Effect of Precursor-Layer Surface Charge on the Layer-by-Layer

Nov 16, 2011 - more uniformly charged surface.18А23 It is believed that this pre- ... However, the surface charge of the precursor layer and its effe...
0 downloads 0 Views 3MB Size
Download PDF
ARTICLE pubs.acs.org/Langmuir

Effect of Precursor-Layer Surface Charge on the Layer-by-Layer Assembly of Polyelectrolyte/Nanoparticle Multilayers Chunqing Peng, Yonathan S. Thio, and Rosario A. Gerhardt* School of Materials Science & Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332-0245, United States

bS Supporting Information ABSTRACT: In this Article, we investigate the effect of a precursor layer, which is composed of four bilayers of polyethyleneimine (PEI) and poly(sodium styrene sulfonate) (PSS), on the subsequent LBL assembly of hybrid films composed of indium tin oxide (ITO) nanoparticles and PSS. A precursor polyelectrolyte layer is usually deposited to minimize interference by the substrate. It is shown here that the “effective” surface charge of the precursor layer can significantly affect the subsequent assembly behavior of [ITO/PSS]9.5 hybrid thin films. Depending on the surface charge of the precursor layer, the subsequent LbL assembly of [ITO/PSS]9.5 hybrid films can exhibit either one or two regimes. When two growth regimes are present, the first one consists of a “recovery regime”, and the second is the expected “linear growth regime.” The length of the “recovery regime” is dependent on how much positive charge the precursor layer possesses and how fast this surface charge can be compensated. This work reveals for the first time that changes in the surface charge of the precursor layer can have a significant effect on the subsequent LBL assembly process. The surface charge of the precursor layer was investigated using ζ-potential measurements on model silica microspheres. These experiments showed that the surface charge of the precursor layer, [PEI/PSS]4, is dependent on the pH of the solution in which it is immersed, and that it can reverse from a negatively charged surface to a positively charged one, at sufficiently low pH due to the protonation of PEI, despite having the negatively charged PSS layer as the outermost layer.

’ INTRODUCTION The preparation of multilayer thin films using layer-by-layer (LbL) assembly has garnered extensive attention in the past decade. In the early 1990s, this method was first introduced to prepare polyelectrolyte multilayers (PEMs).1,2 Since then, it has become one of the most popular and established methods for preparing multifunctional thin films because LbL assembly is a versatile method that is relatively simple to implement. Over the years, multifunctional films have been made from a variety of materials, such as polyelectrolytes,2 proteins,3 DNA,4,5 inorganic nanoparticles,6 9 carbon nanotubes,10 12 nanoplates,13,14 and quantum dots.15,16 LbL-assembled hybrid thin films that consist of both organic and inorganic materials are of special interest due to their special optical and electronic properties.8,9,17 During LbL assembly of hybrid thin films that are composed of polyelectrolytes and inorganic nanoparticles, one often assembles a precursor layer that is composed of one or more bilayers of polyelectrolytes prior to the deposition of the inorganic nanoparticles.12,18 28 The purpose of assembling this precursor layer is to reduce possible interference from the substrate and to render a more uniformly charged surface.18 23 It is believed that this precursor layer can facilitate the subsequent nanopartice deposition.18,19 To date, knowledge about the effect of the precursor layer on the LbL assembly of inorganic nanoparticles is very limited. The precursor layer is most often assembled using weak r 2011 American Chemical Society

polyelectrolytes, such as polyethyleneimine (PEI) and poly(allylamine hydrochloride) (PAH), the charge density of which is dependent on the pH of the polyelectrolyte solution.12,25 28 However, the surface charge of the precursor layer and its effect on the following LbL assembly process has not been studied previously. In this Article, we report on the preparation of hybrid thin films composed of indium tin oxide (ITO) nanoparticles and poly(sodium 4-styrenesulfonate) (PSS), with a precursor layer consisting of four bilayers of PEI and PSS. Hybrid thin films containing ITO nanoparticles are of particular interest to researchers due to their potential for use in optoelectronic devices.29 More specifically, this Article describes the effect of pH on the surface charge of the precursor layer, as well as the surface charge effect on the subsequent LbL assembly of ITO and PSS. To the best of our knowledge, this is the first time that the effect of the solution pH on the surface charge of a precursor layer, as well as its effect on the LbL assembly, has been systematically investigated.

Received: January 25, 2011 Revised: November 14, 2011 Published: November 16, 2011 84

dx.doi.org/10.1021/la203626x | Langmuir 2012, 28, 84–91

Langmuir

ARTICLE

’ EXPERIMENTAL SECTION Materials. Branched PEI (Mw ≈ 750 000 g/mol, pKa ≈ 8 (0.1 M KCl)30), PSS (Mw ≈ 70 000 g/mol, pKa31 ≈ 6.5), 11-mercaptoundecanoic acid (MUA), ITO nanoparticles (particle size = 20 150 nm, surface area ∼27 m2/g), potassium ferricyanide (K3Fe[CN]6), and potassium ferrocyanide (K4Fe[CN]6) were obtained from Sigma Aldrich. Monodispersed silica microspheres in aqueous suspension (∼4 μm, 10 wt %) were obtained from Polyscience Inc. All of the materials were used as received without further purification. Preparation of the Polyelectrolyte Solutions for the Precursor Layer Assembly. Two types of PEI and PSS polyelectrolyte solutions were prepared. The first type was prepared at a concentration of 0.2 wt % and the second at a concentration of 0.02 wt %. The more concentrated solutions were used to assemble the precursor layers on the monodispersed silica microsphere surfaces, because of their high solids content. The lower concentration solutions were used to assemble the precursor layer and the hybrid [ITO/PSS]n thin films onto the QCM crystals. Unless specifically mentioned, both PEI and PSS solutions were prepared at a NaCl concentration of 0.1 M.

