New Ma­te­r­ial Paves the Way to On-​Chip En­ergy Har­vest­ing

Re­searchers from Ger­many, Italy, and the UK have achieved a major ad­vance in the de­vel­op­ment of ma­te­ri­als suit­able for on-​chip en­ergy har­vest­ing. By com­pos­ing an alloy made of sil­i­con, ger­ma­nium and tin, they were able to cre­ate a ther­mo­elec­tric ma­te­r­ial, promis­ing to trans­form the waste heat of com­puter proces­sors back into elec­tric­ity. With all el­e­ments com­ing from the 4th main group of the pe­ri­odic table, these new semi­con­duc­tor alloy can be eas­ily in­te­grated into the CMOS process of chip pro­duc­tion. The re­search find­ings made it onto the cover of the renowned sci­en­tific jour­nal ACS Ap­plied En­ergy Ma­te­ri­als.

The in­creas­ing use of elec­tronic de­vices in all as­pects of our lives is dri­ving up en­ergy con­sump­tion. Most of this en­ergy is dis­si­pated into the en­vi­ron­ment in the form of heat. In Eu­rope, about 1.2 Ex­a­joule of low-​temperature heat is wasted from IT in­fra­struc­tures and de­vices, such as data cen­ters and smart de­vices, per year. This is roughly equiv­a­lent to the pri­mary en­ergy con­sump­tion of Aus­tria or Ro­ma­nia. This low-​grade heat below 80°C is tra­di­tion­ally chal­leng­ing to har­ness due to poor ther­mo­dy­namic ef­fi­ciency and tech­no­log­i­cal con­straints.

There­fore, uti­liz­ing the low-​temperature heat di­rectly for com­puter proces­sors seems to be an ideal so­lu­tion. But there are only very few ma­te­ri­als avail­able to con­vert the heat into elec­tri­cal en­ergy, and none of them are com­pat­i­ble with cur­rent tech­nol­ogy in semi­con­duc­tor fab­ri­ca­tion plants.

A re­search col­lab­o­ra­tion be­tween Forschungszen­trum Jülich and IHP – Leib­niz In­sti­tute for High Per­for­mance Mi­cro­elec­tron­ics in Ger­many, to­gether with the Uni­ver­sity of Pisa, the Uni­ver­sity of Bologna in Italy, and the Uni­ver­sity of Leeds in the UK, now reached a mile­stone in de­vel­op­ing suit­able ma­te­ri­als for on-​chip en­ergy har­vest­ing that are com­pat­i­ble with the CMOS process of chip pro­duc­tion.

"Adding tin to ger­ma­nium sig­nif­i­cantly re­duces the ma­te­r­ial’s ther­mal con­duc­tiv­ity while main­tain­ing its elec­tri­cal prop­er­ties, an ideal com­bi­na­tion for ther­mo­elec­tric ap­pli­ca­tions", ex­plains Dr. Dan Buca, leader of the re­search group at Forschungszen­trum Jülich. The ex­per­i­men­tal con­fir­ma­tion of the low lat­tice ther­mal con­duc­tiv­ity, pub­lished in ACS Ap­plied En­ergy Ma­te­ri­als, high­lights the great po­ten­tial of these GeSn al­loys as ther­mo­elec­tric ma­te­ri­als. The idea be­hind this: By in­te­grat­ing these al­loys into silicon-​based com­puter chips, it is pos­si­ble to utilise the waste heat gen­er­ated dur­ing op­er­a­tion and con­vert it back into elec­tri­cal en­ergy. This on-​chip en­ergy har­vest­ing could sig­nif­i­cantly re­duce the need for ex­ter­nal cool­ing and power, lead­ing to more sus­tain­able and ef­fi­cient IT de­vices.

In ad­di­tion, Group IV el­e­ments, also known as the sil­i­con group, form the basis of any elec­tronic de­vice, and by ex­ploit­ing their al­loy­ing prop­er­ties, the ap­pli­ca­tion areas are now ex­pand­ing to in­clude ther­mo­electrics, pho­ton­ics and spin­tron­ics. The mono­lithic in­te­gra­tion of pho­ton­ics, elec­tron­ics and ther­mo­electrics on the same chip is the am­bi­tious long-​term goal of sil­i­con based tech­nol­ogy. By com­bin­ing these fields, it is pos­si­ble not only to im­prove the per­for­mance of de­vices, but also to sup­port the de­vel­op­ment of more sus­tain­able tech­nolo­gies.

“In the paper we made a very im­por­tant step. We have eval­u­ated one of the most crit­i­cal pa­ra­me­ters for a ther­mo­elec­tric ma­te­r­ial, the ther­mal con­duc­tiv­ity, using a suite of dif­fer­ent ex­per­i­men­tal tech­niques on epi­tax­ial sam­ples with dif­fer­ent alloy com­po­si­tions and thick­nesses”, says Prof. Gio­vanni Capellini, project leader at IHP. "Our joint re­search can have a size­able im­pact in the field of ′Green IT′ in­fra­struc­tures. "

The re­search groups at Forschungszen­trum Jülich and IHP are con­tin­u­ing their suc­cess­ful col­lab­o­ra­tion. They aim to fur­ther de­velop the ma­te­r­ial by ex­tend­ing the alloy com­po­si­tion to SiGeSn and the ul­ti­mate Group IV alloy CSiGeSn, and to fab­ri­cate a func­tional ther­mo­elec­tric de­vice to demon­strate the en­ergy har­vest­ing po­ten­tial of Group IV al­loys. The ac­tiv­ity is fi­nan­cially sup­ported by a newly awarded DFG grant "SiGeSn al­loys for en­ergy har­vest­ing at room tem­per­a­ture". In ad­di­tion, this ac­tiv­ity for FZJ is par­tially sup­ported by the Board of Di­rec­tors via the col­lab­o­ra­tive PhD project "CMOS en­ergy har­vest­ing for big data ap­pli­ca­tions".

More about ther­mo­elec­tric el­e­ments:

A ther­mo­elec­tric el­e­ment con­verts tem­per­a­ture dif­fer­ences di­rectly into elec­tri­cal en­ergy. When there is a tem­per­a­ture gra­di­ent across a ther­mo­elec­tric ma­te­r­ial, it in­duces a flow of charge car­ri­ers, gen­er­at­ing elec­tric­ity. This process can be used to cap­ture and re­cy­cle waste heat in elec­tronic de­vices, con­vert­ing it back into us­able en­ergy and re­duc­ing over­all en­ergy con­sump­tion.

For ther­mo­elec­tric ma­te­ri­als, lower ther­mal con­duc­tiv­ity is de­sir­able be­cause it al­lows for a greater tem­per­a­ture gra­di­ent, which is es­sen­tial for ef­fi­cient en­ergy con­ver­sion. GeSn al­loys, with their re­duced ther­mal con­duc­tiv­ity, excel in cre­at­ing this gra­di­ent, en­hanc­ing their ther­mo­elec­tric per­for­mance.

More about the paper:

• https://pubs.acs.org/doi/10.1021/ac­saem.4c00275

The cur­rent issue:

• pubs.acs.org/toc/aaemcq/7/13

Pre­vi­ous