STUDIES OF CHARGED MOLECULES AT THE AIR/WATER INTERFACE BY SUM FREQUENCY GENERATION VIBRATIONAL SPECTROSCOPY by Dan Hu B.Sc., Jilin University, 2010 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Chemistry) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) June 2016 © Dan Hu, 2016 ii Abstract This dissertation studies the surface chemistry of water and charged molecules using phase-sensitive sum frequency generation (SFG) vibrational spectroscopy. The studied molecules include surfactants, polyelectrolytes, bitumen, and ionic liquid, which are related to technological processes, such as surface modification, catalysis, and bitumen production. Interactions of the polyelectrolyte partially hydrolyzed polyacrylamide with water and cations at air/liquid interfaces were studied. The polyelectrolyte caused water molecules to re-orient with the hydrogen pointing toward the air. The addition of Na+ counteracted the negative charges of the polyelectrolyte. Divalent cation Ca2+ formed a polymer-ion complex with the polymer and completely destroyed the ordered water structure. The addition of polyelectrolytes to a surfactant solution caused a complex behavior of the surface tension. SFG studies showed that the complex surface tension behavior was the result of a surface charge reversal. A better ordered interfacial molecules produced low surface entropy, which counteracted the surface enthalpy decrease and kept the surface tension nearly unchanged at a low surfactant concentration. The ordering of water was found to play a role in surface tension. Four types of surfactants were studied: nonionic, zwitterionic, anionic, and cationic surfactants. Particularly, ionic surfactants decreased the surface entropy to near zero or even negative, which was associated with a surfactant-induced ordering of surface water molecules and iii an increase in hydrogen bond formation. Both effects lead to the reduction of water’s surface entropy. Studies of bitumen/water interfaces using phase-sensitive SFG showed that the bitumen surface carried negative charges, which induced a well-ordered water structure at the bitumen/water interface. The presence of salt neutralized the surface charge and nearly destroyed the ordered water structure. Both anionic and cationic surfactants interacted with the bitumen surface. Finally, the water structure at the air/1-butyl-3-methylimidazolium tetrafluoroborate aqueous solution interface was studied. The orientation of water molecules indicated that a charge reversal occurred at the interface when the concentration of the ionic liquid (IL) changed. The imidazolium cations resided at the water surface at a low IL mole fraction. However, with an increased IL mole fraction, the surface number of anions increased making the surface negatively charged. iv Preface A version of Chapter 3 has been published as: Dan Hu, Z. Yang and Keng C. Chou. “Interactions of Polyelectrolytes with Water and Ions at Air/Water Interfaces Studied by Phase-Sensitive Sum Frequency Generation Vibrational Spectroscopy.” Journal of Physical Chemistry C 2013, 2117 (30), 15698. The project was initiated by Dr. Chou. Dr. Yang constructed the experimental setup. My major contributions are: literature survey, project design, data acquisition, and manuscript composition. A version of Chapter 4 has been published as: Dan Hu and Keng C. Chou. Re-evaluating the Surface Tension Analysis of Polyelectrolyte-Surfactant Mixtures Using Phase-sensitive Sum Frequency Generation Spectroscopy. Journal of American Chemistry Society 2014, 136 (43), 15114. The project was initiated by Dr. Chou. My major contributions are: setup construction, literature survey, project design, data acquisition, and manuscript composition. A version of Chapter 5 will be published as: Dan Hu, Amirhossein Mafi and Keng C. Chou. Revisit the Thermodynamics of Water Surface in the Presence of Surfactants. Accepted by Journal of Physical Chemistry B. The project was initiated by Dr. Chou. Amirhossein Mafi developed computer program for molecular dynamic simulation. My major contributions are: literature survey, SFG spectra acquisition surface tension measurement, and manuscript composition. v A version of Chapter 6 has been published as: Dan Hu and Keng C. Chou. Bitumen/water Interfaces Investigated by Phase-sensitive Sum Frequency Generation Vibrational Spectroscopy - Effects of pH, Ions and Surfactants. Energy & Fuels 2015, 29 (12), 7885. The project was initiated by Dr. Chou. My major contributions are: literature survey, project design, data acquisition, and manuscript composition. A version of Chapter 7 will be submitted for publication. Dan Hu and Keng C. Chou. Surface Charge Reversal of Ionic Liquid