Author: J Date: 28 Versie: 3.
M.Sc. The
Group: T Committe Ing J.W. B Referenc Period: N
S URFA
anarthanan S august 2010 .0
esis University
Transducer Sc ee: Prof.Dr.ir Berenschot, D
e:
Nov. 2009 – au
ACE TE
undaram
y of Twente
cience and Tec r. M. Elwens Dr. J. Snoeijer.
ugust 2010
NSION DI
chnology Gro spoek, Dr.ir.
DRIVEN IMENSI
up
L. Abelmann
N FABRI ONAL M
n, Dr.ir. N.R.
CATION MICROS
Tas, Dr.ir. J
N OF TH STRUCT
J.W. van Ho
HREE ‐
TURES
onschoten,
T ABLE OF CONTENT
1 Introduction... 3
1.1 motivation ... 3
1.2 Outline of this project ... 4
2 Theory ... 5
2.1 Capillary forces ... 5
2.1.1 Folding with capillary forces ... 6
2.1.2 Effects of different contact angles ... 12
2.1.3 Varying the structural dimension ... 14
3 Folding templates ... 15
3.1 Method ... 15
3.1.1 Design of the templates ... 15
3.1.2 Cleanroom processes ... 17
3.1.3 Experimental setup ... 18
3.2 Results ... 19
3.2.1 First results folding structures with more the 2 flaps. ... 19
3.2.2 Results of folding of prism ... 19
4 Folding with different contact angle ... 22
4.1 method ... 22
4.1.1 white light measurement using lateral scanning interferometry ... 22
4.2 Results ... 24
5 Varying design of the prisms ... 25
5.1 Problem definition ... 25
5.1.1 Torsion in folding ... 25
5.1.2 Folding without symmetry ... 26
5.1.3 Varying contact area ... 28
5.1.4 Lifting objects out of the plane ... 28
5.2 Results ... 29
5.2.1 Torsion in folding ... 30
5.2.2 Folding without symmetry ... 30
5.2.3 Varying contact area ... 32
5.2.4 Lifting objects out of the plane ... 34
6 Conclusion ... 36
6.1 Recommendation ... 37
7 Bibliography ... 38
Surface te
1 I NT
1.1 M
The inven compute has more transistor single sili A process technolog scale of t mobile p same tec the figure will be dr
FIGURE 1 S IN SIZE FRO
In this re nitride lik folding m of water We all kn surface te an silicon
ension driven
TRODUCTI
MOTIVATION ntion of the t r. Even more, e capacity tha
r, but also by con chip. This s technology i gy creates sm these mechan hones, which chnology as co
e. This planar ramatically inc
SCANNING ELECT OM 0.5 MM TO 1
eport we will d ke the ancient method has to drops to fold now the smal ension of the n nitride temp
fabrication o
ON
transistor led everyone can an a supercom
the invention s technology is is Micro Elect mall mechanic nical devices.
h react to mo omputer chip r disposition o creased if the
TRON MICROSCO MM
.
(1)discuss a met t Japanese tec be as such th the structure ll insects that water. This sa plate into a str
of three‐dimen to a revoluti n have a comp mputer of the n of the whole s now mature
ro Mechanica cal systems w These system vement, or h s, all the com of the structu re was a way
OPE IMAGE OF A
thod to create chnique called hat a small str
s. This is beca t walk on the ame force can ructure like a p
nsional micros on which ma puter in his po e 70’s. This a e process that e and is being al Systems, als which interact ms are used in
have a compa mponents lay
ures limits the to create ME
SPIDER MITE ON
e three dime d origami. The ructure can be ause capillary e water, this i n be used to fo
pyramid.
structures de possible t ocket nowada all was possib t was needed used to deve so called MEM
electrically.
n sensors and ss function. T planar on the e possibilities MS devices th
N A POLYSILICON
nsional struct e structures a e folded. The forces becom s possible du old structures
hat almost ev ays, the conte ble not only b to make mill lop new types MS. As the na
Figure 1 gives can also be f These structu e surface. Also
of this techn hat stood up fr
N MEMS GEAR‐T
tures by foldin re of course o method we u me dominant a e the capillar s. One single d
very househo emporary mob by the inventi lions of transi s of technolog me already im s an impressi found in cont ures are made o this can bee nology. The po rom the surfa
TRAIN. SPIDER M
ng flat pieces on a micro sca use is the capi at smaller scal ry force also drop is enoug
Page 3 ld owns a bile phone ion of the stors on a gies.
mplies this ion of the temporary e with the en seen in ossibilities ace.
MITES RANGE
s of silicon ale, so the llary force es.
called the
h to fold a
1.2 O UTLINE OF THIS PROJECT
In this project we are going to fold a template made in silicon nitride using standard IC technology and fold this into three dimensional structures. Folding will be done by using the capillary forces of water drops. At first we will determine the theoretical forces working on these structures when a drop of water is placed on them, and assess whether these structures can be made to fold. And if they fold, will these structures stay folded? We will determine the folding rate and how different surface parameters will influence this rate. The structures and theoretical model are already created and determined.
The folding of the structures led to new effects which we will investigate experimentally. Effects that posed new questions were that of the structures stayed folded after folding and that the structures kept folding seamlessly every single time. We will look at the stiffness of hinges, the geometry, the manner in which we dispense water and the surface characteristics of the structures. Getting a clear picture of the forces that play role.
All these experiments will give enough data to give an understanding on the folding of silicon nitride structures.
