DOI: 10.1126/science.1085088
, 79 (2003);
301 Science et al.Ralph T. Yang Conditions
Desulfurization of Transportation Fuels with Zeolites Under Ambient
This copy is for your personal, non-commercial u only.
lleagues, clients, or customers by , you can order high-quality copies for your If you wish to distribute this article to others
here.following the guidelines can be obtained by Permission to republish or repurpo articles or portions of articles
): November 4, 2013 www.sciencemag (this information is current as of The following resources related to this article are available online at
www.sciencemag/content/301/5629/79.full.html version of this article at:
including high-resolution figures, can be found in the online Updated information and rvices, www.sciencemag/content/suppl/2003/07/02/301.5629.79.DC1.html can be found at:
Supporting Online Material 182 article(s) on the ISI Web of Science cited by This article has been www.sciencemag/content/301/5629/79.full.html#related-urls 1 articles hosted by HighWire Press; e:cited by This article has been
www.sciencemag/cgi/collection/chemistry Chemistry
subject collections:This article appears in the following registered trademark of AAAS.
imperial college londonis a Science 2003 by the American Association for the Advancement of Science; all rights rerved. The title Copyright American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the Science o n N o v e m b e r 4, 2013
w w w .s c i e n c e m a g .o r g D o w n l o a d e d f r o m
Desulfurization of Transportation Fuels with Zeolites Under
Ambient Conditions
Ralph T.Yang,*Arturo J.Herna´ndez-Maldonado,Frances H.Yang
Deep desulfurization of transportation fuels(gasoline,diel,and jet fuels)is being mandated by U.S.and foreign governments and is also needed for future fuel cell applications.However,it is extremely difficult and costly to achieve with current technology,which requires catalytic reactors operated at high pressure and temperature.We show that Cuϩand Agϩzeolite Y can adsorb sulfur compounds from commercial fuels lectively and with high sulfur ca-pacities(bycomplexation)at ambient temperature and pressure.Thus,the sulfur content was reduced from430toϽ0.2parts per million by weight in a commercial diel at a sorbent capacity of34cubic centimeters of clean diel produced per gram of sorbent.This sulfur lectivity and capacity are orders of magnitude higher than tho obtained by previously known sorbents.
Sulfur in transportation fuels remains a major
source of air pollution.Becau of govern-
ment mandates worldwide,refiners must pro-
duce increasingly cleaner fuels(1).The pri-
mary focus of the new regulations is the
reduction of sulfur in gasoline and diel.The
U.S.Environmental Protection Agency Tier
II regulations require reductions of sulfur in
diel from the current average of500to15
parts per million by weight(ppmw)by June
2006and reductions of sulfur in gasoline
from350to30ppmw by January2005(1).
Similarly stringent new regulations are being
implemented in Europe and Japan.
Another need for deep desulfurization is
for applications in fuel cells.Gasoline is the
ideal fuel for fuel cells becau of its high
energy density,ready availability,and safety
and ea of storage.However,to avoid poi-
soning the catalysts for the fuel processor and
tho in the electrode of the fuel cell,the
sulfur concentration should preferably be be-
low0.1to0.2ppmw.
Removal of sulfur-containing compounds
is an important operation in petroleum refin-
ing,and is achieved by catalytic process
operated at elevated temperatures(300°to
340°C)and pressures(20to100atm of H
2)
by using the Co-Mo/Al
2O
3
or the Ni-Mo/
Al
2O
3
catalyst(2).The hydrodesulfurization
(HDS)process is highly efficient in removing thiols,sulfides,and disulfides but is less ef-fective for aro
matic thiophenes and thiophene derivatives.Thus,the sulfur compounds that remain in the transportation fuels are thio-phene,benzothiophene,dibenzothiophene, and their alkylated derivatives.To reduce the sulfur content to meet the new regulations,the reactor size needs to be incread by
factors of5to15(3).Faced with the verely
high costs of compliance,a surprising num-
ber of petroleum refiners are riously con-
sidering reducing or eliminating production
of transportation fuels.
