C/N ratio,inoculums concentration,Na2HPO4concentration and Endo nutrient addition on HY and SHPR.The experiments were designed by the Design Expert version7.0Òsoftware,Stat-Ease Inc.,MN, USA.The ranges and levels of independent input variables are shown in Table2.The HY and SHPR were selected as the dependent output variables.For statistical calculations,the test factors(X i)were coded as x i according to the following transformation equation(Eq.(1)):x i¼ðX iÀX0Þ=D X i;(1)where x i is the coded value of the variable X i,X i is the actual value of the i th independent variable,X0is the actual value of X i at the center point and D X i is the step change value.A quadratic model(Eq.(2))[27]was used to evaluate the opti-mization of C/N ratio,inoculums concentration,Na2HPO4 concentration and Endo nutrient addition:Y i¼b0þXb i x iþXb ii x2iþXb ij x i x j(2) where Y i is the predicted responses,x i is the parameters,b0is a constant,b i is the linear coefficients,b ii is the squared coef-ficients,and b ij is the cross-product coefficients.The response variables(Y HY¼the HY response and Y SHPR¼the SHPR response)werefitted using a predictive polynomial quadratic equation(Eq.(2))in order to correlate the response variables to the independent variables[28].The Y HY and Y SHPR values were regressed with respect to C/N ratio(X1),inoculums concen-tration(X2),Na2HPO4concentration(X3)and Endo nutrient addition(X4).Table3illustrates the coded values of the vari-ables,the experimental design and the corresponding results.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y36(2011)14227e14237142292.5.Analytical methodsBiogas composition was measured by a gas chromatography (GC-2014,Shimadzu)equipped with a TCD and a 2m stain-less column packed with Unibeads C (60/80mesh).The operational temperatures of the injection port,the column oven and the detector were 150,145and 150 C,respectively.Argon was used as the carrier gas at a flow rate of 25mL/min.For VFAs and alcohols analysis,the collected effluents were first centrifuged at 6000rpm for 10min then acidified by 0.2N oxalic acid and finally filtered through a 0.45m cellu-lose acetate membrane.The same GC model with an FID and a 30m Â0.25mm Â0.25m capillary column (Stabiwax)was used to analyze the VFAs and alcohols concentrations.The temperatures of the injector and detector were 250 C.The initial temperature of the column oven was 50 C for 2min;this was followed by a ramp of 15 C/min for 12.6min and then raised to a final temperature of 240 C for 1min.Helium was used as the carrier gas with a flow rate of 66mL/min.Concentrations of total nitrogen,total phosphate,magne-sium,manganese,iron,copper,sodium,cobalt,VSS,and vola-tile solid (VS)were measured using the procedures described in standard methods [29].Hydrogen gas production was calcu-lated from the headspace measurement of gas composition andthe total volume of hydrogen produced,at each time interval,using the mass balance equation (Eq.(3))[30]:V H ;i ¼V H ;i À1þC H ;i ÀV G ;i ÀV G ;i À1ÁþV H ;0ÀC H ;i ÀC H ;i À1Á(3)where V H,i and V H,i À1are the cumulative hydrogen gas volumes at the current (i )and previous time interval (i À1),respectively;V G,i and V G,i À1are total biogas volume at the current and previous time interval (i À1);C H,i and C H,i À1are the fraction of hydrogen gas in the headspace at the current and previous time interval (i À1)and V H is the volume of the headspace of the serum bottles (50mL).2.6.Kinetic analysisThe modified Gompertz equation (Eq.(4))was used to deter-mine the cumulative hydrogen production [31].H ¼P exp &Àexp R m eðl Àt Þþ1!'(4)where H is the cumulative volume of hydrogen produced (mL),R m is the maximum hydrogen production rate (mL/h),l is the lag-phase time (h),t is the incubation time (h),P is the hydrogen production potential (mL)and e is 2.718281828.Parameters (P ,R mi n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 36(2011)14227e 1423714230and l)were estimated using the solver function in Microsoft Excel version5.0(Microsoft,Inc.)as explained by Khanal et al.[32].3.Results and discussion3.1.The effects of C/N ratio,inoculums concentration, Na2HPO4concentration and Endo nutrient addition on HYThe effects of independent variables i.e.,C/N ratio(X1),inoc-ulums concentration(X2),Na2HPO4concentration(X3)and Endo nutrient addition(X4)on HY obtained from the co-digestion of food waste and sludge were investigated. Regression analysis of the data from Table3resulted in the quadratic equation(Eq.(5))as follows:Y HY¼96:63þ5:321X1þ15:95X2þ0:43X3À0:19X4þ5:68X1X2þ0:29X1X3þ0:12X1X4þ0:89X2X3À0:23X2X4À0:54X3X4À16:06X21À13:28X22À8:17X23À8:47X24ð5ÞThe ANOVA of the model(Table4)indicated that the model significantly represents the experimental data(P<0.0001).A high determination coefficient(R2)of0.99suggested that the model can explain99%variability of the response variables.In addition,the lack offit of the model was insignificant (P¼1.0000).These results indicated that the effects of inde-pendent variables on HY in this study can be well described by the obtained models.The ANOVA of the model also showed that the quadratic model terms of all variables and the linear model terms of the C/N ratio and inoculums concentration are highly significant(P<0.0001).This indicates that these terms greatly affect the HY.Only the interaction model term of C/N ratio and inoculums concentration had impact on HY indi-cating by the P value less than0.05.Based on the regression analysis of the model(Eq.(5)),the maximum HY of102.63mL H2/g-VS added could be predicted at the optimum conditions of33.14C/N ratio,2.70g-VSS/L inoc-ulums concentration, 6.27g/L Na2HPO4concentration and 7.51mL/L Endo nutrient addition.The response surface plots based on Eq.(5),with two variables kept constant at their optimum values and varia-tions of the other two variables within the experimental range,are depicted in Fig.1.Fig.1a e f is plotted with Na2HPO4 concentration and Endo nutrient addition,inoculums concentration and Endo nutrient addition,inoculums concentration and Na2HPO4concentration,C/N ratio and Endo nutrient addition,C/N ratio and Na2HPO4concentration and C/N ratio and inoculums concentration,respectively,being kept constant.Each response surface plot has a clear peak which suggested that the optimum condition fell well inside the design boundary(Fig.1).HY significantly increased with the increase in the C/N ratio from20to33.14and then HY decreased when the C/N ratio was greater than33.14(Fig.1a e c).The C/N ratio is important in the biological processes by affecting hydrogen fermentation effi-ciency.A suitable C/N ratio could enhance microbial growth and substrate utilization,thus improving the hydrogen production efficiency[33].It is normally found that microor-ganisms utilize carbon25e30times faster than nitrogen.To meet this requirement,a C/N ratio of20e30:1is needed[12]. Our results showed a similar trend in which the optimum C/N ratio was33.14.The optimum C/N ratio varied depending on types of substrate and microorganism used.A C/N ratio of74 was optimum for thermophillic hydrogen production from POME by acclimatized Thermoanaerobacterium-rich sludge[24]i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y36(2011)14227e1423714231while the hydrogen production from sucrose by anaerobic mixed cultures was optimized at the C/N ratio of 47[34].Without a nitrogen source,hydrogen could not be produced by Clostridium butyricum [12].Under an excess of N source (low C/Nratio)condition,the substrate was mostly used for cell growth resulting in a low HY.In contrast,the excess of C/N ratio could decrease hydrogen production efficiency because low nitrogen contents are deficient for cell growth [13].Fig.1e Response surface plots showing the effects of C/N ratio,inoculums concentration,Na 2HPO 4concentration and Endo nutrient addition on HY.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 36(2011)14227e 1423714232Fig.2e Response surface plots showing the effects of C/N ratio,inoculums concentration,Na 2HPO 4concentration and Endo nutrient addition on SHPR.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 36(2011)14227e 1423714233The HY increased when the inoculums concentration increased from 1.48to 2.70g-VSS/L.However,a further increased in inoculums concentration to greater than 2.70g-VSS/L resulted in a decrease in HY (Fig.1a,d and e).Using an inadequate inoculums concentration,the hydrogen producers in the seed inoculums might not be capable of competing with indigenous microflora in the substrate.The indigenous microflora might become dominant and produce the products that do not relate to hydrogen production or that might inhibit the hydrogen producers thus reducing the HY.In addition,the increase in inoculums concentration to a concentration greater than the optimum value might cause a rapid drop of pH in the fermentation broth.This would make the microor-ganisms more favorable for solvent production (second growth phase)and stop the first metabolites production (acids and gases)[35].Therefore,the inoculums concentration should be compatible with the available substrate for maxi-mizing bacterial activity.The effects of the Na 2HPO 4concentration on HY were not significant (P ¼0.4377)(Table 4).The HY increased when Na 2HPO 4concentration increased from 4.00to 6.27g/L indi-cating a positive effect of Na 2HPO 4concentration on HY (Fig.1b,d and f).Na 2HPO 4could improve the hydrogen production due to its buffering capacity because it can reduce the pH fluctuation caused by VFAs accumulated in the fermentation broth,thus enhancing the hydrogen generation and acidogenesis in the first stage of an acid-gas digestion system [36,37].In addition,the P element in Na 2HPO 4is essential for the synthesis of many molecular substances such as DNA,RNA and ATP [38].A further increase in Na 2HPO 4concentration to greater than 6.27g/L resulted in a decrease in HY which may be due to an increase in cytoplasmic osmotic pressure that occurs at high Na 2HPO 4concentration [39].An increase in Endo nutrient addition from 5.00to 7.51mL/L slightly increased the HY (Fig.1c,e and f).Endo nutrient contains elements that are essential for cell synthesis and microbial growth such as Fe,Co 2þ,Cu 2þ,Mg 2þand Mn 2þ.Iron is the most important hydrogen production related trace element because it forms ferredoxin and hydrogenase,which directly relate to the hydrogen production process [36].Co 2þ,Cu 2þ,Mg 2þand Mn 2þare known as the enzyme cofactor [37].NH 4HCO 3in the Endo nutrient can prevent the fluctuation of the pH during hydrogen fermentation [40].However,the results indicate that HY was decreased when the Endo nutrient addition was greater than 7.51mL/L.This may be due to the dissolution of bicarbonate in NH 4HCO 3could increase CO 2in the system which therefore decreases the hydrogencontent in the gas phase.In addition,the high ammonium concentration can cause adverse effect on microorganisms [15].Moreover,Fe and Cu at high concentrations could inhibit and reduce the activity of hydrogen producers [41,42].3.2.The effects of C/N ratio,inoculums concentration,Na 2HPO 4concentration and Endo nutrient addition on SHPREffects of C/N ratio (X 1),inoculums concentration (X 2),Na 2HPO 4concentration (X 3)and Endo nutrient addition (X 4)on SHPR (Y SHPR )during the co-digestion of food waste and sludge were examined.The observed and predicted values of SHPR are presented in Table 3.Multiple regression analysis was applied on the data in Table 3and the obtained second-order polynomial equation (Eq.(6))could well explain the SHPR.Y SHPR ¼55:97þ4:55X 1þ8:32X 2þ3:06X 3þ0:39X 4þ3:26X 1X 2À2:05X 1X 3þ1:04X 1X 4þ1:36X 2X 3À0:52X 2X 4À0:24X 3X 4À9:28X 21À7:59X 22À9:86X 23À5:57X 24ð6ÞIt is evident from ANOVA (Table 4)that the quadraticregression model of Y SHPR was highly significant with a low probability (P <0.0001)and fit the experimental data well with the R 2value of 0.95accompanied by an insignificant lack of fit model (P ¼0.0962).The 3variables i.e.,C/N ratio,inoculums concentration and Na 2HPO 4concentration were significant effect on SHPR (P <0.05)whereas Endo nutrient addition had insignificant effect on SHPR (P ¼0.7211).Results demonstrated that the only significant interaction effect on SHPR was found between C/N ratio and inoculums concentration (P ¼0.0259).The analysis of Eq.(6)provided the optimum conditions for SHPR:a C/N ratio of 33.14,an inoculums concentration of 2.70g-VSS/L,a Na 2HPO 4concentration of 6.27g/L and an Endo nutrient addition of 7.51mL/L.The maximum SHPR of 59.62mL H 2/g-VSS h was obtained under these optimum conditions.The response surface plots base on Eq.(6)with two vari-ables being kept constant at their optimum values and vari-ations of the other 2variables within the experimental range are depicted in Fig.2.Fig.2a e f is plotted with Na 2HPO 4concentration and Endo nutrient addition,inoculums concentration and Endo nutrient addition,inoculums concentration and Na 2HPO 4concentration,C/N ratio and Endo nutrient addition,C/N ratio and Na 2HPO 4concentration and C/N ratio and inoculums concentration,respectively,being kept constant.Results indicated a clear peak for each response surface plot (Fig.2)which suggested that the optimum condition fell well inside the design boundary.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 36(2011)14227e 1423714234Fig.2a e c revealed that the SHPR increased significantly with an increase in the C/N ratio from 20to 33.14.Conversely,the SHPR decreased with the further increase of C/N ratio to over 33.14.The SHPR increased when the inoculums concentration was increased from 1.48to 2.70g-VSS/L and then decreased slightly following a further increase in inoculums concentra-tion (Fig.2a,d and e).An increase in Na 2HPO 4concentration from 4.00to 6.27g/L resulted in an increase in the SHPR to the maximum value.The SHPR decreased when the Na 2HPO 4concentration was greater than 6.27g/L (Fig.2b,d and f).Results showed in Fig.2c,e and f indicated that the SHPR increased when the Endo nutrient addition was increased from 5.00to 7.51mL/L.A further increase in Endo nutrient addition to greater than 7.51mL/L resulted in a decrease of the SHPR.3.3.Optimization and confirmation of the experimentsThe analysis of Y HY (Eq.(5))and Y SHPR (Eq.(6))models indicated that in order to simultaneously obtain the maximum HY and SHPR,the C/N ratio,inoculums concentration,Na 2HPO 4concentration and Endo nutrient addition should be optimized at 33.14, 2.70g-VSS/L, 6.27g/L and 7.51mL/L,respectively (Table 5).Under these optimal conditions,the model predicted an HY of 102.63mL H 2/g-VS added and an SHPR of 59.62mL H 2/g-VSS h In order to confirm the validity of the statisticalexperimental strategy,three replications of batch experiments were conducted under optimal,medium (runs 4,9,11,17,18,28),high (run 5)and worst (run 14)conditions (Table 5).The results of the confirmation experiments are in close agreement with the predicted values of HY and an SHPR.An HY of 101.14mL H 2/g-VS added and SHPR of 59.43mL H 2/g-VSS h were obtained at the optimum condition.These results ensured that the obtained models are satisfactory and accurate.Time (h)20406080100120140C u m u l a t i v e h y d r o g e n p r o d u c t i o n (m L )0100200300400500600700H y d r o g e n y i e l d (m L H 2/g -V S a d d e d )20406080100120140Fig.3e Cumulative hydrogen production and HY in a confirmation experiment at optimum condition.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 36(2011)14227e 1423714235。

