between Composite Mirogels and Gold
Nanopartiles
4.2.1 Introdution
Biologial olloidal systems like virus and proteins often present a omplex and dened
struture with dierent funtionalities whih is the prerequisite of their spei and
tar-geted appliations. They an be onsidered in many ases as omplex polyeletrolytes
with a pathy struture, amosai of negative,positiveor neutralareas.
Obtaining suh strutures synthetially is muh more hallenging. Complex geometries
and dened lusters ould be obtained via pikering emulsion and evaporation of the
solventasinitiatedbyPine etal. onmirometripartiles[257℄and onbidisperseolloids
[258, 259℄. This approah was reently extended by Wagner et al. to olloidal system
[260℄. Another strategy onsistsonthe use of a templateas shown reently by Xiaet al.
[261℄. Nevertheless most ofthe time these lusters are not speiallyfuntionalized and
lak of long ranged interations.
At the same time many eorts have been put into the synthesis and haraterization of
Janus systems [262268℄, whih present a versatile behavior and interesting properties
as stabilizersfor instane [265, 267℄. Most of these new systems relyon their ambivalent
nature,neverthelessitremainsdiulttosynthesizedstableolloidaleletrostatidipoles.
Indeed mixing oppositely harged olloids most of the time leads to the
destabiliza-tion of the system [269℄. In this setion we onsidered the assoiation of ationi
gold nanopartileswiththermosensitive anioniore-shell mirogelspolystyrene/poly(
N
-isopropylarylamide)in the dilute regime.
For this purpose ationi gold nanopartiles were prepared as desribed formerlyby
Ni-idome,where theauthors illustratedtheir appliationinthe funtionalgene deliveryand
anerous ells detetion [270℄. The use of inorganinanopartiles for gene therapy suh
as amino-modied silia nanopartiles and ationi gold nanopartiles was reently
re-ported[270272℄. Thespeial haraterofthese nanopartilesdiers fromthat oforgani
gene arrier moleules, and they are expeted to be a novel base material for the next
generation of funtional gene arriers. Gold nanopartiles have the advantages of easy
preparation and the possibility of hemial modiation of the surfae. Moreover they
havebeenfoundsuitableinmanyappliationsspanningfromatalysisandnano-eletroni
totreatment and detetion of anerogeni ells [273275℄.
On the other hand omposite mirogels have been intensively studied and used inmany
appliationsfrom theprotein adsorption[19,20℄tothe use as templateforthe redution
of metal nano-partiles [2123℄. Inreasing the temperature these mirogels swell and
deswell in a ontinuous or disontinuous fashion following their degree of ross-linking
as shown in the setion 2.1.4. Not only the size of the partiles an be ontrolled with
the temperature, but also their potential. As presented in the former setion, systems
synthesized by seeded emulsion polymerization with an anioni initiator were found to
present someeletrostati originatedfromremaininginitiatorfragmentsmostly based on
the surfae of the ore partiles. This ensures the stability of the system even after the
volume phase transition. Thus the interation potential an be adjusted ontrolling the
size andsaltonentrationof thedispersion. In anexess of salt,the system isnotstable
anymore above the volume phase transition and was found to reversibly oagulate (see
hapter 4.1).
The eletrostati stabilization relatedto the presene of negativeharges ombined with
the steristabilizationensured bythe shell has alreadybeenused toabsorbationigold
nanorods [31, 32℄ and silver nanopartiles [276℄. Whereas most of the studies foussed
on the adsorption of small partiles, we investigate the adsorption of larger partiles
(size ratio in the range 1 to 4). We present a simple way to suessfully adsorbed the
gold on our omposite mirogels to obtain anorgani/organi doublet or triplet keeping
the dipolar harater of the assoiation. The inuene on the size ratio and on the
preparation is rst disussed. Afterwards we investigate the orrelation between both
partiles by turbidimetri titration, UV light spetrosopy, zeta potential measurement,
dynami light sattering and mirosopi methods. The struture and stability of the
assoiationis onsidered atthe end of this setion.
4.2.2 Experimental
Materials
The system used in this study onsists on a polystyrene ore with a rosslinked
PNI-PAM shell ontaining 2.5
mol%
BIS (KS2) (see setion 2.1.2 for further details). Thegold nanopartiles were synthesized following the reipe desribed by Niidome et al.
