Structure and Interactions of Charged Triblock Copolymers Studied by

Dec 2, 2011 - The focusing time was 5 min at a cross-flow of 2 mL/min. ...... Y.; Billingham , N. C.; Armes , S. P.; Tribe , K. Macromolecules 2002, 3...
0 downloads 0 Views 962KB Size
ARTICLE pubs.acs.org/Langmuir

Structure and Interactions of Charged Triblock Copolymers Studied by Small-Angle X-ray Scattering: Dependence on Temperature and Charge Screening Manja Annette Behrens,† Montse Lopez,† Anna-Lena Kjøniksen,‡,§ Kaizheng Zhu,‡ Bo Nystr€om,‡ and Jan Skov Pedersen*,† †

Department of Chemistry and iNANO Interdisciplinary Nanoscience Center, Aarhus University, Langelandsgade 140, DK-8000 Aarhus C, Denmark ‡ Department of Chemistry, University of Oslo, P.O. Box 1033, Blindern, N-0315 Oslo, Norway § Department of Pharmacy, School of Pharmacy, University of Oslo, P.O. Box 1068, Blindern, N-0316 Oslo, Norway ABSTRACT: A series of thermo-responsive cationic triblock copolymers composed of methoxypoly(ethylene glycol) (MPEG, hydrophilic), poly(N-isopropylacrylamide) (PNIPAAM, temperature sensitive), and poly((3-acrylamidopropyl) trimethyl ammonium chloride) (PN(+), cationic) has been investigated as a function of temperature and ionic strength. In the MPEG-b-PNIPAAMb-PN(+) copolymers, the MPEG block length is constant, and the lengths of the PNIPAAM and PN(+) blocks are varied. The solubility of the PNIPAAM block decreases with increasing temperature, and the triblock copolymer thus provides the possibilities of studying micelles with both neutral and charged blocks in the micelle corona as well as the interplay between these two blocks as the electrostatic interactions are varied by addition of salt. Investigation of the systems by densitometry and small-angle X-ray scattering (SAXS) in a temperature range from 20 to 70 °C gave detailed information on the behavior both below and above the critical micelle temperature (CMT). A clear effect of the addition of salt is observed in both the apparent partial specific volume, obtained from the densitometry measurements, and the SAXS data. Below the CMT, the single polymers can be described as Gaussian chains, for which the repulsive interchain interactions, originating from the charged PN(+) block, have to be taken into account in salt-free aqueous solution. Increasing the salt concentration of the solution to 30 mM NaCl leads to an increase in the apparent partial specific volume, and the electrostatic repulsive interchain interactions between the single polymers vanish. Raising the temperature results in micelle formation, except for the copolymer with only 20 NIPAAM units. The SAXS data show that the polymer with the medium PNIPAAM block length forms spherical micelles, whereas the polymer with the longest PNIPAAM block forms cylindrical micelles. Increasing the temperature further above the CMT results in an increase in the micellar aggregation number for both of the polymers forming spherical and cylindrical micelles. The addition of salt to the solution also influences the aggregates formed above the CMT. Overall, the micelles formed in the salt solution have a smaller cross-section radius than those in aqueous solution without added salt.

’ INTRODUCTION Self-assembly in polymeric systems is of interest both from a fundamental and from an application point of view. Understanding the fundamental mechanisms driving micelle formation is important to gain a better control of this process, and hence make the application of such systems more feasible. Amphiphilic block copolymers can self-assemble in aqueous solution to form micelles, where the hydrophilic part composes the outer corona and the hydrophobic part the inner core of the micelle, thereby reducing the contact between the hydrophobic part and the aqueous solution. Amphiphilic polymers are used in a variety of applications, that is, in paints and cosmetics, and additionally as the micellar core constitutes a hydrophobic environment, these systems are investigated as possible drug carriers.13 The polymers poly(ethylene glycol) (PEG) and poly (N-isopropylacrylamide) (PNIPAAM) have attracted a lot of attention, as PEG has a stealth-like character toward the humane immune system and PNIPAAM is a temperaturesensitive polymer with a lower critical solution temperature r 2011 American Chemical Society

(LCST) of approximately 32 °C in water, close to the physiological temperature, where it undergoes a coil-to-globule transition.4,5 It has previously been shown that diblock copolymers containing PEG and PNIPAAM form micelles in aqueous solution.6 These structures can be formed and disrupted reversibly by changing the temperature. Additionally, PEG has a LCST, which is above 100 °C in salt-free aqueous solution,7 so in water PEG is hydrophilic in a broad temperature range. The transition temperature can be shifted to lower temperatures by the addition of relatively large amounts of salt.8 In the present work, a series of charged thermo-responsive triblock copolymers are investigated. The polymers are composed of methoxy-poly(ethylene glycol)-block-poly(N-isopropylacrylamide)-block-poly((3-acrylamidopropyl) trimethyl ammonium chloride) (MPEG-b-PNIPAAM-b-PN(+)) with varying Received: July 22, 2011 Revised: November 14, 2011 Published: December 02, 2011 1105

dx.doi.org/10.1021/la202841q | Langmuir 2012, 28, 1105–1114

Langmuir

Figure 1. (a) Chemical structure of the cationic diblock PNIPAAM48b-PAMPTMA20 and triblock copolymers. MPEGm-b-PNIPAAMn-bPAMPTMAo with values of m/n/o = 45/20/14, 45/50/20, and 45/70/24. (b) Chemical structures and NMR spectra (300 MHz, 25 °C) of the MPEGm-b-PNIPAAMn-b-PAMPTMAo (in D2O). (c) Illustration of the molecular weight distribution of the triblock copolymers in aqueous solution (0.05 M NaCl), with the aid of AFFFF. The number and weight average molar masses, Mn, Mw, and the polydispersity index values are also given in the figure.