In addition to all this we will try to lift the logo of the university out of the plane, and try dispense a drop of water through a channel, instead of dispensing water off hollow fiber used in the experiments
We will discuss the created model of folding in chapter 2. In this chapter the theory behind capillary folding and how this force can be harnessed for folding structures. The setup and the method how we did these experiments are elaborated in chapter 3.1. Different experiments are spread out over different chapters and have sub results. This is the chronological order in which the experiments were done. Results of the Simple folding experiments can be found in chapter3.2, experiments with variation of the structure parameters can be found in chapters 4.2 and 5.2. This is followed by an overall conclusion and recommendations.
Surface te
2 T H
2.1 C A The force capillarie regardles everywhe water str that wet with a for This all ca the same in the flu their sha minimiza
FIGURE 2 LI
Every liqu where th liquid air give an a factor. Th example walk on w inner wal
ension driven
HEORY
APILLARY FO e we use to
s. When thes ss of gravity.
ere in their sy riders to walk glasses stick t rce of 10N, so an be explain e liquid. Molec uid as illustrat
pes to minim tion of the int
IQUID AIR INTER
uid is specific he liquid react
interface. The ngle to the sh hese forces be of this scaling water. Or if w ll of the straw
Air
Liquid
fabrication o ORCES
fold structure e tubes are p This phenom stem. But the k on water. Sa
to flat surface o hold a glass p ed by the fac cules on the s ted in Figure mize the expos
terface manif
FACE, TOP MOLE
how it reacts ts to the solid ese interface hape of the dr ecome domin g is a small in we would look w, and if we loo
of three‐dimen es is the sam placed in a bat menon is use e same force c
ame force is r es. A drop of w
plate of 1 kg ct that molecu surface will ha 2. This is an sed surface. T
ests in formin
ECULE IS NOT HA
to the solid s d is called to s
forces depen roplet on a so ant when we nsect that can k at a straw w ok at a shorel
nsional micros me force that
th of water, t ed by plants can be witness
responsible fo water with a
ules in liquids ave half less m unfavorable e The exposed s ng of droplets.
APPY BECAUSE IT
surfaces. This solid liquid in d on the liqui lid plane. This scale down, a n walk on wat with water we ine we will no
structures can be seen the water will to get wate sed as surface or sticky hair
radius of 1 cm
want to be s molecules surr energy state.
surface is also .
IS NOT SURROU
has a lot to do terface, and id. A contact a s angle relates and thus beco ter using surfa will see that ot see this effe
in really nar tend to flow er with nutri e tension, this when coming m can hold tw
urrounded by rounding them
This is the re o called the li
NDED BY SAME K
o with surface how the liqui angle measure
s directly to th ome favorable ace tension, b the water sta ect. As shown
rrow tubes, a into the tube ients from th s phenomena
g out of the s wo glass plates
y other molec m than other eason that flu
iquid air inter
KIND OF MOLEC
e tension. The d reacts to ai ement of the he solid liquid e In this appli but an elepha ands higher a n in Figure 3.
Page 5 also called e upwards he soil to is used by shower or s together
cules from molecules uids adjust rface. This
ULES
e interface
ir is called
liquid will
d interface
cation. An
ant cannot
gainst the
FIGURE 3 DIFFERENCE BETWEEN WATER IN A TUBE AND LARGE WATER BODIES. THERE IS A CLEAR EFFECT ON THE EDGE OF THE TUBE BUT THERE NONE ON THE EDGE OF A SHORELINE
When drop is placed on a surface the drop can form a thin film or become a drop. This depends on the surface and liquid characteristics. This is quantified by the contact angle θ
c.
When a drop tends to become a thin film on a surface we call this fully wetting. When considering water it will fully wet a SiN surface. This is the material our templates are made of. Other liquids will react differently to this surface and may form a droplet. In this case the contact angle of the drop with the surface can be measured by viewing the drop.
2.1.1 F OLDING WITH CAPILLARY FORCES
It was proven recently (2) that when a drop of water is put on a thin sheet of polymer, the polymer would spontaneously wrap. This is due to that the drop will try to minimize the water air interface, the capillary force will fold the polymer sheet as shown in Figure 4. As you can see in the figure the sheet of plastic is on a millimeter scale. After folding the water will totally evaporate. As already explained in the previous chapter this effect should become larger when we scale down to micrometers.
Water between shorelines Water in a vertically
placed capillary
millimeters
meters
Surface te
FIGURE 4 W FOLDING O
There wa solder. Pl (3). This w hinges. F using wat By using in Figure productio the prod object as most righ same size
ension driven
WRAPPING OF D ON A MILIMETER
as a demons lacing solid so was scaled do or folding it w ter.
water on tem 5(a) to (d). T on by convent uction proces s shown in Fig ht top templat e, namely 50µ
fabrication o
ROP OF WATER SCALE.
tration of as older on the h own to scale o would be gre
mplates the TS The design of tional lithogra sses can be f gure 5(b). The
te results in th µm.