A convenient one-pot synthesis of 2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H -chromen-5(6H )-one derivatives and their further transformationsShaodi Song a ,Liping Song a ,b ,*,Baifan Dai a ,Hai Yi a ,Guifang Jin b ,Shizheng Zhu b ,*,Min Shao caDepartment of Chemistry,School of Science,Shanghai University,No.99,Shangda Road,Shanghai 200444,ChinabKey Laboratory of Organofluorine Chemistry,Shanghai Institute of Organic Chemistry,Chinese Academy of Sciences,354Fenglin Lu,Shanghai 200032,China cInstrumental Analysis and Research Center,Shanghai University,Shanghai 200444,Chinaa r t i c l e i n f oArticle history:Received 24January 2008Received in revised form 2April 2008Accepted 7April 2008Available online 9April 2008Keywords:One-pot multi-component reaction Fluorinated heterocycles Trifluoromethyl-1,3-dione Dehydrationa b s t r a c tThe trifluoromethyl containing heterocycles,2-hydroxy-4-aryl-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H -chromen-5(6H )-one derivatives 4,were synthesized via a one-pot three-compo-nent reaction of aldehyde 1with 1,3-cyclohexanedione 2and 4,4,4-trifluoro-1-(thien-2-yl)butane-1,3-dione 3in the presence of a catalytic amount of Et 3N.The effect of bases and solvents on the reaction efficiency and yield was briefly investigated.Treatment of 4with an excess amount of NH 4OAc in ethanol afforded 2-trifluoromethyl-1H -quinolin-5-one derivatives 5.Refluxing of 4with TsOH in CHCl 3gave the corresponding dehydrated products 8.Ó2008Elsevier Ltd.All rights reserved.1.IntroductionFluorine-containing compounds have attracted much interest because of their unique chemical,physical,and biological activi-ties.1–3In particular,fluorine-containing heterocycles are now widely recognized as important organic molecules showing in-teresting biological activities with potential for applications in the medicinal and agricultural fields.4,5The reactions of fluorinated 1,3-dicarbonyl compounds as fluorine-containing building blocks have been investigated extensively,6,7and well established as synthetic intermediates in heterocyclic chemistry.8In recent years,the syn-thesis of fluorinated heterocyclic compounds has drawn much attention.9Furthermore,the multi-component reactions (MCRs),by virtue of their convergence,productivity,facile execution,and generally high yields of products,have attracted much attention from the vantage point of combinatorial chemistry.10For example,the syn-thesis of polyhydroquinoline derivatives was the classical MCRs involving the three-component coupling of an aldehyde,ethyl acetoacetate,and ammonia in acetic acid or refluxing alcohol.11More recently,several alternate and more efficient methods have been developed for MCRs to polyhydroquinoline derivatives by using microwave,ionic liquid,TMSCl–NaI,metal triflates,I 2,ceric ammonium nitrate,polymer,organo-catalyst,solvent-free,andcatalyst-free reactions.12In order to evaluate the potential appli-cations of MCRs in the field of organofluorine chemistry and to continue our ongoing study on the synthesis of fluorine-containing heterocyclic compounds via MCRs based on the trifluoromethyl-1,3-dicarbonyl compounds,a versatile fluorine-containing building blocks,herein,we wish to report a one-pot,three-component reaction to 2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H -chromen-5(6H )-one derivatives,together with their further chemical transformations including O -heterocyclics to N -heterocyclics interconversion and dehydration reactions of a -hydro-dihydropyrane moieties to 4-H pyrane derivatives.2.Results and discussionInitially,we carried out the one-pot,three-component reaction of benzaldehyde 1a and 1,3-cyclohexanedione 2with 4,4,4-tri-fluoromethyl-1-(thien-2-yl)-butane-1,3-dione 3in the presence of a catalytic amount of triethylamine (10%)in ethanol at room tem-perature,however,no reaction occurred.Subsequently,the mixture was heated to reflux,after stirring for 2h,TLC showed that the reaction proceeded smoothly and gave the product 4a in 60%yield (Scheme 1).Based on the results above,the reaction conditions were opti-mized to improve the yield by changing bases and solvents.The effect of a variety of organic bases and the amount on the reaction efficiency and yield was firstly screened (Table 1).As shown in Table 1,when the reaction was performed in the absence of base,it only*Corresponding authors.E-mail addresses:lpsong@ (L.Song),zhusz@ (S.Zhu).Tetrahedron 64(2008)5728–5735Contents lists available at ScienceDirectTetrahedronjournal homepage:/locate/tet0040-4020/$–see front matter Ó2008Elsevier Ltd.All rights reserved.doi:10.1016/j.tet.2008.04.020afforded product 4a in 10%yield (entry 1,Table 1),however,when 0.25equiv of Et 3N was used,the yield was improved significantly (entry 3,Table 1).The reaction took much longer reaction time to give the 4a in 40%yield when pyridine was involved as catalyst (entry 8,Table 1).Other organic bases such as DABCO,DMAP,and piperidine catalyzed the reactions giving almost the same results as the triethylamine catalyzed reaction in terms of reaction time and yields (entries 5–7,Table 1).Higher amounts of catalyst did not further improve the yield of 4a (entry 4,Table 1).Solvent effect was the next considered factor.As shown in Table 1,generally,the reactions gave the better yields in polar protic solvents such as MeOH or EtOH than that in polar aprotic solvents (entries 7–11,Table 1),even though the latter reactions were carried out in prolonged reaction time.Moreover,the reaction in absolute ethanol under the nitrogen atmosphere gave the same product yield as that in the commercially available ethanol,in-dicating that the trace of water in ethanol did not accelerate the reaction obviously (entries 12and 13,Table 1).With the optimal results in hand as shown in Table 1,entry 3,we investigated the scope and limitation of this one-pot,three-com-ponent reaction with a variety of aldehydes,and the appropriate 4were generally obtained in moderate to good yields (Scheme 1).The reaction results are summarized in Table 2.On one hand,the aryldehydes bearing either electron-donating or electron-with-drawing group have not shown the obvious effort on the formation of the expected products.On the other hand,the substituents lo-cated at para or meta positions of aryldehydes have not shown the much effort on the formation of products,too (entries 8and 9,Table 2).Prolonged reaction time from 2h to 24h slightly increased the yield (entries 2and 3,Table 2).However,the steric effect of sub-stituent was noticeable.For example,ortho substituted aryldehydes such as 2-methoxyl-benzaldehyde or 2-bromo-benzaldehyde gaveno expected product,TLC analysis showed that the starting material remained.Similarly,steric aromatic hindered aryldehydes such as 1-naphthaldehyde 13or 3,4-dimethoxyl-benzaldehyde,gave no expected product,too.The above results have shown that the re-activity of the aryldehyde differs significantly depending on the steric effect,other than the nature of electronic effect of the sub-stituents.Other aldehydes,either the furan-2-carbaldehyde or the aliphatic aldehydes such as acetaldehyde and cinnamic aldehyde were also invested in the reaction,unfortunately,there was no expected product obtained.It should be noted that in the case of 4d (entry 5,Table 2),the yield was lower than that of other arylde-hydes,even after modification of the reaction conditions like prolonged reaction time.The structures of compounds 4were fully confirmed by 1H NMR,19F NMR,MS,IR spectroscopies and elemental analysis.For in-stance,the characteristic features of the 1H NMR in CDCl 3spectra of 4b were the appearances of doublets at d 4.02and 3.80ppm with J H–H ¼11.7Hz for 3-H and 4-H protons,respectively,indicating that a trans configuration of the vicinal two hydrogen atoms.The chemical shift of CF 3group in 19F NMR was a singlet peak at d À80.66ppm (s,3F),which indicated that the CF 3group was bonded to a quaternary carbon atom.In most cases,the products 43EtOH, Ref.Ar OS O OArCHOCF 3OH 1(a-j)234(a-j)F 3CS O OAr = C 6H 5p -CH 3C 6H 4 p -CH 3OC 6H 4 p -HOC 6H 4p -(Me)2NC 6H4p -ClC 6H 4p -NO 2C 6H 4m -NO 2C 6H 4m -PhOC 6H 4m -BrC 6H 4f g h i ja b c d e Scheme 1.Table 2Reaction results of the one-pot,three-component reaction of aryldehyde 1and 1,3-Reaction conditions:aryldehyde 1(1.5mmol),1,3-cyclohexanedione 2(1.5mmol),4,4,4-trifluoro-1-(thien-2-yl)butane-1,3-dione 2(1.5mmol),Et 3N (25%equiv),EtOH (15mL),refluxing.bIsolated yield.Table 1Reaction conditions:aryldehyde 1(1.5mmol),1,3-cyclohexanedione 2(1.5mmol),4,4,4-trifluoro-1-(thien-2-yl)butane-1,3-dione 3(1.5mmol),solvent:15mL,refluxing.bIsolated yield.cWithout base.dThe reaction was carried out in freshly distilled absolute EtOH under nitrogen.Table 3Reaction conditions:1a (1.5mmol),2(1.5mmol),3(1.5mmol),EtOH (12mL),refluxing.bIsolated yield.S.Song et al./Tetrahedron 64(2008)5728–57355729had poor solubility in CDCl 3,therefore,most of 1H NMR spectra of products 4were recorded in DMSO-d 6,however,the characteristic peaks for 3-H and 4-H in DMSO-d 6somehow revealed one broad peak in the range of d 4.1ppm.More recently,ammonium acetate has been used widely as a base or a catalyst in Hantzsch reactions as well as other re-actions.14With this aim in view,we applied ammonium acetate to the above one-pot reaction.It was hoped that the expectedproducts 4would be obtained in the presence of catalytic amount of NH 4OAc,whereas the polyhydroquinoline derivatives should be formed in the presence of stoichiometric amount of NH 4OAc.Thus,we examined the amount of NH 4OAc affecting the formation of products and the yields.The reaction results are listed in Table 3.As expected,the reaction occurred smoothly under refluxing in ethanol in the presence of catalytic amount of NH 4OAc (0.25equiv)and afforded 4a as the major product in 65%yield,along with minorNH 4OAcF 3C O OSOO O CF 3OH +S N HO O CF 3OH S CHO1a 23Scheme 2.Figure 1.Crystal structure (a)and packing map (b)of 5a .S.Song et al./Tetrahedron 64(2008)5728–57355730Hantzsch product 5a in 10%yield (Scheme 2).The low yield of 5a was caused by the insufficiency of NH 3source in reaction medium (entry 1,Table 3).Thus,NH 4OAc was increased to stoichiometric amount,it was noticed that,although the higher amount of NH 4OAc led to the improvement in the yield of Hantzsch product 5a from 10%to 33%,the major product was still 4a (entry 2,Table 3).This result was in agreement with the previous reported work,15in which initial formation of the Hantzsch product appeared low due to large quantities of the intermediate pyran derivative formed instead.Furthermore,the pyran derivatives could be converted by either in situ or a stepwise reaction 16to the corresponding Hantzsch pyridine derivatives.Inspired by the above works,the one-pot reaction was carried out in the presence of 30equiv of NH 4OAc,as expected,the reaction gave the product 5a in 55%yield exclusively,without the formation of 4a .This result suggested that the O-atom ring to N-atom ring interconversion could be achieved by treatment of 4with an excess of NH 4OAc.To verify the sugges-tion,we carried out the ring to ring interconversion reaction by treatment of 4a with an excess of NH 4OAc in refluxing ethanol (Scheme 2).As expected,5a was obtained in 67%yield,whose structure was further confirmed by XRD analysis.It was unambiguous to observe the trans relationship between the vicinal six-membered ring protons in compound 5a .The molecular structure and packing map of 5a are shown in Figure 1.Several selected bond lengths and selected bond angles are listed in Table 4.The crystal date and refinement details are listed in Table 5.The structure solution revealed that the proton attached to oxygen O2of hydroxyl group formed a intramolecular H-bond with the oxygen O3of carbonyl group (D (O3/H2A)¼2.057Å;:(O2–H2A /O3)¼143.70 ).In addition,in the crystal,each of the two molecules were related by two intermolecular H-bonds involving the protons attached nitrogen N1and carbon C7with the oxygen O1of carbonyl group,respectively,(D (O1/H1A)¼2.032Å;:(N1–H1A /O1)¼154.36 ;D (O1/H7A)¼2.510Å;:(O1/H7A–C7)¼139.21 ).We also examined the one-pot,three-component reaction with other 1,3-dicarbonyl compounds such as fluorinated substrates (6a ,6b )or non-fluorinated substrates (6c ,6d ),and similar results were found with these compounds (Scheme 3).The reaction results are summarized in Table 6.However,with 1,3-diketones yields were somewhat lower than the corresponding b -oxo ester (entries 1–4,Table 6).It is very typical for cyclic CF 3-compounds containing hemi-ketal,himi-aminal or even himi-amidal fragment because of the strong electron-withdrawing properties of CF 3group.15How-ever,it should be noted that in our cases of non-fluorinated sub-strates (6c ,6d ),the similar hemi-ketal fragments were also formedinstead of the formation of 4-H pyran moiety (entries 3and 4,Table 6).Finally,we studied the dehydration of compounds 4and 7c .In contrast to the previous works,7b,c in which the a -hydro-dihydropyrane moieties resisted to eliminate water to form the corresponding dehydrated products,compounds 4and non-fluo-rinated analogue 7c were smoothly eliminated water to form the corresponding 4-H pyrane derivatives in the presence of an excess of TsOH in boiling CHCl 3in good yields (Scheme 4,Table 7).The structure of compound 8b was further confirmed by XRDanalysis.Table 4Table 5Et 3N, 25%R 1R 2OOPh OR 2OO CHOR 1OH 7(a-d)126(a-d)a b c R 1 = CF 3, R 2 = Me, R 2 = OEt, R 1= CH 3, R 2 = Me, R 2 = OEt, dScheme 3.Table 6Reaction conditions:1(1.5mmol),2(1.5mmol),6(1.5mmol),Et 3N (0.25equiv),EtOH (12mL),refluxing.bIsolated yield.S.Song et al./Tetrahedron 64(2008)5728–57355731The crystal structure is shown in Figure 2.The selected bond lengths and selected bond angles are listed in Table 4.3.ConclusionsIn conclusion,we have developed a one-pot reaction for the synthesis of 2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H -chromen-5(6H )-one derivatives 4from easily available starting materials.Under the same reaction conditions,non-fluorinated substrates afforded the similar himi-ketal moieties.Besides the organic base catalysts,NH 4OAc catalyzed reaction