"Long before it's in the papers"
January 27, 2015

RETURN TO THE WORLD SCIENCE HOME PAGE


Carbon-nanotube computer could revolutionize electronics, researchers say

Sept. 25, 2013
Courtesy of Stanford University
and World Science staff

En­gi­neers have made a bas­ic com­put­er us­ing car­bon nano­tubes, a ma­te­ri­al they say could launch a new genera­t­ion of faster, more en­er­gy-ef­fi­cient elec­tron­ic de­vices. 

Car­bon nano­tubes are semi­con­duc­tors, ma­te­ri­als that con­duct elec­tri­city in a lim­it­ed way. Semicon­ductors are es­sen­tial to elec­tron­ic de­vices be­cause they al­low elec­trical sig­nals to be con­trolled. Sil­i­con is the semicon­ductor ma­te­ri­al used in most elec­tron­ics now.

Car­bon nano­tubes are a tough, flex­i­ble ma­te­ri­al com­posed of car­bon atoms ar­ranged ge­o­met­ric­ally in­to thin tubes, each thou­sands of times thin­ner than a hair.

Re­search­ers Sub­ha­sish Mi­tra and H.S. Phil­ip Wong of Stan­ford Uni­vers­ity in Cal­i­for­nia led the new proj­ect, de­scribed Sept. 25 in an ar­ti­cle on the cov­er of the jour­nal Na­ture.

“Peo­ple have been talk­ing about a new era of car­bon nano­tube elec­tron­ics mov­ing be­yond sil­i­con,” said Mi­tra, an elec­trical en­gi­neer and com­put­er sci­ent­ist. “But there have been few demon­stra­t­ions of com­plete dig­it­al sys­tems us­ing this ex­cit­ing tech­nol­o­gy. Here is the proof.”

Ex­perts said the work would gal­va­nize ef­forts to find suc­ces­sors to sil­i­con semi­con­duc­tors, which could soon en­coun­ter phys­i­cal lim­its that might pre­vent them from de­liv­er­ing smaller, faster, cheaper elec­tron­ic de­vices.

“Car­bon nano­tubes have long been con­sid­ered as a po­ten­tial suc­ces­sor” to sil­i­con, said Jan Rabaey, an ex­pert on elec­tron­ic cir­cuits and sys­tems at the Uni­vers­ity of Cal­i­for­nia Berke­ley. But un­til now it has­n’t been clear that nano­tubes could ful­fill those ex­pecta­t­ions, he added.

Roughly 15 years ago car­bon nano­tubes were first fash­ioned in­to tran­sis­tors, on-off switches at the heart of dig­it­al elec­tron­ic sys­tems. But im­per­fec­tions in car­bon nano­tube ma­te­ri­als have long frus­trat­ed ef­forts to build com­plex cir­cuits us­ing them. 

The Stan­ford team changed this in two key ways, said Gio­van­ni De Micheli, di­rec­tor of the In­sti­tute of Elec­tri­cal En­gi­neer­ing at École Poly­tech­nique Fédérale de Lau­sanne in Switz­er­land. “First, they put in place a pro­cess for fab­ri­cat­ing [car­bon nano­tube]-based cir­cuits,” De Micheli said. “Sec­ond, they built a sim­ple but ef­fec­tive cir­cuit that shows that com­puta­t­ion is doable” us­ing the ma­te­ri­al.

Why wor­ry about a suc­ces­sor to sil­i­con? Be­cause for dec­ades, prog­ress in elec­tron­ics has meant shrink­ing the size of each tran­sis­tor to pack more tran­sis­tors on a chip. But as tran­sis­tors be­come ti­ni­er they waste more pow­er and gene­rate more heat – all in a smaller and smaller space, which is why lap­tops get hot.

Many re­search­ers think this pow­er-wasting phe­nom­e­non could spell the end of Moore’s Law, named for In­tel Corp. co-founder Gor­don Moore. Moore pre­dicted in 1965 that the dens­ity, or com­pact­ness, of tran­sis­tors would dou­ble roughly eve­ry two years.

Car­bon nano­tubes are very ef­fi­cient at con­trol­ling elec­tri­city. They’re so thin that it takes very lit­tle en­er­gy to switch them off, Wong ex­plained. “Think of it as step­ping on a gar­den hose,” he said. “The thin­ner the hose, the eas­i­er it is to shut off the flow.” In the­o­ry, this com­bina­t­ion of ef­fi­cient con­ducti­vity and low-pow­er switch­ing make car­bon nano­tubes ex­cel­lent can­di­dates to serve as elec­tron­ic tran­sis­tors. The nano­tubes could im­prove per­for­mance by ten­fold or more, Wong said.

But in­her­ent im­per­fec­tions have stood in the way. First, the nano­tubes don’t nec­es­sarily grow in neat par­al­lel lines, a use­ful prop­er­ty. And some car­bon nan­otubes can end up mis­be­hav­ing by act­ing like me­tal­lic wires that al­ways con­duct elec­tri­city, in­stead of like semi­con­duc­tors that can be switched off. Re­search­ers had to find ways to deal with these er­rant nano­tubes with­out hav­ing to hunt for them like nee­dles in a hay­stack, which would choke off mass pro­duc­tion.

