2.3 Crystallization
2.3.3 Eetive volume fration and rystallization
Crystal
Crystal
Glass 0.494 0.545
0.58 0.64
0.74
f F f
M
f G
f eff
Figure2.26: Phase diagramof a hard spheres dispersion [116118℄.
degrees of rosslinking. The nuleation and growthof the rystalliteshas been visualized
by polarizedlightmirosopyand theinuene ontherheologial properties isdisussed.
2.3.2 Experimental
The dierent ore-shell laties presented in the hapter 2.1 were used in this setion.
Following the systems the rosslinking dened by the amount of BIS in the shell was
adjusted to 1.25
mol.%
(KS1), 2.5mol.%
BIS (KS2) and 5mol.%
(KS3) (see setion2.1.2).
The ow urves and the dynami measurements performed onthe KS2 suspensions have
been investigated usingastrain-ontrolledrotationalrheometerRFSIIfromRheometris
Sienti, equipped with a Couette system (up diameter: 34
mm
, bob diameter: 32mm
, bob length: 33mm
). Measurements have been performed on 12mL
solution andthe temperaturewasset with an aurayof 0.05
o
C. A stress ontrolled rheometer MCR
301 (Physia) has been used for the experiment performed onthe KS3.
Polarized mirosopy has been performed with aLeia DMRXE. Sample were lled into
a0.1
mm
thik apillarythermostatedwith anaurayof 0.05o
C. In orderto followthe
rystallizationproessathermostatedellwasdesignedfor0.1and0.5
mm
thikapillary(see g. 2.27). The ellisthermostatedand the temperaturewithinthe ellis ontrolled
by a thermoouple with a preision of 0.1
o C
. The large surfae of ontat between theell and the apillary allows a fast quenhing of the sample and makes it ideal for the
diret observation of the rystallization proess. Images of the samples were taken in a
darkroomwithoutlter.
2.3.3 Eetive volume fration and rystallization
Theestablishedliquid-rystaloexistenedomainforhardspheres laysbetween the
freez-ing volume fration
φ F
atφ ef f
= 0.494 and the melting volume frationφ M
atφ ef f
=0.545asobtained fromomputer simulation[153℄. An experimentalphasediagramould
be ahieved by determining the rystal fration of the samples from the position of the
Capillaries Apertures
Temperature sensor
Cooling
3cm
Figure2.27: Thermostated ell.
oexistene liquid-rystal boundaries after sedimentation. This an be linearly
extrapo-latetodeterminethe beginningandthe endofthe oexistenedomain[117,154℄. Tothis
purpose solutions with weight onentrations rangingbetween 6 and 14
wt.%
have beenprepared. The samples KS1 and KS3 were shaken after preparation to destroy residual
rystallitesandstoredformorethanone monthatroomtemperature20.5
o C ±
0.5o C
. Inthe aseoftheKS2 thesuspensionshavebeenheatedto30
o
Cinordertodestroypossible
rystalsthatmayhaveformedatroomtemperature. Thesesuspensionsaresubsequently
ooled down quikly to 21
o C
and kept at this temperature for a time of typially two months.Aftersreeningofthe eletrostatiinterationsbyadding5.10
−2 mol.L −1
,allthe samplesrystallizeatdened onentrations. Thishintstothe lowpolydispersityofthe partiles.
Thehydrodynamiradiusofthemirogel
R H
anbeusedtoalulatetheeetivevolumefration
φ ef f
fortemperatures below25o C
byφ ef f = φ c
R H
R c
3
(2.33)
where
R c
is the ore radius alulated from the ryo-TEM andφ c
is the volume frationof the ores in the system. The latter quantity an be approximated from the weight
onentration ofthe partiles inthe system and the mass ratiobetween the ore and the
shellofthepartiles. Toavoidpossibleerrorsduetothesmallexperimentalunertaintyof
R H
, averagedvalues have been taken fromthis graphby approximatingR H
by astraightlinein this region of temperatures.