lengths of the PNIPAAM block, constant MPEG block lengths, and a slightly varying PN(+) block length. The PN(+) block is positively charged and is introduced to give a polyelectrolyte character to the polymer. The length of the MPEG block is kept constant at 45 units, whereas the length of the PNIPAAM block is varied from 20 to 70 units. Additionally, the size of the charged PN(+) block is between 14 and 24 units, where the shortest block is in the polymer with the shortest PNIPAAM block. However, the variation of the length of the charged block is not expected to

ARTICLE

influence the character for the copolymer significantly, althought it might affect the strength of the interchain interactions for the single dissolved polymers. Further, a diblock copolymer without MPEG has been investigated, which is composed of 48 PNIPAAM and 20 PN(+) units. The triblock copolymers are denoted by the length of the PNIPAAM block (n = 20, n = 50, and n = 70), and the polymer structure and characteristics are shown in Figure 1. The aim of the current study is to investigate micelle formation in a wide temperature range in aqueous solutions of the cationic triblock copolymers. The variation of temperature changes the solubility of the PNIPAAM block, and the system thus allows one to study micelles with both neutral and charged blocks in the micelle corona. All copolymers are investigated in aqueous solutions at a polymer concentration of 1 wt%. In addition, the self-assembly process is studied in the presence of 30 mM NaCl to obtain information on the impact of electrostatic interactions on the micelle formation. Furthermore, the length of the PNIPAAM block is varied to determine the effect of the thermo-responsive block on self-assembly. Finally, a polymer without an MPEG block has been investigated to study the effect of omitting the hydrophilic block on micelle formation, hereby giving the polymer a more polyelectrolyte-like character. SAXS experiments were conducted in a temperature range from 20 to 70 °C in steps of 5 °C to facilitate investigation of the solution behavior of the unimers and the temperatureinduced self-assembly of the copolymers. In addition, the apparent partial specific volume was determined for the systems in the corresponding temperature interval to provide information of the temperature-induced changes and to obtain contrast factors for SAXS so that the aggregation numbers can be determined. The partial specific volume is denoted apparent as all changes are associated with the solute, although changes in the solvent will contribute to the changes. The partial specific volumes are obtained from measurements of the solution density. From the apparent partial specific volumes, a significant difference is observed between the polymers in the solvent with and without salt. One sees that the partial specific volumes of the systems in the presence of salt are larger than those in water without salt. Modeling the SAXS data, respectively, below and above the CMT gives detailed information about the single polymer behavior and the micellar structures assembled, with different micellar morphologies depending on the length of the PNIPAAM block.

’ MATERIALS AND METHODS Chemicals and Materials. N,N,N0 ,N00 ,N000 ,N000 -Hexamethyltriethylenetetramine (Me6TREN) was prepared according to the procedure described in the literature.9 N-Isopropylacrylamide (NIPAAM) (Acros Organics) was recrystallized from a toluene/hexane mixture and dried under vacuum prior to use. The charged monomer (3-acrylamidopropyl)-trimethylammonium chloride, here abbreviated as AMPTMA (75 wt% in H2O, Aldrich), was purified from the trace inhibitor present in the sample by precipitating into cold acetone, followed by washing with cold acetone and finally drying under vacuum overnight. The synthesis of the MPEG macroinitiator was performed in accordance with a published procedure by the reaction of monomethoxylcapped poly(ethylene glycol) (MPEG45OH and the data of Mn = 2000 were provided by the manufacturer) with 2-bromoisobutyryl bromide in the presence of triethylamine.10 1106

dx.doi.org/10.1021/la202841q |Langmuir 2012, 28, 1105–1114

Langmuir Scheme 1. Synthetic Route for the Preparation of the Cationic Triblock Copolymers via “One-Pot” ATRP Procedurea

a

The three samples have the following compositions: MPEG45-bPNIPAAM20-b-PAMPTMA14, MPEG45-b-PNIPAAM50-b-PAMPTMA20, and MPEG45-b-PNIPAAM70-b-PAMPTMA24.

Copolymer Synthesis and Characterization. The diblock copolymer poly(N-isopropylacrylamide)48-block-poly((3-acrylamidopropyl)trimethylammonium chloride)20 here abbreviated as (PNIPAAM48-bPAMPTMA(+)20) used in this study was synthesized according to previously reported procedures (number-average molar mass Mn = 1.03  104, and polydispersity index Mw/Mn = 1.05).11 The other cationic triblock copolymers methoxyl-capped poly(ethylene glycol)m-block-poly(N-isopropylacrylamide)n-block-poly((3-acrylamidopropyl)trimethylammonium chloride)o, abbreviated as (MPEGm-b-PNIPAAMn-b-PAMPTMA(+)o), were synthesized via a simple “one-pot” two-step ATRP,12,13 which was carried out in a water/DMF, 50:50 (v/v) mixture solvent at 25 °C and with MPEG-MI/CuCl/CuCl2/Me6TREN = 1/1/0.6/1.6 (molar ratio) as the initiator/catalyst system. The details of the preparation and purification of these copolymers were conducted under similar conditions as described in our previous papers.1416 The simplified synthesis route of MPEGm-b-PNIPAAMn-b-PAMPTMA(+)o is illustrated in Scheme 1. In a general procedure, NIPAAM (30 mmol, 3.39 g) and MPEG-MI (0.5 mmol, 1.07 g) were dissolved in 15 mL of water/DMF 50:50 (v/v) solvent mixture ([NIPAAM] = 2.0 M) in a 50 mL Schlenk flask under stirring with a magnetic stirrer. The mixture was degassed by bubbling with argon for at least 1 h, before it was immersed in a water bath that was kept at about 25 °C. A volume of 1 mL of the freshly prepared Cu(I)Cu(II)Me6TREN water stock solution (prepared by adding degassed water (3.2 mL) to CuCl (2 mmol, 0.198 g), CuCl2 (1.2 mmol, 0.16 g), and Me6TREN (3.2 mmol, 0.85 mL)), exposed to vigorous stirring under the influence of argon flow, was withdrawn via a degassed syringe and quickly added to the above mixture; the polymerization reaction was then initiated. When the NIPAAM monomer conversion reaction was approximately 90% (after approximately 30 min), 1 H NMR analysis indicated that more than 90% of the NIPAAM had been polymerized (disappearance of the vinyl signals at δ = 5.5—6.0 ppm), a well-degassed solution of cationic monomer AMPTMA in a water/DMF

