(Preeti  Aghalayam,  Aug  2012)  

¡  Sir  Humphrey  Davy:  1817  –  reaction  of  coal  gas  with  oxygen  on  a  glowing  Pt  wire    

¡  Berzelius:  1836  -­‐  Defined  the  term  ‘catalysis’  ¡  Faraday:  1834  -­‐  Proposed  that  reactants  have  to  adsorb  simultaneously  at  catalyst  surfaces  

(Catalysts  were  being  used  inadvertently,  for  making  beer,  wine,  &  cheese,  in  earlier  times)    

¡  Ostwald:  1900  –  Catalysis,  equilbria  ¡  Haber:  1905  –  Ammonia  production  catalyst  ¡  Langmuir:  1920s  –  Surface  chemistry  ¡  Bosch:  1931  –  High  pressure  reactor  for  ammonia  

¡  Hinshelwood  &  Semenov:  1956  –  Mechanisms  of  chemical  reactions  

¡  Ziegler  &  Natta:  1963  –  Chemistry  &  technology  of  high  polymers  

4 G.A. Somorjai, K. McCrea / Applied Catalysis A: General 222 (2001) 3–18

Fig. 1. Timeline of the progress of heterogeneous catalysis.

the late 1960s/early 1970s, Boudart classified cata-lytic reactions into two groups: structure sensitive andstructure insensitive [5]. The structure sensitivereactions change their rate as a function of particlesize, while structure insensitive reactions remain at aconstant rate as the particle size increases [6]. Thisconcept has withstood the test of times. The conceptof bifunctional catalysis, by which to obtain a desiredproduct one needs two catalysts, was also developedin this period although it was proposed earlier by

Fig. 2. Turnover rates on Rh catalysts for reactions of CO with O2 and NO as a function of particle size exhibiting structure insensitivityand structure sensitivity of catalytic reactions [6].

Haensel. One catalyst produces a reaction intermedi-ate, which then diffuses onto the other catalyst wherethe reaction products form and desorb in the gas or so-lution phase. There are many examples of bifunctionalcatalysis. The selectivity changes due to pore size inzeolites have been shown to be due to the diffusionrate of molecules, which depends on the molecularsize and shape [7]. Using Linde 5A zeolite, Friletteet al. [7] showed that n-butanol could be dehydratedwith a conversion of 60wt.% while maintaining a low

Catalysis  began  by  helping  make  bulk  basic  &  inorganic  chemicals,  several  centuries  ago      

It  transitioned  to  a  huge  volumes  and  investments  business  with  the  advent  of  the  oil  economy  and  the  need  for  petroleum  re<ining  

Further  impetus  was  achieved  via  the  petrochemicals  industy  

Soon,  speciality  chemicals  production  and  enviromentally  relevant  catalysis  became  the  exciting  new  trend  

Today,  in  addition  to  being  a  mainstay  in  the  chemicals  industry,  catalysts  are  expected  to  charter  new  directions  for  the  world  economy!  

¡  80  –  90%  of  current  day  products  emerge  from  some  or  the  other  catalytic  process  

¡  As  our  focus  shifts  in  the  21st  century,  the  challenges,  especially  in  petro  industry,  are  remarkable  

New  materials:  Catalytic  membranes,  hydrogen  storage  

Cheaper  catalysts  for  pollution  control  

Biocatalysts  for  re<ining  of  petroleum  

Fundamental  studies:  kinetics,  characterisations  

The  petroleum  economy  especially  can  expect  widespread  changes  

(from  Marcilly,  2003)  

(from  Vlachos  &  Caratzolous,2010)  

ARTICLE IN PRESS

up to two orders of magnitude higher than that of Li-basedbatteries (Vlachos, 2009).

The above applications underscore the need for developingsmall and efficient chemical plants. While smaller scales canstill be described with the core reaction engineering models,downscaling imposes new challenges that have been detailedelsewhere (Norton et al., 2005; Vlachos, 2009) and are onlybriefly mentioned here. Reactors need to be designed to ensurefast mixing, low pressure drop, high catalyst area for highconversions, and minimal transport (external and internal)resistances. In addition, catalyst requirements, such as activity,selectivity, safety (non-pyrophoric materials), and stability intransient operation, become more stringent. Integration of micro-units into compact, energy-efficient, self-sustained systems iscrucial given the lack of large heat and process integration withnearby plants.