[270℄. The gold nanopartileswere prepared by NaBH
4
redution of HAuCl4
inthepres-ene of 2-aminoethanethiol at a Au/NaBH
4
/ 2-aminoethanethiol ratio of 56 : 0.1 : 85 (mol/mol/mol). Immediately after the NaBH4
redution, the solution was opaque andbeomearedwine olor. The partileswere thenusedforthe assoiationwithoutfurther
puriation.
Thesamplewerepreparedbyslowadditionofthe0.28
gL −1
goldsolutionondilutemiro-gelsolutions. Themirogelonentrationwaseitherset at0.2
gL −1
forthe zetapotentialand for the UV light spetrometry measurement, and between 4.10
−2
and 2.6.10
−2 gL −1
for the sample prepared by titration. The gold nanopartiles were stable for
approxi-mately two weeks, afterwards they started to get adsorbed at the surfae of the glass
ontainer. Thusthe haraterization ofthe gold nanopartilesandthe preparationof the
dierentsolutions were done with a freshgold solution withinone week.
Methods
Sanningforemirosopy(SFM)experimentswerearriedonaommerialSFM(Model
Dimension3100,fromVeeoInstrumentsIn.) (seesetion4.1.3forfurtherdetails).
Field-emission sanning eletron mirosopy (FESEM) was performed using a LEO Gemini
mirosope equipped with a eld emission athode. The samples have been prepared
by spin oating at 2000
rpm
on siliium waver for the FESEM experiments, whereas the SFM samples have been prepared by dipoating on mia. Transmission eletronmirosopy(TEM) andryogenized TransmissionEletronMirosopy(Cryo-TEM) have
0
Figure4.10: Transmission eletron mirograph of a 0.2 wt.aqueous suspension of ationi
gold partiles obtained by NaBH
4
redution of HAuCL4
in the presene ofaminoethanethiol. The average radius of the gold partiles was determined at 22.7
±
3.7nm(see histogram). Thestabilizationof thepartiles by theaminoethanethiol has been evidene by the presene of a 5-8nm
thin shell around the partiles.been performed on a Zeiss EM922 EFTEM (Zeiss NTS GmbH, Oberkohen, Germany).
A few miroliters of the solution were plaed on a bare opper TEM grid (Plano, 600
mesh) and the exess of liquid wasremoved with lter paper.
The turbidity measurement of the adsorption of the gold nanopartilesonthe omposite
mirogel was performed on a titrator (Titrando 809, Metrohm, Herisau, Switzerland)
equipped with aturbidity sensor (
λ 0 = 523 nm
, Spetrosense, Metrohm). The 0.28gL −1
goldsolutionwasaddedatarateof0.25
mL/min
ona16mL
0.04gL −1
mirogelsolutionat25
o C
.Dynami lightsattering was done using a Peters ALV 5000 lightsattering goniometer.
The samples were highly diluted(
c = 2.5.10 −3 wt%
)to prevent multiple sattering. Thedeay rate
Γ
wasobtainedfromtheseondumulantanalysisforsatteringanglefrom30o
to 150
o
with aninrement of 10
o
. The zeta potential was determined with the Zetasizer
withastrit ontrolof thetemperatureset by Peltierelementsat
±
0.1o
C. Forthediretobservationof thestabilityof thedierentsolutions,thesamplewere equilibrated30
min
in athermostated hamber.
4.2.3 Results and Disussion
Cationi gold nanopartiles
Transmissioneletronmirosopyhas beenperformedonthe ationigoldpartiles
stabi-lized by the aminoethanethiol (see g. 4.10). The radius of the partiles alulated from
the TEM has been determined at 22.7
±
3.7nm
. The stabilization atthe surfae of the partileanbediretlyimage,byathinlayerofbetween5and 8nm
aroundthepartiles.The thiolgroups have astronganity withthe gold,under neutralonditions theamine
groups are hydrolyzed into NH
⊕
3
providingthe ationiharater of the partiles.More-over the size of the orona respet to the size of the moleulesupposes the organization
ina multilayer. In ordernot to damagethis layer the piturehave been taken underlow
dose onditions. Nonetheless the surfatant is still very sensitive to the radiation and
get easily damageafter too long exposure. Thusit is not easy todetermine its thikness
preisely. The average hydrodynami radius has been also measured by dynami light
1 2 3 4 5
Figure4.11: Adsorption of the ationi gold nanopartiles on the omposite mirogels. The 0.28
gL −1
goldsolutionwasadded to16mL
4.10−2 gL −1
latexsolution. Theturbidityofthe solution wasmeasured during the addition proess. The turbidity wasexpressed
intermoftheabsorbane
A = − ln(I/I 0 )
whereI 0
orrespondstotheinitialintensityof the latexsolution(see text forfurther details). The dierent number orresponds
to the dierent solutions prepared following the same proedure for dierent gold
onentrations.