of three‐dimen
WITH SQUARE A
sembling mil hinges and me of 100nm. Pro at if we could
ST group alrea the template aphic process found in chap e top center t he cube as sh
nsional micros
AND TRIANGULA
limeter sized elting the sold oblem with th d combine th
ady folded var es are shown es, and there pter 3.1. Fold template resu own in Figure
structures
R SHEETS OF PDM
structures u der made the his technique e scale of the
rious structure in Figure 5(a efore it is desi ing the long ults in the sha e 5 (d). The bla
MS (POLYDIMETH
using capillary e structures st is that the so e last techniq
es (4) with the ). These temp gned in plana structure res pe as shown ack bar in the
HYLSILOXANE ) E
y effects usin tand up from older will be l que and the b
e use of wate plates are des ar manner. M sults in a pris in Figure 5 (c ese figures are
Page 7
EXAMPLE OF
ng molten the plane eft on the benefits of
er. As seen
signed for
ore about
m shaped
c) and the
e all of the
FIGURE 5 (A ZOOM IN O
The prism made. Be just take and the v figure the thickness and then transition different shown in We will m structure progressi model wh angle tha circle. Th the hinge
A) DESIGN TEMP ON THE HINGES (D
m structure is ecause a prism
the front view volume of wa e flaps are be s. Structures c opens, or the ns are shown
angles and t Figure 9 is sh make the follo e is fully wett ion of folding here we again at flaps make
e angle of thi e, t is the thick
PLATES (B) FOLDE D)
s specially de m structure is w. By doing th
ter on the str nt slightly. Th can behave in e hinges are l n in figure 8
he water men hown in Figure owing assump
ting and the we need to d n take the fro when folding s piece of this kness of the h
ED PRISM (C) OP
signed to ver much longer his simplificat ructure. A fron
e hinge is dep two manners loose enough
and Figure 9 niscus is show e 5 (b). An arti
tions, the edg drop covers t define the follo ont view of a p g. The curvatu
s circle is θ. F inge; if we tak
EN AND FOLDED
rify the two d than that it is tion the variab
nt view of the picted as a th s, either the h
so that the s 9. In these fig wn as a thin l ist impression ges of the dro the whole su owing dimens prism to simp ure of the men Figure 6 shows ke l
0as a part
D HALF A DODECA
dimensional t s wide, we can
bles we can d e template of inner line, l
0i hinges are too structure can gures the flap
line. The end n of the folded oplet is pinned urface of the
sions as show plify the mode
niscus on the s the bending of a circle the
AEDER (C) FOLDE
heoretical mo n neglect long
istinguish are a prism is sh s the length o o stiff and the fold complete ps are shown
result of the d prism is show d down on the
structure. To wn in Figure 10 el. W is the w
structure can g of a hinge, w en R is the rad
ED AND OPENED
odel the grou gitudinal dime e the angles o
own in Figure of the hinge a e structure fol ely. These tw n schematical folding prog wn in Figure 7 e edges of the o form a mod 0. This figure s width of the fla n be seen as a where l
0is the dius of this cir
D CUBE WITH
up already ension and f the flaps e 6, in this nd t is the ds al little o types of ly making ression as 7.
e flap. The
del of the
shows the
aps, φ the
a part of a
e length of
rcle.
Surface te
FIGURE 6 B
FIGURE 7 A
ension driven
ENDING A THE H
ARTISTS IMPRESS
R
fabrication o
HINGE, WHERE l0
ION OF A FOLDE
t l
0of three‐dimen
IS THE LENGTH O
D PRISM
nsional micros
OF THE HINGE, T
structures
THE THICKNESS A
AND φ THE ANGGLE THAT THE HIN
Page 9
NGE MAKES
figure 8 schematic overview of the front view of a prism while Folding and opening progresses
Figure 9 schematic overview of the front view of a prism while Folding progresses
Surface te
FIGURE 10 THE MENIS THE MENIS
A droplet bending t interface the bend results in written a
Where volume a with:
FIGURE 11
If we min how the
w
ension driven
THE VARIABLES SCUS AS A PART O
CUS WHEN ENOU
t placed on t the flaps. The of the drople ding stiffness.
a bending en
s
is the and in our sim
PHASE DIAGRAM
nimize the ene equilibrium a
w
θ
fabrication o
FOR THE MODEL OF A CIRCLE THE UGH WATER IS E
the surface w e total energy
et. The bendin E is the Youn nergy per unit
. If we n
balance betw mplification th
M OF THE EVOLUT
ergy in resp angle change φ w
of three‐dimen
L, w IS THE WIDT N θ IS THE ANGL VAPORATED AN
will tend to m of the system ng energy of ng’s modulus t length as:
normalize this
ween mechan he surface S a
sin 2
TION OF EQUILIB
ect to resu depends on t
w
nsional micros
TH OF A FLAP, φ LE OF THE TIP OF D THE STRUCTUR
minimize the m consists of a curved plat and υ the Po
. Th per unit leng
nical bending f as constant, t
and
BRIUM STATES A
lting in an equ the stiffness p
w θ
structures
φ IS THE ANGLE T F THIS PART. RIGH
RE DID NOT FOLD
liquid air inte the elastic en e can be writ oisson’s ratio.
he surface ene
th:
forces and ca the relation o
is the norm
S FUNCTION OF V
uilibrium angl parameter .
φ w
THAT THE FLAP T HT FIGURE SHOW D COMPLETELY
erface and de nergy in the h
ten as and Bending the ergy of the liq
pillary forces.
of the angles malized cross s
VOLUME
e of . As w For a certain
TAKES, WHEN CO WS THE CONCAV
eform the str inges and the d where
flaps over th uid air interfa
. When taking and can section.
we can see in stiffness of t
Page 11
ONSIDERING.
E SHAPE OF.
ructure by e liquid‐air is he angle φ ace can be
g the total be found
Figure 11,
the hinges
the flaps bend due to evaporation of the droplet to a certain angle and then open again after attaining the equilibrium angle , this is shown in Figure 11 as the line where . For a certain lesser stiffness the flaps fold completely into a prism. The in this case is reached when completely folded. This is depicted as the line . The stiffness is the threshold stiffness value where we get the transition from folding and opening into folding completely. For angles smaller then 2 3 the flaps will not close completely and will open again after reaching this angle. At the angle = 2 3 the flaps close completely this corresponds to √3. Reducing any further will result in a not larger than 1.39 rad, leading to a discontinuous transition in the diagram. This means that the structure will abruptly reopen again.
