la1007846_Hierarchical Carbon Foams with Independently Tunable Mesopore and Macropore Size Distribut

/Langmuir©2010American Chemical SocietyHierarchical Carbon Foams with Independently Tunable Mesopore andMacropore Size DistributionsAdam F.Gross*and Andrew P.NowakHRL Laboratories,LLC,3011Malibu Canyon Road,Malibu,California90265Received February23,2010.Revised Manuscript Received April28,2010Hierarchical carbon foams with independently tunable mesopore and macropore size distributions were formed in a high internal phase emulsion(HIPE)template.The HIPE consists of an internal oil phase that controls the macropore dimensions and an aqueous resorcinol-formaldehyde precursor solution external phase that directs the mesopore size distribution.Once the emulsion is formed,the precursor solution is cured,fluid elements are extracted from the monolith via solvent exchange, and then the sample is pyrolyzed to create a hierarchical open-cell foam consisting of macropores with mesoporous carbon xerogel walls.Both mesopore and macropore size distributions may be independently tuned by changing the synthesis parameters.These samples have a peak in the mesopore size distribution that may be tuned to between5and8nm and macropore average diameters that may be tuned to between0.7and2.1μm.Furthermore,the0.7and2.1μm average diameter macropores have0.18and0.53μm diameter macropore windows between adjacent pores,respectively.Pore volumes up to 5.26cm3/g and electrical conductivities as high as0.34S/cm are observed after1200°C carbonization of the framework.These foams may have potential applications as3-D current collectors in batteries and as fuel cell catalyst supports.IntroductionMany porous materials may be formed from emulsion-templated solutions.1,2Emulsions are an advantageous tool for structural control because of their ease of fabrication,wide range of possible pore sizes,and scalability.Furthermore,porous silica,metal oxides, carbons,and metals have been templated with emulsions and these materials demonstrate the flexibility of this technique to form solid materials.1We are particularly interested in forming porous mate-rials in high internal phase emulsions(HIPEs)that have greater than74%internal phase volumes,which is the free space occupied by packed uniform spheres.3,4At higher than74%internal phase volume,the droplets become abutted and the degree of interconnec-tion between resulting pores increases.By forming solid materials in the continuous phase of a HIPE,an open cell foam monolith consisting of polymers,silica,or carbons may be produced.1-3,5-7 Furthermore,when the solid material formed in the external phase of the HIPE is mesoporous silica,a hierarchical macroporous and mesoporous foam with increased surface area and pore volume is obtained.6,7Our goal is to produce hierarchical macroporous carbons with mesoporous carbons walls made from inexpensive and easily removable fluidic precursors templated in HIPEs where both the mesopore and macropore dimensions may be indepen-dently tailored by altering the synthetic composition.Porous carbon materials with hierarchical macropores and mesopores have potential applications as3-D current collectors,electrodes,and catalyst supports.1,2,8-10The carbon framework provides electrical conductivity,and the macropore dimensions and interconnectivity provide control of mass transport and pore volume.1Furthermore,the mesopore structure influences the sur-face area,which is important for catalyst supports,and enables additional ion-transport channels in batteries.Finally,the carbon structure is chemically inert and will survive a variety of reaction conditions.There are many templating approaches that produce hierarchical porous carbons with different degrees of pore structure tunability.A porous carbon may be formed by carbonizing a HIPE in which the continuous phase consists of a polymerized styrene monomer.5This approach produces porous carbons with nan-ometer length scale pores in the walls that are induced as a result of densification during carbonization;however,no peak in the mesopore size distribution was shown because this approach does not actively template mesopores during synthesis.Additionally, hierarchical foams may be formed from agglomerated colloids of a carbonaceous precursor.11,12The packed colloids form macropores whereas surfactant templating in the colloids,or their densification during carbonization,produces mesopores.Pluronic surfactants may also be mixed with water and phenolic resins to form a macroporous and mesoporous hierarchical carbon structured by phase separation during hydrothermal treatment.13In another approach used by many groups,a resin may be polymerized around solid micrometer-sized polymer or silica colloids and then carbonized to form a hierarchical carbon.1,2,8-10,14-18Macropores*Corresponding author.E-mail:afgross@.