Figure 1. ζ-Potential of ITO suspension at pH values ranging from 2 to 12. ITO suspensions at pH of 2.9 and 4 were used for the LbL assembly in this work.

Surface Charge of Precursor Layer on Silica Microspheres.

suspensions at the pH of 2.9 and 4, so that the ζ-potentials of these two solutions were positive and had values of 50.6 and 40.3 mV, respectively. LBL Assembly Procedure for the Hybrid Thin Films. All of the hybrid thin films were prepared in a similar manner: LBL assembly of the precursor [PEI/PSS]4, followed by the assembly of ITO and PSS for 9.5 bilayers. For easier description, the assembled films will be denoted as [PEI/PSS]4 [ITO/PSS]9.5. In this thin film, the [PEI/PSS]4 is the precursor layer for the [ITO/PSS]9.5 hybrid thin film. The hybrid thin films were always finished with an ITO layer, and thus a film with 9.5 bilayers of ITO and PSS is actually composed of 10 ITO layers and 9 PSS layers. During the precursor layer assembly, the substrate was first immersed in the PEI solution (0.02 wt %) for 10 min, washed with pure water, blown dry with pure nitrogen gas, and then immersed in the PSS solution (0.02 wt %) for 10 min, washed with pure water, and then blown dry with pure nitrogen gas. This process represents one assembly cycle of the precursor layer [PEI/PSS]4. After the fourth PSS layer, the [ITO/ PSS]9.5 film was then assembled in the same manner as that of the precursor layers. The substrate was immersed in the PSS solution and ITO solutions, respectively, for 10 and 20 min, and the substrate was also washed with pure water and blown dry with pure nitrogen after each layer. To clearly identify the trends observed as a function of pH for each sample, the pH of the polyelectrolyte solutions and the ITO suspensions used will be indicated in the hybrid specimen names. For example, the thin film prepared from PEI at the pH of 8.5, with PSS at the pH of 6.5, and ITO at the pH of 2.9 will be denoted as [PEI(8.5)/PSS(6.5)]4 [ITO(2.9)/PSS(6.5)]9.5.

To evaluate the surface charge of the precursor layer, we assembled the precursor layer onto monodispersed silica microspheres and compared the ζ-potential of the silica microspheres prior to and after the precursor layer coating. The difference of the ζ-potential between the bare and precursor-layer-coated silica microspheres is thus considered to be contributed by the presence of the precursor layer. The silica microspheres were chosen to be 4 μm in diameter, to facilitate the LbL assembly process. The precursor layer was coated onto the silica microspheres according to the following procedures. The first layer of PEI was deposited by adding 4 mL of PEI solution (0.2 wt %) to 0.2 mL of silica microsphere suspension, occasionally shaking the suspension, allowing 20 min for adsorption, and removing the excess polyelectrolytes by four repeated centrifugation steps (3500 rpm for 5 min)/decantation/redispersion cycles. After the first PEI layer, one layer of PSS was assembled onto the silica microspheres from PSS solution (0.2 wt %) using the same procedure. This assembly cycle was repeated four times to assemble [PEI/ PSS]4 onto the silica microspheres, and the final suspension was denoted as silica [PEI/PSS]4 suspension. In this Article, PEI and PSS solutions at different pH conditions were used to prepare different precursor layers, and the pH values used for each material will be indicated in the final specimen denotation. For example, the precursor layer prepared from PEI at pH of 8.5 and PSS at the pH of 6.5 on silica microspheres will be denoted as [PEI(8.5)/ PSS(6.5)]4. The LbL assembly process on the silica microspheres was followed by measuring the ζ-potential of the silica suspensions after each deposition layer. Both the [PEI/PSS]4-coated silica and the bare silica microspheres were redispersed into pure water to prepare suspensions at a concentration of 0.02 wt %. These suspensions were used for the ζ-potential measurements over the pH range between 2 and 9 to evaluate the surface charge of the precursor layer.