at the Air/Water Interface Observed by Phase-sensitive Sum Frequency Generation Vibrational Spectroscopy. I initiated the project. My major contributions are: literature survey, project design, data acquisition, preparing the manuscripts. vi Table of Contents Abstract..............................................................................................................................ii Preface................................................................................................................................iv Table of Contents..............................................................................................................vi List of Tables.....................................................................................................................ix List of Figures....................................................................................................................x Acknowledgments............................................................................................................xv Dedication………............................................................................................................xvi Chapter 1 Introduction.....................................................................................................1 Chapter 2 Phase-sensitive Sum Frequency Generation ................................................4 2.1 Introduction.............................................................................................................4 2.2 Theory of Phase-sensitive Sum Frequency Generation..........................................5 2.3 Orientation of Molecules and the Sign of Im(χ(2))..................................................9 2.4 Experimental Apparatus.........................................................................................10 Chapter 3 Interactions of Polyelectrolytes with Water and Ions at Ait/Water Interfaces..........................................................................................................................13 3.1 Introduction............................................................................................................13 3.2 Experimental Section.............................................................................................15 vii 3.3 Results and Discussion..........................................................................................16 3.4 Conclusions............................................................................................................25 Chapter 4 Re-evaluating the Surface Tension Analysis of Polyelectrolyte – Surfactant Mixtures.........................................................................................................26 4.1 Introduction............................................................................................................26 4.2 Experimental Section.............................................................................................29 4.3 Results and Discussion..........................................................................................30 4.4 Conclusions............................................................................................................39 Chapter 5 Revisiting the Thermodynamics of Water Surfaces and the Effects of Head Group......................................................................................................................40 5.1 Introduction............................................................................................................40 5.2 Experimental Section.............................................................................................42 5.3 Results and Discussion..........................................................................................45 5.4 Conclusions............................................................................................................51 Chapter 6 Surface Charge at Bitumen/Water Interface – Effects of pH, Ions, and Surfactants........................................................................................................................53 6.1 Introduction............................................................................................................53 6.2 Experimental Section.............................................................................................55 viii 6.3 Results and Discussion..........................................................................................56 6.4 Conclusions............................................................................................................62 Chapter 7 Surface Charge Reversal of Ionic Liquid at Air/Water