(1)Zhang,H;Cooper,A.I.Soft Matter2005,1,107.(2)Yuan,Z.;Su,B.J.Mater.Chem.2006,16,663.(3)Cameron,N.R.Polymer2005,46,1439.(4)Barbetta,A.;Cameron,N.R.;Cooper,mun.2000,221.(5)Wang,D.;Smith,N.L.;Budd,P.M.Polym.Int.2005,54,297.(6)Hu,Y.;Nareen,M.;Humphries,A.;Christian,P.J.Sol-Gel Sci.Technol. 2010,53,300.(7)Carn,F.;Colin,A.;Achard,M.F.;Deleuze,H.;Sellier,E.;Birot,M.;Bakov, R.J.Mater.Chem.2004,14,1370.(8)Yu,J.-S.;Kang,S;Yoon,S.B.;Chai,G.J.Am.Chem.Soc.2002,124,9382.(9)Chai,G.S.;Yoon,S.B.;Yu,J.S.;Choi,J.H.;Sung,Y.E.J.Phys.Chem.B 2004,108,7074.(10)Lee,K.T.;Lytle,J.C.;Ergang,N.S;Oh,S.M.;Stein,A.Adv.Funct. Mater.2005,15,547.(11)Carriazo,D.;Pico,F.;Gutierrez,M.C.;Rubio,F.;Rojo,J.M.;del Monte,F.J.Mater.Chem.2010,20,773.(12)Zou,C.;Wu,D.;Li,M.;Zeng,Q.;Xu,F.;Huang,Z.;Fu,R.J.Mater. Chem.2010,20,731.(13)Huang,Y.;Cai,H.;Feng,D.;Gu,D.;Deng,Y.;Tu,B.;Wang,H.;Webley, P.A.;Zhao,mun.2008,2641.(14)Zhao,J.;Cheng,F.;Yi,C.;Liang,J.;Tao,Z.;Chen,J.J.Mater.Chem. 2009,19,4108.(15)Wang,Z.;Kiesel,E.R.;Stein,A.J.Mater.Chem.2008,18,2194.(16)Lukens,W.W.;Stucky,G.D.Chem.Mater.2002,14,1665.(17)(a)Baumann,T. F.;Satcher,J.H.Chem.Mater.2003,15,3745.(b)Baumann,T.F.;Satcher,J.H.J.Non-Cryst.Solids2004,350,120.(18)Woo,S.W.;Dokko,K.;Sasajima,K.;Takei,T.;Kanamura,K.Chem. Commun.2006,4099.Gross and Nowak Articleform after removal of the colloids,and their dimensions are adjusted by changing the initial colloid diameter.Further-more,very uniform macropores were created by first forming a colloidal crystal(instead of using agglomerated colloids)and then using this ordered solid as a template for the carbona-ceous material.2,8-10,14,17-19An additional level of order in hierarchical carbons may be induced by arranging smaller colloids around larger colloids in a templating colloidal crystal.2,18,19Mesoporosity is induced without a template in many of these colloid-templated carbons as a result of carbo-nization and densification of the resin.In a few reports,the mesopore dimensions were more precisely controlled when larger colloids were surrounded by a mesoporous carbon instead of a resin.16,17One drawback with solid colloidal particle templating is that it requires the initial synthesis of particles and creating porosity in these materials may necessi-tate subsequent etching with hydrofluoric acid if silica-based materials are used.Finally,a macroporous polymer foam may be infiltrated with mesoporous silica in order to form a hierarchical carbon.20The macropore size is controlled by the initial polymer foam pore dimensions,and the mesopore dimensions are controlled by the synthesis parameters of the mesoporous silica.Mesoporous carbon is formed through the vapor deposition of a precursor on the silica,carbonization of the precursor,and the removal of the silica template with HF. Our goal is to fabricate hierarchical porous carbon monoliths with independently tunable mesopore and macropore dimensions without templating from solid materials.By using an oil-in-water soft matter template,we are able to form an open-cell hierarchical carbon foam using inexpensive and easily removable templating materials.We template our materials with a high internal phase emulsion formed by the high-speed mixing of a silicone oil internal phase and an aqueous continuous phase.The aqueous phase contains a resorcinol-formaldehyde precursor solution that enables the tuning of mesoporosity with synthetic variables.21,22 Additionally,the macropore dimensions are tuned with oil viscosity.23A carbon foam is fabricated by blending the oil and aqueous phases,thermally curing the precursor solution,remov-ing the oil phase via solvent extraction,exchanging water for organic solvent,drying the monolith,and finally carbonizing the sample.This process results in a hierarchical macroporous foam made from mesoporous carbon xerogel.Our material differs from carbonized polymer HIPEs because the meosporosity is con-trolled by the xerogel synthetic variables rather than being a byproduct of polymer densification during pyrolysis.We believe that our material is the first hierarchical carbon monolith formed from all fluidic precursors with tunable pore size mesoporous xerogel walls.In this study,the composition of the aqueous and oil phases of the HIPE are varied to demonstrate independent tuning of the mesopore and macropore