In Figure 11 we see that for the interface shape transitions from a convex in to a concave shape. On the transition point the angle of the interface will be zero ( 0), from this we can find . has a minimum for
= 2 3 and this results in √3. Note that there is a gap between after that = 2 3 . According to this model after the gap the structure should fold open.
2.1.2 E FFECTS OF DIFFERENT CONTACT ANGLES
Previous experiments we used the fully wetting surfaces. This means that the contact angle of the water with the surface of the structures is pinned down at the edges and the angle that the water makes with the edge is totally dependent on the volume of water.
For not fully wetting surface the curvature of the meniscus is dependent on the contact angle this is shown in Figure 12. Looking at this figure one might observe that there is an dependency between θ and θ
c. This dependence is clearly illustrated in Figure 13. We can deduce from this figure that α=180
o‐90
o‐φ‐½ θ and α=90
o‐θ
c, this results in ½ θ=θ
c– φ. Substituting this in the energy function formulated in the previous chapter
will result in
. This means that now the energy is not dependent on θ but dependent on θ
c.
FIGURE 12 SCHEMATIC OVERVIEW OF THE MODEL AND VARIABLES WHEN THE LIQUID IS NOT FULLY WETTING . θC IS THE CONTACT ANGLE, W IS THE WIDTH OF A FLAP, φ IS THE ANGLE THAT THE FLAP TAKES, WHEN CONSIDERING THE MENISCUS AS A PART OF A CIRCLE THEN θ IS THE ANGLE OF THE TIP OF THIS PART. RIGHT FIGURE SHOWS THE CONCAVE SHAPE OF THE MENISCUS WHEN ENOUGH WATER IS EVAPORATED AND THE STRUCTURE DID NOT FOLD COMPLETELY.
w φ w w
θ
w φ w w
θ
θ
cθ
cSurface te It was alr fixed con the maxim then see previous
FIGURE 13
FIGURE 14
ension driven ready proven ntact angle it mum point of
how this ma model the qu
HOW CONTACT A
MAXIMUM EQU
φ
eq,maφ
fabrication o that folding s is not necess f the progress ximum equili uestion arises
ANGLE IS RELATE
ILIBRIUM ANGLE
½ θ θ
cw
½ w
ax1
φ
eq,max2φ
eq,max3of three‐dimen structures are ary to perfor sion lies, the m
brium angle c whether the e
ED TO θ
ES FOR DIFFEREN
φ α
nsional micros e following th
m all the ang maximum equ changes for d experimental
NT CONTACT ANG
structures e theoretical gle measurem uilibrium angle
ifferent conta results will fo
GLES
model, so to ents. We nee e φ
eq, as show act angles θ
c. ollow this mod
see the effec ed to determ wn in figure 1 in Figure 15.
del.
Page 13
ct of the a
ine where
4. We can
As in the
FIGURE 15
2.1.3 V
Experime folded. S fact that we need
FIGURE 16 G
Another end up se effect to design in a view on
HOW THE MAXIM
V ARYING THE ents done rev EM picture sh structures sta change the ar
GLUING RESIDU
effect is that eamless over
the limit by i such a way t n the force tha
MUM ANGLE CHA
E STRUCTURA veal new phe how that ther ay folded. To re over which
IN THE SEAM OF
the structure nanometers.
ntroducing ne hat an additio at play a role
ANGES DUE TO T
L DIMENSION nomena whic e are some m get a better i the flaps are
F A FOLDED CUBE
e fold seamle This is a huge ew designs wh onal torsion fo in this folding
THE CHANGE OF C
N
ch need to b material on the
dea about th glued togeth
E
essly every sin e accuracy tha hich will test orce is neede g mechanism.
CONTACT ANGLE
be examined e edge of the e amount of f er.
ngle time. The at is created b
different kind d to fold seam
E
more. The st fold. This mig force that is h
e flaps move y the capillary ds of distortio mlessly. This e
ructures that ght be the ca holding the fla
over microm y force. We pu ons. We will c
extra hurdle w
t fold stay use of the aps folded
meters and
ushed this
hange the
will give us
Surface tension driven fabrication of three‐dimensional microstructures Page 15
3 F OLDING TEMPLATES
3.1 M ETHOD
3.1.1 D ESIGN OF THE TEMPLATES
Now we have explained the theoretical model we need to design and construct the templates to verify the model. Design of the templates were done in CleWin and are depicted in Figure 17 and Figure 18. There were different sets of structures with different hinge types. We varied the hinges so that we can experiment with different stiffness’s. There are two types of hinges, one type is the solid version, this is a hinge made out of 150nm of SiN. The second type of hinge is also made of 150nm SiN but they are not solid, 8 small hinges connect the flap with a length of 10µm. Both these designs are shown in Figure 18. Also the width of these hinges are varied. Every single type of hinge will result in a different . The varieties of this set is listed in Table 1.
The of the perforated hinges are proportional to length of the hinges holding the flaps together divided by the total length of structure. So how bigger the holes in the length, how smaller the factor .
