Hydrodeoxygenation of acetic acid in a microreactor
Narendra Joshi n,Adeniyi Lawal
New Jersey Center for Micro-Chemical Systems,Department of Chemical Engineering and Materials S
cience,Stevens Institute of Technology,Hoboken,NJ07030,USA
H I G H L I G H T S
c Acetic aci
d can b
e converted by HDO below3001C at atmospheric pressure.
c Reaction pathways have been constructe
d based on literatur
e reviews and product analysis.
c External mass transfer resistance was negligible at overallflow velocity of2.54m/s.
c Internal mass transfer resistance was negligible at an average particle size of113m m.
c The conversion of acetic aci
d decreases as internal reactor diameter increases.
a r t i c l e i n f o
Article history:
Received5May2012
Received in revised form
22September2012
Accepted24September2012
Available online1October2012
Keywords:
Microreactor
Heat transfer
Fuel
Energy
Mass transfer
Hydrodeoxygenation
a b s t r a c t
Acetic acid was used as a model compound for pyrolysis oil in a hydrodeoxygenation(HDO)study.HDO
of acetic acid was performed in a packed bed microreactor.The catalyst was reduced sulfided NiMo/
Al2O3.The effects of state of aggregation of acetic acid,temperature,hydrogen partial pressure,liquid
flow rate,reactor diameter,and residence time on conversion,yield,space-time consumption,and
space-time yield were investigated.External and internal mass transfer and heat transfer resistances
were also examined in the microreactor.Temperature was a major factor in HDO of acetic acid.
Many consider hydrodeoxygenation as an unattractive process due to high pressure requirements
(1050–3000psig).In this work,attempt has been made to show that HDO of acetic acid can be
conducted at atmospheric pressure with a significant conversion achieved.More acetic acid was converted
during HDO as temperature was increased at constant pressure of300psig.Conversion was much higher
for vapor phase acetic acid at atmospheric pressure than liquid phase acetic acid.HDO of gas phase acetic
acid in a blank reactor compared to a catalytic HDO showed that thermal decomposition of acetic acid did
not occur appreciably.Partial pressure of hydrogen above240psig had no effect on the conversion of liquid
phase acetic acid.Conversion of vapor phase acetic acid increased as the partial pressure of hydrogen
increased from3psig to15psig.Residence time was0.06s for a maximum conversion of liquid phase
acetic acid,whereas it was0.03s for a maximum conversion of vapor phase acetic acid.The conversion of
acetic acid for both liquid and vapor phases decreases significantly as theflow rate of acetic acid increases.
As reactor diameter increases beyond0.8mm,the conversion reduces significantly.Mass transfer
resistance was negligible at the superficial velocity of2.54m/s and at an average catalyst particle size of
113m m.Radial temperature difference in the microreactor was less than5%.
&2012Elsevier Ltd.All rights reserved.
1.Introduction
Depleting petroleum reserves,rising prices,and environmental
and political concerns have made renewable resources for transpor-
tation fuel attractive.Though there are other renewable sources of
energy such as wind,solar,and hydroelectric,biomass is the only
renewable source of energy that can be used for liquid fuel,fitting
into the current infrastructure of transportation fuel utilization.
Effective utilization of biomass as the transportation fuel may
reduce the world’s dependency on non-renewable fossil fuel.
Another important advantage of the biomass is that it contributes
no new carbon dioxide to the atmosphere(Peter,2002).
Raw pyrolysis oil from biomass can be further processed to
obtain transportation fuel.Processing alternatives include hydro-
deoxygenation,catalytic cracking,and steam reforming followed
by Fischer Tropsch synthesis.Lately more attention has focused
on hydrodeoxygenation as a means of increasing the energy
density of the product fuel.Hydrodeoxygenation of pyrolysis oil
requires about4501C to reduce most of the oxygen content in the
oil.But at this temperature an extreme degradation of oil takes
Contents lists available at SciVerse ScienceDirect
journal homepage:www.elsevier/locate/ces
Chemical Engineering Science
0009-2509/$-see front matter&2012Elsevier Ltd.All rights reserved.