pore size distributions followed by the characterization of the resulting foams.Finally, we measure the electrical conductivity of the carbon foam mono-liths to determine if these materials could be candidates for fuel cell and battery applications.Experimental SectionAll chemicals were purchased from Aldrich and used as received.Resorcinol-formaldehylde precursor solutions were synthesized by mixing a2:1:0.002molar ratio of formaldehyde (37wt%solution in water)to resorcinol(99%)to sodium carbonate(99%)in deionized water.22The resorcinol-formalde-hyde precursor solutions contained30or40wt%organic content (formaldehydeþresorcinol mass).The resorcinol-formaldehyde precursor solutions were stirred until they dissolved and allowed to rest for1h before use in carbon foam fabrication.Carbon foams were fabricated by dissolving0.360g of sodium dodecylbenzenesulfonate surfactant(technical grade)in12.5g of a resorcinol-formaldehyde precursor solution that was then com-bined with37.5g of silicone oil(100and1000cP viscosity,Fluka brand DC200).The mixture was transferred to a Waring laboratory blender and blended for10min in all experiments.A white viscoelastic emulsion was obtained and transferred in equal fractions to three60mL Nalgene polypropylene jars.The jars were sealed and aged for72h at85°C,which resulted in bright orange resorcinol-formaldehyde foam monoliths formed in the shape of the Nalgene containers.The foams were then soaked in chloroform to extract the silicone oil.The chloroform bath was poured off and refilled with fresh solvent twice,and there was at least a4h soak time during each of the first two soaking cycles and 8h during the third cycle.The foams then were then removed from chloroform and placed in an acetone bath to exchange solvent for water in the monolith.The acetone bath was poured off and refilled two times with at least2h between each soaking cycle. Finally,the foam monoliths were removed from the acetone bath, allowed to dry in air,and heated in a tube furnace under flowing nitrogen from room temperature to800°C at2.6°C/min and maintained at800°C for6h to pyrolyze the resorcinol-formalde-hyde gel.In one case,a foam was pyrolyzed under flowing nitrogen and heated from room temperature to1200°C at2°C/ min and maintained at1200°C for6h in order to measure the effect of different pyrolysis conditions.There are no significant chemical hazards encountered during the preparation of the carbon foams described in this study.Samples carbonized at800°C are denoted throughout this article on the basis of the organic loading of the xerogel and the viscosity of the silicone oil.For example,a40%/1000sample was made by blending a40wt%organic resorcinol-formaldehyde precursor solution with1000cP silicone oil.The water/oil volume ratios of the40%/100,40%/1000,and30%/1000foams were 23.5:76.5,23.5:76.5,and23.9:76.1,respectively.The ratio is slightly higher for the30%/1000emulsions as compared to that of the40%/100and40%/1000emulsions because of the lower density of the30wt%resorcinol-formaldehyde precursor solu-tion.If the sample was carbonized at a temperature other than 800°C,then the carbonization temperature is listed after the orga-nic loading of the xerogel and the viscosity of the silicone oil. The sample pore size,surface area,and pore volume were characterized with N2adsorption and Hg intrusion at Micro-meritics Analytical Services(Norcross,GA).N2data were ana-lyzed using the Brunner-Emmett-Teller(BET)and Barrett-Joyner-Halenda(BJH)methods.24Hg intrusion experiments were sensitive to pores between3nm and360μm in diameter. SEM images were acquired on a Hitachi S-4800.Electrical conductivity was measured with an HP34401A multimeter using a four-wire probe.Current and voltage probes were placed on opposite faces of a carbon foam monolith,and the resistance was measured.This resistance was converted to conductivity by inverting the resistance value and dividing by the thickness of the monolith.Results and DiscussionCarbon foams were synthesized as described in the Experi-mental Section from a HIPE template consisting of an external(19)Chai,G.S.;Shin,I.S.;Yu,J.S.Adv.Mater.2004,16,2057.(20) Alvarez,S.;Fuertes,A.B.Mater.Lett.2007,61,2378.(21)Al-Muhtaseb,S.A.;Ritter,J.A.Adv.Mater.2003,15,101.(22)Li,W.;Lu,A.;Weidenthaler,C.;Sch€u th,F.Chem.Mater.2004,16,5676.(23)Mason,T.G.;Wilking,J.N.;Meleson,K.;Chang,C.B.;Graves,S.M. J.Phys.:Condens.Matter2006,18,R635–R666.