4. Hydrogen economy, green hydrogen, and remote/offshoreutilization of natural gas

The concept of hydrogen economy has frequently been in therecent news and a DOE report has been published (Dresselhauset al., 2003; NRC, 2004). The ideal hydrogen cycle entails splittingof H2O to produce (green) hydrogen followed by its electro-chemical (green) combustion. The challenge, like CO2 utilization,is that H2O split is very hard to achieve at reasonable rates andefficiencies.

Hydrogen is the cleanest burning fuel. The much higherefficiency (compared with the current ICEs) and zero emissionsmake the PEM fuel cell running on H2 an appealing technology.Hydrogen addition to fossil fuels can stabilize combustion ofextra-lean mixtures (below their fuel lean flammability limit) inICEs, and thus, it could allow for reduction of temperatures andNOx emissions. Despite these obvious advantages, hydrogen is notavailable as a natural reserve, i.e., hydrogen is not an energysource but rather an energy carrier. It is currently produced andused in chemical industry for, among others, production ofammonia, methanol, and various refinery processes, such assulfur removal.

The strong opposition against a hydrogen economy from someof the scientific community arises in part from the fact that thereis no hydrogen surplus to realize a hydrogen economy and watersplit is too difficult to achieve. However, even if a hydrogeneconomy is never realized, increased hydrogen production fromalternative and renewable sources is still compelling sincehydrogen finds many industrial uses. For instance, the risingprices of crude oil and of natural gas have led to an increased priceof hydrogen. As a result, the price of diesel has increased over thatof gasoline due to more stringent regulations for sulfur removal.These problems can be mitigated if additional hydrogen produc-tion routes are developed.

The hydrogen economy consists of three legs: hydrogenproduction, storage, and use. The CRE community has tradition-ally been involved with the production of hydrogen (Figs. 2 and 3).Fuel cells (use) and catalytic storage/release offer newopportunities. The current production relies mainly on steamreforming of natural gas followed by two (high and lowtemperature) water gas shift (WGS) reactors. Depending on thepurity of hydrogen needed, a final stage of pressure swingadsorption or selective oxidation of CO or methanation may beneeded. This last process is deemed necessary in the case of PEMfuel cells to avoid poisoning of the Pt catalyst by COchemisorption. Coal gasification is another route for H2

production and is currently utilized to a less extent due to coalbeing the least clean fossil fuel. Given the huge coal reserves in the

US, it is expected that coal gasification to syngas and eventually tohydrogen or chemicals will gain substantial momentum in thenear future.

4.1. Green hydrogen production

The exploration of renewables for hydrogen production hasintroduced new processes to produce syngas. One such process isthe short contact time catalytic partial oxidation (CPOX) of liquid(oil) and solid lignocellulosic materials (e.g., wood chips) on Rh-based catalysts (Dauenhauer et al., 2006; Salge et al., 2006; Wanatet al., 2005). As elaborated further in the section on renewables, amajor advantage of this reforming technology is the minimizationor elimination of coke formation. An alternative to this reformingis the aqueous phase reforming (APR) of various oxygenatedmolecules to H2 (Cortright et al., 2002; Davda et al., 2005; Huberand Dumesic, 2006; Huber et al., 2003), e.g.,

C6HxO6(l)+6H2O(l)-6CO2(g)+(12+x)/2H2(g); x=12 or 14 (APR)

Under certain conditions (Davda and Dumesic, 2003), H2/CO2

mixtures (with fractions of CO as low as 100 ppm) are produceddue to the enhanced contribution of the water gas shift (WGS)reaction during low-temperature (!500 K) operation. The low COcontent makes purification of H2 from CO2 (with a membrane) forfuel cells very feasible but is disadvantageous for Fisher–Tropschsynthesis. The CO:H2 ratio can be tuned by changing the support(Chheda et al., 2007a).