/10.s.2012.09.018
n Corresponding author.Tel.:þ12012168314;fax:þ12012168306.
E-mail addresses:njoshi@stevens.edu,njoshipasa@yahoo(N.Joshi).
Chemical Engineering Science84(2012)761–771
place and it becomes a very hard,coke like material.This chemical instability is attributed to unsaturated double bonds such as acetic acid,aldehyde,ketone,and so forth which react through condensation reaction similar to the phenol-formal polymerization(Grange et al.,1996).Therefore,it is very crucial to eliminate these functions before they react to high molecular weight compounds.Experiments with oxygenated model com-pounds show that it can be accomplished by a low temperature (below3001C)hydrotreatment(Elliott and Neuenschwander, 1996;Elliott et al.,1988;Elliott and Oasmaa,1991;Grange et al.,1996).High acetic acid content(upto25%)provides corrosive nature to pyrolysis oil(Milne et al.,1997;Xu et al., 2009)which needs to be reduced for long term storage.It also contains unsaturated double bond functional group which leads to polymerization of pyrolysis oil at higher temperature(above 3001C)during hydrodeoxygenation.Therefore,converting acetic acid by hydrodeoxygenation at temperatures below3001C will increase the stability of pyrolysis oil and in turn,it is expected to reduce polymerization of pyrolysis oil significantly when hydro-treated at
temperatures above3001C.By conducting hydrodeox-ygenation of acetic acid we hope that it will provide an opportunity to investigate upgrading mechanism.
Using acetic acid as a model compound we also hope to study reaction pathways of HDO of acetic acid.An understanding of reaction pathways and kinetics is expected to be of value in process design and modeling(LaVopa and Satterfield,1987).However, studying reaction pathways and kinetics of HDO of PO is very difficult as it contains more than300oxygenated compounds. Therefore,acetic acid is selected as our model compound to investigate the hydrodeoxygenation process and reaction pathways.
Biomass is a bulky material with a very low density and scattered across geographical areas.Research has shown that transporting biomass to a central location and processing for biofuel is not economically feasible due to very high cost of transportation.The research suggest that biofuel production facilities will have to be built close to the sources of feedstock, probably within50miles(David,2008);therefore,a large scale processing plant will not be profitable.Even for a medium size biomass conversion facility,the amount of biomass required is overwhelming.It is calculated that a facility producing50million gallon biofuel per year would require a truck loaded with biomass to arrive every six minutes around the clock(David,2008).The requirements of high production
cost and very large quantity of biomass discourages people to use the biomass conversion facility with conventional macroreactor system.However,on-site and on-demand distributed system could be an alternative solution.In this regard,microreactor system could be a suitable alternative.
Microreactor system consists of devices that are miniature in sizes and scaling up is done by increasing number of channels within a reactor unit as well as by increasing number of reactor units.The conventional method of commercializing a new process starts at the laboratory scale where results are collected which are then used in a pilot plant to obtain critical parameters.The information collected from pilot plant is used to design and build the larger production unit.In microreactor technology,scale-up can be achieved by numbering up straight from laboratory scale bypassing a costly pilot plant(Jenck,2009).There are number of other benefits of using microreactor system.Study showed that heat transfer coefficient reached upto25,000W/m2K in microdevices exceeding those of conventional heat exchangers by at least one order of magnitude(Schubert et al.,1998).In micromixers typicalfluid layer thickness can be set to a few tens of micrometers,consequently,mixing time in micromixers amount to milliseconds which is hardly achievable using stirring equipment and other conventional mixers(Branebjerg et al., 1996;Knight et al.,1998).A comparison of microreactor with conventional trickle-bed reactor for the hydrogenation of styrene showed that the throughput of the liqu
id for microreactor is eight times higher than that of the trickle-bed reactor(Nijhuis et al.,2003).As a result of enhanced mass transfer,the hydrogen concentration on catalyst is significantly higher than on the trickle-bet catalyst(34mol/m3compared to22mol/m3).Micro-reactors possess ultra-low transport resistances,therefore mass diffusion and heat transfer are extremely fast,resulting in rapid thermal and reaction equilibrium(Besser et al.,2003).High yield, improved product quality,better selectivity and safe operation are attainable due to very high heat and mass transfer(Halder et al.,2007;Okafor et al.,2010;Tadepalli et al.,2007;Voloshin and Lawal,2010).Mixing in the microchannels is attainable only by inter-diffusion of reactants due to laminarflow(Adeosun and Lawal,2005),however due to short transverse diffusional dis-tance,rapid and effective mixing is attainable in the microreactor which can quickly bring reactants in contact with catalyst in a heterogeneous reaction(Hessel et al.,2001).
