Paper ID #10406

Introducing Nanotechnology into an Undergraduate Microelectronics Course

Prof. Chung Hoon Lee, Marquette University

Chung Hoon Lee is an Assistant Professor in the Department of Electrical and Computer Engineering atMarquette University, Milwaukee, WI.

Dr. Susan C. Schneider, Marquette University

Susan Schneider is an Associate Professor in the Department of Electrical and Computer Engineeringat Marquette University, Milwaukee, WI. She is also the Director of Undergraduate Laboratories for theElectrical Engineering program. Dr. Schneider is a member of ASEE, the IEEE, Sigma Xi and Eta KappaNu.

Mr. Trevor Thiess, Marquette University

c©American Society for Engineering Education, 2014

Page 24.810.1

Undergraduate Introduction to Micro-fabrication of Memristors

Abstract

In Spring 2012, a pilot project to increase student exposure to nanotechnology was carried out in the first electronic devices course in the electrical engineering program at our university. Students were given the opportunity to build and test memristors in the nano-electronics research laboratory under the supervision of their instructor. In this pilot project, 10% of the students in the class chose to participate. Based on the success of the first trial, this project will be run again in Fall 2014. The topic of “memristors” was chosen for this project motivated in no small part by the fact that these devices are currently “hot” in the microelectronics/nanoelectronics research community. An additional attribute, deemed essential for this student experience, is the ability to create a macro-scale version of the memristor. The project work included mini-lectures and assigned readings on the history, theory of operation and fabrication, and applications of these devices. The project consisted of two sections: macro-scale and micro-scale memristors. During the macro-scale portion of the project, students practiced the memristor fabrication techniques for several different base metals and sulfiding mediums. Then based on the results (either success or failure) determined by the measured current-voltage characteristics of the memristor, the students made choices on the materials and methods to scale down the macro-scale memristor to the micro/nano-scale memristors using an industry standard fabrication techniques. A graduate student working in the nano-electronics laboratory assisted the students during all experimental work including safety training and help on both fabrication and data acquisition. 1. Introduction A memristor is a passive electrical circuit component proposed to explain non-linear circuitry by Leon Chua in 1971 [1]. In 2008, an HP Labs team realized the conceptual fourth component in circuit theory [2] by fabricating a TiO2 based memristor. A memristor is characterized by its ability to ‘remember’ previous resistance states. A memristor is a non-volatile device able to maintain its state when it experiences either no voltage or constant voltage (device off or steady-state). It relates electrical flux and charge as a resistor relates electrical current and voltage [2]. A team of second year undergraduates was assembled to pursue research into memristor design, fabrication, measurement, and analysis. A prototype of the memristor students fabricated is shown in Figure 1. The students were recruited during an introductory lecture on solid-state principles in their microelectronics course by introducing the burgeoning field of memristor research and commercialization. The class was shown a video on the memristor and its potential in the future of electronics was discussed. The basics of the memristor were explained alongside the diode; each placed within its own historical context—both symbolic of transitions in the way engineers imagine circuitry. A conversation progressed over a few weeks of lecture and eventually led to a team of seven students being formed to investigate memristor fabrication. The team included biomedical,

Page 24.810.2

Fig

electricalthree foc

2. Mini Three leccommon early on The first of the medevice hatheoreticthe theor The secoV curvesthe coursthe devicsecond lethe electrwithin thstudents preparati In the fintechniqueevaporatiand devis

gure 1. A mic

l, and compuused lecture

i-Lectures

ctures servedunderstandiin the projec

lecture coveemristor fromad been propal device rel

retical guidel

ond lecture fos. The switchse material aces from theiecture consisron tunneling

he device capwere asked t

ion for the ne

nal meeting, es employedion depositiosed a plan th

cro-scale mem

uter engineers on the mem

d to bring alling amongst ct.

ered the histom theory to tposed by arglating electrilines to desig

ocused on thhing mechans reference, ir course leadsted of how tg phenomenpable of contto carry on text and final

student comd in previouson and sputtehat utilized m

mristor fabric

rs. Arrangemmristor.

l students upthem. This

ory of the mthe prototype

guing that inhical flux andgn a develop

he memristornism of the m

students lead to dramatithe fabricati

na and how thtrolling the etheir own resl lecture.