Interface............63 7.1 Introduction............................................................................................................63 7.2 Experimental Section.............................................................................................63 7.3 Results and Discussion..........................................................................................66 7.4 Conclusions............................................................................................................72 Chapter 8 Conclusion......................................................................................................74 References........................................................................................................................78 ix List of Tables Table 4.1 Surface tension, γ, surface entropy, SS, -TSS, and surface enthalpy HS of SDS solutions measured at T=293 K. -TSS and HS are calculated values using γ and SS. .........34 Table 5.1 Surface tension γ, surface entropy SS, -TSS, and surface enthalpy HS of water at T=293 K with various surfactants at their CMC concentrations. .....................................47     x List of Figures Figure 2.1 (A) Reflection geometry of SFG at an interface. (B) Energy diagram of SFG. The solid lines are resonant states and the dash line is a virtual state. ...............................5 Figure 2.2 Experimental geometry of phase-sensitive SFG. A reference SFG is introduced to the system. ....................................................................................................7 Figure 2.3 (A) Raw interferogram obtained from CCD image. (B) The time domain spectrum where signals at t = -ΔT, 0, +ΔT are separated (where ΔT = ~2.5 ps). (C) The frequency domain spectrum by Fourier transforming the time domain spectrum at t = ΔT. (D) The phase-sensitive spectrum of pure water. ...............................................................8 Figure 2.4 (A) A water molecule in molecular coordinate frame. (B) Definition of Euler angles (θ, ψ, φ). .................................................................................................................10 Figure 2.5 (A) Noncollinear geometry. (B) Collinear geometry. PM is phase modulator. P, F, and L denote a polarizer, a filter and a lens, respectively. .......................................12 Figure 3.1 The structural formula of partially hydrolyzed polyacrylamide. ....................15 Figure 3.2 Im(χ(2)) spectra of air/water interfaces in the OH region with various concentrations of HPAM. (A) pure water, (B) 10-8 M, (C) 10-7 M, (D) 5×10-7 M, (E) 10-6 M, and (F)10-5 M. ..............................................................................................................18 Figure 3.3 Im(χ(2)) spectra of air/water interfaces in the CH region with HPAM concentrations at 10-8 M (□), 10-7 M (▽), 5×10-7 M (◊), 10-6 M (Δ), and 10-5 M (○). .....20 Figure 3.4 (A) Im(χ(2)) spectra in the OH region: 5×10-7 M HPAM solution (Δ), 5×10-7 xi M HPAM with 0.01 M NaCl solution (○), and 5×10-7 M HPAM with 0.01 M CaCl2 solution (◊). (B) Im(χ(2)) spectra in the CH region: 5×10-7 M HPAM solution (Δ), 5×10-7 M HPAM with 0.01 M NaCl solution (○), and 5×10-7 M HPAM with 0.01 M CaCl2 solution (◊). .......................................................................................................................23 Figure 3.5 Illustration of the polyelectrolyte and water structures at air/liquid interfaces for (A) HPAM solution, (B) HPAM in the presence of Na+, and (C) HPAM in the presence of Ca2+. ...............................................................................................................24 Figure 4.1 ST of aqueous PDADMAC solution (50 ppm) with various SDS concentrations. The solid line is a guide to the eye. The colored data points indicate the corresponding colored SFG spectra in Figure 3. The insets are the molecular structures of SDS and PDADMAC. ......................................................................................................28 Figure 4.2 (I) ST of water with various SDS concentrations. (II) Im(χ(2)) spectra of air/water interfaces in the CH stretch region. Deuterated water was used to avoid interference with the OH stretch. The peaks near 2875 cm-1 and 2930 cm-1 were assigned to the CH3 symmetric stretch and Fermi resonance. The CH peaks appear negative when the CH3 pointing up. (III) Im(χ(2)) spectra in the OH stretch region. Data of the same color indicates the same SDS concentration: (a) 0 M (blue), (b) 2×10-5 M (green), (c) 1×10-4 M (orange), (d) 3.2×10-4 M (red), (e) 2×10-3 M (magenta), (f) 1×10-2 M (purple). ..............31 Figure 4.3 (I) Im(χ(2)) spectra of (a) pure water (blue), and