also gave the same reaction results.However,in the presence of an excess of NH 4OAc,the one-pot reaction directly afforded 2-(trifluoromethyl)-1,2,3,4,7,8-hex-ahydroquinolin-5(6H )-one derivative 5,which could also be obtained from 4by O-atom ring to N-atom ring interconversion reaction.Meanwhile,the further chemical transformation in-cluding dehydration of a -hydrodihydropyrane derivatives 4and 7c with p -TsOH in refluxing CHCl 3proceeded smoothly and gave the corresponding fluorinated and non-fluorinated products 8,respectively.4.Experimental section 4.1.GeneralMelting points were measured with digital melting point ap-paratus (WRS-1B,Shanghai precision &scientific instrument Co.,Ltd.)and were uncorrected .1H and 19F NMR spectra were recorded in DMSO-d 6(unless mentioned in text)on Bruker AM-300or AM-500instruments with Me 4Si and CFCl 3(with upfield negative)as the internal and external standards,respectively.IR spectra were obtained with a Nicolet AV-360spectrophotometer.Lower resolu-tion mass spectrum were determined with Finnigan GC–MS 4021using the electron impact ionization technique (70eV).High res-olution mass spectra (HRMS)were run on Ionspec 4.7Tesla FTMS using MALDI/DHB.Elemental analyses were performed in thisInstitute.Table 7Reaction conditions:4a ,4b ,7c (0.5mmol),p -TsOH (2.0mmol,4equiv),solvent:CHCl 3(15mL),refluxing.bIsolated yields.Ar OR 2O O 1OHTsOH (4 eq.)3O R 2O O R 1Ar R 1 = CF 3, R 2 = Thien-2-yl, Ar = C 6H 5, a R 1 = CF3, R 2 = Thien-2-yl, Ar = p -MeC 6H 4, b R 1 = CH 3, R 2 = CH 3, Ar = C 6H 5, c4a, 4b, 7c8(a-c)Scheme 4.Figure 2.Crystal structure (a)and packing map (b)of 8b .S.Song et al./Tetrahedron 64(2008)5728–573557324.2.General procedure for the preparation of4-aryl-2-hydroxy-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4a–4j)To a mixture of aryldehyde1(1.5mmol),1,3-cyclohexanaedione 2(168mg,1.5mmol),and4,4,4-trifluoro-1-(thien-2-yl)butane-1,3-dione3(333mg,1.5mmol)in12mL EtOH was added0.3mmol of Et3N as catalyst under stirring at room temperature.The mixture was refluxed and continuously stirred for specified hour(moni-tored by TLC).After cooling,the resulting solid wasfiltered,washed with10–15mL of cold EtOH and air-dried to afford the crude product.The pure product was obtained by recrystallization from ethanol.4.2.1.2-Hydroxy-4-phenyl-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4a)White solid;mp222–223 C;1H NMR(DMSO-d6,300MHz): d1.91–1.94(m,2H),2.15–2.19(m,2H),2.60–2.63(m,2H),4.06(s, 1H),4.13(s,1H),6.93–7.25(m,6H),7.45–7.48(m,1H),7.83–7.88(m, 1H),8.74–8.77(m,1H);19F NMR(DMSO-d6,282MHz):dÀ80.66(s, 3F);IR(KBr)n max:3431,3100,2789,1702,1659,1609,1520,1496, 1457,1418,1360,1077,729cmÀ1;MS(70eV,EI)m/z(%):422(Mþ, 3.13),311[(MÀC5H3OS)þ,15.00],222(C8H5F3O2Sþ,25.06),199 (C13H11O2þ,100),69(CF3þ,31.34);Anal.Calcd for C21H17F3O4S:C, 59.71;H,4.06.Found:C,59.57;H,3.93.4.2.2.2-Hydroxy-3-(thien-2-oyl)-4-p-tolyl-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4b)White solid;mp216–218 C;1H NMR(CDCl3,300MHz):d2.02–2.10(m,2H),2.16(s,3H),2.28–2.38(m,2H),2.58–2.66(m,2H),3.80 (d,J¼11.7Hz,1H),4.03(d,J¼11.7Hz,1H),6.33(br s,1H),6.87–7.01 (m,6H),7.64(m,1H);19F NMR(DMSO-d6,282MHz):dÀ83.43(s, CF3);IR(KBr)n max:3305,3090,3032,2968,2878,1662,1606,1519, 1458,1419,1377,1073,816,752cmÀ1;MS(70eV,EI)m/z(%):436 (Mþ,1.65),325[(MÀC5H3OS)þ,10.48],222(C8H5F3O2Sþ,9.43),199 (C13H11O2þ,100),69(CF3þ,13.12);Anal.Calcd for C22H19F3O4S:C, 60.54;H,4.39.Found:C,60.38;H,4.24.4.2.3.2-Hydroxy-4-(4-methoxyphenyl)-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4c) White solid;mp226–227 C;1H NMR(DMSO-d6,300MHz): d1.91–1.95(m,2H),2.17–2.21(m,2H),2.48–2.59(m,2H),3.32(s, 1H),3.58(s,3H),4.08–4.11(m,2H),6.61–6.65(m,2H),6.96–7.02 (m,3H),7.50–7.58(m,1H),7.86–7.88(m,1H),8.65–8.75(m,1H);19F NMR(DMSO-d6,282MHz):dÀ80.65(s,3F);IR(KBr)n max:3431, 3090,2967,2835,1660,1605,1513,1459,1418,1375,1255,1072, 834cmÀ1;MS(ESI)m/z:[(MþH)þ,453.3],[(MþNa)þ,475.2], [(MþK)þ,491.2];Anal.Calcd for C22H19F3O5S:C,58.40;H,4.23. Found:C,58.39;H,4.03.4.2.4.2-Hydroxy-4-(4-hydroxyphenyl)-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4d) White solid;mp203–204 C;1H NMR(DMSO-d6,300MHz): d1.88–1.92(m,2H),2.15–2.18(m,2H),2.34–2.70(m,2H),3.98–4.07 (m,2H),6.43–6.46(m,2H),6.86–7.03(m,3H),7.49–7.56(m,1H), 7.87–7.89(m,1H),8.65–8.70(m,1H),8.95–9.01(m,1H);19F NMR (DMSO-d6,282MHz):dÀ80.73(s,3F);IR(KBr)n max:3510,3430, 3096,2981,2768,1664,1655,1605,1517,1448,1419,1185,1072,805, 745cmÀ1;MS(ESI)m/z:[(MþH)þ,439.3],[(Mþ2H)þ,440.2], [(MþNa)þ,461.0],[(MþK)þ,477.0],[(MþKþH2O)þ,493.0];HRMS for C21H17F3O5Sþ1Calcd:438.0749;Found:438.0762.4.2.5.4-(4-(Dimethylamino)phenyl)-2-hydroxy-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4e) Light yellow solid;mp205–207 C;1H NMR(DMSO-d6, 300MHz):d1.90–1.92(m,2H),2.14–2.18(m,2H),2.44–2.53(m,2H),2.73–2.81(s,6H),4.01–4.09(m,2H),6.42–6.45(m,2H),6.88–6.90(m,3H),7.51–7.53(m,1H),7.85–7.88(m,1H),8.62–8.66(m, 1H);19F NMR(DMSO-d6,282MHz):dÀ80.64(s,3F);IR(KBr)n max: 3432,3090,2961,2792,1662,1609,1520,1452,1418,1377,1358, 1071,820,745cmÀ1;MS(ESI)m/z:[(MþH)þ,466.2],[(Mþ2H)þ, 467.3];HRMS for C23H23NO4F3Sþ1Calcd:466.1300;Found: 466.1294.4.2.6.4-(4-Chlorophenyl)-2-hydroxy-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4f) White solid;mp224–225 C;1H NMR(DMSO-d6,300MHz): d1.83–2.05(m,2H),2.15–2.25(m.2H),2.37–2.61(m,2H),4.10–4.14 (m,2H),7.03–7.24(m,5H),7.50–7.60(m,1H),7.89–7.92(m,1H), 8.78–8.82(m,1H);19F NMR(DMSO-d6,282MHz):dÀ80.58(s,3F); IR(KBr)n max:3444,1657,1607,1519,1493,1417,1374,1073,810, 727cmÀ1;MS(ESI)m/z:[(MþH)þ,459.0/457.0];Anal.Calcd for C21H16ClF3O4S:C,55.21;H,3.53.Found:C,54.88;H,3.58.4.2.7.2-Hydroxy-4-(4-nitrophenyl)-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4g) White solid;mp230–231 C;1H NMR(DMSO-d6,300MHz): d1.96–1.98(m,2H),2.18–2.25(m,2H),2.46–2.57(m,2H),4.25–4.27 (m,2H),7.02(s,1H),7.43(d,J¼8.6Hz,2H),7.64–7.66(m,1H),7.90–7.94(m,1H),7.95(d,J¼8.6Hz,2H),8.98–9.00(m,1H);19F NMR (DMSO-d6,282MHz):dÀ80.92(s,3F);IR(KBr)n max:3430,3078, 2792,1655,1606,1520,1458,1416,1074,806,760cmÀ1;MS(ESI)m/ z:[(MþH)þ,468.0],[(Mþ2H)þ,469.0],[(MþHþH2O)þ,486.7];Anal. Calcd for C21H16F3NO6S:C,53.96;H,3.45;N,3.00.Found:C,53.91; H,3.71;N,2.70.4.2.8.2-Hydroxy-4-(3-nitrophenyl)-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4h) Light yellow solid;mp236–237 C;1H NMR(DMSO-d6, 300MHz):d1.97–1.98(m,2H),2.22–2.28(m,2H),2.44–2.66(m, 2H),4.28(br s,2H),7.00(s,1H),7.36–7.39(m,1H),7.57–7.68(m, 2H),7.85–8.00(m,3H),8.90–8.95(m,1H);19F NMR(DMSO-d6, 282MHz):dÀ80.54(s,3F);IR(KBr)n max:3298,3097,2970,2895, 2785,2584,2005,1944,1885,1840,1763,1660,1609,1500,1481, 1457,1417,1077,750,690,711cmÀ1;MS(70eV,EI)m/z(%):467 (Mþ, 5.83),356[(MÀC5H3OS)þ,17.24],245[(MÀC8H5F3O2S)þ, 32.15],228(C13H10NO3þ,100),222(C8H5F3O2Sþ,95.90),199 (C13H11O2þ,45.17),153(C7H5O2Sþ,94.29),111(C5H3OSþ,78.45),69 (CF3þ,96.79);Anal.Calcd for C21H16F3NO6S:C,53.96;H,3.45;N,3.00. Found:C,53.89;H,3.58;N,2.88.4.2.9.2-Hydroxy-4-(3-phenoxyphenyl)-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4i) Light yellow solid;mp196–197 C;1H NMR(DMSO-d6, 300MHz):d1.90–1.98(m,2H),2.21–2.24(m,2H),2.51(s,2H),4.11 (br s,2H),6.59–6.63(m,3H),6.67(s,1H),6.75–6.79(m,1H),7.04–7.15(m,3H),7.29(m,2H),7.58(m,1H),7.98(m,1H),8.81(m,1H); 19F NMR(DMSO-d6,282MHz):dÀ81.13(s,3F);IR(KBr)n max:3310, 3040,2782,2581,1660,1600,1518,1480,1445,1416,1357,1200, 1070,750,727,708cmÀ1;MS(ESI)m/z:[(MþH)þ,515.2], [(Mþ2H)þ,516.3];Anal.Calcd for C27H21F3O5S:C,63.03;H,4.11. Found:C,62.51;H,4.36.4.2.10.4-(3-Bromophenyl)-2-hydroxy-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(4j) Light yellow solid;mp236–237 C;1H NMR(DMSO-d6, 300MHz):d1.95–1.98(m,2H),2.20–2.23(m,2H),2.50–2.56(m, 2H),4.13(br s,2H),7.03–7.04(m,2H),7.12–7.18(m,2H),7.27–7.30 (m,1H),7.61–7.64(m,1H),7.90–7.92(m,1H),8.80–8.85(m,1H);19F NMR(DMSO-d6,282MHz):dÀ80.57(s,3F);IR(KBr)n max:3436, 3079,2964,2891,1658,1608,1518,1476,1457,1416,1362,1189,806, 726cmÀ1;MS(70eV,EI)m/z(%):500(Mþ, 1.58),391/389S.Song et al./Tetrahedron64(2008)5728–57355733[(MÀC5H3OS)þ,4.63/5.10],222(C8H5F3O2Sþ,16.77),199(C13H11O2þ, 100),69(CF3þ,18.50);Anal.Calcd for C21H16BrF3O4S:C,50.31;H,3.22.Found:C,50.01;H,3.35.4.3.Preparation of2-hydroxy-4-phenyl-3-(thien-2-oyl)-2-(trifluoromethyl)-1,2,3,4,7,8-hexahydroquinolin-5(6H)-one5 Method A:To a mixture of benzaldehyde1a(159mg,1.5mmol), 1,3-cyclohexanedione2(168mg,1.5mmol),and4,4,4-trifluoro-1-(thien-2-yl)butane-1,3-dione3(333mg,1.5mmol)in10mL EtOH was added30equiv of NH4OAc under stirring at room temperature. The mixture was refluxed and continually stirred for specified hour (monitored by TLC).After cooling,the mixture was poured into 20mL water,and extracted with CH2Cl2three times(3Â15mL).The combined organic layer was dried with Na2SO4overnight.The solvent was removed by rotavapor and the residue was purified by column chromatography on silica gel using petroleum/ethyl acetate¼2:1(v/v)as eluent to afford pure product5.Method B:Preparation of5from4.The procedure is same as the above except the2-hydroxy-4-phenyl-3-(thien-2-oyl)-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one4a (422mg,1mmol)and30equiv of NH4OAc were engaged as starting materials.4.3.1.2-Hydroxy-4-phenyl-3-(thien-2-oyl)-2-(trifluoromethyl)-1,2,3,4,7,8-hexahydroquinolin-5(6H)-one(5a)Yellow solid;231–232 C;1H NMR(CDCl3,500MHz):d2.18–2.22(m,2H), 2.28–2.41(m,2H), 2.52–2.55(m,2H), 3.83(d, J¼11.5Hz,1H),4.19(d,J¼11.5Hz,1H),5.10(s,1H),6.04(s,1H),6.84–6.86(m,1H),6.98–7.00(m,2H),7.06–7.11(m,4H),7.60–7.62(m, 1H);19F NMR(CDCl3,470MHz):dÀ82.24(s,3F);IR(KBr)n max: 3269,3109,3044,2958,2928,2890,1634,1599,1533,1455,1411, 1034,1192,734,705cmÀ1;MS(ESI)m/z:[(MþH)þ,422],[(Mþ2H)þ, 423];Anal.Calcd for C21H18F3NO3S:C,59.85;H, 4.31;N,3.32. Found:C,60.12;H,4.42;N,3.20.4.4.General procedure for preparation of7a–7dThe procedure is same as the preparation of4except that after cooling,the solvent was removed by rotavapor,and the residue was purified by column chromatography on silica gel using petroleum/ ethyl acetate¼1.5:1(v/v)as eluent to afford the pure products7. 4.4.1.3-Acetyl-2-hydroxy-4-phenyl-2-(trifluoromethyl)-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(7a)White solid;mp203–204 C;1H NMR(CDCl3,500MHz):d1.72 (s,3H),1.97–2.08(m,2H),2.22–2.38(m,2H),2.51–2.66(m,2H), 3.28(d,J¼11.5Hz,1H),3.83(d,J¼11.5Hz,1H),5.73(s,1H),7.12(d, J¼7.0Hz,2H),7.24(t,J¼7.0Hz,1H),7.31(t,J¼2.0Hz,2H);19F NMR (CDCl3,470MHz):dÀ83.52(s,3F);IR(KBr)n max:3447,3051,2950, 2920,2870,1726,1680,1495,1456,1423,1360,1191,714cmÀ1;MS (70eV,EI)m/z(%):354(Mþ,3.09),311[(MÀCH3CO)þ,95.65],199 (C13H11O2þ,198.25),69(CF3þ,32.16),43(CH3COþ,100);Anal.Calcd for C18H17F3O4:C,61.02;H,4.84.Found:C,61.02;H,4.93.4.4.2.Ethyl2-hydroxy-5-oxo-4-phenyl-2-(trifluoromethyl)-3,4,5,6,7,8-hexahydro-2H-chromene-3-carboxylate(7b) White solid;mp220–221 C;1H NMR(CDCl3,300MHz):d1.00 (t,J¼7.2Hz,3H),1.99–2.09(m.2H),2.23–2.39(m,2H),2.53–2.66 (m,2H),2.98(d,J¼11.7Hz,1H),3.94(d,J¼11.7Hz,1H),4.04(q, J¼7.2Hz,2H),5.57(s,1H),7.08–7.26(m,5H);19F NMR(CDCl3, 282MHz):dÀ84.19(s,3F);IR(KBr)n max:3465,3050,2985,2774, 2585,1741,1610,1373,1354,1238,1193,1162,1018,704,619cmÀ1; MS(70eV,EI)m/z(%):384(Mþ,3.46),311[(MÀCO2Et)þ,47.32],241 [(MÀ1ÀCF3ÀCO2Et)þ,100],199[(MÀ1ÀCF3COCH2CO2Et)þ,33.20],69(CF3þ,66.20);Anal.Calcd for C19H19O5F3:C,59.38;H,4.95.Found: C,59.29;H,4.78.4.4.3.3-Acetyl-2-hydroxy-2-methyl-4-phenyl-3,4,7,8-tetrahydro-2H-chromen-5(6H)-one(7c)White solid;mp178–180 C;1H NMR(CDCl3,500MHz):d1.50 (s,3H),1.80(s,3H),1.88(t,J¼5.5Hz,2H),2.24–2.34(m,2H),2.44–2.51(m,2H),3.14(d,J¼8.0Hz,1H),3.91(d,J¼8.0Hz,1H),7.11–7.25 (m,5H);IR(KBr)n max:3202,3029,2963,2925,2880,1719,1631, 1585,1492,1453,1421,1380,1071,727,702cmÀ1;MS(70eV,EI) m/z(%):300(Mþ, 1.52),257[(MÀCH3CO)þ,18.58],215 [(MÀCH3ÀC4H6O)þ,46.61],199(C13H11O2þ,79.54),43(CH3COþ, 100),116(C5H8O3þ,10.39);Anal.Calcd for C18H20O4:C,71.98;H,6.71. Found:C,71.95;H,6.76.4.4.4.Ethyl2-hydroxy-2-methyl-5-oxo-4-phenyl-3,4,5,6,7,8-hexahydro-2H-chromene-3-carboxylate(7d)White solid;mp142–144 C;1H NMR(CDCl3,500MHz):d1.07 (t,J¼7.0Hz,3H),1.52(s,3H),1.99(s,2H),2.59–2.63(m,4H),2.97(d, J¼9.5Hz,1H),4.05(q,J¼7.0Hz,2H),4.09(d,J¼9.5Hz,1H),7.12–7.18 (m,3H),7.21–7.27(m,2H);IR(KBr)n max:3318,3031,2941,2998, 2871,1735,1642,1612,1491,1453,1422,1373,1071,728,699cmÀ1; MS(70eV,EI)m/z(%):330(Mþ,5.50),287[(MÀEtOþ2H)þ,18.10], 241[(MÀEtOCOÀOH)þ,100],199(C13H11O2þ,76.64),129(C6H9O3þ, 16.66),43(CH3COþ,94.78);Anal.Calcd for C19H22O5:C,69.07;H, 6.71.Found:C,69.10;H,6.79.4.5.General procedure for dehydration reactionThe solution of0.5mmol of4or7and4equiv of p-TsOH in15mL of CHCl3was refluxed for several hours until complete of reaction (monitored by TLC).The solvent was removed by rotavapor and the residue was purified by column chromatography using petroleum/ ethyl acetate¼2:1(v/v)as eluent to afford pure product8.4.5.1.4-Phenyl-3-(thien-2-oyl)-2-(trifluoromethyl)-7,8-dihydro-4H-chromen-5(6H)-one(8a)Light red solid;mp134–136 C;1H NMR(CDCl3,500MHz): d2.02–2.17(m,2H),2.39–2.45(m,2H),2.65–2.78(m,2H),4.74(s, 1H),6.91(dd,J1¼4.8Hz,J2¼5.0Hz,1H),7.10(d,J¼3.5Hz,1H),7.12–7.17(m,3H),7.19–7.23(m,2H),7.61(dd,J1¼4.8Hz,J2¼5.0Hz,1H); 19F NMR(CDCl3,470MHz):dÀ66.98(s,3F);IR(KBr)n max:3438, 3088,2961,2921,2867,1699,1640,1518,1489,1454,1376,1152,725, 743,700cmÀ1;MS(70eV,EI)m/z(%):404(Mþ,69.73),335 [(MÀCF3)þ,17.99],111(C5H3OSþ,100);Anal.Calcd for C21H15F3O3S: C,62.37;H,3.74.Found:C,62.39;H,3.86.4.5.2.3-(Thien-2-oyl)-4-p-tolyl-2-(trifluoromethyl)-7,8-dihydro-4H-chromen-5(6H)-one(8b)White solid;mp145–147 C;1H NMR(CDCl3,500MHz):d2.00–2.15(m,2H),2.24(s,3H),2.34–2.44(m,2H),2.63–2.78(m,2H),4.69 (s,1H),6.94(dd,J1¼4.8Hz,J2¼5.0Hz,1H),7.00–7.06(m,4H),7.17 (d,J¼3.5Hz,1H),7.63(dd,J1¼4.8Hz,J2¼5.0Hz,1H);19F NMR (CDCl3,470MHz):dÀ66.99(s,3F);IR(KBr)n max:3439,3090,2964, 2923,2867,1697,1641,1515,1453,1412,1372,1152,818cmÀ1;MS (70eV,EI)m/z(%):418(Mþ,60.18),403[(MÀCH3)þ,47.37],307 [(MÀC5H3OS)þ,22.71],111(C5H3OSþ,100);Anal.Calcd for C22H17F3O3S:C;63.15;H,4.10.Found:C,62.92;H,4.09.4.5.3.3-Acetyl-2-methyl-4-phenyl-7,8-dihydro-4H-chromen-5(6H)-one(8c)Yellow solid;mp134–136 C;1H NMR(CDCl3,500MHz):d1.84–1.93(m,1H),1.96–2.04(m,1H),2.12(s,3H),2.27–2.41(m,5H),2.45–2.58(m,2H),4.80(s,1H),7.14–7.18(m,1H),7.23–7.30(m,4H); IR(KBr)n max:3442,3051,3029,2972,2943,1683,1662,1586,1492, 1455,1424,714,699cmÀ1;MS(70eV,EI)m/z(%):282(Mþ,93.03),S.Song et al./Tetrahedron64(2008)5728–5735 5734。