Type of hinge Hinge width (µm) Flap width (µm) ß
Solid 3 80 5,08
Solid 5 80 3,05
Solid 8 80 1,90
Solid 15 80 1,02
Solid 5 50 4,88
Solid 8 50 3,05
Solid 12 50 0,20
10µm hinges 10 80 0,24
10µm hinges 20 80 0,12
10µm hinges 6 80 0,27
10µm hinges , 5µm gap 25 80 0,07
TABLE 1 TYPES OF DIFFERENT HINGES USED FOR DIFFERENT SETS OF STRUCTURES
FIGURE 17 PYRAMID
FIGURE 18
DIFFERENT STRU
DESIGN PRISM S
UCTURE TEMPLAT
TRUCTURE (A) SO
TES, FIRST ONE I
OLID HINGE (B) H
IS A CUBE, THREE
HINGE WITH HOL
E SIDED PYRAMI
LES
D, HALF A DODEECAEDER AND A FOUR SIDED
Surface tension driven fabrication of three‐dimensional microstructures Page 17
3.1.2 C LEANROOM PROCESSES
As already discussed the templates are created using contemporary cleanroom processes. The process document for creating these structures can be found in the appendix. Here we can see that the following takes place. On a silicon wafer a SiN layer of 1µm is deposited and etched away by plasma etching, leaving grooves for the hinges on the surface of the Silicon wafer. On top of this another SiN layer is deposited of 100 nm. This layer is also deposited between the flaps forming the hinge. Everything but the templates are ethed away, leaving a SiN template on the surface of the silicon wafer. The total structure is then isotropically underetched, the left over Silicon form the pillar on which the template rest. Simplified step by step overview is given in Figure 19.
Blank Silicon wafer
Adding SiN layer
Etching away the hinges
Adding SiN layer
Etching away everything but the template
underetching
FIGURE 19 SIMPLIFIED STEP BY STEP OVERVIEW OF THE PROCESS NEEDED TO CREATE TEMPLATES
3.1.3
To get th just abou created t couple of 50µm and big strong The setup relation t probe m structure
FIGURE 20
Using thi that take measured three tim were com the mode
E XPERIMENTA e structure to ut 300µm in w that can dispe
f hundreds of d a larger one g fiber, to form p consists furt to the table. P akes it possi e. On the micr
PHOTO OF THE S
s setup we w es place. We m
d frame by fr mes the angle, mpensated fo el as discussed
AL SETUP
o fold, one mu width, and th ense small am f micrometers e with an inne m the nozzle.
ther of a tabl Previously me
ble to move oscope a ccd
SETUP WITH THE
will add a drop measured the rame by hand , the average r the viewing d in the previo
ust get a drop e drop must mount of wate
s in diameter.
er diameter of The big fiber e with a micr entioned fiber
the fiber m camera is plac
IMPORTANT CO
p of water on angles that t d using a com
was taken of angle and pl ous chapter u
p of water on t not be much er through a n To do so we f 50µm, the b
is then conne oscope. This r is fixed on a
icrometers a ced, the feed
OMPONENTS HIG
the set of str the flaps make mputer softwa f these three lotted in a gra
sing these exp
the template.
bigger than nozzle which i uses hollow igger one is st ected to a pum
microscope is a probe and fe t a time and is recorded o
HLIGHTED
ructures and c e with the ba are. For every
measuremen aph. The ques perimental re
As explained that. Therefo s small enoug fibers, on wit tronger. The t mp.
s placed at an ed by a pump d maneuver t n the comput
create a set o se of the stru y flap on ever nts as the actu stion is wheth sults.
d before the te ore the setup
gh to generate th an outer di thin fiber is sli
n angle of 60 d p with demi w
the tip just a ter.
of movies of t ucture. The an ry frame we ual angle. The her we can re
emplate is has to be e drops of ameter of id into the
degrees in water. This above the
he folding
ngles were
measured
ese angles
econstruct
Surface te
3.2 R 3.2.1
First expe not repro remarkab Some str Confirmin other wa structure puddle do their orig held its fo If water i the flaps bounce b
FIGURE 21 T
3.2.2
Initial fold heat of th folded co movies ta shown in flap is me measurem structure We did s well. This edges we
ension driven ESULTS F IRST RESULT eriments on t oducible in th ble.
uctures had s ng that the fo y around. Wh e floated on th ownward. So ginal position.
orm even whe is added to th will fold dow back to their o
TRIANGLE PYRAM
R ESULTS OF F ding of prisms he lights mad ompletely. The
aken showed Figure 22. Th easured. The ments were t e folded fully.
see some phe s happened e ere pulled tog
fabrication o TS FOLDING ST these were no he amount of
stiff hinges du olding angle w hen administe he surface of t
we had down This last obs en the water w he cavity the f wn as shown original state.
MID WITH WATE
FOLDING OF P s was a succe e the drop ev e set with the that the ang he measureme
measuremen taken per fla
enomena wort every time. Th ether by the c
of three‐dimen TRUCTURES W ot that succes
time I took t
ue to design a would increase ering an amou
this puddle. W nward folding
ervation cont was evaporate
flaps will floa in Figure 21.
ER IN CAVITY, (B)
PRISM
ss. It is fairly e vaporate it wa e array of 10µ
gle of the flap ent is done at
ts are compe p and averag
th mentioning he meniscus o capillary force
nsional micros WITH MORE TH
ssful. Although to do the exp
and didn’t fold e when the h unt of water in When this pud . After all the tradicts the ob
ed. See Figure t on the surfa . After all the
THE WATER IN T
easy to place as clear that t m hinges fold ps increased t both flaps, a ensated for th ged. Table 2 i
g, the structu of water guid es, and the str
structures HE 2 FLAPS .
h the group h periment. The
d into the full hinges are less nto the cavity ddle dried up water had ev bservation of e 5d.
ace of the wa e water is eva
THE CAVITY IS PU
a drop on the he flaps bend ded all comple until they clo and the angle e angle of the is similar to T
ures that folde ed both edge ructure folds s
had successful ere were som
structures bu s stiff. Folding below the st the flaps follo aporated the the cube that
ater and when aporated in th
ULLING THE FLAP
e structure wit ded and at som etely. Frame b osed seamless
of the flaps in e microscope Table 1 but w
ed completely es of the flaps seamlessly.