Enhancement of mixing in two-phaseflows can also be achieved by selecting appropriate inlet T-orientations to provide a short slug length.Studies have shown that introduction of gas and liquid feeds head to head or perpendicular to each other with the liquid stream parallel to the microchannel markedly improves the mixing(Qian and Lawal,2006).
There are other potential benefits of microreactors regarding application which are listed below(Ehrfeld et al.,2000)
1.Earlier start of production at lower costs.
2.Easier scale-up of production capacity.
3.Smaller plant size for distributed production.
4.Lower costs for transportation,materials and energy.
5.Moreflexible response to market demands.
Many researches have considered microreactor technology as a potential for process intensification.Process intensification involves the development of innovative methods and devices that,in comparison to existing approaches,offer the chance of a dramatic improvement in the quality of production,substantial reduction in the ratio of equipment size to production capacity, and significant drop in the consumption of energy and production of waste.For these reasons,microstructured devices and compo-nents have an important role to play(Matlosz et al.,2009).
Commercially available catalysts such as CoO/MoO3and NiO/MoO3on Al2O3support used for removing sulfur,nitrogen, and oxygen from petrochemical feedstock are generally selected for hydrodeoxygenation process as well.According to research conducted at Pacific Northwest National L
aboratory,the sulfided form of CoO/MoO3and NiO/MoO3are much more active for hydrodeoxygenation than the oxide form(Elliott,2007).Sulfida-tion creates active sites that can play a role in the rupture of carbon-heteroatom bond(Senol,2007).In this study we used the sulfided form of NiO/MoO3on Al2O3.
Model compounds are chosen instead of pyrolysis oil for better understanding and control of the HDO reaction process.Nim-manwudipong et al.used2-methoxyphenol as a model com-pound of lignin-derived pyrolysis oil in the presence of hydrogen to elucidate the reaction network and to predict oxygen removal (Nimmanwudipong et al.,2011).The authors have shown that the catalytic conversion of2-methoxyphenol in the presence of hydrogen catalyzed by Pt/Al2O3involves three major classes of reactions:hydrogenation,hydrodeoxygenation,and transalkyla-tion.Xu et al.has demonstrated that acetic acid can be converted to ethyl acetate using reduced Mo-10Ni/g-Al2O3via hydrodeox-ygenation at473K and3MPa hydrogen pressure(Xu et al.,2009). They have proposed that ethyl acetate was produced from acetic acid via a three step reaction in which produced aldehyde was
N.Joshi,A.Lawal/Chemical Engineering Science84(2012)761–771 762
converted to ethyl alcohol by hydrogenation which then reacted with acetic acid to form ethyl acetate.Similarly,LaVopa and Satterfield studied the hydrodeoxygenation of dibenzofuran on a sulfided
NiMo/Al2O3at3601C and7.0MPa and found that single-ring cyclohexane predominated(LaVopa and Satterfield,1987). They also found that the catalyst in the oxide form had lower activity and double-ring products predominated over single-ring compounds.Also,Li luded that during hydrodeoxygena-tion of1-Naphthol catalyzed by sulfided Ni–Mo/g-Al2O3aromatic ring hydrogenation and direct oxygen extrusion(HDO)occur in parallel(Li et al.,1985).
The objective of the research presented here was to evaluate hydrodeoxygenation of acetic acid in a packed bed microreactor. Effects of various reaction variables such as liquidflow velocity, residence time,hydrogen partial pressure,temperature,state of aggregation of acetic acid,and reactor diameter on conversion, yield,and space-time yield were studied.