mpared notes s research effering. Stude

materials and

ated by secon

ments were m

p to speed onfacilitated te

memristor. The developed herent circui

d charge [1,2pment plan f

r’s principle memristor waarned how thc electronic on techniquehe reductionelectron dynsearch on the

and assembfforts. Unfamnts learned w

d tools to car

nd-year under

made for sho

n the materiaeam commun

he discussiond 37 years latit element sy]. Discussion

for their mem

of operationas explained

he seeminglydifferences es gave rise

n of a metal fnamics througe application

bled a more cmiliar techniqwhich tools rry out their

rgraduate stu

ort after-clas

al as well as nication and

n focused onter. Studentymmetry poin was focuse

mristor.

n and the chad by the I-V cy subtle diffe

in the memrto such prinforms a chargh the two tens of their fie

comprehensiques were exwould be avown memris

udent teams.

ss meetings a

establish a d coordinatio

n the progress learned thainted to a ed on how to

aracteristics curves. Usin

erences betwristor. The

nciples – suchrge boundaryerminals. Theld in

ive list of thexplained, sucvailable in thstor fabricati

and

on

ssion at the

o use

of I-ng

ween

h as y he

e ch as

he lab ion.

Page 24.810.3

3. Preli To explomacro-scfind an oreductioncontact, sbath. Tab The powcontainerwhich allthe metalof aluminshows poaluminummemristo

Althoughthe methouniformitfurnace mquartz tuheated, mappearedmethod. The two fabricatiofabricatiowere impwet chemsulfur coof time, tand silve To confirmeasuredsample. B

iminary han

ore and choocale prototypptimal mate

n of metals wsulfur vapor ble 1 summa

der reductior. The reactiolows studentl appears chanum, no colooor reaction m and sulfuror.

T

h the powderod producedty of the sulfmethod in Tabe furnace (

metals were pd to show the

sulfiding meon. The limiton, a 4-inch practical for mical bath prmpounds (cathe prototypeer.

rm the sulfidd. In order toBy scratchin

nds-on expe

se the apprope was fabricrial and met

with sulfur wproduced by

arizes the me

n method is on between ts to executearcoal-like, mor change is for the other

r, an aluminu

Table 1. Res

r sulfiding md non-uniformfide film, chables 1 and 2the detailed placed near te same type o

ethods (powtation of thediameter wathe 4-inch d

rocedure. Thalled “Liver e was left to

ded film waso make the I-ng a corner o

riments on

priate metalcated. Three hod for micr

was also demy heating suletals and sulf

done by expthe copper a

e the experimmatching thefound, indic

r sulfiding mum block has

sults of macr

method was sm sulfide fil

hemical vapo2). To produprocedure isthe sulfur soof reaction, b

der and vapose two meth

afer would bdiameter wafhis method in

of Sulfur”). dry. This m

s the proper m-V measurem

of the sulfide

a macro-sca

s and the memetals and

ro-scale memmonstrated by

lfur in furnafiding metho

posing bulk mand sulfur ocment in a 24-e appearancecating no or

methods as ws been chose

ro-scale meta

simple and rems across th

or method inuce the sulfurs given in th

ource. By visbut with bett

or) were comhods was the e the standar

fer scale. Annvolved expo After expos

method also p

material for ment, two eled surface, th

ale memrist

etal reductiothree sulfidimristor fabriy three technace, and wet ods and resu

metal directccurs at a slo-hour trial. Ae of copper aan extremely

well. Due to ten to be an e

al sulfiding m

epeatable fohe metal surfn a quartz furr vapor, sulf

he following sual inspectiter film unif

mpatible withsample size

rd size. The nother reductosure of a busure to the bproduced cha

a memristorlectrodes wehe host meta

tor

on (sulfiding)ing methods ication. The niques: direcchemical su

ults.

tly to sulfur pow rate at rooAfter reductioand silver suy slow reactthe poor reacelectrode ma

methods.

or producing faces. To imrnace was usfur powder wsection). Whon the surfac

formity than

h the micro-e. In the casepowder and

tion method ulk material ath for a desarcoal-like f

r, the I-V curere connectedal was expose

) methods, awere used tchemical

ct sulfur powulfur solution

powder in a om temperaton, the surfa

ulfide. In thetion. Aluminction betwee

aterial for the

metal sulfidmprove the sed (indicatewas heated inhile sulfur wce of the methe powder