Fuel

+ Air

Fuel

Steam

Pure H 2Combustor

Fuel + Air

Fuel

Steam

Pure H2

SR

WGSPSA, PROX, methanation

Fig. 2. Schematic of the flow sheet for syngas and H2 production.

Steam reforming CO

H2

CH 4

CO 2

H2O

CHO

H2OSteam reforming

Water-gas shift

H2O

CO

Water-gas shift

CO

H2

CO

H2

CO

H2

O2

O2

Partial o

xidation

Partial oxidation

CH 4

CO 2

H2O

CHO

Dry

ref

orm

ing

Dry reform

ing

CO2

CO2

CH4

CO2

H2OO2

O2

O2

O2

CHO

CH3O

CH2O

CH3OH

CHO

H2O

Combustion

Combustion

Oxidation

Oxidation

Fig. 3. Schematic of the reaction network in converting natural gas to syngas (seeFig. 2). In most of these reactions, more than one overall reactions happen. In CPOXand ATR, combustion, steam reforming, water gas shift, and direct formation ofsyngas are all possible depending on operating conditions and mass transferlimitations.

D.G. Vlachos, S. Caratzoulas / Chemical Engineering Science 65 (2010) 18–2920

The  Hydrogen  Economy  has  to  rely  heavily  on  catalysts  

ARTICLE IN PRESS

This paper focuses mainly on select emerging technologies andprocesses where the CRE community can significantly contributein solving the energy problem, mainly through innovation inheterogeneous catalysis and reactors. Topics touched uponinclude process intensification (PI) and efficiency, hydrogeneconomy, offshore and remote natural gas utilization, andrenewables, such as biomass utilization and transformation ofvarious waste streams. As underscored herein, a crosscuttingtheme emerging for future power generation is processing atscales smaller than those of the conventional refinery andpetrochemical plants. Future research needs are finally outlined.

2. Improving process efficiency

While the quest for alternative and renewable energy sourceswill be very important in meeting the increasing energy needs, akey aspect in overcoming the energy and environmental chal-lenges is to improve process efficiency of existing and newprocesses (Nat.Acad.Press, 2008). This is particularly importantsince fossil fuels will continue to constitute the backbone of ourenergy supply.

Several processes exhibit low energy efficiency. For example,the overall efficiency in converting chemical energy, starting froma power plant running on coal, and ending with light is just 2%:62% of the initial energy is lost in the power plant, 2% intransmission lines, and 34% as heat in light lumps (Nat.Acad.Press,2008). As another example, the efficiency of a typical internalcombustion engine (ICE) is of the order of !20%. This low fractionis despite the progress made in fuel efficiency of automobiles from18 mpg (1978) to 27.5 mpg (1985) to an imposed average of35 mpg (2030) for all cars, SUVs, and light trucks (Nat.Acad.Press,2008), due mainly to improvements in light materials.

How do we improve process efficiency? Improved efficiencyusually entails catalyst and/or reactor/flow sheet optimization.Selectivity is, by and large, the single most important factor.Improved selectivity implies reduced waste and reduction orelimination downstream of the energy-intensive separation units.High throughput experiments and insights gained from computa-tional catalysis promise the development of more selectivecatalysts. An example is the epoxidation of ethylene to ethyleneoxide on Ag-based bimetallic and doped catalysts (Dellamorte etal., 2007) and many more are emerging. Catalyst design willunquestionably play a key role in both conventional processes andbiomass conversion.