Thedensityofthepartilesandtheirsizeisnothighenoughtoensurerapidsedimentation
of the rystalline phase. Only after two months, rystals whih an be seen by eye by
means of the Bragg-reetions,have sedimented. Fig. 2.28 exhibits the liquid-rystalline
region of the dierent samples and the orresponding phase diagram. The experimental
phase diagram was taken from the hange in the position of the oexistene boundary
indiated by the dashed lines. As expeted the experimental points desribe a linear
dependeneinthebiphasiregion. The datahavebeenresaled to
φ F = 0.494
inordertoomparethe dierentexperimentalphasediagrams. FortheKS1,the oexistenedomain
Ks2
Ks3
0 0.2 0.4 0.6 0.8 1.0
0.40 0.45 0.50 0.55 0.60 0.65
0.556
0.494
f eff
C ry st a lF ra ct io n
Ks1
Ks3
1cm
Figure2.28: Dierent ore-shell suspensions in the biphasi region with 1.25
mol.%
(KS1), 2.5mol.%
(KS2) and 5mol.%
(KS3) rosslinking after two months and their orre-sponding experimental phase diagram. The rystal frations determined from theheightof theoexistene boundariesindiatedby thedashed lines onthe photographs
were tted by a linear regression (solid line) for the KS2 (hollow triangles) and
the KS3 (hollow irles). The results obtained for the KS1 are indiated by hollow
squares. Thedata have beenresaled to
φ F =
0.494. The oexistene domain forthe1mm
Figure2.29: Crystallization of a 9.48
wt.%
solution at dierent temperatures orresponding to dierent eetive volume frationsmanifested by the presene of distint rystals has been observed for eetive volume
between 0.494 and 0.535,whih issmalleraswhat isexpeted forhard spheres. This an
berelatedtothe softness ofthe systemasalready observed for PNIPAM mirogel[7℄. At
higher degrees of rosslinking (KS2 and KS3) the resaled oexistene domain has been
found between 0.494 and 0.556. This is inaord with the theoretial values
φ M = 0.545
[153℄.
Fig. 2.28 shows also that the rystallization study of the thermosensitive partiles is
partially hampered by the strong turbidity of the system. This motivates the use of thin
apillary for a diret observation of the rystallites. In this ase polarized mirosopy
an beusedtoinvestigatethe rystallizationkinetis[135℄. Theseexperimentshave been
performed on the KS2, for a onentration of 9.48
wt.%
(see g. 2.29) and 8.22wt.%
(see g. 2.30) at dierent temperatures. The same experiment wasrepeated for the KS3
at a onentration of 13.01
wt%
at 20o C
and will be disussed in the next setion. Thesamples were rst maintained at about 30
o
C in the 0.1
mm
thik apillary and thenquikly ooled-down to thetemperatureof investigationinthe thermostatedell(see g.
2.27).
Below the melting temperature the rystallization proess of olloidal partiles an be
interpreted in the framework of nuleation and growth. The rst rystals have been
observed around
Φ ef f = 0.50
. Considering the time of observation of one hour, this value is in good agreement with the phase diagram. The onset of the rystallization1mm
Figure2.30: Crystallization of a 8.22
wt.%
solution at dierent temperatures orresponding to dierent eetive volume frations0.1 1 10
0 1000 2000 3000 4000
Time [s]
G ', G '' [P a ]
Figure2.31: Visoelasti behavior of rystallizing suspensions vs. glassy systems: The storage
modulus
G ′
(lled symbols) and the loss modulusG ′′
(open symbols) are measuredas funtion of time in the linear visoelasti regime at 1
Hz
and 1%
after 5 min ofshearingat100
s −1
forarystallizingsystem(irles)andaglassysystem(squares).Thetriangles refertotheliquidstate(
φ ef f =
0.49),The irlestoa volumefrationof 0.52 (two-phase regime) whereas the squares give the results for the glassy state
(
φ ef f =
0.65).is manifested by the apparition of large rystals growing on the walls of the apillary.
At lower temperatures orresponding to higher
Φ ef f
most rystals are formed in bulk,the nuleation inreases, whereas the size dereases. Above
Φ ef f = 0.542
no rystalwas observed in the 9.48
wt.%
solution. This is in good agreement with observations performed on olloidal hard spheres assimilated suspension by light sattering [123℄. Itwas demonstrated that when the melting onentration is exeeded, nuleation events
beome orrelated and high nuleation rate densities suppress rystal growth. At higher
eetive volume frations the rystals are indeed strongly ompressed impeding their
growth. Ontheontrarysomerystals anstillbeobserved until
Φ ef f = 0.57
forthe8.22wt.%
solution,whihanbeattributedtoaslightvariationofthesoftnessofthe partilesfor the lower temperatures.