-scale memrie of micro-scd vapor meth

explored wato a solution

signated lengfilm for copp

rve was d to the bulked and used

a to

wder n

ture, ace of e case num en e

des,

ed as n a

was etals

istor cale hods as the n of gth per

k as

Page 24.810.4

one electof the aluthe film. memristocurve. Macro-faexperimewhich comeasurem 4. Fabr With the to the mi

4.1. DTceoatelew

4.2. M

Atsfu

trode. A 20-guminum wireThe measur

or reported in

abrication seental methodould be used ment setup to

rication of a

successful fcro/nano-sca

Design The completcreating one electrode conorientations bare made frothermal evapelectrode by layer is in plelectrode witwas about 5

Figure 2. D

Memristor FAs shown into define a sisilicon waferfabrication. Iused. On eac

guage alumie to the sulfired I-V charan the literatu

erved as an imd. The macroin the microo measure th

a micro/nano

fabrication aale memristo

te design crojunction (th

nfiguration aby choosing

om either silvporator [3]. Treduction deace, the top th 90o degreby 5 micron

Design of misulfiding A

abrication the Figure 2

ingle micro/nr or microscoIn this projecch wafer, twe

num wire wided film waacteristics ofure and a com

mportant guio-scale fabrico-scale memhe I-V chara

o-scale mem

and demonstror fabricatio

osses two paihe device itseallows the deg neighboringver [4] or copThe memristeveloped in electrode is e rotation of

ns square.

icro-scale meAg. (c) depos

2, the first stnano-scale mope cover glct, a 4-inch selve devices

was used for tas made by af the reducedmputer simu

ide for studecation metho

mristor fabricacteristics of

mristor

ration of man and I-V cu

irs of electroelf), where oevice to be mg electrodes pper and aretor layer is crthe macro-scdeposited ov

f the shadow

emristor andsit top-electr

tep of the mimetal wire onlass could besilicon dioxis were accom

the other elea gentle mechd film showeulation of the

ents to carry ods showed ation. Studena memristor

acro-scale meurve measure

odes at a 90°one overlapsmeasured in

in clock-wise created by reated on thecale fabricatver it in the

w mask. The

d process florode in 9 o d

icro-memristn a substratee used as a suide film (SiOmmodated.

ectrode. The hanical touced the charace theoretical

on the micrvarious sulfnts also comr.

emristor, stuements.

° angle as shthe other. Tfour differense rotation. Tmetal depose surface of tion. After thsame way astarget size fo

ow. (a) silverdegree rotatio

tor fabricatioe. A thermalubstrate for

O2) coated si

electrical coch of the wirecteristics of memristor I

ro-scale fiding methompleted the I

udents proce

own in FiguThe four-nt possible The electrodsition in a the bottom

he memristos the bottom

for the memr

r metal wireon.

on process wly oxidized the memristlicon wafer

ontact e to a

I-V

ds, -V

eded

ure 2,

des

r m ristor

(b)

was

tor was P

age 24.810.5

Fs DtHfscpob Ttspsts

4.3. S

Aaffo

4

Figure 3. (a)shadow mask

Defining thetransfer the dHowever, in fabrication sstudents. Thechemical etcproject for unonce it had bbottom) for e

The shadow technique utisilicon etchinplaced in theshadow maskthe substrateshown in Fig

Sulfide methAs briefly deall three metfabrication, tfabrication. Ton the substr

.3.1. FurnaThe sulfas showmelting

) shadow mak process.

metal wire design to subthis project,implicity. Ae use of a shhing of metandergraduate

been created each device

mask used iilizing opticang. The shade thermal evak was detach. The shadow

gure 3.

hods escribed in shods led to athe furnace aThe powder rate which in

ace Method fur vapor men in Figure 4point of sulf

ask (b) patter

at micro/nanbstrate and w, a shadow m

A shadow mahadow mask al layer, whies. Another it could be rfabrication.

in this projecal lithographdow mask waporator. Afthed from thew mask and

ection 3, stua successful and wet chemmethod was

nterfered wit

ethod uses a 4. The furnafur, 115 °C.