药物化学基本名词（1）2010-01-20 20:221、药物（drug）：药物是人类用来预防、治疗、诊断疾病、或为了调节人体功能，提高生活质量，保持身体健康的特殊化学品。

2、药物化学（medicinal chemistry）：药物化学是一门发现与发明新药、研究化学药物的合成、阐明药物的化学性质、研究药物分子与机体细胞（生物大分子）之间相互作用规律的综合性学科，是药学领域中重要的带头学科以及极具朝气的朝阳学科。

3、国际非专有药名（international non-proprietary names for pharmaceutical substance，INN）：是新药开发者在新药研究时向世界卫生组织（WHO）申请，由世界卫生组织批准的药物的正式名称并推荐使用。

该名称不能取得任何知识产权的保护，任何该产品的生产者都可使用，也是文献、教材及资料中以及在药品说明书中标明的有效成分的名称。

在复方制剂中只能用它作为复方组分的名称。

目前，INN名称已被世界各国采用。

4、中国药品通用名称（Chinese Approved Drug Names，CADN）：依据INN的原则，中华人民共和国的药政部门组织编写了《中国药品通用名称》（CADN），制定了药品的通用名。

通用名是中国药品命名的依据，是中文的INN。

CADN主要有以下的一些规则：中文名使用的词干与英文INN对应，音译为主，长音节可简缩，且顺口；简单有机化合物可用其化学名称。

5、巴比妥类药物（barbiturates agents）：具有5，5二取代基的环丙酰脲结构的一类镇静催眠药。

20世纪初问市的一类药物，主要由于5，5取代基的不同，有数十个各具药效学和药动学特色的药物供使用。

因毒副反应较大，其应用已逐渐减少。

6、非经典的抗精神病药物（atypical antipsychotic agents）：近年来问世的一些抗精神病药物。

和传统的吩噻嗪类和氟哌啶醇药物不同，其拮抗多巴胺受体的作用较弱，可能是产生多巴胺和5-羟色胺受体的双相调节作用，其锥体外系的副反应较少，具有明显治疗精神病阳性和阴性症状的作用。