sattering between 25 and 45
o
C in bak sattering at 173
o
with the Zetasizer. No
signif-iant dependene on the temperature was observed. The average hydrodynami radius
obtainedfromtheseondumulantanalysiswasfoundequalto27
nm
witharelativehighpolydispersity index (0.28). On the ontrary of the evaluation of the size by TEM, the
DLSismoresensitivetothe preseneofaggregatesinthesolution. Nonetheless thisvalue
presents anieagreementwiththe TEMif weonsider the ontributionofthe surfatant
layer. The eletrophoreti mobility has been expressed by the way of a Zeta potential
alulated from the Smoluhowski approximation between 25 and 45
o
C. This value was
also not sensitive to the temperature and the average zeta potential was found equal to
27
mV
. This onrms the ationi harater of the partiles, nevertheless this value is lowerthan the onereportedin theliterature forthe same samplepreparation(36.2mV
).Turbidimetri titration
Whereas a slow addition of latex on an exess of gold partiles diretly leaded to the
destabilizationof the solution,itwas possible toadd goldon anexess of latexpartiles.
Forthisreasonthe solutionshavebeenpreparedbyslowadditionofthe goldonthe
ore-shell partiles. Weperformed turbidimetrititrationin order tofollowthe assoiationof
theationigoldnanopartilesontotheompositemigrogels. Tothispurposeweprepared
a16
mL
latexsolutionwithaninitialonentrationof4.10−2 gL −1
andaddedatarateof0.25
mL/min
a0.28gL −1
goldsolution. Theturbidityofthesolutionwasmeasuredduringthe wholeadditionproess. Thegold solutionasastrongabsorption at525
nm
losed tothe wavelength ofthe probe(523
nm)
. Forthis reasonwe used the transmittedintensityI 0
ofthelatexsolutionasreferenetonormalizethetransmittedintensityI
ofthesolutionand we followed the evolution of the absorbane dened by
A = − ln(I/I 0 )
as funtionof the gold onentration (see g. 4.11). As expeted, the absorbane inreases rst
-50
Figure4.12: Zeta potential of the gold nanopartiles (hollow squares), of ore (full irles) and
Core-Shell (full squares) partiles alulated with the Smoluhowski approximation
as funtion of the temperature with and without addition of gold for a 0.2
gL −1
latex solution. Remaining eletrostati interations explain the stability of the latex
at high temperature. Addition of the gold solution is sreening the Zeta potential
(see the dierent onentrations in the legend). An aggregation an be observed for
a onentration of 0.160
gL −1
at 35o C
.linearly untila gold onentration of 0.03
gL −1
. After this onentration the absorbane still inreased linearly but with a lower slope. This deviation was attributed to a fastaggregation proess followed by the destabilization of the solution, to the shift of the
plasmon maximum towards higher wavelengthes onrmed by a hange of the solution
from red to blue and to the presene of free gold partiles. Four dierent samples were
preparedfollowingthesameproedure(seeg. 4.11). Foraonentrationhigherthan0.03
gL −1
,the solutionwere not stableand sedimented withinanhour. Weapproximatedthe orresponding number ratio between the gold and latexpartilesonsidering the densityof gold (19.3
gcm −3
), the size of the gold partiles determined from the transmission eletron mirosopy and the number of omposite mirogel partile present in solution(asdesribed in setion 2.1.2). We found out a ratio of 1.05, whihbasially means that
in average no more than one gold partiles ould be adsorbed onto the mirogelsunder
this onditions. This nding was onrmed afterkeeping the solutionsfor more than six
months. Below this ritial onentration the solution ould be very easily redispersed
andnogoldwasabsorbedonthesurfaeoftheglassontainerwhihisastrongindiation
oftheorrelationofthegoldwiththemirogel. Ontheontraryforhigheronentrations
the solutionsould not be redispersed anymore after two months and a part of the gold
was absorbed on the surfae of the glass ontainer, whih onrmed the assumption of
free gold partilesin the solution.