lly folded a cu me phenomen
ut they bende g action also
ructure the fl owed the surf flaps jumped t was folded.
n the water e he cavity the
PS OF THE TEMPL
th the setup.
me type of hi by frame analy
sly. Measurem n relation to t . On every fra we added wh
y, folded seam s to each oth
Page 19 ube it was a that are
ed a little.
works the aps of the face of the d back into This cube
vaporates flaps will
LATE DOWN
While the nges even ysis of the ments are the center ame three hether the
mlessly as
er. So the
FIGURE 22
Other hi measurem structure
FIGURE 23
2 4 6 8 10 12 14 16
angles in degrees
0 5 10 15 20 25 30 35
angles in degrees
MEASUREMENT
nges which ment done o e behave in wh
MEASUREMENT
0 20 40 60 80 00 20 40 60
0
0 5 0 5 0 5 0 5
20 25 3
OF FULLY FOLDIN
were solid a n a structure hich manner.
OF FOLDING AN
20
30 35 40 45
NG, TAKEN FROM
nd were nar e that doesn’t In this case a
D OPENING, TAK
40 frame
5 50 55 60 frame
M A MOVIE THAT
rrow were to t fold comple
maximum an
KEN FROM A MO
60 es
65 70 75 es
T PROGRESSED 5
oo stiff for f etely. The tab gle of 32 degr
VIE THAT PROGR
80
80 85 90 9
FRAMES PER SEC
fully folding.
le beneath th rees is achieve
RESSED 5 FRAME
100
95
COND.
Figure 23 s his figure sho ed by both fla
ES PER SECOND
angle ri angle le
angle angle
hows the ows which aps.
ght flap eft flap
rightflap left flap
Surface tension driven fabrication of three‐dimensional microstructures Page 21
Type of hinge Hinge width Flap width (µm) ß folding
Solid 3 80 5,08
Solid 5 80 3,05
Solid 8 80 1,90
Solid 15 80 1,02 Fully
Solid 5 50 4,88
Solid 8 50 3,05
Solid 12 50 0,20 Fully
10µm hinges 10 80 0,24 Fully
10µm hinges 20 80 0,12 Fully
10µm hinges 6 80 0,27 Fully
10µm hinges , 5µm
gap 25 80 0,07 Fully
TABLE 2 FOLDING OF DIFFERENT STRUCTURES
We can convert the measurements done frame by frame, a time progression, in to a measurement done on variation of the water volume. This is exactly how we setup the theoretical model and these measurements will give the measurement point for constructing the graph shown in Figure 24.
FIGURE 24 MEASURED PHASE DIAGRAM OF THE EVOLUTION OF EQUILIBRIUM STATES AS FUNCTION OF VOLUME (4)
It is striking that the simplification of the model has so little influence that the actually measured values correspond almost exactly to the model. Simplifications like totally ignoring a whole dimension do not make the theoretical model deviate from the measured values.
0 0.5 1 1.5 2 2.5
0 0.5 1 1.5 2
normalised volume
eq (ra d )
= 1.8
= 1.5
= 0.07
= 0.8
4 F OLDING WITH DIFFERENT CONTACT ANGLE
4.1 METHOD
To determine the effects of a defined contact on the folding of the structure we need to use liquids that have a defined contact angle on SiN. Or we need to change the surface of the SiN so that water has a contact angle on the this surface. After drafting a list of liquids with different contact angles we came to a conclusion that not all liquids were appropriate to use in my setup due to safety reasons.
Changing the property of a surface can be done by depositing a layer of Fluor Carbon. Fluor Carbon is hydrophobic and the thickness of the layer determines the amount of hydrophobeness and thus the contact angle of the liquid on this surface. A layer of FC can be deposited by using a Reactive Ion Etching chamber (RIE) (5).
The results of some reference experiments done on an empty silicon en SiN surface are depicted in Figure 25.
We decided to do the folding measurements by putting the samples in the RIE chamber between 0 and 10 minutes.
FIGURE 25 VARYING OF CONTACT ANGLE DUE TO TIME OF FC DEPOSITION
Changing the surface into hydrophobic means that the drop of water coming from the fiber might cling to the fiber and not want to stick to the surface. To overcome this we have coated the fiber also with FC, this is done by submersing the fiber into liquid FC. The fiber did not clog by this layer of FC.
4.1.1 WHITE LIGHT MEASUREMENT USING LATERAL SCANNING INTERFEROMETRY
To change the contact angle we chose to deposit a FC layer onto the structures. The question rised whether the flaps on the structure would bend to the extra force asserted by the elasticity of the FC layer. To measure whether this is the case, whitelight measurements using lateral scanning interferometry were conducted on the structures. This will give a clear indication of the orientation of the flaps. The interferometer measures the horizontal distance of the surface.
0 20 40 60 80 100 120
60 30 10
Angle (degrees)
Time (minutes)
Si : Contact angle water
SiN : Contact angle water
Surface te
FIGURE 26
Figure 26 angle of t not addit
ension driven
INTERFEROMETE
6 shows one o the flaps of 0 tionally increa
fabrication o
ER MEASUREMEN
of the actual m .06 degrees. T ase the angle o
of three‐dimen
NT OF THE FLAPS
measurement This we concl of the flaps.
nsional micros
S
t done on the luded as negl
structures structures. T ectable. This
here was an a means that a
average incre adding a layer
Page 23
ase of the
of FC will
4.2 R ESULTS
We did the same folding experiments as in the previous experiments, and this time we looked only to the maximum of each set. We got similar results as shown in Figure 22 and Figure 23, and from these results we looked at the maximum angle attained. We managed to do this experiment for two different hinge stiffness’s.