Process intensification (PI) entails the enhancement of theeffective rate by increasing transport rates and/or impartingmultifunctionality into devices (Stankiewicz, 2007). The net resultcan be improvement of process efficiency, reduction in size (withan associated reduction in capital cost), and/or in operation cost.Several concepts for PI have been developed over the past fewyears. The overall idea of multifunctional reactors was madepopular about two decades ago (Agar and Ruppel, 1988; Wester-terp, 1992). A popular example of PI is reactive separation, withthe reactive distillation (Malone et al., 2003) of MTBE by EastmanChemicals being a successful commercial test bed. Membranereactors for equilibrium-limited reactions or for selectivelyremoving a product that inhibits catalysts is another example ofreactive separation (Harale et al., 2007). Reactive adsorption forCO2 capture during the water gas shift reaction is a recentlyexplored application (Dadwhal et al., 2008; Martavaltzi andLemonidou, 2008). Miniaturization of chemical processes leadsto enhanced transport rates and concomitant size reduction. Asdiscussed further below, this size reduction is deemed essentialfor smaller or distributed scale (remote, offshore, transportation,portable power) applications. Integration of heat exchangers with

reactors in making efficient, compact systems has also beenintensively studied (Kolios et al., 2005, 2007). The parallel-platecatalytic reactor (where an endothermic and an exothermicreaction take place on opposite sides of a wall that serves as aheat exchanger) is a fairly common configuration of spatialcoupling (Fig. 1). Heat recuperation strategies via recirculationor regeneration are essential strategies (Federici and Vlachos,2008; Jones et al., 1978; Lloyd and Weinberg, 1974; Matros andBunimovich, 1996; Neumann and Veser, 2005) for energy lossminimization. Additional examples of PI will be discussed below.

3. Distributed power generation and downscaling ofchemical processes

In a recent report, it was suggested that ‘building small plantsnear customers, known as distributed generation, may becomemore important in order to meet demand and maintain reliability’(Nat.Acad.Press, 2008). Processing at smaller scales can meetmultiple objectives, namely increased reliability, overcomingexpensive or impossible transportation from remote and offshorelocations, improved PI and enhanced efficiency, and the need forH2 production for PEM fuel cells for transportation and portabledevices.

Miniaturization will be central to several energy efforts in thefuture. As discussed below, if we were to realize a hydrogeneconomy in the short term, we would need onboard reforming.This will entail !108 reactors to runs all cars in the US alone. Inthe mid term, and assuming that suitable nanomaterials forhydrogen storage are developed, hydrogen may be produced ingas stations (!105 in the US) to take advantage of the liquid fueldistribution infrastructure. When compared with the operating149 refineries in the US, the aforementioned numbers indicatereduction in reactor volume and increase in the number ofreactors needed for H2-based transportation economy.

Remote and offshore utilization of natural gas will requiresmaller-scale systems (Lerou, 2006). One could envision futuresupertankers, being filled offshore with crude oil, to be smallchemical plants transforming natural gas into liquids via gas toliquids (GTL) or easy-to-liquefy gases, e.g., ammonia, which arethen transported to mainland. Achieving this goal will requiredownscaling the currently bulky steam reforming process. In thecase of biomass, feedstock utilization will be localized. This is dueto the large water content of biomass, which makes thetransportation cost high. It is anticipated that the optimal scaleplants (biorefineries) will be of the order of !1000 barrels perday, much smaller than the current refinery and petrochemicalplants.

Electronics (1–100 W) currently rely on batteries whoseefficiency is low and their weight is high (their mass energydensity is low). One could replace batteries with microchemicalsystems since the mass energy density of common liquid fuels is

catalyst

Fuel + O2 ! CO2 + H2O

CH4 + H2O ! CO+3H2

Wall

Fig. 1. Schematic of multifunctional catalytic parallel-plate microreactor. Catalyticcombustion occurs on a Pt washcoat catalyst in one channel and steam reformingof methane on a Rh washcoat catalyst in the other channel. The (thin) separatingwall serves as a compact and efficient heat exchanger.

D.G. Vlachos, S. Caratzoulas / Chemical Engineering Science 65 (2010) 18–29 19

Energy  ef<icient  processes  will  require  new  ideas-­‐  such  as  this  combined  reactor/heat  exchanger  in    micro-­‐scale  

The  current  challenge  is  to  design  catalysts  to  explore  these  new  vistas  

First  implemented  in  the  USA  in  1975  

In  Japan  &  Europe  in  1986  

Currently,  all  over  the  world!  