rned metal w

no-scale is tywet chemicalmask processask for the meliminates c

ich were beyadvantage oreused to pat

ct was fabrichy, reactive ias then affix

fter metal evae silicon wafa slice of a w

udents develometal sulfid

mical methos eliminated th the follow

quartz tube ce temperatuAn alumina

wire with con

ypically donl etching to s to define th

metal wire pacomplicationyond the capof the shadowttern the wir

cated by a tyion etch (RIExed to the silaporation wifer leaving thwafer contai

oped three suded film in mds were usedbecause the

wing top-elec

furnace to aure, 200 °C,

a boat of sulf

ntact electro

ne by optical define the phe wire was atterning wasns of optical ability and t

w mask methre and electro

ypical micromE), anisotroplicon wafer (ith desired thhe patterned ining the pat

ulfiding metmacro-scale md in micro-s method left

ctrode depos

accelerate sulis set higher

fur is placed

odes by the

lithographyattern of thechosen for t

s provided tolithography

time scale inhod was thatodes (top &

maching pic wet chem(substrate) ahickness, themetal wires

tterned wires

thods. Althomemristor cale memrist many particsition.

lfur sublimar than the in the cente

y to e wire. the o the and

n this t

mical nd e s on s are

ugh

stor cles

ation

er of

Page 24.810.6

4

the furnatemperasulfur sodirectly the furnaTable 2. The sulfmost funtuning bthis use.process.

.3.2. ChemChemicaimportansubmergthe meta The chevariabledisplayethe intenmetal subehaviorsulfidedsuccessfinvestigconcentrsection 3Table 2.

Figure 4.

ace’s quartz ature effect oource and theadjacent to tace’s insulat.

fur vapor menctional devbefore it coul. A major dra

mical Bath al reduction ntly, its abiliged in a solual reacted wi

mical bath ms than the fu

ed memristivnded reactionurface after sr was due to

d. Several attful to producation on the ration might3. The sulfid.

Quartz tube

tube and deof the sulfidine temperaturthe sulfur sotion (1 & 4 i

ethod producices. This inld be a reliabawback to th

via sulfur soity to be scal

ution containith the sulfur

method was aurnace methove behavior in seemed to sulfiding. It to the initial thempts with vce devices shchemical co

t produce theding parame

e furnace for

evices are pong. The posire is varied a

ource (2 & 3 in Figure 4).

ced a wide vndicated that ble process, his method w

olution was fled up to ful

ning sulfur ior ions creatin

a more easilod. Howeverin this studyoccur as jud

turned out tohin metal layvarious conchowing the mompositions e memristiveeters used in

r sulfur vapo

ositioned in mitioning of dacross the triin Figure 4) A record o

variety of conthe method but that it di

was that it w

favored for ill-wafer prodons for a speng sulfided s

ly controlledr this method. Microscopdged by the co be that the yer. The thincentrations amemristive bof the “Live

e film as the chemical ba

or sulfiding m

multiple placdevices with ials. Devices) and above f trial runs is

nditions, butneeded muc

id have greawas an individ

its simplicityduction. Thecific duratiosurface.

d process witd produced nic investigatcolor and grabsence of m

n metal layerand reaction behavior. Fuer of Sulfur”macro-scale

ath method a

method

ces to study respect to ths are placed the boundars summarize

t also yieldedch more careat potential fodual-device

y and most e device was on. The surfa

th fewer no devices thtion showed ranularity of memristive r was entireltimes were n

urther ” and e case in the are listed in

the he

ry of ed in

d the eful or

ace of

hat that the

ly not

Page 24.810.7

4.4. T

OmsTuTtjj

5. I-V C

Prototypetaped to tacrylic plbonding

Table 2.furnace

Top ElectrodOnce the memask with 90second metaThe second eup optical imTo determinethe electricalunction wasunction resi

Fig

Characterist

es were prepthe electrodelate as showmethod, but

List of eachand chemica

de Depositioetal sulfided 0° rotation wl deposition evaporation

mage of the de the qualityl resistance bs about 3.5 kstance value

gure 5. A clo

tics of the m

pared on an ie contacts an

wn in Figure 6 here a mech

h sulfiding coal bath meth

on layer was cr

with respect tto create theof top-electr

device after ty of the junctbetween the

kΩ. Among tes and the oth

ose-up image

micro-scale m

insulating acnd the other 6. Typicallyhanical conta

ondition parhod.

reated, the deto the initiale top-electrorode complethe second mtion created electrodes w

the 12 deviceher two devi

e of the junct

memristor

crylic plate. Tend of the w, the terminaact with a K

rameters for

evices werel pattern of thode was doneeted the memmetal evaporbetween the

was measurees, 10 deviceices were ele

tion, metal-m

Thin electricwires was claal contacts w

Kapton tape s

both Ag and

mounted to he metal laye in the therm

mristor fabricration is showe overlappinged. Typical rees showed thectrically sh

metal sulfide

cal wires (24amped on thewere created simplified th

d Cu for both

the shadow yer. Then themal evaporacation. A clown in Figureg micro-wireesistance forhe typical orted.

e-metal.