The new challenge is to u adsorption to
lectively remove the sulfur compounds
from transportation fuels.Becau adsorption
would be accomplished at ambient tempera-
ture and pressure,success in this develop-
ment would lead to major advances in pe-
troleum refining as well as for fuel cell
applications.However,all of the commer-
cially available sorbents that have been
tested for desulfurization have proven inef-
fective(3).Studies of nonconventional sor-
bents have not yielded better results(3).
A class of highly sulfur-lective and high–
sulfur-capacity sorbents have been discovered
and are reported here.This class of sorbents
binds thiophenic compounds lectively by
complexation.The new sorbents are zeolites
containing cuprous or silver cations and were
prepared by ion exchange of zeolites using
known ion-exchange procedures(4).First,
the candidates were identified in a screening
study that ud molecular orbital(MO)theory
to arch for sorbents that would bond thio-
phene more strongly than benzene(5).Here,
benzene was ud as a model compound for
aromatics,whereas thiophene reprented the
basic compound for sulfur.The results for the
adsorption bond energies are given in Table1.
The adsorption bond energies were also mea-
sured from equilibrium adsorption isotherms at
different temperatures(6)and are also given in
Table1.The agreement between the experi-
mental and theoretical values was excellent.
The results indicate that the Cuϩand Agϩ
zeolites could adsorb thiophene preferentially
over benzene and that Cuϩis stronger than
Agϩin bonding with thiophene.The natural
bond orbital analysis showed that the bonding
followed the classical picture ofcomplex-
ation:There was some donation of electron
charges from theorbital of thiophene to the
vacant s orbital of metals(known asdona-
tion)and,simultaneously,back-donation of
electron charges from the d orbitals of metals to
the*,the antibondingorbital)of
thiophene(known asback-donation).The
results also showed thatdonation is more
predominant for thiophene and theback-
donation is more important for benzene.
Then we proceeded to perform adsorption
experiments for desulfurization of commer-
cial gasoline and diel in a fixed-bed ad-
sorber that contained particles of CuY or
AgY zeolite,at ambient temperature and
pressure(7).The zeolites were prepared by
the ion exchange of NaY zeolite(Si/Alϭ
2.43)with Agϩor Cu2ϩ.The Cu(II)-
exchanged form was subquently subjected
marissa mayer
to autoreduction(8)to form Cu(I)Y.
The results with the commercial diel
(containing430ppmw sulfur)are summa-
rized in Fig.1for CuY as the sorbent in the
main bed.A thin layer of activated carbon
(15%of the bed)was ud as the guard bed
that extended the sorbent capacity of the main
heal
bed by adsorbing the largest molecules from
the fuels.The sulfur capacity was thus in-
cread by about20%by the guard bed.
However,the concentration of sulfur in the
effluent(before sulfur breakthrough)re-
mained the same without the guard bed.The
detailed sulfur breakthrough behavior is
shown in Fig.2.The detailed sulfur analysis
showed that the earliest sulfur breakthrough
appeared at a cumulative effluent volume of
15cm3/g,with approximately equal concen-
tration of4-methyl dibenzothiophene(DBT)
and4,6-dimethyl DBT.The total concentra-
tion of sulfur at this point was0.11ppmw
sulfur.The sulfur contents in effluent sam-
ples collected before the cumulative volume
of10cm3/g were below the detection limit;
0.02ppmw sulfur.Figure3summarizes the
sulfur breakthrough performance of CuY ad-
sorber for commercial gasoline and diel.
The content of aromatics was higher in the
gasoline than in the diel(approximately
30%compared with20%).The higher con-
centration of aromatics in the gasoline result-
上海网球培训
Department of Chemical Engineering,University of Michigan,Ann Arbor,MI48109,USA.
*To whom correspondence should be addresd.E-mail:yang@umich.edu Table1.Adsorption bond energies(in kcal/mol) from ab initio MO theory and from the experiment.