Investigation of the LBL Assembly of [ITO/PSS]n Films by QCM and AFM. The LBL assembly process under various conditions was investigated using a quartz crystal microbalance (QCM, Stanford Research System, Inc., QCM 200). The quartz crystals used were AT cut crystals, coated with chrome/gold electrodes and resonating at 5 MHz at room temperature. When a rigid thin film is deposited onto the electrodes, the resulting frequency change (i.e., Δf) is linearly proportional to the thin film mass according to the Sauerbrey equation.32 Therefore, we can evaluate the thin film assembly rate by measuring the resonant frequency after each assembly step. The quartz crystals were treated prior to the LBL assembly using a procedure similar to what we have described before.9,33 In brief, the crystals were first cleaned with piranha solution (2:1 H2SO4:H2O2) for 1 h. (Caution! Piranha solution is extremely reactive and exothermic, and it should be handled with extreme care.) The cleaned crystals were then immersed into 1 mM of 11-mercaptoundecanoic acid (MUA) ethanol solution for over 24 h to construct a self-assembled monolayer

Preparation of the ITO Colloidal Suspensions for Hybrid Film Assembly. ITO suspensions were prepared at a concentration of 0.1 wt %, with 0.1 mM NaCl. The surface charge of the ITO nanoparticles was evaluated by measuring the ζ-potential of the ITO suspensions for the pH range from 2 to 12 using zetasizer (Malvern Instruments, Nano-ZS). As shown in Figure 1, the ITO nanoparticles exhibited a positively charged surface at a pH below 6, and a negatively charged surface at a pH above 9.5. The suspensions between pH of 6 and pH of 9.5 were quite unstable, and the surface charge was hard to be determined. Therefore, the isoelectric point was considered to be between the pH of 6 and 9.5. To carry out the assembly, we prepared ITO 85

dx.doi.org/10.1021/la203626x |Langmuir 2012, 28, 84–91

Langmuir

ARTICLE

(SAM) of MUA on the gold electrode of the crystals. The SAM-modified crystals were washed with pure ethanol, blown dry with pure nitrogen gas, and immersed in 1 mM NaOH solution for 3 min to ionize the crystal surface. The NaOH-treated crystal surface is negatively charged at this point, and it is considered to be ready for the precursor layer assembly. The LBL assembly was then performed by using the QCM crystals as substrates, and the resonant frequency of crystals was recorded after every assembled layer, making sure that it was in a dry condition. The thin film buildup process on the quartz crystals was also followed by taking images of the samples at different stages of the assembly using atomic force microscopy (AFM, Park Systems, XE-100E). All of the AFM images were obtained under noncontact mode in air. The AFM tips used (Nanosensors, NCHR) exhibit a nominal tip radius smaller than 10 nm, resonate at approximately 300 kHz, and have a spring constant around 40 N/m.

’ RESULTS AND DISCUSSION Surface Charge of the Precursor Layer. The surface charge of the precursor layer [PEI/PSS]4 was characterized by comparing the ζ-potential of precursor-layer-coated silica microspheres, silica [PEI/PSS]4, with that of the bare silica microspheres. The coating of the each of the layers of the 4-bilayer precursor film on the silica microsphere surface was confirmed by the reversal of the ζ-potential of the silica microsphere suspensions after every assembly step. As shown in Figure 2a, the bare silica microspheres exhibited a negatively charged surface at neutral pH conditions. This surface charge reversed to a positive surface charge after the first PEI assembly layer, and it changed between negative and positive after every LbL assembly step. This phenomenon suggests that each of the polyelectrolyte layers was successfully coated onto the microspheres. The ζ-potential of silica [PEI/PSS]4 and bare silica microspheres is compared in Figure 2b. As is shown in the figure, the bare silica microspheres exhibited a negatively charged surface at a pH above 4.4, and this surface charge increased to a value close to zero at a pH below 3.5. When the silica microspheres were coated with precursor layer [PEI(4.5)/PSS(4.5)]4, almost no difference could be observed between the ζ-potential of bare silica and silica [PEI(4.5)/PSS(4.5)]4 over the whole pH range investigated. However, a significant difference in the surface charge between the bare silica and [PEI(8.5)/PSS(6.5)]4 coated silica microspheres was observed. At a pH above 5.9, the [PEI(8.5)/ PSS(6.5)]4-coated silica microspheres exhibited a slightly more negatively charged surface. At a pH below 5.9, the [PEI(8.5)/ PSS(6.5)]4-coated silica started to gain positive surface charge and eventually reached a positive value as large as 69.5 mV at a pH of 2.4. Obviously, this positive surface charge of [PEI(8.5)/ PSS(6.5)]4 coated silica obtained at low pH was completely contributed by the positive surface charge gain of the precursor layer [PEI(8.5)/PSS(6.5)]4 at the low pH conditions used. The positive surface charge gain of the precursor layer [PEI(8.5)/ PSS(6.5)]4 is due to the protonation of PEI at low pH.30,34,35 The protonation degree of PEI chains will increase as the pH of the solution decreases.30 During the assembly of the precursor layer, the degree of the protonation of the PEI chains depends on the last pH condition that the PEI chains were exposed to. Therefore, the protonation degree of PEI chains in the [PEI(8.5)/PSS(6.5)]4 precursor layer is equal to the protonation degree of PEI at a pH of 6.5, which is around 60%.30 This degree of protonation of PEI increases to almost 100% at a pH below 2.5.30 It is believed that this drastic increase in the degree of protonation of PEI

Figure 2. (a) ζ-Potential of 4 μm silica microspheres after each LbL assembly step; and (b) ζ-potential for uncoated silica and silica coated with [PEI(8.5)/PSS(6.5)]4 and [PEI(4.5)/PSS(4.5)]4.