4-gauge) were side of the using a wiree procedure.

h the

e tor.

ose-e 5. es, r the

re

e-. The P

age 24.810.8

wire-bonand Au e The I-V cwith the generatormemristorepresentoutput cuelectrodeand is fed With the 8. All woa copper characterthe I-V cthe samethe both

Fi

nding was avevaporation f

characteristifunction genr. The functior electrodests the x-axis urrent is conve is connected to channel

setup, a typorking devicdevice redu

ristic curve ourve is due tmetal and th

electrodes, t

igure 6. Prep

voided becaufor the conta

ic measuremnerator via Gion generatos. The AC vo(applied volverted to a v

ed to a resisto2 of the osc

ical I-V chares show simced by the fu

of the memrito the differehe junction cthe I-V curve

Fi

paration of d

use of furtheract pads.

ment setup is GPIB. A cust

r is set to ouoltage is alsoltage). To re

voltage and por (1 kΩ). A

cilloscope, re

racteristic cumilar trend offurnace methistor simulatent metal elecondition bee will be sym

igure 7. Mem

device for I-

r process com

shown in Fitom Labviewutput AC volo fed to chanepresent the tplotted with

A probe is conepresenting t

urve of the mf the curves. hod for ten mted by a PSPectrodes on tetween the mmmetrical.

mristor I-V m

V characteri

mplications

gure 7. The w program coltage (sine wnnel 1 on thetypical memthe input vonnected betwthe y-axis.

micro-scale mThe device

minutes at 20PICE model. the memristo

metal and sul

measuremen

istic measur

such as opti

computer coontrols the f

wave) on onee oscilloscop

mristor I-V choltage. To doween the sec

memristor isshown in th

00 °C. Figure The asymmor. If the eleclfided metal

t setup

rement.

ical lithograp

ommunicatefunction e of the pe, which haracteristico this, the seccond electrod

s shown in Fe Figure 9 (ae 9 (b) is the

metric behavictrode materis identical o

phy

s

s, the cond de

igure a) is e I-V ior of rial is on

Page 24.810.9

6. Discu The mostprototypeentire macurves wdevices d The chemPossibly is occurriof sulfur could be Figure 9 demonstrpositioneBoth devreduced i& 9) indiof the meChemicashown in A commosimilar ‘sbut after collapse from the

Figure 8. I

ussion

t promising es were madaterial being

were observeddemonstrated

mical bath mthe solutioning. In thesecompounds added to the

shows the vrate the diffeed the farthesvices 4 & 5 ain the same ticate that lesetals. Similaral Bath methon devices 2 &

on trend appsweet spot’ fa few minutand becomemeasuremen

I-V charactermemristor. (

results camede with a rela

reduced witd on almost d the most st

method needsn needs to bee trials a Live

and could lee technique s

variety of reserence in plast from the h

appeared to ntrial, exhibitss heat and mr reduction ood yielded a

&3.

peared duringfor functionates of expect a flat resistant setup for a

(a) ristics of a m(b)simulated

e from the coatively thinnthout leavingall devices (tability and r

s further refin more finelyer of Sulfur ead to differsuch as heati

ults observeacement withheater, it demneed further ted what appmore time in occurred in ta more comp

g the measurality, mostlyted behaviorance, indicata few minute

memristor (a)d I-V curve o

opper samplener initial depg electrode m(90 %) from repeatability

ning in ordey controlled tsolution was

rent sulfides ing the solut

ed over severhin the quartmonstrated aexposure tim

peared to be hthe furnace

the silver furlete and unif

ring of untouy proportionar, the area witing an openes, the devic

a)measured Iof the memris

es. This is pposition, leavmetal. Memrdifferent me

y.

r to be a viabto ensure thas implementdeveloping.tion before r

ral reductiontz furnace tua more deep rme to complheat damagecould lead t

rnace trials (form reducti

uched prototal to their iniithin the loop

n circuit. Aftce shows the

I-V curve of istor by PSPI

ossibly becaaving them toristor I-V chethods, altho

able means oat only the inted, which co In the futur

reaction.