ARTICLE

(50/50, v/v) mixture was then added quickly ([AMPTMA]/[MPEGMI] = 30/1) to the reaction mixture via a syringe under an atmosphere of argon. After 1 h, the polymerization reaction was stopped by exposing it to air, and the sample was diluted with water and further dialyzed first against 0.1 N NaCl and then against distilled water at least three times using a regenerated cellulose dialysis membrane with a molecular weight cutoff of 3500. The white solid triblock copolymer MPEG-b-PNIPAAM-b-PAMPTMA was then finally isolated by lyophilization. The chemical structure and composition of the triblock copolymers were ascertained by their 1H NMR spectra (Figure 1b) using a Bruker AVANCE DPX 300 NMR spectrometer (Bruker Biospin, F€allanden, Switzerland), operating at 300.13 MHz at 25.0 °C by using heavy water (D2O) as the solvent. The 1H chemical shift in D2O is referred to the residual HDO proton (δ = 4.70 ppm) in D2O. The number-average molecular weight and the unit numbers of m, n, and o, in MPEGm-bP(NIPAAM)n-b-P(AMPTMA)o were then evaluated by comparing the integral area of the methenyl proton peak (2) of EG (OCH2CH2O, δ = 3.70 ppm), the methyne proton peak (6) of PNIPAAM (NCH(CH 3 )2 , δ = 3.85 ppm), and the methyl proton (13) of PAMPTMA (N(CH3)3, δ = 3.1 ppm) obtained from their NMR spectra (Figure 1b). Three triblock copolymers M(PEG)m-b-P(NIPAAM)n-b-P(AMPTMA)o have been successfully synthesized with different lengths of the NIPAAM (n = 20, n = 50, and n = 70) block, but keeping the same numbers of PEG and nearly the same cationic AMPTMA units in the copolymers, by carefully adjusting the molar ratio of monomers of NIPAAM, AMPTMA, with the MPEG macroinitiator, as well as the polymerization time. The monomers/initiator (NIPAAM/AMPTMA/MPEG-MI, molar ratio) are 25/30/1, 60/30/1, and 80/30/1 for the samples with n = 20, n = 50, and n = 70 of NIPAAM units, respectively. The compositions of the triblock copolymers are estimated to be m/n/o = 45/20/14, 45/50/20, and 45/70/ 24. In the rest of the paper PAMPTMA(+) is denoted PN(+). The polymer solutions for densitometry and SAXS were prepared by dissolving an appropriate amount in the solvent to obtain a 1 wt% polymer concentration. All four polymers were prepared in water and in an aqueous solution of 30 mM NaCl. Asymmetric Flow Field-Flow Fractionation. The asymmetric flow field-flow fractionation (AFFFF) experiments were conducted on an AF2000 FOCUS system (Postnova Analytics, Landsberg, Germany) equipped with an RI detector (PN3140, Postnova) and a multiangle (seven detectors in the range 35145°) light scattering detector (PN3070, λ = 635 nm, Postnova). The triblock samples (1.0 wt % in 0.05 M NaCl) were measured using a 500 μm spacer, a regenerated cellulose membrane with a cutoff of 1000 (Z-MEM-AQU-425N, Postnova), and an injection volume of 20 μL. The measurements were performed by employing a constant detector flow rate of 0.9 mL/min. The focusing time was 5 min at a cross-flow of 2 mL/min. The cross-flow was then reduced to zero during a period of time of 1015 min depending on the sample. Processing of the measured data was achieved by the Postnova software (AF2000 Control, version 1.1.011). The molecular weights of the samples were determined using this software with a Zimm-type fit, and a refractive index increment (dn/dc) of 0.169 for n = 20, 0.166 for n = 50, and 0.169 for n = 70 (determined by using the RI detector at 32 °C). The results from the AFFFF experiments are displayed in Figure 1c. Densitometry. Solution density measurements were performed on a DMA5000 densitometer from Anton Paar. The densities are determined by an oscillating tube technique that exploits the relationship between the period of oscillation and density. The relation holds when the viscosity of the sample is relative low, as it is for the samples in the present work. The apparent partial specific volume, vsolute, of the solute is 1107

dx.doi.org/10.1021/la202841q |Langmuir 2012, 28, 1105–1114

Langmuir

ARTICLE

determined from the density measurements of a solution with solute and of the pure solvent as  vsolute ¼

1

csolute



! 1 Fsolute



1  csolute  csolute



! 1 Fsolvent

2½expð xÞ  1 þ x x2

ð3Þ

½1  expð xÞ x

ð4Þ

The expressions for the structure factors SHS(q,RHS,ηHS) can be found in Kinning and Thomas.20 The structures formed by the polymers when they self-assembled at higher temperature were modeled using either a spherical or a cylindrical coreshell model, as described in the following. The form factor for a spherical coreshell micelle structure with a graded outer interface can for a large number of chains in the corona, N, be approximated by:21 ! q2 σ 2 Ps ðqÞ ¼ ½ΔFshell Vtot ΦðqRout Þ exp  2  ðΔFshell  ΔFcore ÞVcore ΦðqRcore Þ2 þ NðΔFchain Vchain Þ2 Pchain ðqÞ