2.Experimental
2.1.Materials
Presulfided NiO/MoO3/Al2O3catalyst obtained from Albemarle (sulfided and supplied by Eurecat USA),Houston,Texas was ground and sieved to obtain particles with diameters in the range of 75–150m m.An average surface area of the sieved catalysts was 164m2/g and average pore diameter was106˚A.The surface area and pore diameter were obtained by using multipoint BET techni-que and t
he instrument used was Quantochrome Autosorb-1. The catalyst was reduced with5.0sccm of hydrogen at593K and 3.45MPa for2h.The average surface area of the reduced catalysts was209.0m2/g and average pore diameter was92.0˚A.ACS regent grade(conc.99.7%)acetic acid was purchased from Pharmco Inc. The gas used was extra dry hydrogen from Praxair.Nitrogen was used as tracer to perform a material balance.
2.2.Experimental setup
A HPLC pump(Laboratory Alliance Series III)was used to control theflow rate of acetic acid.Massflow controllers(Porter Model201)were used to control theflow rates of hydrogen and nitrogen.Ranges of superficial velocities of acetic acid,hydrogen and nitrogen gases used were0.0011–0.0065m/s,0.54–4.71m/s, and0.36–2.75m/s respectively.The liquid and gas phases were combined in a T-junction mixer(Upchurch)with508Â10À6m through-hole.Reynolds number for the combinedflow was less than100for all experiments indicating laminarflow.Thefluids exiting from the T-junction exhibited a Taylorflow pattern with a liquid slug length in the range0.001–0.003m,whereas gas bubble length varied from0.001to0.005m.A microreactor was prepared from a0.0016m(1/16in.)316stainless steel tubing with 762Â10À6m internal diameter,and was gravityfilled with cata-lyst.The total length of the packed bed microreactor varied from 0.025to0.18m.Hastelloy micronfilter-cloth(200Â1150mesh, Uniqu
e Wire Weaving Co.,Hillside,New Jersey)was placed at the ends of the reactor to retain the catalyst.The reactor system was pressurized using a back pressure regulator(GO Regulator Co.). The entrance and the exit pressures of thefluids(liquid and gas combined)in the reactor were measured.The pressure drop along the reactor varied from0.07MPa to0.2MPa depending upon reactor length.Acetic acid was vaporized prior to entering to a reactor using a vaporizer heated to a temperature above the boiling point(1181C)of acetic acid.The product stream was condensed using a chiller(model NESLA
B RTE7,Thermo Fisher Scientific).2.3.Analysis
The water content in the HDO product was analyzed using Volumetric Karl Fischer(KF)Titration Workstation(Model375, Denver Instrument)with the use of hydranal reagents obtained from Sigma-Aldrich.The titrant used was hydranal titrant-2while the working medium was hydranal solvent.Before the actual titration,the hydranal solvent was titrated to dryness in a drift determination step.This step removes moisture from solvent, electrode,and titration vessel.The HDO product was added using a syringe through the septum on the KF cell.Analysis of liquid product was conducted in a HPLC(Shimadzu series:mobile phase degasser[DGU_20A5],pump station[LC-20AT],auto-sampler [SIL-20AC],and reflective index detector[RID-10A])equipped with BioRad Aminex HPX-87H column.Th
e mobile phase con-sisted of0.007N aqueous H3PO4with an isocraticflow(flow rate of6Â10À7m3/min).The gas phase was analyzed for CO2,CO, and CH4using Varian450GC with Hayesep N and Mol sieve5A columns in series.