Challenges  today:  •  Selectivity  •  Cold  start  emissions  •  Fuel-­‐lean  engine  •  Biofuel  engines  •  CO2  &  PM    •  Bi-­‐functional  catalysts  

This  catalyst  can  reduce  NO  in  fuel-­‐lean  engines  without  problems  of  

NH3  slip  

New  age  engines  and  new  age  ground  rules  have  kept  researchers  on  their  toes!  (from  Schauer  et  al.,2012)  

¡  Picture  this:  metallic  gold  is  typically  chemically  inert.  But  prepared  in  a  special  way  –  depositing  only  particles  that  are  a  few  nanometers  in  size  –  makes  it  a  viable  catalyst  for  several  reactions!  

settled, it appears that mild reduction treatments may be

preferred in many instances, even if those leave some ofthe organic matter behind [30]. In fact, it is possible that

the remaining carbonaceous deposits may actually help

with the performance of the catalyst, at least in the pro-motion of mild reactions such as olefin hydrogenations

[37]. Access to the metal inside dendritic or colloidal

structures may also be possible in liquid solutions [38], inwhich case the catalyst may not even require special acti-

vating treatments. The issue of the activation of heteroge-

neous catalysts prepared by these new self-assemblymethodologies requires further studies.

When considering nanoparticle size in heterogeneous

catalysts, one extreme is catalysis by one single atom, orperhaps by a small number of atoms in a well-defined

molecular cluster. The behavior of the catalyst in suchcases may resemble more closely that of homogeneous

catalysts, where selectivity can sometimes be controlled at

a molecular level. In fact, the heterogeneous catalysts canbe prepared by starting with the corresponding discrete

molecular clusters [39]. However, the interaction of the

atoms of the catalytic phase with the support is rarelynegligible, and needs to be considered. The final structure

of the surface species may also dynamically change as the

pretreatment or reaction conditions are changed, and thefinal active phase may exhibit very different characteristics

to those of the original organometallic precursors. An

interesting example of a change in the structure of thecatalyst leading to changes in reaction selectivity has been

recently reported for the conversion of ethylene on sup-

ported rhodium catalysts [40]. In that case, the initialRh(C2H4)2 complexes bonded to a crystalline zeolite HY

support could be made to remain isolated and to display

high selectivity for the dimerization of ethylene to butenesand butane under most conditions, except upon exposure to

highly reducing environments, after which they were seen

to form small metal clusters and to preferentially promotehydrogenation to ethane instead. Curiously, this transfor-

mation was shown to be reversible: the isolated-Rh

dimerization sites could be regenerated upon exposure ofthe catalyst to ethylene-rich mixtures. In general, the use of

small molecular clusters as precursors for the preparation

of heterogeneous catalyst could be quite useful if issues ofstability and selectivity can be worked out.

3 Nanoparticle Shape

Perhaps more interesting than controlling the performanceof catalysts by controlling the size of the nanoparticles of

the active phase is the idea of exerting that control via theselection of their shape. It has been long known that some

catalytic processes are structure sensitive, which in tradi-

tional catalysis has come to mean that their performance interms of activity or selectivity changes significantly with

the method used for their preparation. However, this

behavior has been justified on the basis of the associatedchanges in the distribution of particle size in the resulting

catalysts [41]. It has only been recently, with the incor-

poration of methods to better control particle size andshape independently of each other, that the effects of those

two parameters have started to be untangled.

In surface science studies using model system, structuresensitivity has traditionally been probed by comparing

chemical reactivity on single crystals exposing surfaces

with different orientations [5, 8]. Initial studies on chemi-sorption were later extended to catalytic rate measurements

using so-called ‘‘high pressure cells’’ [42–44]. Those

studies have been quite useful, but also revealed someintrinsic limitations, in particular the fact that they cannot

Fig. 2 Pyrrole hydrogenation selectivity at 413 K (4 torr pyrrole,400 torr H2, 2 % conversion) as a function of the size of the Ptnanoparticles, dispersed on a HY zeolite, used as catalysts [32].Hydrogenation to pyrrolidine is facile in all cases, but further

hydrogenolysis to n-butylamine can only be partially avoided if smallnanoparticles, of diameters on the order of *1 nm, are used. Figurecourtesy of Jeong Park and Gabor Somorjai, reproduced fromRef. [32] with permission. Copyright 2009 American Chemical Society

504 F. Zaera

123

We  are  now  able  to  really  control  size  of  the  catalyst  particles  we  make  (from  Zaera,2012)  

¡  How  does  Zeolite  microporosity  control  reactivity?      àby  selectively  allowing  access  to  an  active  site  within  the  zeolite  cavity!  