Electronic properties and hydrogen storageapplication of designed porous nanotubes from a polyphenylene networkDewei Rao a ,b ,Ruifeng Lu a ,*,Zhaoshun Meng a ,Yunhui Wang a ,Zelin Lu a ,Yuzhen Liu a ,Xuan Chen c ,Erjun Kan a ,Chuanyun Xiao a ,d ,Kaiming Deng a ,d ,Haiping Wu a ,*aDepartment of Applied Physics,Nanjing University of Science and Technology,Nanjing 210094,PR China b Institute for Advanced Materials,Jiangsu University,Zhenjiang 212013,PR China cCollege of Science,PLA University of Science and Technology,Nanjing 211101,PR China dKey Laboratory of Soft Chemistry and Functional Materials,Ministry of Education,Nanjing University of Science and Technology,Nanjing 210094,PR Chinaa r t i c l e i n f oArticle history:Received 24June 2014Received in revised form 17September 2014Accepted 21September 2014Available online 11October 2014Keywords:Curvature effect Porous graphene Hydrogen storage NanotubeDensity functional theory Metal dopinga b s t r a c tBased on a polyphenylene network,a series of porous graphene nanotubes (PGNTs)are created and optimised via density functional theory calculations.The calculated band dispersion of the two-dimensional porous graphene can be tuned by rolling it into nano-tube form.To explore the energy application of PGNTs,we studied H 2adsorptions on metal (Li,Ca,and Na)decorated structures of PGNTs as well as B-substituted PGNTs.The results indicate that both the curvature effect and B substitution can strengthen the metal binding and prevent the metal atoms from clustering.Particularly for H 2adsorption,modification of the electronic property by the curvature effect is beneficial to provide more accessible space,leading to much higher adsorption energies of H 2on PGNTs than that on planar porous graphene,which is promising for the practical application of hydrogen storage.Copyright ©2014,Hydrogen Energy Publications,LLC.Published by Elsevier Ltd.All rightsreserved.IntroductionCurrently,some new energy resources are recommended as alternatives,such as hydrogen energy,solar energy,wind energy,nuclear energy,etc.[1,2].Hydrogen gas (H 2),one of the most ideal energy carriers,has attracted extensive attention because it possesses many crucial advantages,includingenvironmental cleanliness,renewability,abundance,and high gravimetric energy density.In recent years,a consider-able number of researchers have devoted themselves to exploring the utilisation of H 2using a safe and economic method under realistic environmental conditions.Unfortu-nately,the H 2storage capacities in current host materials are insufficient to realise commercialisation [3].Towards achieving the “Hydrogen Economy ”,it is necessary to enhance*Corresponding authors .E-mail addresses:refine6@ ,rflu@ (R.Lu).Available online at ScienceDirectjournal homepage:/locate/hei n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 39(2014)18966e 18975/10.1016/j.ijhydene.2014.09.1120360-3199/Copyright ©2014,Hydrogen Energy Publications,LLC.Published by Elsevier Ltd.All rights reserved.the capacities of H2storage of materials or to search/design new materials with high H2uptakes at ambient temperature and safe pressures.As interesting candidates,porous materials(PMs)have been widely studied for H2storage in the past decade[4e8] because they have high surface areas and large void spaces for H2guesting.In addition,the mechanism of H2storage in PMs is primarily physical sorption,which,compared with the chemical sorption,has advantages of fast kinetics and com-plete reversibility without extreme conditions.Recently, many PMs have been proposed for use in gas storage, including metal organic frameworks(MOFs)[9e12],covalent organic frameworks[13e16],conjugated microporous poly-mers[17e19],porous carbon-based materials[20e23],and so on.Note that all of the above-mentioned PMs have a carbo-naceous component,which greatly influences the H2storage capacities in PMs due to its high surface area[24].Importantly, carbon materials(CMs),including fullerenes[25],carbon nanotubes[26],graphene[27,28]and graphyne[29,30],have been tested as H2carriers,and some investigations have demonstrated that CMs are promising as a H2storage medium when various modifications are performed[25e32].Overall, functionalised CMs could play an important role in H2storage in future works.Graphene[33],a famous two-dimensional(2D)CM,has increasingly become the focus of research due to its fantastic properties and potential applications in sustainable energy areas.Subsequently,graphene-related materials,even though non-carbon2D materials,have become popular topics of sci-entific research in academia and in industry[34e40].Porous graphene(PG),a2D graphene-like material with periodic pore distribution,was successfully synthesised on metal surface via a coupling reaction of pre-designed molecular building blocks[41].PG has been extensively studied due to its versatile applications,such as gas purification or separation[42e44] and H2storage[2,23].Due to the large surface area,2D mate-rials have been proposed as one of the most promising can-didates for H2adsorption.Nevertheless,it has been reported that the interaction between H2and the carbon framework is very weak,and the adsorption energy(AE)of H2on a PG sur-face is too low to satisfy the requirements of practical applications.To improve the H2storage capacity of PG,the surface has been chemically modified through introducing heteroatoms, such as metal doping and boron/nitrogen substituting.Du et al.[2]studied the adsorption behaviour of an H2molecule in a Li-decorated PG usingfirst-principles calculations based on the local density approximation,and they found that each Li atom can adsorb three H2molecules,with an average AE of approximately0.25eV on PG,resulting in a theoretical ca-pacity of12wt%,which is much higher than the target of 5.5wt%set by the U.S.DOE in2011[45].Reunchan et al.[46] also theoretically studied the interaction between H2and metal-dispersed PG and demonstrated that Li,Na,and Ca atoms can be stably adsorbed on PG at the centre of six-carbon rings in the form of ionised-metal as a result of charge transfer from metal to PG,while multiple H2molecules can be further adsorbed around metal atoms,based on static multiple Coulombic interactions.Similarly,our previous work noted that the mechanism of H2adsorption on Li-MOFs is mainly due to the charge transfer between Li and H2[11].Recently, Huang et al.[47]found that Li-doped oxidised PG is highly attractive for use in H2storage.The above-mentioned results are based on standard den-sity functional theory(DFT),which is not adequate to describe the weak interaction between H2and host materials,and in practical utilisation,the H2uptake at approximately room temperature will be quite reduced compared to the simulated results.Moreover,pristine PG has small binding energies to metal atoms,much smaller than the cohesive energy of metals,which cannot completely prevent metal atoms from clustering.Therefore,we considered the influence of B sub-stitution on metal doping and on the following H2adsorption using multiscale simulations.The results indicate that the B-substituted PG increased the AE of both Li and Ca,as well as that of H2[23].Grand canonical ensemble Monte Carlo(GCMC) simulations indicate that B-substituted PG combined with Li/ Ca doping exhibits a good performance of H2storage at room temperature,exceeding6wt%[23].On the one hand,although the performance of PG material on H2storage has been claimed to be good,the storage ca-pacity will decrease significantly when introduced in storage equipment used in practical application.Thus,we must continue to improve the H2storage capacity of PG.On the other hand,our work regarding graphyne[29]indicated that the numbers of hydrogen molecules from GCMC and ab initio molecular dynamics at room temperature and from dispersion-corrected DFT calculations are almost the same, which should be caused by the lack of spatial positioning for H2adsorption.We also noticed that pioneering studies regarding the H2uptake on metal-dispersed PG surface encounter this deficiency,that is,the distance between doped metal atoms is too short for them to fully play their role in H2 adsorption.Consequently,we infer that lengthening this distance could be an effective way to enhance H2adsorption. Herein,we attempt to extend the distance between doped metals on PG through bending the2D structures.To this end, we roll PG into nanotubes(denoted as PGNTs),as displayed in Scheme1,and the name rule follows the transition of gra-phene into carbon nanotubes.Many studies have explored hydrogen storage in various tubes[26,48e52],and interestingly,similar carbon nitride nanotubes have been produced and their structural and op-tical properties for selected chiralities were investigated[39]. Additionally,some applications including Li storage and hydrogen storage have been explored in graphitic carbon nitride nanotubes[40].Motivated by the hydrogen adsorption investigations of nanotubes and PG materials,we will high-light the combination of pore-containing and tube-forming conformations for significant improvement of metal doping and hydrogen storage properties.To the best of our knowl-edge,no experimental evidence exists regarding hydrogen adsorption on the PGNTs,which have yet to be synthesised in experiments.The presented theoretical design will provide inspiration for future experimental endeavours to utilise PG (and other2D materials)as a gas storage carrier.In this work, Li,Ca,and Na atoms are selected as doping atoms,from which we can explore the curvature effect on the distance of metal atoms as well as on H2adsorptions.Because B substitution has been proven to be a useful method to enhance metal bindingi n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y39(2014)18966e1897518967and avoid clustering,in this work,B-substituted PGNTs (B-PGNTs)are also studied.Computational methodsThe Vienna Ab-initio Simulation Package [53]was used in all of the calculations.The geometric structures of PGNTs and B-doped PGNTs,along with their metal-doped configurations and hydrogen adsorptions,were optimised without any symmetric constraint.In general,standard DFT methods cannot accurately describe the weak interaction between H 2and the adsorbent.To improve the computational precision,the Grimme DFT-D2dispersion correction method [54]was used.The ion e electron interaction is described with the pro-jected augmented wave method [55]with the cut-off energy of 500eV.The exchange-correlation functional used is Perdew-Burke-Ernzerhof [56]in the generalised gradient approxima-tion,and the k -mesh grids are 8Â8Â1and 1Â1Â8for PG and PGNTs,respectively.The convergence thresholds are set to10À5eV in energy and 10À2eV/A˚in force.Results and discussionGeometric and electronic structuresThe properties and applications of materials are mainlydetermined by their geometric structures.Our optimised lat-tice constant of PG is 7.51A˚,which is in good agreement with experimental [57]and theoretical [2]values.The distance be-tween two carbon atoms in six-carbon rings (donated asC e C e 1)is 1.40A˚,and the bond length of C e H is 1.09A ˚,which is nearly equal to those of C e C and C e H in benzene,while thedistance between two aromatic rings (donated as C e C e 2)inPG is 1.49A˚.Four different PGNTs are constructed,which are PGNT(9,0),PGNT(6,6),PGNT(12,0)and PGNT(9,9),and the cor-responding diameters are roughly measured as 7.3A˚,8.0A ˚,9.1A˚and 12.2A ˚,respectively.The DFT results indicate that in PGNTs,the bond lengths of C e C e 1and C e H increase very slightly,and the shortest and longest distances of C e C e 2are1.47A˚and 1.50A ˚,respectively.These structural changes are caused by the redistribution of atoms from PG to PGNTs.More details of all optimised structures are provided in the Supplementary Materials (SM).As is well known,graphene has no band gap [58],and semiconducting carbon nanotubes (a series of rolled up gra-phene sheets)have different band gaps according to the crimping methods [59],which indicate that electronic prop-erties of 2D materials can be remarkably altered through bending.Thus,we first examine the electronic structures in selected PGNTs.Fig.1presents the side views of the optimised PG and PGNTs,as well as their band structures along the high-symmetry points.It is clear that compared to PG,the band dispersions of PGNTs have been changed as expected:the band dispersions of PGNTs are much more flat than the band dispersion of PG.This result is due to another quantum confinement effect (in addition to quantum confinement from intrinsic pores)in 1D PGNTs from the 2D planar form.The direct band gap of PG computed in this work is 2.40eV,which agrees well with the published results [2,42].For PGNTs,the direct band gap values for PGNT(9,0),PGNT(6,6),and PGNT(12,0)are 2.87eV,2.85eV,and 2.91eV,respectively,while for PGNT(9,9),there is an indirect gap of approximately 2.86eV.Interestingly,with varying the diameter and chirality of PGNTs under study,their band gaps are almostunchanged,Scheme 1e Rolling the PG into PGNTs,including the (m,0)and (n,n)chiralities similar to the formation of carbon nanotubes from graphene.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 39(2014)18966e 1897518968as shown in Fig.1.The physics behind this behaviour also originate from quantum confinement,which leads to a weaker p -type orbital interaction in both the valance band and the conduction band of these PGNTs.However,we cannot make a qualitative conclusion regarding the band gap because other chirality and diameters are not considered here due to the huge computational cost.Metal dopingTable 1lists the binding energies of Li,Ca,and Na atoms on PG,PGNTs,and B-PGNTs.The calculated values are 1.38eV,0.74eV,and 0.52eV for one Li,Ca,and Na,respectively,adsorbed on PG,which are in good agreement with previous results [46].Focusing on PGNTs,it is clear that the binding energies of metal atoms on PGNTs are all higher than the corresponding values on PG.This result means that the coiled PG with certain curvature can strengthen the metal adsorp-tion,which is due to the change in electronic structures (the strengthened hybridisations between the p orbitals of C and the s orbitals of the metal atoms,see the partial density of states in the SM)and the reduction of the repulsive interaction between two metal atoms.For different metals,the in-crements of binding energies are discrepant.The binding en-ergy of Li on PGNT(6,6)is the highest one,approximately 11%larger than that on PG.Particularly for Ca adsorbed on PGNTs,the binding is doubly enhanced and is largest on PGNT(9,0).For Na,the most significant case is a 48%increase on PGNT(12,0).Previous works [11,60]have demonstrated that charge can transfer from the metal to the carbonaceous part due to the difference in electronegativity,and the transferred charge can greatly enhance the electrostatic interaction,which is the reason that metals can be adsorbed on CMs.The electronegativity of phenyl is larger than those of alkali and alkaline-earth metals.Therefore,Li,Ca and Na atoms on PG or PGNTs lose some electrons.Another aspect is that the Ca easily loses its 4s electron due to its orbital radius being larger than those of Na and Li.The increase in binding energy in B-substituted species is also reasonable on the basis of the increase in electrostatic interaction by much more charge transfer from the metal to the host structure.As we reported,the B-substituted PG and graphyne are electron-deficient materials [23,29],i.e.,they can more easily adsorb electron-rich atoms,such as metal atoms.The present results are consistent with our previous work and other works [23,29].Comparing the results,the binding en-ergies of Li and Ca on B-PGNTs are larger than those on B-PG [23].In Table 1,different data can be observed for different B-PGNTs,implying that the curvature effect on metal doping still remains after B substituting in PGNTs.The charge ana-lyses and partial density of states illustrate tiny differences among different PGNTs or B-PGNTs,which are not presented here because these are difficult to attribute to the curvature effect;note that charge transfer and orbital hybridisationFig.1e Band structures and side views of PG and PGNTs.The dotted line is the Fermi level.