n methods. Dube; althoughreduction thlete reductione. These threto a more eff(as shown byion of the su

types. They itially measuped I-V curver disconnec

e characterist

(b) the micro-scICE.

ause the silveoo prone to t

haracteristic ough the cop

f productionntended reacontains a varre more varia

Devices 4 & h device 4 whan device 5.n. Device 9,ee devices (4fective reducy device 6). urface metal,

would have ured resistanve would cting the devtic I-V curve

cale

er the

pper

n. ction riety ables

5 was

. , 4, 5, ction The , as

a nce,

vice e

Page 24.810.10

123456789

again whtemperatuFurther atransport The reseacleared thimprovemthe failur

: Electrode b: Same elect: Exposed to: Exposed to: Same trial : Exposed to: Reduced by: Probe dam: Deformatio

hen connecteure increase

annealing stet.

arch has demhe path for hment. Fabricre of chemic

Figure 9: Pbefore reductrode as in 1o sulfur soluto sulfur vapoas 4, but pla

o sulfur vapoy direct expo

mage observeons in coppe

ed to the mea due to the J

eps after dev

monstrated anhigh-potentiacation technial bath proce

Physical Resction , post-reducttion (chemicor at 200 °C aced in betweor at 200 °C osure to sulfd on 100 nm

er observed w

asurement seJoule heatingvice is fabrica

n effective fal advancemques can beess, the poss

sults of Redu

tion (30 secocal bath) for for 10 minueen the centfor 10 minu

fur powder m Ag post-rewhen placed

etup. This phg at the junctated may sta

fabrication tements to the p

further refinsibility of us

uction by Var

ond Chemica20 seconds

utes (placed aer and the en

utes on bound

duction (15 d directly on

henomenon ition interferiabilize the te

echnique forprocedure byned to addresing annealin

rious Metho

al Bath)

at end of quand (boundarydary

second chemheater (10 m

is probably ding with the emperature e

r micro-scaley uncoveringss electrode

ng process, a

ods.

artz tube) y)

mical bath) minutes)

due to the ion transpor

effect of the i

e memristorsg areas for adhesion iss

and poor

rt. ion

s and

sues,

Page 24.810.11

response probing mmeasurem The iteraproject ballowed sresults. The projealready fthe silico

7. Stud Students otherwisetime, get Students methods phase in developm

in silver-bamethod, conment jig to e

ative approacecause no onstudents to le

ect proved tofound in typion wafers. A

ent Evaluat

found this pe in undergrhands-on exfound analyto be very eeach method

ment.

sed devices.trolling for “

ensure consis

ch implemenne involved earn and exp

o be very lowical electronibreakdown

Table 3: Lis

Table 4: Lis

tion

project very eaduate studixperience wiyzing the samducational bd. It is uniqu

Analysis te“burning” oustency.

nted by the tehad a particu

plore a wider

w-cost as weics labs. Theof necessary

st of materia

t of equipme

engaging bees. Studentsith new toolsmples after rebecause it allue in that stu

chniques canut devices du

eam was fouularly deep ur range of po

ell. The pricee most signify fabrication

als used with

ent used with

ecause the su liked being s, and learn eduction andlowed themdents implem

n be improvuring measu

und to be espunderstandinossibilities a

e was kept dficant costs wn supplies is

h suggested a

h suggested a

ubject matter able to put iabout microd comparingto see the mment a full c

ved by furtheurements, and

pecially usefng of the proand see a bro

down by usinwere the shalisted in Tab

alternatives.

alternatives.

r was not comin sizeable p

omachining tg them acros

molecular chacycle of rese

er developingd design of a

ful for such aocess initialloader spectru

ng many tooadow masks bles 3 & 4.