(24)Lowell,S.;Shields,J.E.;Thomas,M.A.;Thommes,M.Characterization of Porous Solids and Powders:Surface Area,Pore Size and Density;Springer: Dordrecht,The Netherlands,2006.Article Gross and Nowak phase resorcinol-formaldehyde precursor solution and an internaloil phase.Carbon xerogel is formed in the external phase,and themesoporosity is controlled by the resorcinol-formaldehyde pre-cursor solution composition.The water fraction of the resorcinol-formaldehyde precursor solution acts as a porogen and templatesthe mesopores.Additionally,the organic fraction and amountof catalyst in the resorcinol-formaldehyde precursor solutioncontrol the micropore structure and shrinkage during dryingand pyrolysis.(Smaller amounts of catalyst relative to resorcinolcreate microporosity in the carbon particles that make up thexerogel and better resist shrinkage during drying.21)Althoughthe cured resorcinol-formaldehyde gel from the continuousphase of the HIPE provides structural rigidity in the car-bon foam,the oil phase creates regions where the resorcinol-formaldehyde precursor solution is excluded to produce macro-pores.3Additionally,the oil droplets are abutted,which resultsin windows forming between the macropores and creates an open-cell foam.We refer to these openings between macropores as“macropore windows”to differentiate them from the largermacropores.A conceptual diagram of the HIPE is shown inFigure S1.The HIPE that templates the carbon foam is formed using highshear mixing,and both the macropore and macropore windowdimensions are controlled by the oil viscosity.Higher oil viscosityresults in increased viscous resistance against oil droplet deforma-tion and separation into larger droplets.Thus a lower interfacialsurface area is created between oil and water for a given amountof mechanical energy deposited into the system(with the samemixing speed used in all experiments).25This reduced surface arearesults in larger droplets,larger macropores,and larger macro-pore windows.We varied both the oil viscosity and resorcinol-formaldehyde precursor solution composition to demonstratecontrol of both mesopore and macropore textural properties.Sample Fabrication.The monolithic structures of an as-synthesized resorcinol-formaldehyde HIPE-templated foam andof a subsequently800°C pyrolyzed carbon foam are shown inFigure S2.The foams take the shape of the cylindrical jars inwhich the emulsions are cured and maintain this shape with someshrinkage during carbonization.SEM images of the carbon foamsare shown in Figures1(top)and2and demonstrate that the foamretained the HIPE structure through carbonization.The macro-pores reproduce the morphology of the templating silicone oilphase with macropore windows resulting from abutted droplets.Similar openings between macropores are also observed incolloidally templated carbons because adjacent templates aretouching.2,8-14,17,18The higher-magnification image of themacropore wall in Figure1(bottom)clearly shows the hierarchi-cal structure of the mesoporous carbon xerogel forming the wallsof the macropores.SEM images in Figures2and S3of the30%/1000and40%/1000materials show larger macropores than observed in the40%/100sample and demonstrate that the macropore size may be tuned with the silicone oil viscosity. Higher-resolution images of the xerogel structure,such as in Figure1(bottom),show no discernible differences for all three samples.Additionally,multiple SEM images for each sample were used to calculate the mean macropore diameter by averaging the diameters of at least60macropores.When100cP silicone oil is used in the40%/100foam,the macropores have a mean diameter of0.7μm,whereas both the30%/1000and40%/1000samples have a mean macropore diameter of2.1μm.This represents a 3-fold increase in the macropore mean diameter from a10-fold increase in oil viscosity.To determine if the macropore pore diameters between samples were distinct,the statistical signifi-cance between the means of macropore distributions was mea-sured by applying Welch’s t test for unequal variances for all sample pairs.Differences in the means for each macropore distribution with differing oil viscosity(40%/100vs30%/1000 and40%/100vs40%/1000)were both found to be statistically different(two-tailed p<0.0001)whereas macropore populations between identical oil viscosity samples(30%/1000vs40%/1000) were not(two-tailed p=0.97).Porosimetry.The carbon foams fabricated in this study were characterized with Hg intrusion and N2adsorption to understand the pore structures.Hg intrusion is most sensitive to pores with diameters>50nm whereas the small size and gaseous delivery of N2atoms make their adsorption best suited to characterizing mesopores and micropores.24Pore size distributions for all800°C Figure1.SEM images of a40%/100carbon foam showing the HIPE-templated structure of the carbon foam.Scale bars are shown in the images.Macropores and macropore windows formed from abutted silicone oil droplets are observed in the top image. This image is indistinguishable from a HIPE-templated polymer, demonstrating that the original emulsion structure is retained.3 The mesoporous xerogel structure of the walls is observed under higher magnification(bottom).Figure2.SEM images of40%/1000(top)and30%/1000(bottom) carbon foams.Scale bars are shown in the images.The structure of a high internal phase emulsion is observed in both samples.