HDO of acetic acid involves complex reactions consisting of a series of reactions forming acetaldehyde,ethanol,and ethyl acetate and parallel reactions forming acetone,carbon dioxide, carbon monoxide,and methane.Acetic acid reacts with hydrogen forming acetaldehyde and H2O.A model for the formation of acetaldehyde is suggested by(Grootendorst et al.,1994)with following pathway:
CH3COOHþH2¼4CH3CHOþH2O,
D H1¼1.27kJ/mol(Reaction1)
and(Xu et al.,2009)suggested the formation of ethyl acetate with following reaction mechanism:
CH3CHOþH2¼4CH3CH2OH,D H1¼À80.98kJ/mol(Reaction2) CH3CH2OHþCH3COOH¼4CH3COOCH2CH3þH2O,
D H1¼À4.95kJ/mol(Reaction3)
Acetone,CO2,CH4,and CO are formed according to assump-tions of following reaction mechanisms as provided by(Blake and Jackson,1968;Nguyen et al.,1995;Pestman et al.,1997)
2CH3COOH¼4CH3COCH3þCO2þH2O,reaction kinetics mechanism期刊
D H1¼37.64kJ/mol(Reaction4) CH3COOH¼4CH4þCO2,D H1¼15.13kJ/mol(Reaction5) CH3COOHþH2¼4CH4þCOþH2O,
D H1¼12.3kJ/mol(Reaction6)
Based on the analysis of the product and the literature reviews, reaction pathways of HDO of acetic acid are proposed as shown in Fig.1.
The reaction of acetic acid was characterized in terms of conversion,space-time yield(STY),space-time consumption (STC),and yield which are defined as follows:
Conversion%ðÞ¼
Amount of acetic acid reacted
Amount of acetic acid fed
Â100%ð1ÞYield%ðÞ¼
Amount of products formed
Theoretical amount of product that could be formed
Â100%
ð2ÞSpaceÀtimeyield STY
ðÞrate of formation of product,g=g cat h
ÀÁ¼
Amount of product formed
Amount of catalystÂtime
ð3Þ
N.Joshi,A.Lawal/Chemical Engineering Science84(2012)761–771763
3.Results and discussions
3.1.Dependence of conversion and yield on temperature
In hydrodeoxygenation of acetic acid,temperature was found to be the most important parameter in removing oxygen.A series of experiments was conducted to study the dependence of conversion of liquid phase acetic acid and yield of products on temperature at constant pressure of 300psig.Residence time was kept constant by varying reactor length (catalyst loading)to compensate the change in gas velocity.The results shown in Fig.2indicate that the conversion of acetic acid and yields of acetaldehyde and ethyl acetate increase as temperature increases.It was observed that upto 2101C only acetaldehyde was formed but as the temperature increased above 2101C formations of ethyl acetate and ethanol were detected.A significant amount of acetone was detected at 4001C and 4501C.
A series of experiments was conducted with acetic acid in a vapor phase at atmospheric pressure.Acetic acid was vaporized by heating above its boiling point (1181C)prior to entering a reactor.The results shown in Fig.3indicate that the conversion of vapor phase acetic acid increases to 60percent as temperature increases from 2001C to 4501C compared to 47percent conver-sion for liqui
d phase acetic acid.A work by Xu hydrodeoxygenation of acetic acid showed that ethyl acetate was the final product at 2001C and proposed that ethyl acetate was produced via three step reactions in which acetaldehyde and ethanol were the intermediates (Xu et al.,2009).From analysis of the product we have confirmed that HDO of acetic acid produced all these compounds.Our result also showed the formation of acetone which was not mentioned in the literature by Xu et al.It is also found that the rate of conversion of vapor phase acetic acid at 2001C is comparable to the literature value mentioned by Xu et al.
Analysis of gaseous products showed that methane,carbon dioxide,and carbon monoxide were formed as the temperature increased to 4501C for catalytic HDO of vapor phase acetic acid as shown in Fig.4.Another set of experiments was conducted with acetic acid in the vapor phase without catalyst in the reactor.As the temperature increased from 2001C to 4501C formation of carbon dioxide,carbon monoxide,and methane was not observed except for a trace amount of carbon dioxide at 4001C and 4501C.Therefore,thermal decomposition did not play a role in the formation of carbon dioxide,carbon monoxide,and methane when experiments were conducted with catalytic hydrodeoxy-genation upto 4501C.An increase in conversion as the tempera-ture increases is indicative of HDO of acetic acid being highly influenced by kinetics.As temperature increases more
reactants
Fig.1.Reaction pathways of HDO of acetic acid (Based on literature review and analysis of
products).