ð5Þ

where ΔFcore, ΔFshell, and ΔFchain are the excess scattering length density of the core, shell, and chain, respectively, Vout = 4πRout3/3 is the volume related to the position of the outer surface Rout, and Vcore = 4πRcore3/3 is the volume of the core with radius Rcore, Vchain is the volume of a corona chain, N is the number of such chains, and σ describes the width of the outer interface. Φ(qR) is the form factor amplitude for a spherical particle given by Rayleigh:22 ΦðqRÞ ¼

3½sinðqRÞ  qRðcos qRÞ ðqRÞ3

ð6Þ

For a more detailed derivation, the reader is referred to Pedersen and Gerstenberg.19,23 The form factor for a cylindrical micelle consists of the cross section and longitudinal term, Pcyl(q) = Pcross‑section(q)  Plongt(q), when L . R. The longitudinal term of the form factor is that of an infinitely thin rod,24 and the form factor of the coreshell micelle with core radius Rcore and outer radius Rout can for large aggregation numbers and a graded outer interface be described as:25 2 ! 2J1 ðqRout Þ q2 σ2 2 4 Pcyl ðqÞ ¼ Plongt ðqÞ ΔFshell πRout L exp  2 qRout

ð2Þ

where x = q2R2g , and Rg is the root-mean-square ensemble average of the radius of gyration. In pure water, the charged block copolymers display a peak in the scattering data originating from electrostatic repulsion between these. A model was introduced to describe the interchain interactions that originate from repulsion between the charged block situated at one end of the polymer. It was assumed that the interactions could be approximated by a potential centered at the end of the polymer chain and that the effects could be described by an effective hard-sphere potential. The interactions are thus described by a certain effective hard-sphere radius, RHS, and a certain effective hard-sphere volume fraction ηHS. The scattering intensity then becomes:19 IðqÞ ¼ Pchain ðqÞ þ A2chain ðqÞ½SHS ðq, RHS , ηHS Þ  1

Achain ðqÞ ¼

ð1Þ

where csolute is the weight fraction of the solute, and Fsolute and Fsolvent are the measured densities of the solution with the solute and of the pure solvent, respectively. The density measurements were carried out in a temperature range from 20 to 70 °C in steps of 5 °C. Densities were measured for all four block copolymers with concentrations of 1 and 2 wt% in both water and an aqueous solution of 30 mM NaCl. In the temperature range from 20 to 50 °C, the accuracy of measurement is 0.000020 and 0.000050 g/cm3 from 55 to 70 °C. The apparent partial specific volumes have an accuracy of about 0.10.3% due to the relatively low concentration of polymer. Small-Angle X-ray Scattering. The SAXS experiments were performed in the laboratory-based facility at the Department of Chemistry at Aarhus University, Denmark. The camera is a modified NanoSTAR from Bruker AXS optimized for solution scattering.17 Samples were measured in reusable home-built quartz capillary holders that were thermostatted by a Peltier element (Anton Paar). The block copolymers were investigated at a concentration of 1 wt% within a temperature range from 20 to 70 °C in steps of 5 °C. The triblock copolymers were investigated in water both without added salt and with 30 mM NaCl. All data and subsequent backgrounds were collected for 3600 s at each temperature. Background subtraction and conversion of the data to absolute scale by use of water as a primary standard was carried out using the SUPERSAXS program package (Oliveira, C. L. P.; Pedersen, J. S., unpublished). The final intensity is displayed as a function of the scattering vector modulus, q = 4π sin θ/λ, where the X-ray wavelength, λ, is 1.54 Å, and 2θ is the angle between the incident and scattered X-rays. SAXS Models. The scattering data were fitted using a nonlinear least-squares method, and several models were implemented to describe the data over the entire temperature range investigated. The scattering intensity for a monodisperse sample of centrosymmetric particles can be written as I(q) = nP(q)  S(q), where n is the number density of particles, and P(q) and S(q) are the form and structure factor, respectively. Single polymer chain scattering below the CMT is described using the expression for a Gaussian chain given by the Debye function:18 Pchain ðqÞ ¼

where the form factor Pchain(q) is that of a Gaussian chain given in eq 2,18 and the form factor amplitude is:

2J1 ðqRcore Þ 2 ðΔFshell  ΔFcore ÞπRcore L qRcore þ NðΔFchain Vchain Þ2 Pchain ðqÞ

#2

ð7Þ

where J1(x) is the first-order Bessel function of the first kind. To describe the intermicellar correlation in the system of cylinders, a structure factor is included. A random phase approximation is used, and the structure factor is described by: SðqÞ ¼

1 1 þ νRPA Plongt ðqÞ

ð8Þ

where νRPA is proportional to the concentration of the cylinders.26 The data were not fitted on absolute scale because we judged that it would be very difficult to obtain reliable values for the scattering length density of the individual blocks. It also became evident from fitting the scattering data that the core and corona could not be distinguished, 1108

dx.doi.org/10.1021/la202841q |Langmuir 2012, 28, 1105–1114

Langmuir

ARTICLE

where Acs(q) is the scattering amplitude for the cylinder cross section as given by the term in the square brackets in eq 7, and J0(x) is the zeroorder Bessel function of first kind.