¡  para-­‐xylene  can  gain  access  to  the  inside  of  the  zeolite  channel,  whereas  the  meta-­‐  and  ortho-­‐  forms  of  xylene  are  sterically  hindered  from  doing  so.  

We  are  now  able  to  really  understand  how  these  magical  things  work!  (from  Bill  Vining’s  work)  

New  experimental  techniques  help  us  with  molecular  level  pictures!    (from  Somorjai  &  Wang  1997,)  

Electron  based  microscopy  

Molecule/ion  based  spectroscopy  

Photo-­‐mediated  spectroscopy  

Scanning  tunneling  microscopy  

( )G.A. Somorjai, M.X. YangrJournal of Molecular Catalysis A: Chemical 115 1997 389–403 393

� .Pt 110 surface exposed to ambient pressures ofhydrogen, oxygen and carbon monoxide at 425K. Under 1.6 atm of hydrogen pressure, thesurface presents various sizes of missing-rowreconstruction. In 1 atm of oxygen, however,

� .enlarged 111 microfacets can be observed.The surface in 1 atom of carbon monoxideappears to have large scale terraces separated bymultiple height steps.The surface reconstruction is a reversible

process. The platinum surface was exposed todifferent gases alternatively and the surface

w xstructure changed accordingly 18 . In 1.6 atmH , chemisorbed oxygen reacts to form water2and desorbs from the surface. In CO environ-ment, the binding energy of hydrogen atoms onthe surface is reduced and surface hydrogens arereplaced by CO molecules. Under atmosphericoxygen pressure, surface CO molecules are oxi-dized to CO and the surface is switched to be2covered by oxygen. The conversion of surfacestructures is indicative of the adsorbate compo-sition change on the surface.

2.1.2. Coadsorption-induced reconstruction ofadsorbate oÕerlayerAdsorbate overlayer as well as substrate

atoms can be rearranged by adsorption of coad-sorbates. Surface species are highly mobile onthe surface. In many cases, adsorbate structuresare reorganized in order to accommodate othersurface species. A reconstruction of adsorbateoverlayers by coadsorption has been demon-strated in a STM and LEED study of sulfur

� . � .chemisorption on Re 0001 and Pt 111 sur-w xfaces 19 . Surface structures imaged by STM

are consistent with electron diffraction patternsobtained in complementary LEED studies. At a

�sulfur coverage of 0.25 monolayer one atom. � .per four substrate metal atoms , a 2=2 or-

dered sulfur structure can be observed, as shown� . � . � .in Fig. 4 a . On either Re 0001 and Pt 111

surfaces, coadsorption of carbon monoxidemolecules induces a reordering of sulfur struc-ture. The sulfur overlayer is compressed, creat-ing empty space on the surface for carbon

� . � .Fig. 4. STM images a before and b after the reordering of� . � .sulfur overlayer on Re 0001 induced by CO exposure. a The

� .round maxima are due to individual p 2=2 ordered sulfur atomsadsorbed at the hcp hollow site of the surface. Image size: 40=40˚ � . � .A. b A hole has formed in the p 2=2 layer where CO has

�adsorbed CO molecules are not visible in the STM images,.presumably due to their facile diffusion on the surface . The sulfur

atoms which resided previously in the hole have been compressedto form trimers of three atoms which appear as bright spots

˚surrounding the hole. Image size: 55=55 A.

monoxide adsorption. The new sulfur overlayer� .presents a 3 63= 363 R308 structure on