From
MO theory
From
experiment
Thio-
phene Benzene
Thio-
phene Benzene Cu
zeolite
21.420.522.120.5 Ag
zeolite
20.019.121.419.5
R E P O R T S
www.sciencemag SCIENCE VOL3014JULY200379
ed in more competition for the sulfur com-pounds and thus showed earlier sulfur break-through.Nonetheless,the effluent products from both gasoline and diel were suitable for fuel cell applications.
Experiments on sorbent regeneration were also performed,which showed that CuY can be effectively regenerated either thermally or with solvents.CuY was regenerated by first treating with air at 350°C (to burn off sulfur)and then by
autoreduction (of Cu 2ϩto Cu ϩ)at 450°C,and only 5%sulfur capacity was lost upon regener-ation (4).For thermal regeneration,activated carbon would not be suitable for the guard bed;activated alumina would be effective.On regen-eration with solvents,total sulfur recovery was achieved at room temperature with both sol-vents that were tested [dimethylformamide (DMF)and carbon tetrachloride (CCl 4)].How-ever,the solvents needed to be effectively re-moved by heating (at 200°C for DMF and 350°C for CCl 4).
All major commercial sorbents have been subjected to the same adsorption experiment.None of the sorbents could be ud to produce gasoline or diel at the parts per million level.For example,at a cumulative eluent volume of 5cm 3/g,the sulfur contents were 40ppmw for ZSM-5zeolite,10ppmw for activated carbon,and 5ppmw for Selexsorb activated alumina.The latter is a commercialized form of activated alumina designed specifically for liquid fuel desulfurization.The corresponding values at a cumulative eluent volume of 10cm 3/g were 190ppmw for ZSM-5,25ppmw for activated car-bon,and 6ppmw for alumina.The total sulfur capacities of the commercial sorbents are also substantially lower than tho of CuY and AgY zeolites.More recently,a sulfur-lective sor-bent was developed by Ma et al .(9)using transition metals supported on silica gel.Using a synthetic diel (containing eight compounds,including two sulfur compounds),the eluent contained Ͻ1ppmw sulfur.However,sulfur breakthrough occurred at 0.85cm 3/g eluent vol-ume,compared with the capacity of our sorbent at 34cm 3/g.Synthetic diels are much easier to desulfurize than real diel (which contains Ͼ150compounds).Thus,our sorbent is not only more sulfur-lective but also has a much higher sulfur capacity (by a factor of Ͼ40).Fuel desulfurizer is a critical element for liguid-powered fuel cell vehicles.The tar-get design of such vehicles (at 50kW)established by the U.S.Department of En-ergy (10)ts a limit of 12.5kg weight for the desulfurizer.Using 10kg of the sorbent discovered in this work would produce 340liters of diel (at Ͻ0.2ppmw sulfur),which would
provide a driving range of 5000miles.This sorbent can also be readily ud for liquid fuel desulfurization for stationary applications.
References and Notes
1.A.Avidan,M.Cullen,paper AM-01-55,prented at the National Petroleum and Refiners Association An-nual Meeting,Washington,DC,18to 20March 2001.
2.B.C.Gates,J.R.Katzer,G.C.A.Schuit,Chemistry of Catalytic Process (McGraw-Hill,New York,1979),chap.5.
3.R.T.Yang,Adsorbents:Fundamentals and Applica-tions (Wiley,New York,2003),chap.10.This refer-ence contains a detailed review of the literature of all commercial and nontraditional sorbents that have been tested,as well as the different approaches for deep desulfurization of transportation
fuels.
Fig.1.GC-FPD chromatogram of a commercial diel fuel for sulfur analysis and progression of sulfur breakthrough during diel adsorption in the bed of AC/Cu(I)-Y.Also shown is cumulative effluent volume normalized by the total weight of adsorbent.a.u.,arbitrary
units.Fig.2.Detailed sulfur breakthrough of commercial diel (with 430ppmw sul-fur)in an adsorber of AC/CuY zeolite,where C (t )is sulfur concentration at time t
.
Fig.3.Breakthrough of total sulfur from an adsorber of AC/CuY for commercial gasoline (squares)and diel (circles),where C (t )is sulfur concentration at time t and C i is the initial sulfur concentration.