.

mmonly covperiods of labtechniques. s different anges of matarch and

g the a

a y; it

um of

ls and

vered b-

terial Page 24.810.12

Making tenvironmbringing amongst they lackproject gconfidentsmall gro The groustudent tomanageapossibilitspent moplans & dof the teaknowledgmany fac& gradua This projnumbers student insome stuundergrathey are a This projcurriculucompleteconsists ocommitmthree minthe fabricposed a lavailablecut downgroups careduction Table 5:

the introductment, which i

all students them. The s

ked understanoing and evet contributin

oups, allowin

up size (3 or o lessen the

able. Althougty of creating

ore time in thdetails). If alam and lead geable studecets of the prate student a

ject was a voof participan

nput upon codents found

aduate electroalready prep

ject could beum with some the fabricatof 3 large pr

ment once a wni-lectures incation procelarge probleme electronicaln by preparatan be assignn parameters

breakdown

tory lecture mis conduciveto the same tudents whonding in the entually dom

ng to the projng each stud

4 students intime dedicat

gh it had timg parallel unhe lab had mll students wto dispersion

ents participaroject: the nuaid, and the c

oluntary suppnts and leveompletion ofthe time com

onics laboratpared to plan

e easily inserme modificati

tion of multirojects, each week. Followntroducing adure and tesm, lectures clly through otion by a labed different

s (e.g. time, t

of approxim

meetings lige to team builevel of und

o participatedmaterial. Th

minated the pject over tim

dent to partic

n each grouption they had

me benefits, snderstanding

more feedbackwere in the lan of responsating the moumber & avacomfort of th

plement to thls of participf the projectmmitment totory course,

n for that amo

rted into a stions. Table 5iple device sspanning 5

wing the guiand discussinsting would bcould be brokour universitboratory T.A

materials antemperature,

mate time com

ght and discuilding. The mderstanding ad least in thihe more wellproject. This

me. This probcipate in a sp

p) had its advd to commit,splitting the tgs between grk and had a

ab at the samsibility, oftenst. Ultimatelailability of mhe team with

he required mpation are ex. Although so be too demit could be m

ount of time

tandard unde5 lays out thesamples. At oweeks and cdelines of th

ng the devicebe performedken down inty course ma

A. When convnd encourage, dilution, etc

mmitment of

ussion-basedmini-lecturesand creating s project wel-versed stud

s led the lessblem can alspecific, focus

vantages. Th, which oftenteam into sh

groups of studifferent und

me time thougn leading to tly, the numbmachines/to

h each other

microelectroxpected if ofstudents wer

manding. If inmore manag

e in lab.

ergraduate ele hours of coour universitconsisting ofhis structure,e. The followd in the fina

nto smaller management sverted into aed to investic.).

f each step in

d fosters a fris are also ve a common l

ere more likedents were e familiar stuo be helped sed way.

his group sizn made thing

hifts also opedents (i.e. thderstanding gh, it could cthe most conber of studenools, the amo& the subjec

onics courseffered in the re interested nserted into geable for stu

lectronics laommitment nty, the analof a three-hou, the first labwing three laal lab. If timemini-lecturesystem and la

a larger classigate modific

n the fabrica

iendly ery importanlanguage ely to feel theager to get tudents to feel

by working

ze allowed eags more ened up the he students wof the projecrowd the fonfident or nts depends oount of lab spct matter.

. Similar future basedin the projeca standard udents becau

aboratory necessary to

og electronicur time b would coveabs would coe constraintss and made ab time can bsroom settingcations to th

ation process

nt for

at the l less in

ach

who ct

ocus

on pace

d on ct,

use

o cs lab

er the over

be g, lab e

s

Page 24.810.13

8. Conclusion The two teams successfully developed a fabrication technique that produced functional memristors. Fabrication research concludes that reducing copper to copper sulfide to form the insulation layer of a memristor by means of a sulfur vapor method is an effective way to realize memristive behavior. The overall structure of the project was successful and valuable to students. This project demonstrates an effective path for introducing nano- and micro-electronics engineering to undergraduate students in a personal, experience-driven way.

9. References

[1] Chua, L. O., "Memristor—The Missing Circuit Element", IEEE Transactions on Circuit Theory, CT-18 (5): 507–519, 1971

[2] Strukov, Dmitri B.; Snider, Gregory S.; Stewart, Duncan R.; Williams, R. Stanley, "The missing memristor found", Nature, 453(7191): 80–3, 2008

[3] Madou, Marc J., Taylor & Francis, Fundamentals of Microfabrication and Nanotechnology, Volume Two: Manufacturing Techniques for Microfabrication and Nanotechnology, Third Edition, CRC Press, July 2011

[4] Terabe, K., Hasegawa, T., Nakayama, T., Aono, M. TI, Quantized conductance atomic switch, Nature 433, 47-50, 2005

Page 24.810.14