PDADMAC solutions (50 ppm) with (b) 0 M (cyan), (c) 7×10-5 M (green), (d) 2.5×10-4 M (orange), (e) 8×10-4 M (red), (f) 1.6 ×10-3 M (magenta), and (g) 10-2 M (purple) of SDS. (II) Spectra of xii PDADMAC. (III) Spectra of SDS. Spectra of the same color have the same SDS concentration. ....................................................................................................................35 Figure 4.4 Illustration of the proposed models. (a) A small amount of SDS attracts the PDADMAC to the surface, making the surface overall positively-charged and the net orientation of water's OHs points down. (b) A larger amount of SDS on the surface makes the surface negatively-charged, and the net orientation of OHs points up. ...........37 Figure 4.5 (I) Surface tension of poly(sodium styrene sulfonate) (PSS) solution (50 ppm) with various concentrations of cetyltrimethylammonium bromide (CTAB). The Insets are the molecular structures of CTAB and PSS. (II) Im(χ(2)) spectrum of the PSS solution with 1×10-4 M of CTAB. (III) Im(χ(2)) spectrum of the PSS solution with 6×10-4 M of CTAB. ...............................................................................................................................39 Figure 5.1 Scheme of the attractive forces commonly used to explain the origin of surface tension. While the forces (red arrows) on an inner molecule are balanced, a molecule at the surface experiences a net inward force. In the presence of surfactants (green spheres with tails), the intermolecular attractive forces at the surface are weakened, and the surface tension consequently decreases. ..............................................................41 Figure 5.2 Surface tension of water surface with various concentrations of C12E4 (!), DDAPS (▲), SDS (■), and DTAC (●). The insets are the molecular structures of the surfactants. ........................................................................................................................46 Figure 5.3 (a) Phase-sensitive SFG spectra of various surfactants on water at their CMCs. xiii The peaks near 2875 cm-1 and 2930 cm-1 are the CH3 symmetric stretch and Fermi resonance. The CH peaks appear negative when the CH3 pointing up. (b) The corresponding SFG spectra of water. The OH peaks appear negative when the OH bonds pointing down and positive when the OH bonds pointing up. ..........................................49 Figure 5.4 (a) Structures of water and SDS in the MD simulation. (b) Water dipole order parameter where water’s surface normal is defined as θ = 0. (c) The averaged number of hydrogen bonds per water molecule vs. the depth. .........................................51 Figure 6.1 Im(𝜒!""(!) ) spectra of bitumen/water interfaces in the OH vibrational region at (a) pH = 6 and (b) pH = 9. ................................................................................................58 Figure 6.2 Im(𝜒!""(!) ) spectra of bitumen/water interface in the OH vibrational region with (a) NaCl concentrations at 0 (1), 0.2 mM (2), 10 mM (3) , and 100 mM (4) and (b) CaCl2 concentrations at 0 (1), 0.1 mM (2), 5 mM (3), and 50 mM (4). All spectra were taken at pH = 9. ..............................................................................................................................60 Figure 6.3 Im(𝜒!""(!) ) spectra of bitumen/water interface in OH vibrational region with (a) water, (b) 1 mM SDS, (c) 1 mM DTAC in the aqueous subphase. All spectra were taken at pH = 9. ..........................................................................................................................62 Figure 7.1 Molecular structure of 1-butyl-3-methylimidazolium tetrafluoroborate ([BMIM][BF4]). ................................................................................................................65 Figure 7.2 Water spectra of air/IL aqueous solution interface at IL mole fraction = 0.00018 (magneta), 0.0005 (purple), 0.004 (orange), 0.02 (cyan), 0.04 (blue), 0.2 (green). The color of each concentration corresponds to the same concentration in Figure 7.3. ...68 xiv Figure 7.3 CH vibrational spectra of air/IL aqueous solution interface at IL mole fraction = 0.00018 (magneta), 0.0005 (purple), 0.004 (orange), 0.02 (cyan), 0.04 (blue), 0.2 (green). ..............................................................................................................................69 Figure 7.4 Illustrative scheme on the water molecules orientation and population of imidazolium cations and BF4- anions at (a) low and (b) high IL mole fraction. ...............71 xv Acknowledgements First and foremost, I would like to express my sincere gratitude to my supervisor, Dr. Keng-Chang Chou, for his