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 39(2014)18966e 1897518969between the metal and C/B atoms are responsible for their interactions,which were already elucidated in our previous work[23].Hydrogen adsorptionWe know that the experimental cohesive energies[61]for Li, Ca,and Na are1.63eV,1.84eV,and1.11eV,respectively; therefore,the enhanced binding energies of Ca on PGNT(9,0) and the enhanced binding energies of all three considered metals on B-PGNTs could help to prevent metal clustering and disperse them on surfaces atomically,which is a prerequisite for a metal-modified hydrogen storage strategy.Next,we further investigate1e4H2adsorption on metal dispersed PGNTs and B-PGNTs.From the optimised structures,wefind that all of the bond lengths of H2molecules are in the range of 0.75e0.79A˚,which means that H2maintains its molecular form and is adsorbed in the physisorption type.Fig.2displays n H2(n¼1e4)adsorption around Na on B-PGNT(12,0)as a representative example(other structures are provided in the SM).Clearly,all of the adsorbed H2molecules are in molecular form.The distance between H2and Ca or Na is in the range of 2.3e2.4A˚,whereas the distance between H2and Li is 2.2e2.6A˚.These data and adsorption patterns are similar to the results of a published work[46],which indicates that H2is exclusively bound by a physical sorption that has the advan-tages of fast kinetics and good reversibility without extreme conditions.The requirement of practical application related to the average AE of H2on hydrogen storage materials should be in the range of0.2e0.4eV[62],and at this level,H2does not exhibit chemical sorption.On metal-doped PGNTs and B-PGNTs,the calculated AEs of H2are collected in Fig.3.Obvi-ously,around Li,the AE of thefirst H2is high,with a value of up to0.3eV.In the case of4H2adsorption,the average AEs are approximately0.2eV,except for Li@PGNT(6,6)(0.11eV),which are much higher than the AE of H2on metal-undoped PG (0.03eV)[42]and that of H2on Li-doped PG(0.09eV)[2,23].On Li@B-PGNTs,the AEs were found to decrease;a similar situa-tion occurs in Li-doped B-PG in comparison with Li-doped PG [23].With the increased loading of H2around Li,the average AE decreases in a linear fashion,with the exceptions of Li@PGNT(6,6)and Li@B-PGNT(12,0).Regarding Na,the varia-tion of AE is rapid but relatively mild,and the average AEs of 4H2adsorption on all chosen PGNTs and B-PGNTs are lower than0.2eV,which means that at room temperature,the H2 storage capacities of Na-doped nanotubes are not better than those of Li-decorated tubes.Most notably,Ca-preadsorbed structures have amazingly different behaviours of multiple H2adsorption,that is,the average AEs continue to increase with increasing H2loading.This behaviour can be observed in almost all Ca@PGNTs and Ca@B-PGNTs,and even in4H2 adsorption,the average AEs of H2are as high as0.3eV on Ca@PGNTs and0.24eV on Ca@B-PGNTs.This result can be attributed to the hybridisation of Ca s orbitals with the H2s state,and two4s electrons can match with multiple H2load-ings in a suitable space.In a previous study,the average AEs of H2on Ca-doped PG and Ca-doped B-PG were found to be less than0.2eV[23],which again confirms that the curvature ef-fect enhances H2adsorption.Because4H2can be stably adsorbed around Li,Na or Ca on PGNTs,the corresponding gravimetric densities are high,with values up to9.0wt%, 7.6wt%and6.5wt%,based on the support composition of C6H3e M,where M¼Li,Na,and Ca,respectively.For B-PGNTs, the gravimetric densities are slightly higher because of the lighter formula of C5BH3-M.In fact,the number of H2around Ca is not limited,and5H2 per Ca with average AEs of0.35eV for PGNT(9,9)and0.25eV for B-PGNT(9,9)correspond to saturated adsorption,which results in a high gravimetric density of8.0wt%for Ca@PGNT(9,9)and8.1wt%for Ca@B-PGNT(9,9).The fully relaxed geometries for5-H2uptakes are shown in Fig.S5, which indicates that thefifth dihydrogen is just above the quasi-square pattern formed by the other four hydrogen molecules.Similar conclusions have been obtained from Ca-doped carbon materials[48,49,63].Overall,the curvature ef-fect definitely enhances the hydrogen adsorption,as observed when straightforwardly comparing the AEs of H2on metal-doped or metal-undoped PGNTs(B-PGNTs)with the AEs in the corresponding PG(or B-PG)[23,46],and Ca doping is strongly recommended for further study because it has the best performance for H2storage,as suggested in our previous work[23].Fortunately,the experimental results demonstrated that the hydrogen storage capacity can be improved in the multi-walled carbon nanotubes decorated by Ca[64].To understand the H2adsorption mechanism,we plot the charge density difference for1e4H2adsorption on Ca@PGNT(9,9)in Fig.4.One can see that with an increaseinFig.2e1e4H2adsorption around Na on B-PGNT(12,0).Green balls,the adsorbed H2;white balls,H;purple balls,Na;pink balls,B.(For interpretation of the references to colour in thisfigure legend,the reader is referred to the web version of this article.)i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y39(2014)18966e1897518970the H 2number,the charge accumulation between Ca and H 2is clearly increased,which can improve the electrostatic inter-action between them,thereby enhancing the H 2adsorption.We emphasise here that the rolled PGNT offers a greater amount of room for gas loading because more accessible volume over a bent surface can be accommodated than that over a planar surface of PG.Thus,the increase in AE is due to a synergistic effect of the polarization and the curvature mechanisms.To understand how the curvature affects H 2adsorption,Fig.5presents the charge density difference for 2H 2adsorp-tions around Ca on four selected PGNTs withdifferentFig.3e Average adsorption energies for 1e 4H 2around the metal atom on PGNTs andB-PGNTs.Fig.4e Plots of the charge density differences for (a)1H 2,(b)2H 2,(c)3H 2,and (d)4H 2molecules adsorbed on Ca-PGNT(9,9).Red ball,Ca;grey ball,C;white ball,H;blue,charge depletion;yellow,charge accumulation.(For interpretation of the references to colour in this figure legend,the reader is referred to the web version of this article.)i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 39(2014)18966e 1897518971diameters (i.e.,different curvatures).Fig.5(d)for PGNT(9,9)exhibits the most charge accumulation,and the average AE of H 2is the highest one (0.323eV),whereas the smallest AE is 0.286eV on PGNT(9,0),which has the lowest amount of charge accumulation,as shown in Fig.5(a).The curvature can greatly influence H 2adsorption on metal-decorated PGNTs.For example,in the cases of 2H 2or 3H 2adsorption on Ca-doped PGNTs,the larger the diameter (i.e.,the smaller curvature)of the nanotube,the higher the AE of H 2will be.However,it is difficult to establish a qualitative relationship between the H 2binding strength and the curvature of PGNTs.One definite point is that the crimping of PG indeed enhances the hydrogen adsorption.In addition to the H 2adsorption on the outer surface of M@PGNTs (M ¼Li,Na,Ca),we found that H 2can also be adsorbed inside the channel of PGNTs.In Fig.6,we show the optimised geometry of 1H 2in the inner cavity of M@B-PGNT(9,9).In this case,the bond length of H 2is 0.75A˚,being stretched slightly in contrast to a free H 2molecule.The AEs ofH 2are 0.093eV,0.085eV,and 0.098eV for Li-,Ca-,and Na-doped B-PGNT(9,9),respectively.To the best of our knowl-edge,these AE values are still larger than the benchmark value of 0.075eV in Li-coated C 60calculated from standard DFT [25].Thus,we can infer that metal-decorated PGNTs (or B-PGNTs)can store hydrogen on both the outer and the inner surfaces,and desorption of H 2inside the tube channel can occur via H 2passing through the pores of the tube walls.ConclusionIn summary,we designed some novel pore-containing nano-tubes based on the rolling up of 2D polyphenylene and have schematically studied their electronic properties as well as their application in hydrogen storage.The main findings are as follows.(1)The band gap of PG is increasedbyFig.5e Plots of the charge density differences for 2H 2adsorption around Ca on (a)PGNT(9,0),(b)PGNT(6,6),(c)PGNT(12,0),and (d)PGNT(9,9).Red ball,Ca;grey ball,C;white ball,H;blue,charge depletion;yellow,charge accumulation.(For interpretation of the references to colour in this figure legend,the reader is referred to the web version of thisarticle.)Fig.6e One H 2adsorbed in the channel of M@B-PGNT(9,9).(a)M ¼Li,(b)M ¼Ca,(c)M ¼Na.Li,blue ball;Ca,red ball;Na,purple ball;H,white ball;C,black ball;B,pink ball;adsorbed H 2,green balls.(For interpretation of the references to colour in this figure legend,the reader is referred to the web version of this article.)i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 39(2014)18966e 1897518972approximately0.4eV in its nanotube forms;however,the band gaps of the designed PGNTs are nearly unchanged with varying diameter and chirality.(2)The curvature effect can strengthen the binding of Li,Ca,and Na on PGNTs(or B-PGNTs)compared to planar PG(or B-PG),whereas B substitu-tion plays a much more important role in metal doping and effectively prevents metal atoms from clustering on the sur-face.(3)The curvature effect greatly enhances the H2 adsorption on metal-doped PGNTs(or B-PGNTs)compared to planar PG(or B-PG).More specifically,for H2storage,Li-doped PGNTs exhibit a high adsorption energy of H2,but the low binding energy of Li on PGNTs is disappointing.Li doping can be solved on B-PGNTs,but the adsorption energy of H2is not satisfied.In terms of Na doping,although the H2adsorption is not weak-ened with more H2loading as quickly as that in Li-doped nanotubes,the average adsorption energy of H2is even smaller than that in Li-doped nanotubes.Fascinating results were obtained for Ca-doped nanotubes,which exhibit a completely different trend from those of Li and Na regarding the adsorption energy versus the number of H2.The adsorp-tion up to5H2per Ca on both PGNTs and B-PGNTs with a high hydrogen storage capacity of approximately8wt%meets the requirement for practical applications.We also found that the inner channels of the designed nanotubes are useful for hydrogen storage.It is expected that our effective scheme of utilising metal pre-adsorbing,boron doping,and rolling up planar structures will be helpful in searching for new hydrogen storage materials.Note from theoretical works [65e67]that oxygen may compete with hydrogen in the pro-cess of adsorption,so oxygen interference(and probably also water in the case of alkali metals)should be carefully taken into account to realise an experimental system.AcknowledgementsThis work was supported by NSF of China Grant(No.21373113, 11304394,11304155),the Fundamental Research Funds for the Central Universities(No.30920140111008,30920140132037), Jiangsu Province Science Foundation for Youths(No. BK2012394,BK20140526),China Postdoctoral Science Foun-dation funded project with Grant No.2014M561576,and the Research Foundation for Advanced Talents of Jiangsu Uni-versity with Grant No.13JDG100.Appendix A.Supplementary dataSupplementary data related to this article can be found at /10.1016/j.ijhydene.2014.09.112.r e f e r e n c e s[1]Chu S,Majumdar A.Opportunities and challenges for asustainable energy future.Nature2012;488:294e303.[2]Du AJ,Zhu ZH,Smith SC.Multifunctional porous graphenefor nanoelectronics and hydrogen storage:new propertiesrevealed byfirst principle calculations.J Am Chem Soc2010;132:2876e7.[3]Durbin DJ,Malardier-Jugroot C.Review of hydrogen storagetechniques for on board vehicle applications.Int J Hydrogen Energy2013;38:14595e617.[4]Cao DP,Lan JH,Wang WC,Smit B.Lithium-doped3Dcovalent organic frameworks:high-capacity hydrogenstorage materials.Angew Chem Int Ed2009;48:4730e3. [5]Lan JH,Cao DP,Wang WC,Ben T,Zhu GS.High-capacityhydrogen storage in porous aromatic frameworks withdiamond-like structure.J Phys Chem Lett2010;1:978e81. [6]Wang L,Lee K,Sun YY,Lucking M,Chen Z,Zhao JJ,et al.Graphene oxide as an ideal substrate for hydrogen storage.ACS Nano2009;3:2995e3000.[7]Wang XJ,Li PZ,Chen Y,Zhang Q,Zhang H,Chan XX,et al.Rationally designed nitrogen-rich metal-organic framework and its exceptionally high CO2and H2uptake capability.Sci Rep2013;3:1149.[8]Reardon H,Hanlon JM,Hughes RW,Godula-Jopek A,Mandal TK,Gregory DH.Emerging concepts in solid-statehydrogen storage:the role of nanomaterials design.Energy Environ Sci2012;5:5951e79.[9]Rosi NL,Eckert J,Eddaoudi M,Vodak DT,Kim J,O'Keeffe M,et al.Hydrogen storage in microporous metal-organicframeworks.Science2003;300:1127e9.[10]Furukawa H,Ko N,Go YB,Aratani N,Choi SB,Choi E,et al.Ultrahigh porosity in metal-organic frameworks.Science2010;329:424e8.[11]Rao DW,Lu RF,Xiao CY,Kan EJ,Deng KM.Lithium-dopedMOF impregnated with lithium-coated fullerenes:ahydrogen storage route for high gravimetric and volumetric uptakes at ambient temperatures.Chem Commun2011;47:7698e700.[12]Meng ZS,Lu RF,Rao DW,Kan EJ,Xiao CY,Deng KM.Catenated metal-organic frameworks:promising hydrogen purification materials and high hydrogen storage mediumwith further lithium doping.Int J Hydrogen Energy2013;38:9811e8.[13]El-Kaderi HM,Hunt JR,Mendoza-Cortes JL,Cote AP,Taylor RE,O'Keeffe M,et al.Designed synthesis of3Dcovalent organic frameworks.Science2007;316:268e72. [14]Cote AP,Benin AI,Ockwig NW,O'Keeffe M,Matzger AJ,Yaghi OM.Porous crystalline covalent organic frameworks.Science2005;310:1166e70.[15]Srepusharawoot P,Swatsitang E,Amornkitbamrung V,Pinsook U,Ahuja R.Hydrogen adsorption of Li functionalized covalent organic framework-366:an ab initio study.Int JHydrogen Energy2013;38:14276e80.[16]Li F,Zhao JJ,Johansson B,Sun LX.Improving hydrogenstorage properties of covalent organic frameworks bysubstitutional doping.Int J Hydrogen Energy2010;35:266e71.[17]Cooper AI.Conjugated microporous polymers.Adv Mater2009;21:1291e5.[18]Li A,Lu RF,Wang Y,Wang X,Han KL,Deng WQ.Lithium-doped conjugated microporous polymers for reversiblehydrogen storage.Angew Chem Int Ed2010;49:3330e3. [19]Lu RF,Li A,Deng WQ.First-principles investigation of Liþ-doped conjugated microporous polymer as a potentialhydrogen storage mun Comput Chem2013;1:27e39.[20]Shevlin SA,Guo ZX.Density functional theory simulations ofcomplex hydride and carbon-based hydrogen storagematerials.Chem Soc Rev2009;38:211e25.[21]Yurum Y,Taralp A,Veziroglu TN.Storage of hydrogen innanostructured carbon materials.Int J Hydrogen Energy2009;34:3784e98.[22]Lu RF,Meng ZS,Kan EJ,Li F,Rao DW,Lu ZL,et al.Tunableband gap and hydrogen adsorption property of a two-i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y39(2014)18966e1897518973。