Fig.2.Dependence of conversion and yield on temperature.Reaction conditions:2.07MPa (300psig);gas phase:hydrogen (80%)and nitrogen;liquid phase:acetic acid (Conc.99.7%);sulfided NiMo/Al 2O 3catalyst;liquid flow rate:8.33Â10À10m 3/s (0.05ml/min).
N.Joshi,A.Lawal /Chemical Engineering Science 84(2012)761–771
764
collide with high enough energy to overcome activation energy barrier to form products.Furthermore,high conversion of acetic acid can be achieved at atmospheric pressure in contrast to high pressure (70–200atm.)used by other groups for DHO of bio-oil (Bulushev and Ross,2011;Furimsky,2000;Huber et al.,2006).
A carbon balance around the reactor showed that 14%of carbon was unaccounted for which was assumed lost as uncon-densed acetaldehyde,ethanol,ethyl acetate,and acetone.From an oxygen balance,97%of the oxygen is accounted for,of which 47%is in carbon dioxide,carbon monoxide,and water.As a result of HDO of acetic acid,oxygen to carbon ratio reduced from 1.33in the feed to 1.0in the product.
3.2.Dependence of conversion and yield on hydrogen partial pressure
Very high hydrogen pressure (1050psig–3000psig)require-ment due to low reactivity of some of the oxygenates in pyrolysis oil discourages many to consider hydrodeoxygenation as a viable alternative f
or biofuel production.In this work,hydrodeoxygena-tion of acetic acid at low pressure was examined for the conver-sion of acetic acid.A set of experiments was carried out at 1501C with liquid phase acetic acid in the range of 300–600psig total pressures to study the effect of inlet hydrogen partial pressure on conversion,yield and STY.Reaction temperature and residence time were kept constant.The residence time was kept constant by varying reactor length (catalyst loading).The H 2partial pressure was varied by changing total pressure.The results in Fig.5
indicate that increasing hydrogen partial pressure had no effect on conversion and yield for the selected pressure range,indicating that adsorbed hydrogen on the catalyst surface had reached a maximum (saturated)value.The increase in STY with increase in hydrogen partial pressure was due to an increase in hydrogen concentration (due to constant residence time)resulting in a higher reaction rate.
Another set of experiments was carried out at 4501C with vapor phase acetic acid in the range of 3–15psig to study the effect of inlet hydrogen partial pressure on conversion and yield.The reactions were carried out at constant temperature and residence time with acetic acid in vapor phase.Partial pressure of hydrogen was varied by changing the composition of hydrogen with nitrogen.The results in Fig.6indicate that conversion of acetic acid increases as the partial pressure of H 2increases and reac
hes a maximum value at a pressure close to 15psig.The increase in conversion could be due to increase in concentration of hydrogen on the catalyst surface.
3.3.Dependence of conversion and yield on flow rate of acetic acid Concentration of hydrogen on the catalyst surface affects the conversion of acetic acid during HDO.As mass transfer of hydrogen to catalyst surface occurs via liquid slugs,dependence of conversion and yield on flow rate of acetic acid was examined.Experiments were conducted to study the dependence of yield and conversion on flow rate of acetic acid by varying the liquid flow rate from 0.03–0.15ml/min.Other operating conditions
such
Fig.3.Dependence of conversion and yield on temperature.Reaction conditions:15psig;gas phase:hydrogen (80%),nitrogen,and acetic acid (Conc.99.7%);sulfided NiMo/Al 2O 3catalyst;acetic acid flow rate:8.33Â10À10m 3/s (0.05
ml/min).
Fig.4.Dependence of gaseous products formation on temperature.Reaction conditions:15psig;gas phase:hydrogen (80%),nitrogen,and acetic acid (Conc.99.7%);sulfided NiMo/Al 2O 3catalyst;acetic acid flow rate:0.05ml/min.
N.Joshi,A.Lawal /Chemical Engineering Science 84(2012)761–771765
版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。
发表评论