(25)Meleson,K.;Graves,S.;Mason,T.G.Soft Mater.2004,2,109.Gross and Nowak Articlepyrolyzed sample were derived from Hg intrusion data(Figure S4)and are shown in Figure3.The peak positions for the30%/ 1000and40%/1000samples are both at0.53μm,and the peak for the40%/100sample is at0.18μm.Additionally,the pore size distributions for30%/1000and40%/1000samples overlap almost perfectly and are insensitive to resorcinol-formaldehyde precursor solution composition.We note that these pore dimen-sions are much smaller than those observed in SEM,which is due to Hg intrusion measuring the smallest openings between pores.24 These openings are the macropore windows seen in Figures1,2, and S3,and the macropore window dimensions observed in SEM images are in good agreement with the Hg intrusion data.In addition,the macropore window diameters have the same∼3-fold increase in diameter from a10-fold increase in oil viscosity observed for macropores in SEM.Because macropore and macropore window dimensions were tuned with silicone oil viscosity but not with the resorcinol-formaldehyde precursor solution,we conclude that the silicone oil viscosity alone mainly controls macropore and macropore window structure.All textural parameters from Hg intrusion are presented in Table1.All carbon foams have large pore volumes and porosities; however,the40%/100sample is the most distinct from the other samples.The40%/100sample is templated with a volume of silicone oil that is equal to that used for other samples;however,it is formed from lower-viscosity oil that is sheared into smaller droplets with an increased surface area to volume ratio during HIPE formation.The greater surface area of the smaller oil droplets should produce a carbon foam with greater surface area, which agrees with the data in Table1.Additionally,the smaller droplets result in the aqueous phase being distributed over a larger interfacial area,which yields thinner xerogel walls.The larger Hg intrusion pore volume of the40%/100sample may result from abutted droplets more easily producing macropore windows between thinner walls and creating more open space as compared to the thicker walls in the40%/1000and30%/1000samples.N2adsorption BJH pore size distributions calculated from N2 adsorption isotherms(Figure S5)are shown in Figure S6for all 800°C pyrolyzed samples.The40%/100and40%/1000samples have very similar pore size distribution peak positions,and larger pores are found in the30%/1000sample as shown by the1.33-fold larger pore size distribution peak position.This demonstrates that the composition of the aqueous phase,and not the templating oil, mainly controls mesopore pore size distribution peak positions. We note that the oil viscosity has some effect on mesopore structure as shown by comparing the100and1000cP oil-templated samples in Figure S6.The40%/100sample has some mesopore volume above20nm,but the pore volumes of the40%/ 1000and30%/1000samples are negligible in this region.This may be explained by the thinner macropore walls in the40%/100 sample having less internal volume as compared to that in the 40%/1000and30%/1000samples.There will be more xerogel material exposed to the oil-water interface in the40%/100 sample,and with fewer surrounding carbon particles on the interface,some surface pores may have diameters greater than 20nm.The calculated N2adsorption textural parameters of all samples are presented in Table2.All foams are microporous as well as mesoporous.The presence of micropores is confirmed by the t-plot micropore volume derived from the N2adsorption data(Table2).This micropore volume is formed during pyrolysis in the carbon particles that make up the mesopore walls.26Thus,these carbon foams have two levels of hierarchy:(1)the microporous carbon particles that make up the mesoporous xerogel and(2)the mesoporous xerogel composing the walls of the macropores. The data in Tables1and2demonstrates that the resorcinol-formaldehyde