’ RESULTS

Figure 2. Apparent partial specific volume for the triblock copolymers n = 70 (blue square), 50 (red triangle), 20 (black circle), and the diblock copolymer (magenta star) in pure water (top) and aqueous solution with 30 mM NaCl (bottom). All lines are guides to the eye. The gray lines represent a linear increase with temperature for the different polymers, and the black lines represent the actual behavior of the apparent partial specific volume as a function of temperature. because no sharp coreshell interface was observed. This in turn led to simplified models where the core was neglected and the micellar structure was described by the corona only with the addition of the polymer term with a radius of gyration and a scale factor as fit parameters. However, for completeness, we have given the expressions for the corecorona model. The aggregation numbers for both the spherical and the cylindrical micelles, Nagg, were determined from the forward scattering, I(q = 0). The aggregation number was determined by taking the ratio between the forward scattering for a single polymer at 20 °C and that of the aggregated structure at the considered temperatures above the CMT: Nagg ¼ α

I agg ðq ¼ 0Þ polymer I20o C ðq

ð9Þ

¼ 0Þ

Here, α is a correction factor introduced to account for the variation of contrast with temperature. It was obtained from the apparent partial specific volumes as a function of temperature. To obtain the average radial scattering density distribution, ΔFmic(r), for the spherical micelles, the form factor amplitude for the micelle, Amic(q) = Acore(q) + Acorona(q), was Fourier transformed:21 ΔFmic ðrÞ ¼

1 Z 2π2

∞ 0

Amic ðqÞ

sinðqrÞ 2 q dq qr

ð10Þ

For the cylindrical micelles, a similar calculation using a Hankel transform and the form factor amplitude of the cross section gives the radial profile of the cross section of the micelles: ΔFcs ðrÞ ¼ 2π

Z ∞ 0

Acs ðqÞJ0 ðqrÞ dq

ð11Þ

Densitometry. The densities were converted to apparent partial specific volumes as described in the Materials and Methods. The measurements were performed on 1 and 2 wt % solutions; a similar trend was observed for both concentrations, and data for 1 wt % are shown in Figure 2. All of the polymers display an increase in the apparent partial specific volume with increasing temperature. The triblock copolymer with the shortest PNIPAAM block (n = 20) displays a nearly perfect linear increase with temperature, and no change, as found for the PNIPAAM homopolymer at the transition temperature, is seen.4 Increasing the length of PNIPAAM blocks to n = 50 or 70 introduces a nonlinear increase of the partial specific volume between 45 and 50 °C that can be associated with the transition of the copolymer, due to the PNIPAAM block. The transition temperature of the PNIPAAM homopolymer is located at approximately 32 °C;5 however, the transition is shifted to higher temperatures when more water-soluble polymer blocks are connected to PNIPAAM.27,28 For the diblock copolymer, a pronounced change of v is found at a lower temperature, between 35 and 40 °C. Furthermore, a significant difference of the temperature dependence of v is seen between the polymers in the salt-free and salt-containing solvent. The partial specific volumes in a solvent of 30 mM NaCl are larger by approximately 1%; however, the same trend is seen as without salt. Scattering Data and Model Fits. The block copolymers are studied at a concentration of 1 wt% in a temperature range from 20 to 70 °C in steps of 5 °C. A subset of the data is shown in Figure 3 at 25 and 70 °C. It is seen that the polymers selfassemble at elevated temperatures, apart from the n = 20 polymer, which is observed to exist as single polymers in the entire temperature domain. In aqueous solution without added salt, a correlation peak is present for the single polymers, which can be ascribed to interchain interactions. As the length of the PNIPAAM block decreases, the maximum of the correlation peak shifts to the right; thus the average distance between the polymers decreases. However, the scattering intensity for the copolymer without a MPEG chain displays a correlation peak at the same q-value as the copolymer with n = 70. Addition of salt to the polymer solutions causes the interchain interactions to vanish, which is apparent from the insets in Figure 3. The critical micelle temperature (CMT) was estimated from the SAXS data (Table 1). It is observed that the polymer with the shortest PNIPAAM block (n = 20) reveals single polymer behavior over the entire temperature range investigated. Gradually increasing the length of the PNIPAAM block introduces a CMT in the measured temperature range. The diblock copolymer without a MPEG block has the lowest CMT. The CMTs are alike for the copolymers in water without salt and the corresponding solutions in the presence of salt, except for the copolymer with n = 50 (Table 1). Single Polymers and Interchain Interactions. The scattering data for the polymers in aqueous solution without added salt at temperatures below the CMT are satisfactorily described by a Gaussian chain including an effective hard-sphere structure factor with the center of the interaction potential placed at one 1109

dx.doi.org/10.1021/la202841q |Langmuir 2012, 28, 1105–1114

Langmuir

ARTICLE

Figure 3. Scattering data and fits for all polymer systems. (a) Scattering data for solutions of triblock copolymer with n = 70 for single chains (closed symbols) and micellar state (open symbols). (b) Scattering data for solutions of triblock copolymer with n = 50 for single chains (closed symbols) and micellar state (open symbols). (c) Scattering data for solutions of triblock copolymer with n = 20 for single chains at 25 °C (closed symbols) and 70 °C (open symbols). (d) Scattering data for solutions of the diblock copolymer for single polymer chains (closed symbols) and aggregated state (open symbols). Inserted plots: Scattering data for the polymers in water without added salt (black) and in aqueous solution with 30 mM NaCl (gray) at 25 °C.

Table 1. Critical Micelle Temperature (CMT) Estimated from SAXS for the Four Polymers in Water and in Aqueous Solutions with 30 mM NaCl polymer/solution

water

30 mM NaCl

n = 20 n = 50

no CMT ∼50 °C

no CMT ∼55 °C

n = 70

∼50 °C

∼50 °C

diblock copolymer

∼40 °C

∼40 °C

end of the polymer to account for the repulsion between the charged part of the polymers. Several characteristic parameters are obtained from the modeling, the radius of gyration, Rg, the effective hard-sphere radius, RHS, and the effective hard-sphere volume fraction, ηHS. Figure 4 shows that Rg of the polymer with n = 20 is constant over the entire temperature range and only slightly affected by addition of salt. With a radius of gyration of approximately 20 Å, the polymer size is significantly smaller than the Rg of the other investigated polymers with longer PNIPAAM blocks. The polymers with n = 50 and 70 are similar in size with an Rg of about 34 Å. Moreover, no noticeable effect is observed upon addition of salt to the solution. The diblock copolymer has a radius of gyration of approximately 31 Å, which is only slightly less than that of n = 50 and 70. Repulsion between the polymers is observed in aqueous solution without salt, and this effect can be described by an effective hardsphere structure factor. The effective hard-sphere radius, RHS, increases with increasing size of the PNIPAAM block, apart from the diblock copolymer, which displayed the largest RHS (Figure 5). The CMT is reflected in the variation of the effective hard-sphere volume