Zeta potential measurement
Wemeasured theZetapotentialofsolutiononsistingon0.2
gL −1
latexsolutionwithdif-ferentonentrationsofgoldvaryingbetween5.4.10
−3
and1.6.10
−1 gL −1
. Thepolystyrene orepartilesusedforthe synthesisof theompositemirogelshas beenmeasured aswellasthepure goldnanopartilessolutionand thepure ore-shellpartilesasfuntionofthe
temperature(see g. 4.12). The orepartilespresented arelativelyonstant zeta
poten-0 0.2 0.4 0.6 0.8
300 400 500 600 700 800
l [nm]
300 400 500 600 700 800
45°C
Figure4.13: A) UV visible extintion spetrumof a pure 0.2
gL −1
ore-shell dispersion (dashedline) and of 0.2
gL −1
ore-shell dispersion with 0.0280, 0.120 and 0.160gL −1
ationi gold measured at 20
o C
. The inset display the adsorption of the pure gold(dashed line) and of the three solutions after substration of the latex ontribution.
The position of the maximum adsorption
λ max
of the gold shifts from 525 to 535nm
after adsorption on the omposite mirogels as shown by the thin dotted lines).B) Adsorption of a 0.02
gL −1
ore-shell dispersion with 0.120gL −1
ationi goldsolution measured with inreasing temperature. C) Maximum of the orresponding
plasmon band of the gold nanopatiles as a funtion of temperature (full irles)
ompared with the evolution of polymer volume fration of the PNIPAMshell
(hol-low squares). The full line presents the t following the Flory-Rehner theory (see
setion 2.1.4).
tialat -43
mV
over allthe temperaturerange. Conerningthe pure ore-shell dispersion below32o
C, theeletrophoretimobilitywasfoundtoaround-15
mV
,reetingboththelow surfae harge density and the high frition oeient of the swollen partiles.
How-ever, a dramatihange inthe eletrophoreti mobilityversus temperaturewas observed
above the volume transition temperature. As expeted, the absolute values of the zeta
potentialinreasedwithinreasingtemperatureduetothe thermalsensitivityand
shrink-ing of the PNIPAM shell to reah approximately the value of the ore zeta potential at
high temperatures asdesribed by López-León et al. [72℄. The temperature dependene
ofthezetapotentialouldbeinterpretedbytheinrease inthesurfaehargedensitydue
to the redution in the partile size, the enhanement of the loal harge onentration
onthe partile's surfae and the fritional oeientredution of the ollapsed partiles.
This eet ould bediretly visualizedonsidering the strong bukling up of the shell at
low temperature and the ollapsed form of the shell at higher temperature as shown in
gure 2.2.
Addinggold induedthe derease of the absolutevalueof the zetapotentialtobealmost
equal to zero for a onentration for a onentration of 1.6.10
−1 gL −1
below the volumephase transition. At higher temperatures the thermosensitivity was maintained with a
transition around 32
o C
. All the measurements presented a mononodale distribution of the zeta values onrming the assoiation of the gold with the mirogel. For the higheronentrationorrespondingtoanumberratioofone goldforonemirogelaoaggulation
proesshas beenobservedfortemperatureshigherthan40
o C
. Thediminutionofthezetapotentialouldbeinterpretedasanindiationoftheassoiationofthe oppositelyharged
partiles leading rst to the distortion of the harge distribution, to the inrease in the
size of the partilesand tothe inrease of the surfatant onentration in the solution.
Thermoresponsive optial properties
The assoiation of the gold nanopartiles with the mirogelswas investigated by UV-vis
spetrosopy. The dependene ofthe plasmonadsorption onthe size and temperatureof
olloidalgoldpartilesinaqueoussolutionhasbeenalreadydisussedbyLinketal. [277℄.
They foundout thatmonodisperse 21.7
nm
goldpartileshaveamaximum adsorptionat521
nm
. The maximuminthe adsorption wasobserved at525nm
forour solutionwhihisingoodresultswith the literatureif weonsiderthe inuene ofthe polydispersityand
of the surfatant adsorbed atthe surfae of the gold partiles.