immense guidance, support, and encouragement throughout my study. His enthusiasm to science, extensive knowledge, and timely help has made every research project an enjoyable experience. He gave me freedom to carry out research projects on my own, which helped me become an independent researcher. His kind advice and optimism is beneficial for my study as well as my future career. I am grateful to Dr. Zheng Yang for his generous assistance and suggestions in my experiments. He also offered many help in my daily life. I enjoy the moments with him and his family. I would like to thank Reza Tafteh, Kaitlin Lovering, and Amirhossein Mafi, who are great colleagues to work with and are always willing to help. I would like to thank Dr. Qifeng Li for his kind guidance, suggestion, and inspiration for the experimental setup construction. Deepest gratitude is owed to my husband Bo Liu, for his enduring love and unconditional support, for his every word of comfort and encouragement, for his company in those difficult moments. I also thank my parents for everything they have done for me. They are amazing and I will be always proud of them. Finally, I give thanks to the Lord. He saved me, loved me, and blessed me. Glory to God. xvi Dedication To my parents, my husband Bo Liu, and our son Ezra Liu. 1 Chapter 1 Introduction The air/water interface has attracted wide attention for its ubiquity in industry and nature. While a pure air/water interface is rather rare in the real world, an interface often contains charged molecules, such as, surfactants, lipids, ions, polyelectrolytes, and proteins. In bulk water, a water molecule can form hydrogen bonds with surrounding water molecules. But at an interface, this structure is disrupted, which gives rise to surface properties that are substantially different from the bulk. For instance, at the air/water interface, water molecules were observed to have free OH groups pointing into the air.1 When charged molecules occupied the interface, they produced an electrostatic field that applied to the interfacial water molecules. Because a water molecule has a small size and a high polarity, it is very sensitive to the properties of the present molecules. A clear understanding of interfacial water molecules not only reveals the water’s role in interface structure but also provides valuable information about the charged molecules at the air/water interface. Various techniques have been developed to study the air/water interface in the presence of charged molecules. Surface tension is the most used technique, as it can provide macroscopic properties of the water surface.2 Neutron reflection has been adopted to investigate the thickness and the surface number density of surfactants, however this technique does not provide information of water.3 Optical spectroscopy, such as IR reflection spectroscopy, has been used to study the molecular structure of the 2 air/water interface, but it is difficult to distinguish between the signals from the bulk and those from the surface.4 Phase-sensitive sum frequency generation (SFG) has been proven to be a powerful tool for studying the air/water interface.5 It is a surface-specific analytical tool with sub-monolayer sensitivity.6,7 In addition, the phase-sensitive SFG technique probes not only the vibrational information but also the orientation and ordering of the functional groups at the interface. These information obtained by phase-sensitive SFG can provide insight to the air/water interface in the presence of charged molecules. In this dissertation, phase-sensitive SFG vibrational spectroscopy is used to study the structure and properties of water at the air/water interface in the presence of various charged molecules, including surfactants, polyelectrolytes, ionic liquid, and bitumen. Chapter 2 presents a theoretical background of phase-sensitive SFG vibrational spectroscopy. In Chapter 3, the interaction of water and partially hydrolyzed acrylamide is studied. Partially hydrolyzed acrylamide has been extensively studied in order to resolves problems related to enhanced oil recovery and oil sands tailing treatment. Chapter 4 provides experimental evidence of a surface charge reversal occurring at the surface of the polyelectrolyte/surfactant mixtures. In addition, the water ordering has been related to surface entropy to provide a thermodynamic picture of water’s surface tension variation in the presence of surfactants. Chapter 5 discusses the surface enthalpy and surface entropy of water in the presence of surfactants. Chapter 6 presents a study of 3 the bitumen/water interface using the phase-sensitive SFG, providing information about the surface charge of the bitumen/water interface and the effects of water pH, salts, and surfactants. Chapter 7 examines the structure of water molecules at the air/liquid interface of water-ionic liquid (IL) mixtures. SFG spectra revealed that the distribution of ions at the air/water interface from an IL-dilute state to an IL-rich state. 