precursor solution controls the mesopore diameter but has little effect on the macropore and macropore window structure.Although the silicone oil viscosity controls the macro-pore volume and size,it minimally affects the peak position of the mesopore pore size distribution.Thus,the mesopore and macro-pore/macropore window dimensions in these hierarchical carbon foams may be independently adjusted.Furthermore,the macro-pore and macropore window average diameters should be tunable between0.7and2.1μm and0.18and0.53μm,respectively,by incorporating oil into the HIPE with the appropriate viscosity between100and1000cP(with200,350,and500cP currently available).Larger or smaller macropores and macropore win-dows than those made in this study are likely accessible using silicone oils with viscosity greater than1000cP or less than100cP. The mesopore size should also be tunable over a wider range of values than those shown in this study by using alternative aerogel or xerogel materials as discussed later in this article.We note that the mesopore pore size distribution peak posi-tions,pore volumes,and surface areas are significantly smaller than those observed in neat xerogels made with the same recipe. (These materials have also been referred to as aerogels in the literature,and they have sample compositions and processing conditions that are similar to those of our materials.22,27)The mesopore textural parameters may have been altered by the curing temperature profile or the presence of surfactant.The carbon foams in this study were cured for3days at85°C in contrast to the1day at room temperature,1day at50°C,and1 day at90°C previously used for a neat xerogel in ref22.When a neat40wt%organic xerogel control solution was cured for3 days at85°C and solvent exchanged with chloroform and acetone in an identical manner to that for the carbon foams,we obtained the same pore volume of1.39cm3/g and a similar surface area of 641m2/g as found in a traditionally cured sample.28However,the peak position in the pore size distribution increased from25 to51nm.Thus,the different curing temperature profile does alterFigure3.Hg intrusion pore size distributions of800°C pyrolyzedcarbon foams.Samples are indicated by the legend on the graph.Hg intrusion is sensitive to the smallest opening between pores,andthis data is indicative of the macropore windows.The40%/1000and30%/1000carbon foams were templated with higher-viscositysilicone oil than was the40%/100sample,and this results in largermacropores and macropore windows.This data demonstrates thatthe macropore window pore size is insensitive to the composition ofthe resorcinol-formaldehyde precursor solution and is tunable withsilicone oil viscosity.(26)Gavalda,S.;Gubbins,K.E.;Hanzawa,Y.;Kaneko,K.;Thomson,K.T.Langmuir2002,18,2141.(27)Wu,D.;Fu,R.;Dresselhaus,M.S.;Dresselhaus,G.Carbon2006,44,675.(28)Gross,A.F.;Vajo,J.J.;Van Atta,S.L.;Olson,G.L.J.Phys.Chem.C2008,112,5651.Article Gross and Nowakthe pore structure and probably contributes to differences in morphology,but it cannot explain the decrease in the pore volume,surface area,or pore size distribution peak position found in Table2.Another possible reason for the differences in mesopore morphology is that our samples contain2.8wt% sodium dodecylbenzenesulfonate surfactant in the resorcinol-formaldehyde precursor solution,which is not present in neat carbon xerogel syntheses.Equal and lower surfactant levels are known to densify aerogels/xerogels and result in lower pore volumes,pore diameters,and surface areas than observed in unmodified materials.27,29Thus,the large amount of surfactant in the HIPEs is likely the primary cause of the morphological changes that we observe in the xerogel framework.In the HIPEs, a significant fraction of surfactant is isolated at the oil-water interface and is not available to alter xerogel structure.Because of uncertainty about the distribution of surfactant between the oil/ water interface and the aqueous phase,we chose not to run a control experiment using only surfactant and our resorcinol-formaldehyde precursor solution because the interpretation of results would be not be meaningful.In addition to the previously discussed samples,we attempted to form a30%/100material for this study and obtained a resorcinol-formaldehyde foam that partially collapsed after sol-vent exchange and ambient drying.We measured a Hg intrusion pore volume of only1.30cm3/g in the30%/100carbonized foam, and the macropore walls buckled during drying as shown in the SEM image in Figure S7.Because lower organic loading xerogels contain