Figure 4. Radius of gyration, Rg, obtained for the polymers with n = 70 (circle), 50 (triangle), 20 (square), and the diblock copolymer (star) in water (closed symbols) without salt and aqueous solutions with 30 mM NaCl (open symbols). The lines are guides to the eye.

fraction, ηHS. There is a gradual transition for the polymers with n = 50 and 70 as the effective hard-sphere volume fraction decreases with temperature over a range of 10 °C. In contrast, the transition is abrupt for the diblock copolymer as no change in ηHS is observed until the transition temperature. Furthermore, RHS and ηHS of the polymer with n = 20 are constant over the entire temperature interval investigated (Figure 5). Micelle Structure. Above the CMT, the polymers selfassemble, apart from the polymer with n = 20. Using micelle models of either spherical or cylindrical geometry, all scattering data above CMT can be described satisfactory for polymers with n = 50 and 70. The scattering length density of the core and corona was similar, thus making it impossible to identify the distribution of the different blocks, and therefore models with a 1110

dx.doi.org/10.1021/la202841q |Langmuir 2012, 28, 1105–1114

Langmuir

Figure 5. Effective hard-sphere radius, RHS, and volume fractions, ηHS (top and bottom, respectively), obtained for aqueous solutions without salt for the triblock copolymers with n = 70 (open circle), 50 (open triangle), 20 (closed square), and for the diblock copolymer (closed star). All lines are guides to the eye.

distinct core and shell were not used. One sees that the polymer with n = 50 forms spherical micelles in solutions both with and without salt. The polymer with n = 70 cannot be described as spherical micelles; however, by employing the cylindrical micelle model, the data are satisfactory described. It is also found that the diblock copolymer self-assembles above the CMT as can be seen from the scattering data displayed in Figure 2. However, the scattering data are poorly defined at low q with no indication of a Guinier region; hence, it was not possible to determine with any confidence the structure and morphology of the polymer assemblies formed by the polymer, even though the pronounced increase in intensity at low q shows that the polymers do self-assemble. The radial profiles of the spherical micelles formed by the polymer with n = 50 are displayed in Figure 6 (top), and a decrease of the micellar radius with both increasing temperature and salt addition is observed. The variation of the radius of the micelles as a function of temperature is large with a decrease of ∼35 and ∼10% in aqueous solution without salt and with 30 mM NaCl, respectively (Figure 6 (top)). The aggregation number, Nagg, for the polymer with n = 50 is smaller in the salt solution by 66% as compared to that of the polymer in water without added salt at 70 °C. The radial cross-section profiles of the cylinders formed by the polymer with n = 70 show little variation with temperature and salt content (Figure 6 (bottom)); however, it is clear that the addition of salt decreases the cross-section radius of the micelles, which is also reflected in the cross-section radii in Figure 7a. Only a slight decrease is observed with temperature, ca. 25% in aqueous solution and ca. 10% in aqueous solution with 30 mM NaCl. Moreover, at 70 °C, the cross-section radius

ARTICLE

Figure 6. Average radial density distribution profiles. Top: Radial profiles for spherical micelles of the polymer with n = 50 in aqueous solution without salt at 55 () and 70 °C ( 3  3 ) and aqueous solutions with 30 mM NaCl at 55 ( 3 3 3 ) and 70 °C (  ). Bottom: Radial profiles for the cross section of cylindrical micelles from the polymer with n = 70 in aqueous solution without salt at 55 () and 70 °C ( 3  3 ) and in aqueous solution with 30 mM NaCl at 55 ( 3 3 3 ) and 70 °C (  ). Note that the radial profiles at 70 °C in aqueous solution without added salt and 55 °C with 30 mM NaCl coincide. Thus, only the radial profile at 55 °C with 30 mM NaCl is visible.

in aqueous solution with 30 mM NaCl is ∼10% lower than that in water without added salt. The aggregation number per unit length, Nagg, shows that the micelles at 70 °C in salt solution have an aggregation number per unit length of ∼80% of that in water without added salt (Figure 7c).

’ DISCUSSION The combined densitometry and SAXS study has provided detailed information over a wide temperature range about the single polymer behavior and micelle formation of the complex thermoresponsive block copolymers investigated. From the apparent partial specific volumes of the polymers, a clear difference between solutions with and without added salt was evident, although the trend is similar. The observed overall linear increase of the partial specific volume of the polymer, v, with increasing temperature has its origin in increased thermal motions. Additionally, the pronounced increase of v that is seen for the polymers with n = 50, 70, and 48, as the temperature is raised, which does not follow the linear increase, is attributed to the collapse of the PNIPAAM chains, as is consistent with the CMT obtained by SAXS. The transition has been observed above the lower critical solution temperature of homopolymer PNIPAAM,4,5 which is due to increased hydrophobicity of PNIPAAM at higher temperature. A transition is clearly seen for the diblock copolymer, without an MPEG block, and the transition temperature is relatively close 1111

dx.doi.org/10.1021/la202841q |Langmuir 2012, 28, 1105–1114

Langmuir

Figure 7. (a) Rout (micelle cross-section radius) obtained for the polymers with n = 70 (circle) and 50 (triangle) in water without added salt (closed symbols) and in solution with 30 mM NaCl (open symbols). (b) Aggregation number in water without added salt (closed symbols) and in solution with 30 mM NaCl (open symbols) for the polymer with n = 50 polymer. (c) Aggregation number per unit length in water without added salt (closed symbols) and in aqueous solution with 30 mM NaCl (open symbols) for the cylindrical micelles of the polymer with n = 70. All lines are guides to the eye in aqueous solution in the absence of salt (black line) and in aqueous solution with 30 mM NaCl (gray line).