4 Chapter 2 Phase-sensitive Sum Frequency Generation 2.1 Introduction Since the first demonstration of sum frequency generation (SFG) vibrational spectroscopy in 1987, the technique has been used in various research fields, including polymer science, atmospheric chemistry, and biophysical chemistry.8 It is known for its surface specificity and high sensitivity, which rely on the principle that the bulk and surface possess different structural symmetries.9 Under the electric-dipole approximation, a SFG process is only allowed in the media with broken centrosymmetry, such as interfaces, but is forbidden in the media with centrosymmetry, such as the bulk water.10 To measure the SFG signal from an interface, one near-IR laser beam (ω1) and one mid-IR (ω2) laser beam are overlapped spatially and temporally at the interface, as shown in Figure 2.1. When the molecular vibration is resonant with the mid-IR frequency (ω2), an enhanced SFG signal (ω3 =ω1 +ω2) can be detected. Compared to the traditional SFG technique stated above, the recently developed phase-sensitive SFG is able to provide not only the vibrational information but also the molecular orientation of the interfacial molecules.11 In phase-sensitive SFG, the SFG signal from a sample interferes with a reference SFG signal, generating an interference pattern from which the orientation information can be extracted. The following sections present a brief description of the theory and experimental apparatus of the phase-sensitive 5 SFG vibrational spectroscopy. 2.2 Theory of Phase-sensitive Sum Frequency Generation Figure 2.1. (A) Reflection geometry of SFG at an interface. (B) Energy diagram of SFG. The solid lines are resonant states and the dash line is a virtual state. As shown in Figure 2.1A, two incoming fields, 𝑬! 𝝎! and 𝑬! 𝝎! , overlap at the interface of interest, 𝑬! 𝝎! = 𝐸!𝑒𝑥𝑝 𝑖𝒌! ∙ 𝒓! − 𝑖𝝎!𝑡 , (1) 𝑬! 𝝎! = 𝐸!𝑒𝑥𝑝 𝑖𝒌! ∙ 𝒓! − 𝑖𝝎!𝑡 , (2) where ki is the wavevector of the incident electric field i with the frequency ωi. The induced second-order nonlinear polarization P(2) is 𝑷(!) = 𝜀!𝜒(!):𝑬! 𝝎! 𝑬! 𝝎! ,           (3)  where  the   𝜀!   is  the permittivity of the vacuum,  and   𝜒(!)   is  the  second-­‐order  nonlinear  susceptibility.  The  sum  frequency  field  generated  by  the  nonlinear  polarization  is   6 𝐸!" 𝜔!" = !!"!!"!!!𝒌!" !"#!!" 𝜒!""! 𝐸!(𝝎!)𝐸!(𝝎!)𝑒𝑥𝑝 𝑖𝒌!"𝑡 − 𝑖𝝎!"𝑡  , (4)  𝜒!""(!) = 𝐿 𝜔!" ∙ 𝑒!" :𝜒(!): 𝐿 𝜔! ∙ 𝑒! 𝐿 𝜔! ∙ 𝑒! , (5) with the output frequency 𝝎!" = 𝝎! +𝝎! and the output momentum 𝒌!" = 𝒌! + 𝒌!. 𝜒!""(!) is the effective second-order susceptibility. In the reflection geometry shown in Figure 1A, the reflection angle 𝜃!" of the sum frequency field follows 𝑘!" sin𝜃!" = 𝑘! sin𝜃! + 𝑘! sin𝜃!, (6) where 𝜃! is the incident and reflection angle at the interface. The intensity of the SF field measured by a detector is 𝐼!" = 𝐸!" 𝜔!" ! = !!"!!!!!! !"#! !!" 𝜒!""(!) !𝐼!𝐼!, (7) where c is the speed of light, and I1 and I2 are the intensities of incident light fields. ieˆdenotes  the  unit  vector  of  the  light  field  at iω ,  and )( iωL   is  the  tensorial  Fresnel  factor.  The 𝜒(!) is a third-order tensor with 27 elements. In a medium with centrosymmetry, 𝜒(!) = 0. Therefore, the SFG process is forbidden in centrosymmetric media, such as bulk water, under the electric-dipole approximation. A nonzero 𝜒(!) consists of a non-resonance contribution and a resonance contribution and can be expressed as 𝜒!""(!) = 𝜒!"(!) + !! !!!!!!!!!!!! , (8) 𝐼𝑚𝜒!""(!) = !!!! !!(!!!!!)!!!!!! , (9) where 𝜒!" is the non-resonant background. Aq, Γq and ωq denote the amplitude, damping constant and frequency of qth vibrational resonance, respectively. When 𝜔! is 7 resonant with 𝜔!, as shown in Figure 2.1B, 𝜒!""(!) will be enhanced, yielding an SFG spectrum. However, with the conventional geometry illustrated in Figure 2.1A, only 𝜒!""(!) can be obtained whereas the phase of 𝜒!""(!) is missing. The sign of the imaginary part of complex χ(2) is particularly important as it reveals the orientation of the dipole. In order to extract the phase of 𝜒!""(!) , a reference SF field is introduced into the system, as shown in Figure 2.2. The reference SF field interferes with the SF field from the interface of interest, yielding an interferometric pattern that contains the phase information. Figure 2.2. Experimental geometry of phase-sensitive SFG. A reference SFG is introduced into the system and produce an interferometric pattern. Extracting the phase information of a sample, i.e., the Im(χ(2)), requires two spectral scans. One scan is to obtain the raw SFG interference fringes of the sample, as shown in Fig 2.3A, the other one is to obtain the raw interference fringes of a dry quarts, which is non resonant within the wavenumber being measured. The total SFG intensity can be expressed as 𝐼 =   𝑬!"