R E P O R T S
4JULY 2003VOL 301SCIENCE www.sciencemag
80
4.A.J.Herna´ndez-Maldonado,R.T.Yang,Ind.Eng.
Chem.Res.42,123(2003).
5.The calculations were performed at the Hartree-Fock
and density functional theory level using effective core potentials(11).The restricted Hartree-Fock the-ory at the LanL2DZ level basis t(12)was ud to determine the geometries and the adsorption bond-ing energies(8).Natural bond orbital analysis at the B3LYP/LanL2DZ level was ud for studying the elec-tron density distribution of the adsorption system
(8).A cluster model was ud to reprent zeolite
framework structure to which Agϩand Cuϩcations were bonded.
6.A.Takahashi,F.H.Yang,R.T.Yang,Ind.Eng.Chem.
lamiaRes.41,2487(2002).
7.The adsorber bed contained1to2g of zeolite,and the
feedflow rate was maintained at0.5cm3/min.Effluent
(or eluent)samples were collected at regular intervals
until saturation was reached,and the samples were
subquently analyzed for sulfur-containing compounds
with a gas chromatograph(GC)equipped with aflame
photometric detector(FPD).The FPD was operated at a
nsitivity(or detection limit)of0.02ppmw sulfur.其实他没那么喜欢你 下载>deem什么意思
Fourier transform infrared spectroscopy was ud for
analysis of aromatic and aliphatic contents by means of
the C-H stretching bands.
8.S.C.Larson,A.Aylor,A.T.Bell,J.A.Reimer,J.Phys.
Chem.98,11533(1994).
9.X.Ma,S.Lu,C.Song,Catal.Today77,107(2002).
10.Fuel Cells for Transportation,2001Annual Report
accom
(U.S.Department of Energy,Office of Transportation
Technologies,Washington,DC,2001).
11.P.J.Hay,W.R.Wadt,J.Chem.Phys.82,299(1985).
12.T.V.Russo,R.L.Martin,P.J.Hay,J.Phys.Chem.99,
17085(1995).
13.This work was funded by NSF and the U.S.Depart-
ment of Energy.U.S.and foreign patents are pending.
Supporting Online Material
www.sciencemag/cgi/content/full/301/5629/79/DC1
Materials and Methods
Figs.S1and S2
References
31March2003;accepted5June2003
Effects of Basal Debris on Glacier
Flow
Neal R.Iverson,1*Denis Cohen,2Thomas S.Hooyer,3
Urs H.Fischer,4Miriam Jackson,5Peter L.Moore,1
Gaute Lappegard,6Jack Kohler7
Glacier movement is resisted partially by debris,either within glaciers or under glaciers in water-saturated layers.In experiments beneath a thick,sliding gla-cier,ice containing2to11%debris exerted shear traction of60to200 kilopascals on a smooth rock bed,comparable to the total shear traction beneath glaciers and contrary to the usual assumption that debris-bed friction is negligible.Impod pore-water pressure that was60to100%of the normal stress in a subglacial debris layer reduced shear traction on the debris suffi-ciently to halt its deformation and cau slip of ice over the debris.Slip resistance was thus less than debris shearing resistance.
Fast flow of glaciers,including ice sheets,is usually associated with process that cau rapid movement at glacier beds(1).Such movement,which is exhibited only by gla-ciers that are melting at their bas,can have profound effects on a level and climate (1–3).The character of glacier beds controls how movement is resisted.Becau ice-bed friction is negligible due to a water film between ice and the bed,resistance to basal movement of hard-bedded glaciers—tho that rest directly on bedrock—is commonly assumed to depend only on viscous resistance to ice flow past irregularities on the rock surface(4).Thus,a glacier supported only by its bed would accelerate catastrophically if the bed were to become perfectly smooth.Basal resistance to movement of soft-bedded
glaciers—tho that are parated from rock
by a layer of wet diment—is usually
thought to equal the deformation resistance of
the diment as it shears beneath the weight
of the glacier(4).Basal slip,in this ca,is
assumed to occur within the diment bed
rather than at its surface.