The results of the measurement are depicted In Figure 27. The solid line is the theoretical and expected progression of to variation of the contact angle . The crosses are the measured points.
FIGURE 27 MAXIMUM ANGLE FOR DIFFERENT CONTACT ANGLES (4)
Again it is striking that we get measurements that fit the model despite of the simplifications. We actually conclude that capillary properties doe follow the model, and thus this process is predictable and useful for fabrication of three dimensional structures on this scale.
0 0.5 1 1.5 2
0 0.1 0.2 0.3 0.4 0.5 0.6
c (rad)
ma x (ra d )
Surface te
5 V A
5.1 P 5.1.1 T
We have hurdles. O lateral di shown in structure rear whe folding p close sea
FIGURE 28 S
ension driven
ARYING DE
ROBLEM DE T ORSION IN F seen that ev One of these rection. In ot Figure 28 wil e would close re the structu rocess Other mlessly, and m
STRUCTURE WIT
D1
D2
fabrication o
SIGN OF T
EFINITION
FOLDING
very time the hurdles is to her words if l the structur seamlessly, a ure was D1 in scenario for may open aga
TH DECREASING F
of three‐dimen
THE PRISM
prism fold th see whether we design fla e fold seamle at the front w width, the h this design is ain.
FLAP WIDTHS
nsional micros
MS
e fold seamle the capillary aps of the pris ssly. This wou where the wid inge would b s that the hin
structures essly. To put t
forces will ov sm with decre uld result in a dth is D2 the h
e pressed dow nge would no
this to the tes vercome the r easing flap wi folding like sh hinges would wnwards. add ot give way an
st we have m rotational for idth over the hown in Figur stretch out a ding extra stra nd the struct
Page 25
ade some
rces in the
length as
e 29 if the
and at the
ain on the
ure won’t
FIGURE 29
The diffe not be an
5.1.2
Next set shown in And if th meet hal same, in t
DISTORTION OF
rence betwee ny influence o
F OLDING WIT
of structures Figure 30, th ey do they ca fway of the o the second ca Rear vie
Fro
THE HINGES
en D1 and D2 f other effect THOUT SYMM have flaps of he width D2 is an fold in two other flap as s
ase the angles ew
ont view
are not more s other than t
ETRY
different wid s larger than D o ways, the e shown in Figu s will be differ
e than the hin the bending a
dth, the left fla D1. Again we
nds of the fla ure 31. In the rent, ɸ1 is not
ge width of 5 and stretching
ap will have a will test whe aps meet each
first case bot t equal to ɸ2.
µm. This is ch g of hinges.
a smaller widt ther these fla ht other, or o th the angles
hosen so that
th then the rig aps will fold se one end of th
of the flaps w
there will
ght one as eamlessly.
e flap will
will be the
Surface te
FIGURE 30 S
FIGURE 32 ( FIGURE 31
ension driven
STRUCTURE WIT
(A) STANDARD 4
POSSIBLE OUTCO
D1 D
fabrication o
TH DIFFERENT FLA
45 ANGLE EDGE (B
Option
OME FOR FOLDIN
D2
of three‐dimen
AP WIDTHS
B) FLEXIBLE EDGE
n 1
NG A STRUCTURE
nsional micros
E
E WITH DIFFEREN
structures
NT FLAP WIDTHS
optiion 2
Page 27
5.1.3 V
Third typ structure the conta hinge stif structure structure We will c is varied f
FIGURE 33 S
FIGURE 34 S
d d
V ARYING CON pe of structur es tend to stay
act surface of ffness and co es folding and es together.
create three se from 20 µm to
STRUCTURE WIT
STRUCTURE WIT
NTACT AREA
res will have y folded desp f the flaps and ontact area co d opening aga
ets of three d o the full 390µ
TH VARYING CON
TH VARYING CON
d
d
a varying co pite of the ela d we will vary ombination w ain. This will g
different conta µm.
NTACT SURFACE
NTACT SURFACE F
d
d
ntact surface astic forces in y the elasticit will be on the give us inform
act angles. Ev
FOLDED
e. Previous ex the hinges. T ty of the hing transition of mation of the
very set consis
d
d d
d
xperiments ha To test this ph es . We will t structures al e quantity of t
sts of twenty v
ave shown th henomena we try to determ
ways staying the forces ke
variations of d
d
d
hat folded e will vary mine which folded to eeping the
d where d
Surface te Last but n created s from the will be et This desig to perfor etching n the etcha underetc
FIGURE 35 S
FIGURE 36 W
5.2 R
ension driven not least we w structures wit
plane and we tched free fro
gn has a smal rm well. With not to free the
ant to reach i hing. This stru
STRUCTURE WIT
WHERE IT CAN G
ESULTS
fabrication o will try to lift h the letters o e will have a m m the underly ll hole in the e hout the hole
e right flap fro n the center ucture suppor
TH A LETTER
GO WRONG WITH
of three‐dimen a structure o of the univers micro billboar ying silicon, an
extension tha the addition om the cente of the arm. I rts the flaps th
H UNDERETCHING
nsional micros of the surface
sity on the fla rd. In the sam nd thus hang at is holding t to the struct r flap as show n this figure t hat lay on top
G
structures using this fol ps. When the e time that th freely like the he letter. This ture which ho wn in Figure 3
the white sha p.
lding action. F e flaps fold the he flaps are u e flaps.
s hole is made olds the lette 36. The hole w
ape is the stru
For this exper ese letters wi
nder etched t
e for the und er can cause t will make it po ucture that is
Page 29 riment we ll be lifted the letters
er etching
the under
ossible for
s left after
5.2.1 T ORSION IN FOLDING
Folding experiments confirm seamless folding. This means that the hinges stretch and fold where needed to facilitate this. We can conclude that the capillary forces are large enough to force the hinges.