resulted in the significant positive surface charge increase of the precursor layer [PEI(8.5)/PSS(6.5)]4 observed in Figure 2b. On the other hand, the protonation degree of the PEI chains in the [PEI(4.5)/PSS(4.5)]4 precursor layer is already around 80% at a pH of 4.5.30 Therefore, the gain of positive surface charge of this precursor layer is not as much as that for the [PEI(8.5)/PSS(6.5)]4 precursor layer, thus explaining the different response to both precursor layers consisting of PEI and PSS as a function of the pH. It is worth noting that the ζ-potential of the [PEI/PSS]4-coated silica suspensions exhibited a smaller standard deviation than that of the bare silica microsphere suspension. This indicates an increase in the stability of the microsphere suspensions due to the polyelectrolyte precursor modification.9 Besides the ζ-potential measurement described above, the surface charge of the [PEI(8.5)/PSS(6.5)]4 precursor layer was also investigated by electrochemical impedance spectroscopy (EIS) measurements on the surface of an ITO glass film using ferricyanide ([Fe(CN)6] 3) and ferrocyanide([Fe(CN)6] 4) negative ions. Using this method, it was shown that this precursor layer gained positive charge when the pH of the solution that the precursor layer was immersed into was changed from pH 7 to pH 5 (see Figure S1, Figure S2, and Table S1, Supporting Information). Effect of Surface Charge of the [PEI/PSS]4 Precursor Layer on the Assembly of [ITO/PSS]9.5. The effect of the precursor layer surface charge on the assembly of hybrid thin film [ITO/ PSS]9.5 was first investigated by using three different precursor layers as follows: [PEI(8.5)/PSS(6.5)]4, [PEI(4.5)/PSS(4.5)]4, 86

dx.doi.org/10.1021/la203626x |Langmuir 2012, 28, 84–91

Langmuir

ARTICLE

Figure 4. The effect of changing the pH of the ITO suspensions on the Δ( Δf) of the ITO assembly steps using the same precursor layer [PEI(8.5)/PSS(6.5)]4 [ITO/PSS(6.5)]n, where the pH of the ITO solutions used was 2.9 and 4.0. (The data for the ITO solution at 2.9 are repeated from Figure 3c here for easier comparison.) The transition points between the “recovery regime” and the “linear growth regime” are indicated by the arrows.

frequency after each PSS step is minimal as compared to the ITO steps (marked using horizontal curly brackets). The Δf during the assembly of the precursor layer is displayed in Figure 3b. As shown in the figure, the precursor layers [PEI(4.5)/PSS(4.5)]4 and [PEI(2.9)/PSS(2.9)]4 exhibited typical linear growth with the same growth rate, whereas [PEI(8.5)/ PSS(6.5)]4 exhibited a faster growth rate, suggesting the possibility of exponential growth. The exponential growth of precursor layer [PEI(8.5)/PSS(6.5)]4 can be attributed to the pH difference between the PEI and PSS solutions, as has been previously observed by us and other researchers.33,36,37 It has been shown that exponential film growth can occur by depositing a positively charged weak polyelectrolyte (e.g., PEI) at a high pH and another polyelectrolyte or nanoparticle suspension at a low pH.33,36,37 Because of the exponential growth, the [PEI(8.5)/PSS(6.5)]4 precursor layer was about 1.5 times thicker than the other two precursor layers (see curves in Figure 3b). In other words, more PEI chains were deposited using [PEI(8.5)/PSS(6.5)]4 than were deposited for [PEI(4.5)/PSS(4.5)]4. This larger amount of PEI chains in the precursor [PEI(8.5)/PSS(6.5)]4 is also believed to have contributed to its stronger positive surface charge than that of precursor [PEI(4.5)/PSS(4.5)]4 obtained at low pH. Because the majority of the Δf increase during the assembly of [ITO/PSS]9.5 hybrid thin film was contributed by the ITO assembly step (shown in Figure 3a, where the growth for each PSS and ITO was highlighted), we only plotted the change of Δf during each ITO assembly step, that is, Δ( Δf), in Figure 3c. A constant value of Δ( Δf) during the assembly process would indicate a linearly grown thin film. As shown in the figure, the assembly of [ITO(2.9)/PSS(6.5)]9.5 on precursor layer [PEI(8.5)/PSS(6.5)]4 can be divided into two regimes. Prior to the ninth ITO assembly step, the Δ( Δf) increased as the number of bilayers was increased. The region with increasing Δ( Δf) is called the “recovery regime”, which will be explained in more detail later. After the ninth step, the Δ( Δf) reached a constant value around 550 Hz, and this region is designated as the “linear growth regime”. The assembly of [ITO(2.9)/PSS(4.5)]9.5 on the [PEI(4.5)/PSS(4.5)]4 precursor layer also exhibited two regimes. However, the “recovery regime” in this case

Figure 3. (a) The Δf of the QCM crystals during the LbL assembly of [PEI/PSS]4 [ITO(2.9)/PSS]9.5 using polyelectrolyte solutions at different pH values. (b) The Δf of the QCM crystals during the assembly of [PEI/PSS]4 precursor layers (enlarged from (a)). (c) The Δ( Δf) of the QCM crystals during the ITO assembly steps only. The transition points between the “recovery regime” and the “linear growth regime” are indicated by the arrows.