电影的英文单词The assumptions,however,are uncer-
tain due to other possible sources of flow
resistance.In particular,slip of hard-bedded
glaciers,in addition to being resisted by bed
irregularities,must also be resisted by debris
within basal ice and held in frictional contact
with the rock bed.Although this friction has
not been measured and theoretical predictions
of its magnitude are uncertain(4–7),it is
usually neglected in modeling basal motion
(4).Also,soft-bedded glaciers may slip over
their beds rather than shear them,with slip
resistance dictated by process at the ice-
diment interface(8,9).Thus,equating re-
sistance to basal motion with bed shearing
resistance may be incorrect.
Definitive tests of the hypothes are dif-
ficult to execute,due largely to the inaccessi-
bility of glacier beds.Interpretations of data
gathered remotely at the bottoms of boreholes
drilled through glaciers are uncertain due to bed
disturbance by drilling,poorly known bed con-
ditions,and unsure instrument placement.
Moreover,the scope of instrumentation deploy-
able through narrow(ϳ0.1m)boreholes is
limited.Measurements made from tunnels in
glacier margins(10)are difficult to interpret
becau ice-marginal stress states are not rep-
rentative of glaciers in general(4).Studies in
natural subglacial cavities where glaciers pa-
rate from hard beds have been made only under
very thin ice or without the goal of studying
debris-bed friction.
We have conducted experiments with hard
and soft beds at the Svartin Subglacial Lab-
oratory in northern Norway,where human ac-
cess to the ba of the Svartin ice cap pro-
vides an unusual opportunity for experiments
beneath210m of sliding,temperate ice(11).As
part of a hydropower installation,tunnels have
been excavated in rock beneath the glacier.One
tunnel extends upward to the ba of the glacier,
where experiments were conducted during
March2001and repeated in April2002.
In one experiment,aimed at asssing
friction between debris in ice and a hard bed,
an insulated panel was installed at the mouth
of a vertical shaft in the rock bed(12)(fig.
S1).The panel was mounted flush with the
bed surface and contained,at its upper sur-
face,a smooth granite tablet(0.09m2).Sed-
iment-laden ice slid across the tablet,and the
associated shear traction on it was measured.
Sliding speed,pressure in the water film at
the tablet surface,and temperatures in the
panel were also measured.Before installing
the panel,basal ice samples were collected at
the mouth of the shaft;they contained2to
11%debris by volume,which consisted pri-
marily of sand and gravel.
A cond experiment,designed to study de-
formation of a soft bed,was conducted concur-
rently(12).From an ice tunnel,a trough(ϳ2.0
m long,1.5m wide,and0.5m deep)wascarbonfiber
blasted in the bedrock and filled with simulated
till(fig.S2).Instruments in the till recorded
shear deformation,total stress normal to the till
surface,pore-water pressure,and other vari-
ables at multiple locations.After preparing the
till wedge,ice deformation clod the tunnel,
bringing the glacier into contact with the till.
Thereafter,water was pumped for veral-hour
periods to the till to bring its pore-water pres-
sure clo to the total normal stress,as is ob-
rved beneath soft-bedded glaciers(9).
1Department of Geological and Atmospheric Sciences,
Iowa State University,Ames,IA50011,USA.2Depart-
ment of Geology and Geophysics,Yale University,
New Haven,CT06520,USA.3Wisconsin Geological
and Natural History Survey,3817Mineral Point Road,
Madison,WI53705,USA.4Laboratory of Hydraulics,
Hydrology and Glaciology,ETH-Zentrum,CH-8092
Zu¨rich,Switzerland.5Norwegian Water Resources and
Energy Directorate,Middelthuns Gate29,Post Office
Box5091,N-0301Oslo,Norway.6Department of
Geography,University of Oslo,Post Office Box1042,
N-0316Oslo,Norway.7Norwegian Polar Institute,
Polar Environmental Center,N-9296Tromsø,Nor-
way.
*To whom correspondence should be addresd.E-
mail:niverson@iastate.edu
R E P O R T S
www.sciencemag SCIENCE VOL3014JULY200381