专利名称：OPTIMIZED PASSIVATION ON TI-/ZR-BASIS FOR METAL SURFACES发明人：BROUWER, Jan-Willem,KRÖMER,Jens,CORNEN, Sophie,FRANK,Michael,HEISCHKAMP, Nicole,CZIKA, Franz-Adolf,TEUBERT, Nicole申请号：EP2009053109申请日：20090317公开号：WO09/115504P1公开日：20090924专利内容由知识产权出版社提供摘要：The present invention relates to a chromium-free aqueous agent based on water-soluble compounds of titanium and/or zirconium and a source of fluoride ions, copper ions and metal ions selected from the group consisting of calcium, magnesium, aluminum, boron, zinc, iron, manganese and/or tungstene and to a method for the anti-corrosive conversion treatment of metal surfaces. The chromium-free aqueous agent is suitable for the treatment of various metal materials, joined in composite structures, amongst others of steel or galvanized steel or the alloys thereof or any combinations of said materials. Furthermore, surfaces of aluminum and alloys thereof can be treated in an anti-corrosive manner using the agent according to the invention. The anti-corrosive treatment is intended in particular as a pretreatment for a subsequent dip-coating. The invention further relates to a metallic substrate that was treated according to a predefined method sequence with the chromium-free agent according to the invention and to the use thereof, particularly in the automotive production of vehicle bodies.申请人：HENKEL AG & CO. KGAA,NIHON PARKERIZING,BROUWER, Jan-Willem,KRÖMER, Jens,CORNEN, Sophie,FRANK, Michael,HEISCHKAMP, Nicole,CZIKA, Franz-Adolf,TEUBERT, Nicole地址：DE,JP,DE,DE,DE,DE,DE,DE,DE国籍：DE,JP,DE,DE,DE,DE,DE,DE,DE更多信息请下载全文后查看。