and [PEI(2.9)/PSS(2.9)]4, while using the same ITO(2.9) suspension. The pH of the PSS solution for the hybrid thin film [ITO/PSS]9.5 was kept the same as that of the PSS solution used for the precursor layer. The changes in the resonant frequency of QCM crystals, Δf, during the LbL assembly of three different thin films [PEI(8.5)/ PSS(6.5)]4 [ITO(2.9)/PSS(6.5)]9.5, [PEI(4.5)/PSS(4.5)]4 [ITO(2.9)/PSS(4.5)]9.5, and [PEI(2.9)/PSS(2.9)]4 [ITO(2.9)/ PSS(2.9)]9.5 are shown in Figure 3a. As shown in the figure, the frequency change of QCM crystal, the Δf, changed dramatically when the pH values of the assembly solutions were changed. In this graph, we can also see that the change in resonant 87

dx.doi.org/10.1021/la203626x |Langmuir 2012, 28, 84–91

Langmuir

ARTICLE

Table 1. Summary of the pH of the Solutions Used for the Assembly of Hybrid ITO PSS Thin Films and the Resultant Surface Charge and the Number of Assembly Steps Needed to Reach the Linear Growth Region assembly steps for “recovery regime”/

average Δ( Δf) during the

precursor layer

[ITO/PSS]n

hybrid films surface charge of precursor layer

“linear growth regime”

“linear growth regime”

[PEI(8.5)/PSS(6.5)]4 [PEI(8.5)/PSS(6.5)]4

[ITO(2.9)/PSS(6.5)]9.5 [ITO(4)/PSS(6.5)]9.5

strongly positive positive

9/1 3/7

570 Hz 209 Hz

[PEI(4.5)/PSS(4.5)]4

[ITO(2.9)/PSS(4.5)]9.5

close to zero

2/8

449 Hz

[PEI(2.9)/PSS(2.9)]4

[ITO(2.9)/PSS(2.9)]9.5

negative

0/10

582 Hz

Figure 5. The AFM images of LbL films with 1, 3, 5, and 10 layers of ITO during the assembly of (a) [PEI(8.5)/PSS(6.5)] [ITO(2.9)/PSS(6.5)]9.5, (b) [PEI(8.5)/PSS(6.5)] [ITO(4)/PSS(6.5)]9.5, and (c) PEI(2.9)/PSS(2.9)] [ITO(2.9)/PSS(2.9)]9.5. The precursor layer exhibited (a) strong positive, (b) positive, and (c) negative surface charge, respectively. All of the AFM images were obtained using the same scanning scale of 5  5 μm2 and are shown with the same height scale.

only consisted of two ITO assembly steps, which quickly changed to the “linear growth regime” after the second ITO assembly step. The Δ( Δf) was seen to increase somewhat in the later stage of the “linear growth regime”, probably due to the increased surface roughness. Unlike these two thin films, the assembly of [ITO(2.9)/PSS(2.9)]9.5 on precursor layer [PEI(2.9)/PSS(2.9)]4 only exhibited one regime, the “linear growth regime”. Interestingly, the Δ( Δf) values of these three films in the “linear growth regime” are all around 550 Hz, further confirming that the growth rate for the linear region is the same in all cases. The effect of the surface charge of the precursor layer [PEI/ PSS]4 on the assembly of the [ITO/PSS]9.5 thin films was also investigated by using the same precursor layer [PEI(8.5)/PSS(6.5)]4 with ITO suspensions at pH values of 4. As shown in Figure 4, the “recovery regime” during the assembly of [ITO(4)/ PSS(6.5)]9.5 is much shorter than that observed during the assembly of [ITO(2.9)/PSS(6.5)]9.5 on the [PEI(8.5)/PSS(6.5)]4 precursor layer. The QCM results during the different assembly conditions, as well as the surface charge of the precursor layer in the different ITO suspensions used, are summarized in Table 1. According to Table 1, it is obvious that the length of the “recovery regime” is

dependent on the positive surface charge density of the precursor layer. The higher is the positive surface charge, the longer is the “recovery regime”. As shown in the table, [PEI(8.5)/PSS(6.5)]4 exhibited a strong positive surface charge in the ITO(2.9) suspension, and as many as 9 out of 10 assembly steps belonged to the “recovery regime”. On the other hand, the precursor layer [PEI(2.9)/PSS(2.9)]4 exhibited a negative surface charge when exposed to the ITO(2.9) suspension, and, therefore, none of the 10 assembly steps were found to belong to the “recovery regime”. The effect of the surface charge of the precursor layer on the LBL assembly was also examined by AFM imaging. As shown in Figure 5a, when the precursor layer exhibited a strongly positive surface charge during the assembly of [PEI(8.5)/PSS(6.5)]4 [ITO(2.9)/PSS(6.5)]9.5, only a few relatively large agglomerates were deposited onto the surface after the first ITO layer, and the deposited particles could not cover the whole surface even after the fifth ITO layer. On the other hand, as shown in Figure 5c, when the precursor layer exhibited a negative surface charge during the assembly of [PEI(2.9)/PSS(2.9)]4 [ITO(2.9)/ PSS(2.9)]9.5, many more ITO nanoparticles were deposited in the first ITO assembly step, and only 5 ITO layers were sufficient to fully cover the substrate surface. As shown in Figure 5b, when 88

dx.doi.org/10.1021/la203626x |Langmuir 2012, 28, 84–91

Langmuir

ARTICLE

Figure 6. Schematics of the assembly of the first ITO(2.9) layer on different precursor layers: (a) on positively charged [PEI(8.5)/PSS(6.5)]4 precursor layer, positively charged ITO cannot easily deposit onto the surface due to the repelling electrostatic forces; and (b) on negatively charged [PEI(2.9)/ PSS(2.9)]4 precursor layer, the positively charged ITO can easily deposit onto the surface due to the attractive forces.