The re­search­ers took a two-pronged ap­proach to deal with this. To elim­i­nate the me­tal­lic-like nan­otubes, they switched off all the “good” nano­tubes and pumped the semicon­ductor cir­cuit full of elec­tri­city, which heat­ed and lite­rally va­por­ized the “bad” nano­tubes. They al­so built a sys­tem that maps out a cir­cuit lay­out de­signed to work no mat­ter wheth­er or where nano­tubes might be out of place.

The team used the de­sign to as­sem­ble a bas­ic com­put­er with 178 tran­sis­tors, which per­formed tasks such as count­ing and num­ber sort­ing. It runs a bas­ic ope­rating sys­tem that al­lows it to swap be­tween these pro­cesses. To demon­strate its po­ten­tial, the re­search­ers al­so showed that the com­put­er could run MIPS, a com­mer­cial in­struc­tion set de­vel­oped in the early 1980s by then-Stan­ford en­gi­neering pro­fes­sor and now uni­vers­ity Pres­ident John Hen­nessy.


* * *

Send us a comment on this story, or send it to a friend

Sign up for
e-newsletter
   
 
subscribe
 
cancel

On Home Page         

LATEST

  • St­ar found to have lit­tle plan­ets over twice as old as our own

  • “Kind­ness curricu­lum” may bo­ost suc­cess in pre­schoolers

EXCLUSIVES

  • Smart­er mice with a “hum­anized” gene?

  • Was black­mail essen­tial for marr­iage to evolve?

  • Plu­to has even cold­er “twin” of sim­ilar size, studies find

  • Could simple an­ger have taught people to coop­erate?

MORE NEWS

  • F­rog said to de­scribe its home through song

  • Even r­ats will lend a help­ing paw: study

  • D­rug may undo aging-assoc­iated brain changes in ani­mals

Engineers have made a basic computer using carbon nanotubes, a material they say could launch a new generation of faster, more energy-efficient electronic devices. Carbon nanotubes are semiconductors, materials that conduct electricity in a limited way. Semiconductors are essential to electronic devices because they allow electrical signals to be controlled. Silicon is the semiconductor material used in most electronics now. Carbon nanotubes are a tough, flexible material composed of carbon atoms arranged geometrically into thin tubes thousands of times thinner than a hair. Researchers Subhasish Mitra and H.S. Philip Wong of Stanford University in California led the new project, described Sept. 25 in an article on the cover of the journal Nature. “People have been talking about a new era of carbon nanotube electronics moving beyond silicon,” said Mitra, an electrical engineer and computer scientist. “But there have been few demonstrations of complete digital systems using this exciting technology. Here is the proof.” Experts said the work would galvanize efforts to find successors to silicon semiconductors, which could soon encounter physical limits that might prevent them from delivering smaller, faster, cheaper electronic devices. “Carbon nanotubes have long been considered as a potential successor” to silicon, said Jan Rabaey, an expert on electronic circuits and systems at the University of California Berkeley. But until now it hasn’t been clear that nanotbues could fulfill those expectations, he added. For roughly 15 years carbon nanotubes were first fashioned into transistors, on-off switches at the heart of digital electronic systems. But imperfections in carbon nanotube materials have long frustrated efforts to build complex circuits using them. The Stanford team changed this in two key ways, said Giovanni De Micheli, director of the Institute of Electrical Engineering at École Polytechnique Fédérale de Lausanne in Switzerland. “First, they put in place a process for fabricating [carbon nanotube]-based circuits,” De Micheli said. “Second, they built a simple but effective circuit that shows that computation is doable” using the material. Why worry about a successor to silicon? Because for decades, progress in electronics has meant shrinking the size of each transistor to pack more transistors on a chip. But as transistors become tinier they waste more power and generate more heat – all in a smaller and smaller space, which is why laptops get hot. Many researchers think this power-wasting phenomenon could spell the end of Moore’s Law, named for Intel Corp. co-founder Gordon Moore. Moore predicted in 1965 that the density, or compactness, of transistors would double roughly every two years. Carbon nanotubes are very efficient at controlling electricity. They’re so thin that it takes very little energy to switch them off, Wong explained. “Think of it as stepping on a garden hose,” he said. “The thinner the hose, the easier it is to shut off the flow.” In theory, this combination of efficient conductivity and low-power switching make carbon nanotubes excellent candidates to serve as electronic transistors. The nanotubes could improve performance by tenfold or more, Wong said. But inherent imperfections have stood in the way. First, the nanotubes don’t necessarily grow in neat parallel lines, a useful property. And some carbon nanotubes can end up misbehaving by acting like metallic wires that always conduct electricity, instead of like semiconductors that can be switched off. Researchers had to find ways to deal with these errant nanotubes without having to hunt for them like needles in a haystack, which would choke off mass production. The researchers took a two-pronged approach to deal with this. To eliminate the metallic-like nanotubes, they switched off all the “good” nanotubes and pumped the semiconductor circuit full of electricity, which heated and literally vaporized the “bad” nanotubes. They also built a system that maps out a circuit layout designed to work no matter whether or where nanotubes might be out of place. The team used the design to assemble a basic computer with 178 transistors, which performed tasks such as counting and number sorting. It runs a basic operating system that allows it to swap between these processes. To demonstrate its potential, the researchers also showed that the computer could run MIPS, a commercial instruction set developed in the early 1980s by then Stanford engineering professor and now university President John Hennessy. scientists say