� . � � .. � .Re 0001 Fig. 4 b and a 63=63 R308 struc-� .ture on Pt 111 . The CO molecules have a high

mobility on the surface and are not visible inSTM experiments. The change of sulfur over-layer structure is reversible and the original� .2=2 sulfur structure can be restored afterdesorbing CO molecules at high temperature.Competitive adsorption and mobility of ad-

sorbates on the surface attribute to the coadsorp-tion-induced reconstruction of adsorbate over-layers. If surface species are immobile becauseof a high activation energy for surface diffusion,coadsorption cannot take place. On the other

( )G.A. Somorjai, M.X. YangrJournal of Molecular Catalysis A: Chemical 115 1997 389–403400

catalyst. An optimization of catalyst perfor-mance can be accomplished.UV light and X-ray radiation have also been

used in lithography studies. The main advantageof electron beam lithography over the othertechniques is its exceptional high resolution. Itcan generate features as small as a few nanome-

w xters 30 . Notice that the average particle size ofindustrial catalysts is 1–100 nm, electron beamlithography is, at present, the best choice inmodel catalyst preparation.

3.2. ReactiÕity tests

An initial reactivity test of metal clustersprepared by electron beam lithography has

w xyielded encouraging results 31 . A metal clustersample was prepared by Dr. S.J. Wind at IBMresearch center, Yorktown Heights. Platinumparticles of 50 nm diameter and 15 nm heightwith 200 nm periodicity have been prepared ona 0.5=0.8 cm oxidized silicon wafer. Scanning

� .electron microscopy SEM pictures of the metalcluster array are shown in Fig. 15.The cluster sample shows a remarkable sta-

bility upon annealing and exposure to ions andelectrons. It allows us to remove surface con-taminants by low energy ion sputtering andoxygen treatment. The sample can be character-ized by electron- and ion-scattering surface sci-ence spectroscopies. AFM studies indicate thatthe sample structure remains intact after surfacecleaning, catalytic reaction and sample charac-terization.The rate of ethylene hydrogenation over this

new model catalyst was measured in aUHVrhigh pressure system. The surface area ofmetal cluster arrays is one to two orders ofmagnitude smaller than a single crystal surfaceof comparable sample size. Fig. 16 shows theethane yield as a function of time at roomtemperature, along with a background signaltaken on a blank silicon wafer. The measuredturnover rate is in good agreement with previ-ous results obtained on conventional supportedcatalysts and single crystals. As shown in Fig.

Fig. 15. SEM micrographs of platinum cluster array fabricated byelectron beam lithography. The sample has a cluster size of 50 nmand a periodicity of 200 nm.

16, an increase of reaction rate is also observedupon increasing sample temperature.The saturation coverage of adsorbates on the

platinum cluster sample can be determined frompeak areas in thermal desorption studies. Thethermal desorption spectra of D and 13C18O2from the cluster sample are displayed in Fig. 17,along with reference spectra collected on a plat-inum foil. The ratio of deuterium desorbingfrom the cluster sample to that desorbing fromthe reference foil sample is 2–4 times greaterthan the same ratio for carbon monoxide. Thisindicates a spillover of deuterium from platinummetal clusters onto silicon oxide support, whichis a characteristic of dispersed metal catalysts.Through the collaboration with IBM, we have

accumulated valuable experience on electronbeam lithography, sample handling, surfacecleaning and reactivity studies of nanoscale

STM  &  SEM  images  of    catalyst  surfaces  

¡  Catalysis  is  old  science    §  Industrial  catalysis  is  even  older!  

¡  Nevertheless,  its  contextually  immensely  relevant  today  in  several  walks  of  life  

HOME  ASSIGNMENT:  Identify  a  modern  day  (alive  as  yet)  scientist  working  in  catalysis.    Read  some  abstracts/summaries  of  their  work.  Bring  me  a  name,  affiliation,  and  1-­‐2  line  descriptions  of  their  research.  Note  down  the  source  of  your  information.      (We  will  read  these  out  in  class  tomorrow)  

Already  completed