the precursor layer exhibited a medium positive surface charge during the assembly of [PEI(8.5)/PSS(6.5)]4 [ITO(4)/PSS(6.5)]9.5, the thin film buildup rate was found to be higher than that when the precursor layer was strongly positively charged (see Figure 5a), but lower than when the precursor layer was negatively charged (see Figure 5c). Explanation of the Effect of the Surface Charge of Precursor Layer [PEI/PSS]4 on the Growth of [ITO/PSS]9.5. The assembly of hybrid thin film [ITO/PSS]9.5 is significantly dependent on the surface charge of the precursor layer [PEI/ PSS]4. The surface charge of the precursor layer [PEI(8.5)/ PSS(6.5)]4 and [PEI(2.9)/PSS(2.9)]4, as well as the assembly of the first ITO(2.9) layer on these two precursor layers, are schematically illustrated in Figure 6. As depicted in Figure 6a, the precursor layer [PEI(8.5)/PSS(6.5)]4 initially exhibits a negatively charged surface due to the outermost PSS layer. Because PEI is assembled at the pH of 8.5, the ionization of PEI chains is low, and the PEI chains form a collapsed structure.33 However, when this precursor layer is immersed in an ITO suspension at a pH as low as 2.9, the PEI chains will become highly ionized. These PEI chains will extend and diffuse out to the outer surface and render a positively charged surface for precursor layer [PEI(8.5)/PSS(6.5)]4.33,36 Because both the precursor layer and the ITO(2.9) solution are positively charged in this case, strong repulsive forces between the precursor layer and the ITO nanoparticles will prevent the deposition of the ITO nanoparticles. On the other hand, as illustrated in Figure 6b, when the PEI and PSS are both assembled at a pH of 2.9 for precursor layer [PEI(2.9)/PSS(2.9)]4, this precursor layer will exhibit a negatively charged surface in the ITO(2.9) suspension. The attractive forces between negatively charged precursor layer [PEI(2.9)/PSS(2.9)]4 and the positively charged ITO(2.9) nanoparticles will lead to deposition of a large number of ITO(2.9) nanoparticles during the first ITO assembly step. Therefore, our results indicate that the initial assembly of the

ITO nanoparticles onto the precursor layer largely depends on the surface charge of the precursor layer. This is why the assembly of [ITO(2.9)/PSS(6.5)]9.5 on [PEI(8.5)/PSS(6.5)]4 started with a very small Δ( Δf) value in the “recovery regime”, whereas the assembly of [ITO(2.9)/PSS(2.9)]9.5 on [PEI(2.9)/PSS(2.9)]4 started with a large Δ( Δf) value in the “linear growth regime”. After the first ITO(2.9) nanoparticle layer is deposited onto a substrate coated with a [PEI(8.5)/PSS(6.5)]4 precursor layer, the positive surface charge of the precursor layer [PEI(8.5)/ PSS(6.5)]4 can be compensated by PSS during the following assembly steps of [ITO(2.9)/PSS(6.5)]9.5. However, this compensation could not be finished in one single step of the PSS assembly (as suggested by Figure 5a), because the PSS was assembled at the pH of 6.5, at which pH the amount of the positive surface charge of the precursor layer was not as high as that at the pH of 2.9. Therefore, multiple steps of PSS assembly were necessary to overcompensate the positive surface charge of the precursor layer. This process of the compensation of the positive surface charge by the PSS layers resulted in the “recovery regime” described earlier (shown in Figure 3a and c as regions of positive or increasing slope). Apparently, the length of this “recovery regime” will be dependent on the amount of positive surface charge that needs to be compensated. Because the precursor layer is strongly positively charged at the pH of 2.9, and weakly positively charged at the pH of 4, the length of the “recovery regime” during the assembly of [ITO(4)/PSS(6.5)]9.5 is shorter than that during the assembly of the [ITO(2.9)/PSS(6.5)]9.5 (see Figure 4 and Table 1). The Effect of Ionic Strength of the PSS Solutions on the “Recovery Regime”. In addition to the amount of the positive surface charge, the length of the “recovery regime” should also depend on how fast the positive surface charge of the precursor layer is compensated. This can be demonstrated by changing the 89