5.2.2 F OLDING WITHOUT SYMMETRY
Folding with flaps of different widths resulted in some remarkable results. As already stated in the previous chapter the folding could have gone in two ways. From the measurements of the optical microscope it looked like that the folding occurs in the manner as depicted as option 2 in Figure 32b. Further experiments revealed the full extent of the mechanics in this folding.
In the very first measurements we saw that the flaps will take a maximum angle and fold back when the hinges were stiff enough. This is due the torque on the flaps. We have determined that the capillary effect administers a force on the flaps, the folding occurs not when the this force is larger than the counter acting elastic force of the hinges, but when the width of the flaps times the capillary force is larger than the elastic force. This means a fully folding structure won’t fully fold when the flaps are made less wide. This is shown in Figure 37.
Another way to explain this is by using the model we used in the previous chapter. We have a factor which is dependant on the width of the flaps w: . This means that when we change the w we change the folding property of the flap. In this case in such a way that it will not bend through to close.
FIGURE 37 CAPILARY FORCE VS ELASTICITY FORCE, THIS SIMPLIFIED MODEL SHOWS HOW TORQUE IS RELATED TO THE LENGTH OF THE FLAP.
We have witnessed that a smaller flap takes a maximum angle, while the normal and larger flap keeps folding and then crashing into the flap with the maximum angle. The larger flap doesn’t have the width to fold seamlessly over the a flap that would stop the folding. This shown in frames taken from a measurement in Figure 38. Here you can see that the right flap stays on the same angle while the left flap keeps on folding, in the end it will collapse and not stay folded.
T=F
capilary∙L
1L
1L
2T=F
elastic∙L
2Central flap
Folding flap
Surface te
FIGURE 38 A
Although which ap previous maximum
FIGURE 39 S
ension driven
ASYMETRIC FOLD
the width dif pear to fold n example, und m angle, this is
SEM PICTURE OF
fabrication o
DING, RIGHT FLA
fference betw normally (Figu der a SEM the s shown in Fig
F ASYMETRIC FOL
of three‐dimen
AP REACHES A MA
ween flaps in t ure 39), where
it is clear tha gure 40.
LDED STRUCTUR
nsional micros
AXIMUM ANGLE
his figure is d e the differen at the left flap
E
structures
AND STOPS FOL
ramatic, this p ce in width ar crashed into
DING
phenomena o re not that dr the right flap
occurs also in ramatic as sho p that stood st
Page 31
structures
own in the
teady on a
FIGURE 40 FLAP.
5.2.3 V
This expe contact a stayed fo solid hing µm
2. Onl structure µm
2the l area of 7 used in th This mea
Where µm
this i µm leads and gives SEM pict middle th
CRASHING PHEN
V ARYING CON
eriment consis area and every olded for all c
ge resulted in ly three type es revealed tha
ast 7 structur 720 µm
2and w
he calculation n that for the
is an equilibri s to 4.4∙10
‐5J s us a measure
ures of folded he flaps will no
NOMENA, RIGHT
NTACT AREA
sted of three y variation wa ontact areas.
n all structure es of structur at due to the res open at 68 we assume th ns.
energy
10 and w um point, this J/m
2. This is t e for the force d structure w ot stay closed
T FLAP REACHED
sets, with thre as repeated te
So we can co es opening fo res fell in thi manufacturin 80 µm
2. Becau hat the deviat
with θ = 2/3 s means that the minimum es that the glu with reduced c
.
A MAXIMUM A
ee different h en times in te onclude that t or all contact
is category. R ng process the use the majori tion is due to
1 2 π ≈2 we get a structure th m energy per s
uing substanc contact area
ANGLE AND THE
hinge types. Ev n rows. The h the hinges we
area’s except Repeating thi ere is a deviat ity of the stru production p
U
b=4∙10
‐5J/m hat folds seam square area n
e has to asser is shown in F
LEFT FLAP CLOSE
very set had t hinges with ho
eren’t stiff en t for contact s experiment ion. The first t cture only sta process we wi
m
2. For a conta mlessly over t
eeded to kee rt.
Figure 41. For
ED INTO THE ST
twenty variati oles in them f nough. The se area’s larger t for the ten three rows op ay closed with ill take this va
act area of 72 he whole leng ep the structu
r any larger h
EADY RIGHT
ons of the folded and t with the than 720 n identical pen at 720 h a contact alue to be
20 µm x 1 gth of 800 ure folded,
ole in the
Surface te
FIGURE 41
There wa middle of need a la
FIGURE 42
We made
Where w length) . height wh The disto requiring expressio the lengt
dimensio the same
ension driven
FOLDED STRUCT
as an unexpec f the folded f rge force.
BENDING IN THE
e an estimatio
w(x) describes E is the Youn hen taking the ortion is appr
g 0 0,
on for h of the flaps
. We meas ons we get 602 e of the whole
fabrication o
URE WITH SMAL
cted effect wh flaps for very
E MIDDLE OF FOL
on by taking u
the deformat g’s modulus e crosssection roximated by
s. Figure 43 sh sure a maxim 2 N/m for q. T e width. If yo
of three‐dimen
LLER CONTACT AR
hile doing the large gaps. A
LDED STRUCTURE
sing the Euler
tion along the and I is the s n of a flap, 1 µ a 4
thorder p
0,
. Here u is t hows the w(x) mum distortio This model ass
u look carefu
nsional micros
REA
ese experimen A SiN flap of 8
ES