In order toonrm the assoiationbetween the gold and the mirogelwe investigate the
dierent solutions by UV-vis spetrosopy. The pure gold solution was rst measured.
A maximum in the adsorption was obtained at 525
nm
whih is ommon for partilesin this size range [277℄. On the ontrary the latex solution did not present any spei
adsorption. Afteradsorption of the gold (g. 4.13 A)), the plasmonmaximum measured
aftersubstrationofthelatexontributionshifted from525to535
nm
(see insetg. 4.13A)).
Theadsorptionwasmeasuredfora0.2
gL − 1
ore-shelldispersionwith0.120gL −1
ationigoldvaryingthetemperaturebetween10and45
o C
(seeg. 4.13B))asdesribedreently[31,32℄. Theadsorptionwasfoundtoinreasewiththetemperaturefollowedbyaredshift
from535to543
nm
between10and45o C
whihhavebeenattributedtotheinreaseoftheloal refrative index upon mirogel ollapse for low surfae overage [31℄. The position
ofthe maximum
λ max
has been monitoredfor thedierenttemperatureandompared to the variation of polymer volume frationφ
of the shell of unoated omposite mirogeltaken from the setion 2.1.4 (see g. 4.13 C)). The linedisplays the theoretial tusing
the Flory-Rehner theory. The transition observed in the variation of
λ max
follows thesame feature asthe transition inthe PNIPAM network exept that the transitionours
about2
o C
beforethe LCSTofthepure mirogel. Thisagainorroboratesthe assoiation of the gold with the mirogels.Colloidal stability and oagulation
The stabilityof the partileshas been heked by adding dierent onentrationsof gold
on a 2
gL −1
mirogels solution (see g. 4.14). At 32o C
the system was perfetly stable.Inreasingthe temperatureat35
o C
leads tothe oagulation ofthe system foronentra-tionshigherthan0.107
gL −1
. Darkred aggregateswereobserved, whihquiklysedimentlettingalear solutioninomparison tothe puregold and ore-shellsolutions. This
on-rmstheassoiationofthe goldwiththe mirogel. Thesystem reovereditsoriginalform
25
o C
after a small redispersion. The understanding of the aggregation proess is quite hallenging. First we hek if it was orrelated to the ratio between gold and mirogels.For the same partileratio but aftera dilutionby ten the system did not aggregate. We
onsidered the inueneof the gold onentrationfor dierent latexonentration. Fora
onentration of 0.160
gL −1
the system was found to aggregate for 2 and 0.2gL −1
latexonentrations. Thus, the onentration ofgold andnot the numberratiogold/mirogels
seems to be the determinant fator for the reversible oagulation at high temperatures.
The omposite mirogelsare sterially stabilized by the PNIPAM shell and
eletrostati-allydue tothe use of ananioni initiatorand to the remainingSDSproviding fromthe
32°C
35°C
25°C
Figure4.14: Stabilityandoagulationofthesolutionsasfuntionofthegoldonentrationandof
thetemperature. The dierentsolutions arefromtherighttotheleft: apure 2
gL −1
ore-shell solution, a2
gL −1
ore-shell solution+ 0.054, 0.107, 0.160,0.220, 0.280gL
−1
goldrespetivelyandapure 0.2
gL −1
goldsolution. Allsolutionswereobservedrstat32
o
C,thenat35
o
Cand25
o
C.Thesamplesat25
o C
wereredispersed topoint out the reversibility of the oagulation proess.synthesis. As disussed in the setion 4.1, addition of salt redues the Debye length and
thestabilityofthesystemathightemperature. Fig. 4.12learlyshows thiseet interm
of a diminution of the zeta potential with inreasing gold onentration. It is still not
fully understood how the gold nanopartiles and the remaining surfatant ontribute to
the sreening of the eletrostatis.
Eletri dipoles formation and dynami lusters
Theformerexperimentslearlyshowtheassoiationofthegoldwiththemirogels.
Never-thelessthestrutureofthe omplexhasnotbeen investigateduntilnow. Forthispurpose
Never-thelessthestrutureofthe omplexhasnotbeen investigateduntilnow. Forthispurpose