dx.doi.org/10.1021/la203626x |Langmuir 2012, 28, 84–91

Langmuir

ARTICLE

charged surface below a certain value of pH, even though it could remain negatively charged at high pH. The ability of the surface charge of the precursor layer to change sign as a function of the pH of the solution was shown to significantly affect the subsequent assembly behavior of [ITO/ PSS]9.5 hybrid thin films. When the precursor layer surface was positively charged during the assembly of the ITO nanoparticles, which are also positively charged, the assembly process was slower and was divided into two growth regions: the “recovery regime” in which the positive surface charge is eventually compensated by PSS and recovered to a negative surface, followed by the regular “linear growth regime”. The length of the “recovery regime” was found to be dependent on the amount of positive surface charges that need to be compensated for and how fast this surface charge is able to be compensated. The more positive charges the precursor layer has, the longer the “recovery regime” will be. In contrast, when the surface charge is compensated by increasing the ionic strength, the “recovery regime” was found to be shorter than without it. Finally, when the precursor layer surface was negatively charged during the assembly of the first ITO layer, only the “linear growth regime” was observed. To the best of our knowledge, this is the first report on the systematic evaluation of the effect of the surface charge of a polyelectrolyte precursor layer, as well as its significant effect on the efficiency of the subsequent LbL assembly of hybrid thin films.

Figure 7. The effect of adding 0.1 M NaCl to the PSS solution during the growth of two different hybrid films deposited onto the same two precursor layers. The change in Δ( Δf) does not matter for precursor layer [PEI(2.9)/PSS(2.9)]4, while a large difference is seen when hybrid film [ITO(2.9)/PSS(6.5)]9.5 is deposited onto the precursor layer [PEI(8.5)/PSS(6.5)]4, when 0.1 M NaCl was added into the PSS solutions.

ionic strength in the PSS solution. Only the ionic strength of the PSS solution for the assembly of [ITO/PSS]9.5 was changed from 0 to 0.1 M NaCl, and the ionic strength of the PSS solution for the assembly of the precursor layer [PEI/PSS]4 was kept as 0.1 M. As shown in Figure 7, the ionic strength in the PSS solution of the hybrid film significantly affected the “recovery regime”. When 0.1 M NaCl was added into the PSS solution for the hybrid film assembly of the [ITO(2.9)/PSS(6.5)]9.5 thin film onto the positively charged [PEI(8.5)/PSS(6.5)]4 precursor layer, the film showed a “recovery regime” and reached the “linear growth regime” with an assembly rate of 570 Hz/step after the ninth ITO layer. However, when no NaCl was added into the PSS solution of the hybrid film growth, the assembly process started from the “recovery regime”, and it never reached the “linear growth regime” within the 10 ITO layers shown in Figure 7, and the Δ( Δf) of every ITO step was well below 100 Hz/step. In other words, the length of the “recovery regime” was significantly extended when the ionic strength of the PSS solution was decreased. This is due to the lower assembly rate of PSS at low ionic strength,38 which will lower the compensation rate of the positive surface charge of the [PEI(8.5)/PSS(6.5)]4 precursor layer. On the other hand, the ionic strength in the PSS solution did not affect the “linear growth regime” during the assembly of [ITO(2.9)/PSS(2.9)]9.5 on the precursor layer [PEI(2.9)/PSS(2.9)]4.

’ ASSOCIATED CONTENT

bS

Supporting Information. Electrochemical spectroscopy study on the precursor layer surface charge. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT This work was funded by an Otto Kress Scholarship from the Institute of Paper Science and Technology at the Georgia Institute of Technology and was also partially supported by U.S. DOE grant (DE-FG-02-03-ER4603S). ’ REFERENCES (1) Decher, G.; Hong, J. D.; Schmitt, J. Thin Solid Films 1992, 210, 831–835. (2) Decher, G. Science 1997, 277, 1232–1237. (3) Lvov, Y.; Ariga, K.; Ichinose, I.; Kunitake, T. J. Am. Chem. Soc. 1995, 117, 6117–6123. (4) Lvov, Y.; Decher, G.; Sukhorukov, G. Macromolecules 1993, 26, 5396–5399. (5) Ren, K. F.; Ji, J.; Shen, J. C. Biomaterials 2006, 27, 1152–1159. (6) Lee, D.; Rubner, M. F.; Cohen, R. E. Nano Lett. 2006, 6, 2305–2312. (7) Hammond, P. T. Adv. Mater. 2004, 16, 1271–1293. (8) Peng, C. Q.; Thio, Y. S.; Gerhardt, R. A. Nanotechnology 2008, 19, 505603. (9) Peng, C. Q.; Thio, Y. S.; Gerhardt, R. A. J. Phys. Chem. C 2010, 114, 9685–9692. (10) Olek, M.; Ostrander, J.; Jurga, S.; Mohwald, H.; Kotov, N.; Kempa, K.; Giersig, M. Nano Lett. 2004, 4, 1889–1895.