less material to resist shrinkage and because the macro-pore walls are thinner in samples made with100cP silicon oil,it is understandable that the30%/100sample collapsed.This indi-cates that our method of forming carbon foams is limited by the strength of the xerogel walls.We believe our method could overcome these limitations and produce macroporous carbon foams with a larger range of densities and average mesopore diameters by incorporating supercritically dried carbon aerogel formulations or by reducing the amount of surfactant during HIPE fabrication.27,30Supercritically dried carbon aerogel for-mulations may allow the formation of lower-density carbon foams because of their lower organic content and gentler drying procedure.Additionally,reducing the surfactant concentration may allow larger mesopores.According to ref27,less surfactant in our xerogel formulation should increase the peak position of the mesopore size distribution up to values as large as25nm.This approach will work only if the surfactant preferentially migrates to the oil-water interface and sufficient surfactant is present to enable HIPE formation.A40%/100sample was pyrolyzed at1200°C where partial graphitization and sintering of the framework occur to determine the maximum possible electrical conductivity of our materials.21 After1200°C pyrolysis,the sample monoliths remained intact and SEM images in Figure S8shows that the macropore, macropore window,and xerogel structure were all maintained. Elevated firing temperatures are known to degrade the surface area and pore volume of xerogels and we characterized the porosity of this foam to quantify any reduced porosity.31-33 The macropore textural characteristics are shown in Table1, and a comparison of the macropore pore size distributions from 40%/100foams pyrolyzed at800and1200°C is shown in Figure S9.The40%/1001200°C sample shows a reduced macropore window diameter,which is expected because of the sintering of the carbon framework,and a slightly decreased foam density,which has been observed in another report on high-temperature-fired porous carbons.34However,the sample showed an unexpected large increase in macropore volume and surface area.We hy-pothesize that closed macropores originally formed around non-abutted oil droplets and may have opened as a result of the1200°C treatment of the carbon framework.The mesopore textural data for the40%/1001200°C sample is shown in Table2,and a comparison of the mesopore pore size distributions from40%/100foams pyrolyzed at800and1200°C is shown in Figure S10.The increase in the peak position of the mesopore size distribution and the reduced micropore volume fraction are both expected from the sintering of carbon particles that template the mesopores.21,31-33However,the vastly increased surface area and pore volume do not agree with known trends for surfactant-free carbon aerogels and xerogels.31,33The40%/100 1200°C sample is different from materials in previous reports on sintering effects because it contains surfactant in the synthesis solution,which results in denser networks of carbon particles forming the mesoporous framework.27,29The more closely packed carbon particles may form inaccessible pores,and the modifica-tions in the carbon framework that occur at1200°C may enableTable1.Calculated Textural Parameters of Carbon Foams from Hg Intrusion DatasampleHg intrusionsurface area(m2/g)Hg intrusionpore volume(cm3/g)foam density(g/cm3)skeletaldensity(g/cm3)porosity(%)Hg intrusion pore sizedistribution peak position(μm)40%/100155 4.370.197 1.41860.18 40%/1000102 3.270.263 1.88860.53 30%/1000118 3.220.262 1.68840.53 40%/1001200°C254 5.260.166 1.34870.15 Table2.Calculated Textural Parameters of Carbon Foams from N2Adsorption DatasampleN2adsorptionsurface area(m2/g)N2adsorption single-pointadsorption total porevolume(cm3/g)N2adsorption t-plotmicropore volume(cm3/g)N2adsorptionV micropores/V totalN2adsorption BJH poresize distribution peakposition(nm)40%/1001870.1930.0330.17 5.3 40%/10001740.1440.0350.24 5.7 30%/10002470.2490.0420.177.6 40%/1001200°C5930.5770.0680.127.6(29)Worsley,M.A.;Satcher,J.H.;Baumann,T.F.J.Non-Cryst.Solids2010, 356,172.(30)Liu,N.;Zhang,S.;Fu,R.;Dresselhaus,M.S.;Dresselhaus,G.J.Appl. Polym.Sci.2007,104,2849.(31)Zanto,E.J.;Al-Muhtaseb,S.A.;Ritter,J.A.Ind.Eng.Chem.Res.2002,41, 3151.(32)Gross,J.;Alviso,C.T.;Pekala,R.W.Mater.Res.Soc.Symp.Proc.1996,431, 123(also available at /bridge/product.biblio.jsp?osti_id=231318).(33)Lin,C.;Ritter,J.A.Carbon2000,38,849.(34)Wiener,M.;Reichenauer,G.;Hemberger,F.;Ebert,H.-P.Int.J.Thermo-phys.2006,27,1826.。