Combined effects of hydrogen back-pressure and NbF 5addition on the dehydrogenation and rehydrogenation kinetics of the LiBH 4e MgH 2composite systemJianfeng Mao a ,Zaiping Guo a ,b ,*,Xuebin Yu c ,**,Huakun Liu aaInstitute for Superconducting and Electronic Materials,University of Wollongong,NSW 2522,AustraliabSchool of Mechanical,Materials and Mechatronics Engineering,University of Wollongong,NSW 2522,Australia cDepartment of Materials Science,Fudan University,Shanghai 200433,Chinaa r t i c l e i n f oArticle history:Received 30July 2012Received in revised form 14December 2012Accepted 23December 2012Available online 5February 2013Keywords:Hydrogen storage MgH 2LiBH 4Additivea b s t r a c tIt is well known that the dehydrogenation pathway of the LiBH 4e MgH 2composite sys-tem is highly reliant on whether decomposition is performed under vacuum or a hydrogen back-pressure.In this work,the effects of hydrogen back-pressure and NbF 5addition on the dehydrogenation kinetics of the LiBH 4e MgH 2system are studied under either vacuum or hydrogen back-pressure,as well as the subsequent rehydrogenation and cycling.For the pristine sample,faster desorption kinetics was obtained under vacuum,but the performance is compromised by slow absorption kinetics.In contrast,hydrogen back-pressure remarkably promotes the absorption kinetics and increases the reversible hydrogen storage capacity,but with the penalty of much slower desorption kinetics.These drawbacks were overcome after doping with NbF 5,with which the dehydrogenation and rehydrogenation kinetics was significantly improved.In partic-ular,the enhanced kinetics was observed to persist well,even after 9cycles,in the case of the NbF 5doped sample under hydrogen back-pressure,as well as the suppression of forming Li 2B 12H 12.Furthermore,the mechanism that is behind these effects of NbF 5additive on the reversible dehydrogenation reaction of the LiBH 4e MgH 2system is discussed.Copyright ª2013,Hydrogen Energy Publications,LLC.Published by Elsevier Ltd.All rightsreserved.1.IntroductionHydrogen is one of the most promising energy carriers,as no pollutants are produced when it is burned or used in fuel cells [1].Hydrogen storage still faces a major technical barrier,however,in terms of its on-board application in hydrogen fuel cell vehicles [2].For mobile applications,a system target of 5.5wt %H 2capacity for 2015has been set by the US Department of Energy [3].Due to their relatively high volumetric and gravimetric capacities,complex hydrides,especially light*Corresponding author .Institute for Superconducting and Electronic Materials,University of Wollongong,NSW 2522,Australia.**Corresponding author .E-mail addresses:zguo@.au (Z.Guo),yuxuebin@ (X.Yu).Available online atjournal homepage:/locate/hei n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 38(2013)3650e 36600360-3199/$e see front matter Copyright ª2013,Hydrogen Energy Publications,LLC.Published by Elsevier Ltd.All rights reserved./10.1016/j.ijhydene.2012.12.106metal borohydrides,have been attracting intense interest[4]. For example,LiBH4is an attractive system that stores13.9wt% hydrogen according to LiBH44LiHþBþ3/2H2.Its rigorous thermodynamic and kinetic properties,however,limit its ability to cycle hydrogen under moderate conditions[5e8]. Various methods have been aimed at modifying LiBH4to promote hydrogen desorption at lower temperatures and increase the reversible reaction kinetics,including confine-ment in nanoporous scaffolds[9,10],doping with catalysts [11],and destabilization with additives such as metals[12,13], metal oxides[14],metal halides[15],and metal hydrides [16e18].Among these attempts,destabilization shows the most promise because the addition of a destabilizing reagent can react exothermally with the host material to form more stable end products,which will reduce the enthalpy for dehydrogenation,thereby reducing the decomposition temperature.For destabilization of LiBH4,one of the most prominent destabilization reagents is MgH2.This is because the addition of MgH2can reduce the dehydrogenation temperature of LiBH4 to nearly as low as350 C and promotes complete decom-position below500 C[19].It is already well known that the LiBH4e MgH2combined system displays two-step dehydro-genation features,corresponding to the individual decom-position of MgH2and LiBH4,with increasing temperature.The reaction path for the second-step dehydrogenation,however, i.e.,the decomposition of LiBH4,is dependent on whether decomposition is performed under vacuum or under a hydro-gen pressure(typically!3bar).For decomposition performed under a hydrogen back-pressure,the system forms MgB2,as shown in reaction(1)[19e22].2LiBH4þMgH2/2LiBH4þMgþH2/2LiHþMgB2þ4H2(1)In a dynamic vacuum environment,the system forms Li e Mg alloys according to reaction(2)[23e26].0.3LiBH4þMgH2/0.3LiBH4þMgþH2/0.78Mg0.816Li0.184þ0.52Mg0.70Li0.30þ0.3Bþ1.60H2(2)For both reactions(1)and(2),the observed reaction kinetics for the LiBH4e MgH2composite system is slow,and tempera-tures above400 C are needed to attain a reasonable desorp-tion rate.Therefore,many methods have been successfully researched to enhance the dehydrogenation and rehydroge-nation properties of the LiBH4e MgH2composite system,such as reducing the particle size[27e29]and using catalysts or additives,such as carbon,Ni,Al,TiCl3,and TiF3[30e45].For example,Fan et al.[36]found that the reaction kinetics of the LiBH4e MgH2composite system can be improved by doping with Nb2O5.More recently,Xiao et al.[44]and Sabitu et al.[45] found that NbF5exhibits superior catalytic activity compared to the other catalysts such as Nb2O5and TiF3.The catalytic mechanisms in the NbF5doped LiBH4e MgH2system are un-clear,however.For example,Xiao et al.[44]observed that the F anion may partially substitute for the anionic H in LiBH4, resulting in favorable thermodynamic modification.There was no direct observation of any Nb related phase,however, which should play an important role in the enhancement of the kinetics.These results indicate that more detailed studies are needed to understand the catalytic mechanism and to identify the effects on the destabilization reaction and rever-sibility.On the other hand,although there have been a few studies on the effects of hydrogen back-pressure and vacuum on the reaction pathway and products of the LiBH4e MgH2 composite system,the effects of hydrogen back-pressure and vacuum on the dehydrogenation and rehydrogenation ki-netics of this system have not yet been fully clarified.There-fore,a comparative study of the hydrogen back-pressure and static vacuum effects on the dehydrogenation and rehy-drogenation kinetics of the LiBH4e MgH2system with and without catalyst is highly desirable.In this work,the effects of hydrogen back-pressure and vacuum on the hydrogen desorption and absorption kinetics of the LiBH4e MgH2system,with and without NbF5addition, have been comparatively investigated with a series of hydro-gen release and uptake experiments.As expected,the NbF5 catalyst exhibits pronounced effects toward promoting the dehydrogenation kinetics of LiBH4e MgH2under either hydrogen back-pressure or static vacuum,as well as on the subsequent rehydrogenation.In particular,the enhanced ki-netics was observed to persist well,even after9cycles,in the case of the NbF5doped sample when the dehydrogenation was conducted under hydrogen back-pressure.The reaction and catalytic mechanisms are also elucidated in depth,based on examination by means of X-ray diffraction(XRD),Fourier transform infrared(FTIR)spectroscopy,and X-ray photo-electron spectroscopy(XPS).2.Experimental proceduresThe chemicals MgH2(98%purity),LiBH4(hydrogen storage grade,!90%purity),NbF5(98%purity),Nb(99.98%purity, <45m m),and LiF(!99.98%purity)were all purchased from Sigma e Aldrich and used directly without pretreatment.All sample storage and handling were performed in an Ar-filled glove box(MBraun Unilab).Two mixed samples of 2LiBH4e MgH2and2LiBH4e MgH2e0.1NbF5were ball milled for 2h at a rate of400rpm in a QM-2SP planetary ball mill, respectively.The ball-to-powder ratio was30:1.The hydrogen desorption/absorption properties were measured in a Sieverts apparatus(Advanced Materials Cor-poration,USA),where the temperatures and pressures of the samples and the gas reservoirs were monitored and recorded by GrcLV-LabVIEW-based control program software during the sorption process.The hydrogen desorption kinetic measurements were performed at400 C under both static vacuum and a hydrogen back-pressure of p(H2)¼5.5bar. Before the measurement,the sample chamber wasfilled with hydrogen at55bar pressure,and the temperature was then raised to and kept at the desired temperature.Then,the chamber pressure was quickly tuned to vacuum or5.5bar hydrogen back-pressure,respectively,before the onset of measurements.The hydrogen absorption measurements were performed at400 C and about55bar hydrogen pres-sure.Unless otherwise specified,the H-capacity was calcu-lated using the weight of the samples containing additives to allow for an evaluation of the practical hydrogen storage properties.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y38(2013)3650e36603651Thermogravimetry(TG)and differential scanning calo-rimetry(DSC)analyses of the dehydrogenation process were carried out on a Mettler ToledoTGA/DSC1.About6mg of sample was loaded into an alumina crucible in the glove box. The crucible was then placed in a sealed glass bottle in order to prevent oxidation during transportation from the glove box to the TG/DSC apparatus.An empty alumina crucible was used as the reference material.The samples were heated from room temperature to500 C under1atmflowing argon at-mosphere,and the heating rate was10 C/min.X-ray diffraction(XRD)measurements on a GBC MMA dif-fractometer(Cu K a radiation)were carried out to identify the product phases in the sample after ball milling and de/rehy-drogenation.To avoid oxidation during the XRD measure-ment,samples were mounted onto a glass slide1mm in thickness in the Ar-filled glove box and sealed with an airtight hood composed of polyvinylchloride(PVC)tape,which has a broad peak around2q of20 .The vibration spectra of the species were identified using a Fourier transform infrared(FTIR)spectrometer(Shimadzu Prestige21).The obtained samples were ground with potas-sium bromide(KBr)powder at a weight ratio of1:100and pressed into the sample holder.The measurements were in transmission mode.X-ray photoelectron spectroscopy(XPS)of the NbF5doped LiBH4e MgH2after ball milling and of the milled sample after de/rehydrogenation was conducted using a SPECS PHOIBOS 100Analyzer installed in a high-vacuum chamber with the base pressure below10À11bar;X-ray excitation was provided by Al Ka radiation with photon energy h n¼1486.6eV at the high voltage of12kV and power of120W.XPS binding energy spectra were collected at the pass energy of20eV infixed analyzer transmission mode,and the XPS spectra of the doped samples were collected after bombardment of the samples using an Ar ion source with ion energy of5keV.Samples were prepared inside an Ar glove box,by dusting the powders onto adhesive carbon tape.The samples were then placed in a sealed container in order to reduce oxidation during trans-portation from the glove box to the XPS apparatus.XPS depthprofiles were acquired by ion beam sputtering(Ar ions),fol-lowed by XPS acquisition of elemental lines.The binding en-ergy reference was taken as the main component of the C1s peak at284.5eV.3.Results and discussion3.1.Effects of NbF5on the non-isothermal dehydrogenation of the LiBH4e MgH2composite system Fig.1(a)presents the XRD patterns for the LiBH4e MgH2and LiBH4e MgH2e NbF5composite samples after2h of ball milling. The initial phases LiBH4and MgH2were observed in the pristine LiBH4e MgH2mixture,indicating that it is just a physical mixture during ball milling.In the case of LiBH4e MgH2e NbF5,after ball milling,however,the LiBH4 phase disappeared.In addition to the initial phase MgH2,new phases corresponding to Nb and LiF were detected in the LiBH4e MgH2e NbF5composite sample,indicating that a chemical reaction between LiBH4and NbF5may take place during ball milling,in good agreement with our observation that the pressure inside the jar increased after ball milling. This reaction describes the reduction of the halide by the borohydride.Similar reduction reactions have been well studied for LiBH4when it is ball milled or heated together with metal halides[46].In further FTIR examination,the signature B e H vibrations of LiBH4at1125,2221,2296,and2380cmÀ1 were observed in both the pristine and the doped samples (Fig.1(b)),indicating the presence of LiBH4in the post-milled samples.So,the disappearance of LiBH4phase in the LiBH4e MgH2e NbF5composite sample during the XRD meas-urement may be attributed to the disorder or amorphous phases produced during ball milling.The peak at1633cmÀ1 belongs to the O e H bond,which is probably due to air and moisture contamination during the measurement[47].Com-bining the XRD and FTIR results,we believe that the as-milled pristine sample is merely a physical mixture of the two phases of LiBH4and MgH2.In contrast,the NbF5can react with LiBH4 to form Nb and LiF during ball milling in the case of the NbF5 doped sample.The Nb and F containing species formed in situ may contribute to the enhancement of thedehydrogenation (a)Fig.1e(a)X-ray diffraction(XRD)patterns and(b)Fourier transform infrared(FTIR)spectra of the(I)LiBH4e MgH2and (II)LiBH4e MgH2e NbF5samples after2h ball milling.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y38(2013)3650e3660 3652and rehydrogenation kinetics of the LiBH 4e MgH 2composite system and will be discussed below.Fig.2(a)contains non-isothermal dehydrogenation profiles of the LiBH 4e MgH 2system,with and without NbF 5,as obtained by thermogravimetry (TG).For the pristine LiBH 4e MgH 2sample,the dehydrogenation profile shows clear two-step features,which can be attributed to the decom-position of MgH 2and LiBH 4with increasing temperature.The first hydrogen desorption started at around 360 C.Further heating led to a second decomposition at 384 C,and a total-hydrogen release capacity of 11.3wt%was obtained below 500 C.Unlike the pristine sample,the NbF 5doped LiBH 4e MgH 2sample started to release hydrogen at a low temperature (w 90 C),which may be due to a possible reaction between LiBH 4and NbF 5,or reduced Nb species produced during ball milling,that forms more stable reaction products such as LiF and Nb,as indicated by the XRD results on the as-milled sample.Further heating caused its second dehydro-genation at 250 C,corresponding to the decomposition of MgH 2,which is 110 C lower than for the pristine sample.Still further heating leads to a third dehydrogenation corre-sponding to the decomposition of LiBH 4,and the total dehy-drogenation can be completed at 390 C,which is much lowerthan for the pristine sample (w 500 C).These results indicate that the first and second-step dehydrogenation of the MgH 2e LiBH 4system can be significantly improved by doping with NbF 5.The DSC curves in Fig.2(b)reflect the thermodynamic behavior of the LiBH 4e MgH 2samples with and without NbF 5doping.Four endothermic peaks were found on the DSC pro-file for the two samples.These features can be assigned to the phase transition of LiBH 4,the melting of LiBH 4,the decom-position of MgH 2,and the hydrogen desorption of LiBH 4,respectively [20].Interestingly,the LiBH 4e MgH 2e NbF 5sample has a lower phase transition and melting temperature com-pared to the pristine LiBH 4e MgH 2sample.In particular,the peak temperatures for the dehydrogenation of MgH 2and LiBH 4in the LiBH 4e MgH 2sample are significantly reduced.The peak temperature for the dehydrogenation of MgH 2and LiBH 4can be reduced from 387and 463 C to 323and 387 C,respectively.These results agree well with the TG results (Fig.2(a))and further confirm the effects of NbF 5on the decomposition of the LiBH 4e MgH 2system.3.2.Effects of hydrogen back-pressure and NbF 5on the isothermal dehydrogenation and subsequentrehydrogenation kinetics of the LiBH 4e MgH 2composite systemFig.3presents the isothermal dehydrogenation profiles of the LiBH 4e MgH 2system,with and without NbF 5doping,under static vacuum and 5.5bar hydrogen back-pressure at 400 C,respectively.Clearly,a two-step dehydrogenation reaction was observed in all cases,which would correspond to the decomposition of MgH 2and LiBH 4,respectively,agreeing well with a previous report that no simultaneous desorption of H 2from MgH 2and LiBH 4without intermediate formation of metallic Mg could be observed [22].Meanwhile,the first dehydrogenation for the pure or doped LiBH 4e MgH 2system,889092949698100(a)W e i g h t (w t %)Temperature (°C)100200300400500(b)Temperature (°C)Fig.2e (a)Thermogravimetry (TG)and (b)differential scanning calorimetry (DSC)profiles of the (I)LiBH 4e MgH 2and (II)LiBH 4e MgH 2e NbF 5samples.1020300246810H y d r o g e n r e l e a s e d (w t %)Time (h)Fig.3e Isothermal dehydrogenation profiles of theLiBH 4e MgH 2sample under (I)hydrogen back-pressure and (II)static vacuum,respectively,and LiBH 4e MgH 2e NbF 5sample under (III)hydrogen back-pressure and (IV)static vacuum,respectively,at 400 C,with the inset showing a shorter time frame.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 38(2013)3650e 36603653whether under a5.5bar hydrogen back-pressure or under vacuum,is almost the same,which may be due to the fact that 400 C is above the decomposition temperature of MgH2(Fig.2) [48],resulting in fast kinetics.The second-step dehydrogen-ation behavior of the system is somewhat different,however. For the pristine sample,the second dehydrogenation starts quickly after thefirst dehydrogenation hasfinished,and sat-uration of the dehydrogenation process can be limited to within3h if the dehydrogenation is conducted under vacuum. In contrast,if the dehydrogenation is conducted under5.5bar hydrogen back-pressure,the dehydrogenation nearly stops after the completion of thefirst-step dehydrogenation,and an incubation period of about9.5h was observed before the beginning of second-step dehydrogenation,with most hydrogen desorbed between9.5and24h.Obviously,the sys-tem presents much faster dehydrogenation kinetics when starting from vacuum.After doping with NbF5,the dehydro-genation was enhanced under either vacuum or 5.5bar hydrogen back-pressure.In the case of static vacuum,the total dehydrogenation of the doped sample can be completed within0.5h,which is2.5h shorter than for the pristine sam-ple.In particular,the addition of NbF5has a significant effect on the incubation period,as well as on the reaction kinetics of the second reaction step under 5.5bar hydrogen back-pressure,in that the second-step dehydrogenation quickly started as soon as thefirst-step dehydrogenation wasfin-ished,and almost no incubation period was observed for the NbF5doped sample.Saturation of the hydrogen release pro-cess can be limited to within1h for the NbF5doped sample, which is much shorter than for the pristine sample(w24h).To understand the dehydrogenation mechanism,struc-tural changes in the LiBH4e MgH2sample with and without NbF5doping were examined after dehydrogenation under vacuum and5.5bar hydrogen back-pressure by means of XRD and FTIR,as shown in Fig.4.It can be seen from Fig.4(a)that the major diffraction peaks for the pristine and doped samples dehydrogenated under vacuum at400 C are Mg or Li e Mg alloy(with Li e Mg alloy and Mg peaks overlapping in the XRD patterns).The formation of elemental boron is very hard to identify by means of XRD,due to the fact that the boron phase in the dehydrogenated samples that contained LiBH4is usu-ally amorphous[6].Meanwhile,some residual LiBH4,MgH2, and small traces corresponding to LiH and MgO for the pris-tine sample,and Nb and LiF for the doped sample,are also present.On the other hand,dehydrogenation under hydrogen back-pressure at400 C results in completely different fea-tures.The major phase for the pristine and doped samples after dehydrogenation is MgB2,as well as some residual LiBH4, MgH2,and small traces corresponding to LiH and MgO for the former,and Nb and LiF for the latter,which is different from the case of the samples after dehydrogenation under dynamic vacuum.Apparently,the main phases after dehydrogenation for the LiBH4e MgH2sample with and without NbF5doping depend on the environment,agreeing well with a previous report that the presence of hydrogen back-pressure evidently drives the formation of MgB2[19e22].The formation of MgB2 is kinetically limited,however,in the case of the pristine LiBH4e MgH2sample,as shown in Fig.3.An interface-controlled one-dimensional growth mechanism has been proposed for the formation of MgB2,where nucleation restriction will occur due to the significant mass transport in the composite during the reaction,the low mobility of atoms and ions,and the length scales of the phase separation[40]. Indeed,this makes sense,if one considers the long-range diffusion of the species as one of the main kinetic barriers. Moreover,the kinetic barriers were successfully overcome by introducing NbF5as shown in Fig.3.The changes in the structure during dehydrogenation for the different cases were further investigated by FTIR exami-nation,as shown in Fig.4(b).It can be seen that for the pristine sample dehydrogenated under dynamic vacuum or hydrogen back-pressure,the peaks for the B e H vibrations of LiBH4are still observed,agreeing well with the XRD results(Fig.4(a)). The B e H vibrations of LiBH4for the doped sample have almost disappeared,however,indicating that the dehydrogenation of LiBH4is promoted by the catalyst,agreeing well with the XRD results(Fig.2(a)).In addition,a new weak signal around 2475cmÀ1was also observed in the pristine sample after dehydrogenation,whether under hydrogen back-pressureor (a)Fig.4e(a)XRD patterns and(b)FTIR spectra of theLiBH4e MgH2samples after dehydrogenation under(I)hydrogen back-pressure and(III)static vacuum,and the LiBH4e MgH2e NbF5samples after dehydrogenation under (II)hydrogen back-pressure and(IV)static vacuum.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y38(2013)3650e3660 3654dynamic vacuum,and in the doped sample after dehydro-genation under dynamic vacuum.The infrared (IR)peak occurring at round 2500cm À1was also seen by Muetterties et al.from studying the decomposition products of pure LiBH 4[49].Recent theoretical and experimental investigations revealed the formation of Li 2B 12H 12as an intermediate step or as a side reaction during the decomposition of LiBH 4,whether pure or in the composite,according to LiBH 4/1/12Li 2B 12H 12þ5/6LiH þ13/12H 2/LiH þB þ3/2H 2[50,51].The-refore,this IR activity may possibly be due to trace amounts of Li 2B 12H 12.Interestingly,no Li 2B 12H 12phase was observed in the doped sample after dehydrogenation under hydrogen back-pressure.The results indicate that a synergistic effect may be present that suppresses the formation of Li 2B 12H 12.It was reported that Li 2B 12H 12was formed under hydrogen pressure 10bar but not formed under 20bar [51].Moreover,the formation of Li 2B 12H 12also influences subsequent forma-tion of MgB 2:the more Li 2B 12H 12that is formed,the less MgB 2is obtained.In this regard,the hydrogen back-pressure of 5.5bar used in this present work was not enough to suppress the formation of Li 2B 12H 12,in agreement with the result that Li 2B 12H 12was observed in the pristine sample under hydrogen back-pressure.The addition of NbF 5,however,can promote the formation of MgB 2,which also reduces the formation of Li 2B 12H 12.Therefore,the combination of NbF 5and hydrogen back-pressure can create a synergistic effect to suppress the formation of Li 2B 12H 12.Fig.5compares the isothermal hydrogenation profiles of the pristine and doped LiBH 4e MgH 2samples after dehydro-genation under static vacuum and 5.5bar hydrogen back-pressure.The rehydrogenation of the samples was con-ducted under w 55bar H 2at 400 C.Unlike dehydrogenation,rehydrogenation with faster kinetics and higher capacity was observed in the case of hydrogen back-pressure.For the pristine sample,the hydrogenation process for the case in which the dehydrogenation was conducted under 5.5bar hydrogen back-pressure could be limited to within 2.5h,and7.4wt %hydrogen was absorbed.In contrast,the hydroge-nation process in the case of vacuum was still not finished after 10h,and the absorbed hydrogen only amounted to 4.4wt %,which is 3wt %lower than in the former case.After doping with NbF 5,the rehydrogenation kinetics for the LiBH 4e MgH 2sample was also enhanced both under the hydrogen back-pressure and under dynamic vacuum.For the doped sample,6.0wt %hydrogen can be absorbed within 10h in the case of dynamic vacuum,which is larger than for the pristine sample under the same conditions,but is lower than for the doped sample if dehydrogenation was conducted under 5.5bar hydrogen back-pressure.As also shown in Fig.5,for the case of hydrogen back-pressure,a hydrogen absorption capacity of 6.8wt %was reached in 15min,and saturation of the ab-sorption could be achieved within 1h with a capacity of 7.6wt %,which is shorter than for the pristine sample.These results indicate that the presence of hydrogen back-pressure during dehydrogenation is favorable for the hydrogenation kinetics and capacity of the LiBH 4e MgH 2system thereafter,and the kinetics and capacity can be further improved with catalyst doping.XRD and FTIR analyses were carried out on the dehydro-genated LiBH 4e MgH 2samples with and without NbF 5doping under hydrogen back-pressure and static vacuum after hy-drogenation at 55bar and 400 C to determine the rehy-drogenation products,as shown in Fig.6.For the doped sample,Nb,LiF,and MgH 2can be observed in the cases of both vacuum and hydrogen back-pressure.LiBH 4was not detected in the former case,however,although it was observed in the latter (Fig.6(a)).Typical features of the [BH 4]Àgroup can be observed in the spectra for both cases,where the B e H bending vibration (w 1125cm À1)and B e H stretching vibrations (2221,2296,and 2380cm À1)were detected,indicating the recombi-nation of MgH 2and LiBH 4(Fig.6(b)).In addition,the vibration corresponding to the [B 12H 12]2Àgroup was observed in the case of vacuum,but disappeared in the case of hydrogen back-pressure.For the pristine sample,LiBH 4and MgH 2phases were observed in the cases of both vacuum and hydrogen back-pressure.Meanwhile,the vibrations of the [BH 4]Àgroup were further confirmed in the FTIR spectra (Fig.6(b)).In addition,the vibration corresponding to the [B 12H 12]2Àgroup was also observed in both cases.It is not clear that Li 2B 12H 12was generated during rehydrogenation,because the features of the [B 12H 12]2Àgroup were also observed in the dehydro-genated sample.The rehydrogenated sample may consist of two species,however,viz.,MgH 2and LiBH 4,indicating that part of the dehydrogenated product reverses back to the starting reactants with hydrogen absorption.bined effects of hydrogen back-pressure and NbF 5on the reversible cycling of the LiBH 4e MgH 2composite systemAs discussed above,it has been clearly shown that the hydrogen back-pressure facilitates the formation of MgB 2,which plays a crucial role in enhancing the absorption kinetics and increasing the reversible hydrogen storage capacity of the LiBH 4e MgH 2system.Therefore,the cycling performance of the LiBH 4e MgH 2system with and without NbF 5was further characterized by dehydriding the samples at 400 C under2468H y d r o g e n U p t a k e (w t %)Time (h)Fig.5e Isothermal hydrogen absorption profiles of LiBH 4e MgH 2samples after dehydrogenation under(I)static vacuum and (III)hydrogen back-pressure,and the LiBH 4e MgH 2e NbF 5samples after dehydrogenation under (II)static vacuum and (IV)hydrogen back-pressure.i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 38(2013)3650e 36603655。