’ CONCLUSIONS In this work, we reported on the effect of the surface charge of the polyelectrolyte precursor layer [PEI/PSS]4 on the subsequent LBL assembly of [ITO/PSS]9.5 hybrid films. The surface charge of the precursor layer was determined by measuring the ζ-potential of precursor-coated monosize silica microspheres. It was found that the surface charge of the precursor layer [PEI/ PSS]4 was not always negatively charged as might be expected due to the presence of the PSS on the outermost layer, but it was dependent on the pH of the PEI and PSS solutions, as well as the pH of the ITO solution that this precursor layer was immersed into. Because of the protonation of PEI chains at low pH, the precursor layer [PEI/PSS]4 was shown to convert to a positively 90

dx.doi.org/10.1021/la203626x |Langmuir 2012, 28, 84–91

Langmuir

ARTICLE

(11) Lee, S. W.; Kim, B. S.; Chen, S.; Shao-Horn, Y.; Hammond, P. T. J. Am. Chem. Soc. 2009, 131, 671–679. (12) Su, P. G.; Lee, C. T.; Chou, C. Y.; Cheng, K. H.; Chuang, Y. S. Sens. Actuators, B 2009, 139, 488–493. (13) Podsiadlo, P.; Kaushik, A. K.; Arruda, E. M.; Waas, A. M.; Shim, B. S.; Xu, J. D.; Nandivada, H.; Pumplin, B. G.; Lahann, J.; Ramamoorthy, A.; Kotov, N. A. Science 2007, 318, 80–83. (14) Alazemi, M.; Dutta, I.; Wang, F.; Blunk, R. H.; Angelopoulos, A. P. Adv. Funct. Mater. 2009, 19, 1118–1129. (15) Cassagneau, T.; Mallouk, T. E.; Fendler, J. H. J. Am. Chem. Soc. 1998, 120, 7848–7859. (16) Jaffar, S.; Nam, K. T.; Khademhosseini, A.; Xing, J.; Langer, R. S.; Belcher, A. M. Nano Lett. 2004, 4, 1421–1425. (17) Kotov, N. A. Layer-by-Layer Assembly of Nanoparticles and Nanocolloids: Intermolecular Interactions, Structure and Materials Perspectives. In Multilayer Thin Films: Sequential Assembly of Nanocomposite Materials, 1st ed.; Decher, G., Schlenoff, J. B., Eds.; Wiley-VCH: Weinheim, 2003. (18) Caruso, F.; Mohwald, H. Langmuir 1999, 15, 8276–8281. (19) Cho, J. H.; Caruso, F. Chem. Mater. 2005, 17, 4547–4553. (20) Caruso, F.; Spasova, M.; Saigueirino-Maceira, V.; Liz-Marzan, L. M. Adv. Mater. 2001, 13, 1090–1094. (21) Lee, D.; Gemici, Z.; Rubner, M. F.; Cohen, R. E. Langmuir 2007, 23, 8833–8837. (22) Bravo, J.; Zhai, L.; Wu, Z. Z.; Cohen, R. E.; Rubner, M. F. Langmuir 2007, 23, 7293–7298. (23) Lvov, Y.; Ariga, K.; Onda, M.; Ichinose, I.; Kunitake, T. Langmuir 1997, 13, 6195–6203. (24) Caruso, F.; Lichtenfeld, H.; Giersig, M.; Mohwald, H. J. Am. Chem. Soc. 1998, 120, 8523–8524. (25) Ai, H.; Meng, H. D.; Ichinose, I.; Jones, S. A.; Mills, D. K.; Lvov, Y. M.; Qiao, X. X. J. Neurosci. Methods 2003, 128, 1–8. (26) Zhang, Y. M.; Liu, L. J.; Xi, F. N.; Wu, T. X.; Lin, X. F. Electroanalysis 2010, 22, 277–285. (27) Xu, L.; Zhu, Z. C.; Sukhishvili, S. A. Langmuir 2011, 27, 409– 415. (28) Patricio, S.; Cruz, A. I.; Biernacki, K.; Ventura, J.; Eaton, P.; Magalhaes, A. L.; Moura, C.; Hillman, A. R.; Freire, C. Langmuir 2010, 26, 10842–10853. (29) Ginley, D. S.; Bright, C. MRS Bull. 2000, 25, 15–18. (30) Shepherd, E. J.; Kitchener, J. A. J. Chem. Soc. 1956, 2448–2452. (31) Smith, M. March’s Advanced Organic Chemistry: Reactions, Mechanisms, and Structure, 6th ed.; Wiley-Interscience: Hoboken, NJ, 2007; pp 359 364. (32) Sauerbrey, G. Z. Phys. Chem. 1959, 155, 206–222. (33) Peng, C. Q.; Thio, Y. S.; Gerhardt, R. A.; Ambaye, H.; Lauter, V. Chem. Mater. 2011, 23, 4548–4556. (34) Meszaros, R.; Thompson, L.; Bos, M.; de Groot, P. Langmuir 2002, 18, 6164–6169. (35) Adamczyk, Z.; Michna, A.; Szaraniec, M.; Bratek, A.; Barbasz, J. J. Colloid Interface Sci. 2007, 313, 86–96. (36) Fu, J. H.; Ji, J.; Shen, L. Y.; Kueller, A.; Rosenhahn, A.; Shen, J. C.; Grunze, M. Langmuir 2009, 25, 672–675. (37) Podsiadlo, P.; Michel, M.; Lee, J.; Verploegen, E.; Kam, N. W. S.; Ball, V.; Qi, Y.; Hart, A. J.; Hammond, P. T.; Kotov, N. A. Nano Lett. 2008, 8, 1762–1770. (38) Schlenoff, J. B.; Ly, H.; Li, M. J. Am. Chem. Soc. 1998, 120, 7626–7634.

91

dx.doi.org/10.1021/la203626x |Langmuir 2012, 28, 84–91