to that of the homopolymer PNIPAAM. For the polymers with n = 50 and 70, the introduction of the MPEG block leads to an increase of the transition temperature. The increase of the volume at the transition can be explained by poorer and “looser” hydrogen bonding to the polymer at higher temperature, which in turn also results in an increased hydrophobicity. The difference between the partial specific volumes of the polymers in the solvent with and without salt can also originate from a rearrangement of the hydration layer, so that water is bound more loosely when salt is added due to perturbation of the hydrogen pattern in the water caused by the ions. However, it should be noted that there is an intriguing interplay of many effects including also ion binding to the polyelectrolyte chain,29 and that the various effects cannot be separated, as the local

ARTICLE

concentration of the various polymer blocks also influences the partial volumes of the others. The SAXS data below the CMT show a clear effect of addition of salt, as the repulsive interchain interactions are no longer observed at a salt concentration of 30 mM. However, Rg does not change upon addition of salt; thus the chains maintain their conformations, and only the interchain interactions change significantly. The average distance between the chains increases with increasing PNIPAAM block length, which is seen from the shift of the maximum of the correlation peak to lower q values. As the samples have the same concentration by weight, an increase in polymer mass gives a lower number density of polymer chains and thus a larger average distance between the polymer chains. The charged diblock copolymer without an MPEG block has the largest distance between the chains, which is inconsistent with the above consideration regarding the number density. However, the behavior can probably be explained by the absence of the MPEG block, which gives the polymer a more pronounced polyeletrolyte character and thus results in stronger electrostatic interactions. The repulsive interchain interactions show that the polymers with n = 50 and 70 have a gradual transition, as the hardsphere volume fraction, ηHS, gradually decreases over 1015 °C. In contrast to this, the diblock copolymer exhibits an abrupt transition as ηHS is constant until the interchain interactions vanished. The transition temperature is closer to that of the homopolymer PNIPAAM4,5 for the diblock copolymer than for the triblock copolymers with n = 50 and 70. Both phenomena are attributed to the absence of a MPEG block that makes the contribution from the PNIPAAM block more pronounced, which is also reflected in the partial specific volume variations. The CMT is observed to depend on the polymer composition and slightly on the solvent conditions. The CMT is highly affected by the presence of the MPEG block, resulting in an increase of the LCST by 10 °C as compared to when no MPEG block is present in the copolymer. The micelles formed above the CMT display a clear dependence on the polymer composition, in solutions both with and without added salt for the polymers with n = 50 and 70 forming, respectively, spherical and cylindrical micelles. Further, the diblock copolymer is found to self-assemble in aqueous solution; however, the structures formed cannot be determined from the SAXS data, because of the absence of a Guinier region in the scattering data at low q. The spherical micelles formed in solution of the polymer with n = 50 decrease significantly in radius with increasing temperature, by ∼35% in aqueous solution without salt and by ∼10% in aqueous solution with 30 mM NaCl. In both solutions, the aggregation number increases with increasing temperature from 7 to 10 molecules to 30 and 45 with and without added salt, respectively. A similar trend is evident for the polymer with n = 70, which formed cylindrical micelles at elevated temperatures. Here, the decrease in cross-section radius with temperature of the micelles is ∼25% in aqueous solution and ∼10% in 30 mM NaCl. The aggregation number is given per unit length, due to the large size (effectively infinite with the resolution of the SAXS instrument) of the micelles, and it is also observed to increase with temperature in both solvents. The decrease in micelle radius and increase in aggregation number, as a function of temperature, is associated with the decreased solvent quality of water to MPEG, as the hydration state of PNIPAAM above its transition temperature has been 1112

dx.doi.org/10.1021/la202841q |Langmuir 2012, 28, 1105–1114

Langmuir reported not to change significantly.30,31 PNIPAAM still contains a large amount of water even in its fully collapsed state,32 and one study has shown it contains as much as 66% water.33 It can be estimated that PNIPAAM associates approximately 10 water molecules per repeat unit below the transition temperature; this number decreases to between two and three molecules above the transition.31 For PEG, two to three water molecules are associated per EO repeat unit, hereby rendering PEG water-soluble. At elevated temperatures, the hydrogen bonds between PEG and water are disrupted, which is caused by the conformational changes of the chain, thus making PEG less soluble in water34 at higher temperatures. In addition, the solvent quality of water to PEG can also be changed by addition of salt to the solution, having a similar influence as raising the temperature. Less information exists about MPEG; however, it is expected to exhibit a behavior similar to PEG. We may note that Florin et al. showed that addition of 100 mM NaCl to a solution of PEG would only lower the transition temperature by 3 °C;35 thus the addition of salt should mainly influence the charge screening of the PN(+) block and not have a significant effect on the MPEG block. The relative lower effect of temperature on the micelle size in aqueous solution of 30 mM NaCl is attributed to the screening of the charges, thus allowing the polymer chains to pack closer at lower temperatures as compared to an aqueous solution without added salt. It has previously been shown that diblock copolymers of PEG and PNIPAAM formed spherical micelles in aqueous solution,36 corresponding to the observations for n = 50 in the current study. Additionally, it was found that the polymer with the smallest NIPAAM/EO ratio did not change aggregation state upon increasing the total polymer concentration.6 This agrees with the findings in the present study where changing the PNIPAAM block length promoted structural changes above a certain PNIPAAM block length.

’ CONCLUSION A series of triblock copolymers composed of a hydrophilic block (MPEG), a temperature-sensitive block (PNIPAAM), and a cationic block (PN(+)) has been investigated as a function of temperature and ionic strength. It has been shown that the micelles that form, which have the two hydrophilic blocks, the neutral MPEG and the charged PN(+) block, are affected by many factors, such as temperature, charge screening, and varying length of the PNIPAAM block. Below the CMT, all polymers displayed repulsive interchain interactions in aqueous solution, and addition of 30 mM NaCl led to charge screening and removed the effects of interchain interactions in the systems. With respect to micelle formation in the systems, the length of the PNIPAAM block was observed to govern micelle formation and shape; increasing the block length first favors formation of spherical micelles, and further increase of the PNIPAAM length changes the micellar shape to cylindrical. The addition of 30 mM NaCl gives rise to a closer packing of the polymer chains in the micelles as the charges of the PN(+) block are screened. The effect of increasing temperature is evident above the CMT, where the micelles cross-section radius decreased as the blocks in the outer part of the polymers became less soluble in water. The absence of an MPEG block in the polymer facilitated formation of structures with a size larger than can be characterized by the SAXS instrument used in the present study.

ARTICLE

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT M.A.B. and J.S.P. acknowledge the Danish Council for Independent Research for Natural Sciences for supporting this work. B.N. gratefully acknowledges the support from the Norwegian Research Council for the PETROMAKS project. ’ REFERENCES (1) Sommer, C.; Pedersen, J. S.; Garamus, V. M. Langmuir 2005, 21, 2137. (2) Chuang, C.-Y.; Don, T.-M.; Chiu, W.-Y. J. Polym. Sci., Part A: Polym. Chem. 2009, 47, 5126. (3) Muthukumar, M.; Ober, C. K.; Thomas, E. L. Science 1997, 277, 1225. (4) Stieger, M.; Richtering, W.; Pedersen, J. S.; Lindner, P. J. Chem. Phys. 2004, 120, 6197. (5) Schild, H. G. Prog. Polym. Sci. 1992, 17, 163. (6) Virtanen, J.; Holappa, S.; Lemmetyinen, H.; Tenhu, H. Macromolecules 2002, 35, 4763. (7) Pedersen, J. S.; Sommer, C. Prog. Colloid Polym. Sci. 2005, 130, 70. (8) Ataman, M. Colloid Polym. Sci. 1987, 265, 19. (9) Ciampolini, M.; Nardi, N. Inorg. Chem. 1966, 5, 41. (10) Liu, S.; Weaver, J. V. M; Tang, Y.; Billingham, N. C.; Armes, S. P.; Tribe, K. Macromolecules 2002, 35, 6121. (11) Dedinaite, A.; Thormann, E.; Olanya, G.; Claesson, P. M.; Nystr€om, B.; Kjøniksen, A.-L.; Zhu, K. Soft Matter 2010, 6, 2489. (12) Wang, J.-S.; Matyjaszewski, K. J. Am. Chem. Soc. 1995, 117, 5614. (13) Matyjaszewski, K.; Tsarevsky, N. V. Nat. Chem. 2009, 1, 276. (14) Zhu, K.; Jin, H.; Kjøniksen, A.-L.; Nystr€om, B. J. Phys. Chem. B 2007, 111, 10862. (15) Kjøniksen, A.-L.; Zhu, K.; Karlsson, G.; Nystr€om, B. Colloids Surf., A 2009, 333, 32. (16) Beheshti, N.; Zhu, K.; Kjøniksen, A.-L.; Knudsen, K. D.; Nystr€om, B. Soft Matter 2011, 7, 1168. (17) Pedersen, J. J. Appl. Crystallogr. 2004, 37, 369. (18) Debye, P. J. Phys. Colloid Chem. 1947, 51, 18. (19) Pedersen, J. S. J. Chem. Phys. 2001, 114, 2839. (20) Kinning, D. J.; Thomas, E. L. Macromolecules 1984, 17, 1712. (21) Pedersen, J. S.; Svaneborg, C. Curr. Opin. Colloid Interface Sci. 2002, 7, 158. (22) Rayleigh, L. Proc. R. Soc. London, Ser. A 1910, 84, 25. (23) Pedersen, J. S.; Gerstenberg, M. C. Macromolecules 1996, 29, 1363. (24) Neugebauer, T. Ann. Phys. 1942, 42, 509. (25) Pedersen, J. J. Appl. Crystallogr. 2000, 33, 637. (26) Edwards, S. F. Proc. Phys. Soc. 1966, 88, 265. (27) Kjøniksen, A.-L.; Zhu, K.; Behrens, M. A.; Pedersen, J. S.; Nystr€om, B. J. Phys. Chem. B 2011, 115, 2125. (28) Pamies, R.; Zhu, K.; Kjøniksen, A.-L.; Nystr€om, B. Polym. Bull. 2009, 62, 487. (29) Tondre, C.; Zana, R. J. Phys. Chem. 1972, 76, 3451. (30) Lele, A. K.; Hirve, M. M.; Badiger, M. V.; Mashelkar, R. A. Macromolecules 1997, 30, 157. (31) Dong, L. C.; Hoffmann, A. S. Proc. Int. Symp. Controlled Release Bioact. Mater. 1990, 17, 116. (32) Maeda, Y.; Higuchi, T.; Ikeda, I. Langmuir 2000, 16, 7503. (33) Wang, X.; Qiu, X.; Wu, C. Macromolecules 1998, 31, 2972. (34) Holmberg, K.; J€onsson, B.; Kronberg, B.; Lindman, B. Surfactants and Polymers in Aqueous Solution; Wiley: New York, 2002; 1113

dx.doi.org/10.1021/la202841q |Langmuir 2012, 28, 1105–1114

Langmuir

ARTICLE

Chapter 4: Physicochemical Properties of Surfactants and Polymers Containing Oxyethylene Groups, pp 97118. (35) Florin, E.; Kjellander, R.; Eriksson, J. C. J. Chem. Soc., Faraday Trans. 1 1984, 80, 2889. (36) Topp, M. D. C.; Dijkstra, P. J.; Talsma, H.; Feijen, J. Macromolecules 1997, 30, 8518.

1114

dx.doi.org/10.1021/la202841q |Langmuir 2012, 28, 1105–1114