!#$ 𝜔!" ! = 8 𝑬!"!"#$ ! + 𝑬!"!"# ! + 𝑬!"!"#$𝑬!"!"#∗ exp 𝑖𝜔Δ𝑇 + 𝑬!"!"#$∗𝑬!"!"# exp −𝑖𝜔Δ𝑇 , (10) where 𝑬!"!"#$ and 𝑬!"!"#are the SF electric fields generated from the sample and the reference. After inverse Fourier transformation of the original interference pattern, as shown in Figure 2.3B, the cross terms, 𝑬!"!"#$𝑬!"!"#𝑒𝑥𝑝 𝑖𝜔Δ𝑇 and 𝑬!"!"#$𝑬!"!"#𝑒𝑥𝑝 −𝑖𝜔Δ𝑇 , correspond to the patterns at t = -ΔT and t = +ΔT, respectively. The fringes at t = 0 are from 𝑬!"!"#$ ! + 𝑬!"!"# !. A high-pass filter function is applied to extract the fringes at t = +ΔT of both the sample and the quartz, followed by a Fourier transformation from the time domain to the frequency domain, as shown in Figure 2.3C. The real part Re(χ(2)) and the imaginary part Im(χ(2)) of the sample are obtained by normalizing χ(2) of sample to that of the quartz, 𝜒(!) ∝ !!"#$(!)!!"#$%&(!) . 9 Figure 2.3. (A) Raw interferogram obtained from CCD image. (B) The time domain spectrum where signals at t = -ΔT, 0, +ΔT are separated (where ΔT = ~2.5 ps). (C) The frequency domain spectrum by Fourier transforming the time domain spectrum at t = ΔT. (D) The phase-sensitive spectrum of pure water. 2.3 Orientation of Molecules and the Sign of Im(χ(2)) 𝜒!"#(!)is the average contribution of hyperpolarizability 𝛽!"#(!) of molecules at an interface, and can be expressed as ! 10 𝜒!"#(!) = 𝑁! 𝛽!"#(!) 𝚤 ∙ 𝑙 𝚥 ∙𝑚 𝑘 ∙ 𝑛!,!,! , (11) where Ns is the molecular number density at the interface, 𝜒!"#(!)   and 𝛽!"#(!)   are in the lab coordinate and the molecular coordinate, respectively, 𝚤 ∙ 𝑙 𝚥 ∙𝑚 𝑘 ∙ 𝑛 is used to transform the molecular coordinate to the lab coordinate, and the angle bracket denotes the average orientation of molecules. 𝛽!"#(!)  is related to the symmetry of a molecule. For a molecule with C2v symmetry, such as H2O, which has an OH symmetric stretch (ss) mode with an ssp polarization combination, the second-order nonlinear optical susceptibility can be written as 𝜒!!",!"!!(!) =!!𝑁!𝛽!!! 1.29+ 0.557 !"#!!!"#! cos!𝜃 + 0.557+ 1.29 !"#!!!"#! sin𝜃 +1− !"#!!!"#! cos𝜃 , (12) where (θ, ψ, φ) are the Euler angles, (a, b, c) are the molecular axes, and the angle brackets indicate the ensemble averages (see Figure 2.4). Figure 2.4. (A) A water molecule in molecular coordinate frame. (B) Definition of Euler angles (θ, ψ, φ). 11 The orientation angle θ, as shown in Figure 2.4, is the angle between the surface normal (z axis) and the molecular c axis, which is also the direction of the transition dipole moment of the symmetric OH stretching. The term inside the curly brackets in Eq. (12) is positive regardless of the value of θ, and Im(βccc) is positive for the symmetric OH stretching mode. Consequently the sign of )Im( )2( ,OHssyyzχ is the same as that of θcos . Therefore, the Im(χ(2)) can be positive or negative for the OH symmetric stretch, depending on the sign of the average OH projection with respect to the surface normal; a positive peak ( θcos > 0) indicates the average orientation of water pointing toward the air (up), and a negative peak ( θcos < 0) indicates the average orientation of the water pointing toward the bulk (down). 2.4 Experimental Apparatus A femtosecond Ti-sapphire laser (120 fs, 800 nm, 1 kHz and 2 mJ/pulse) was used to produce a picosecond 800 nm beam and to pump an optical parametric amplifier to generate a femtosecond IR beam. The input power is 5 µJ for the IR beam (p-polarized) and 20 µJ for the 800 nm beam (s-polarized). In this dissertation, two experimental geometries, collinear and non-collinear, have been used.12 In the non-collinear SFG geometry shown in Figure 2.5A, the IR beam and the 800 nm beam were first overlapped on the sample stage. The incident angles of the IR and the 800 nm beams were 60° and 45°, respectively. The generated SFG signal and the reflected IR and 800 nm beams are then focused onto the reference stage to generate 12 reference SFG. The SFG from the sample stage was delayed with a phase modulator by ~2.5 ps in order to generate an interferometric pattern. Figure 2.5B shows a collinear SFG geometry. Both IR and 800 nm beams, with incident angles at 60°, were focused on a reference stage to generate a reference SFG. The reference SFG, transmitted IR, and transmitted 800 nm then went through a phase modulator, which delayed the reference SFG by ~2.5 ps, and were focused on the sample stage. Two SFG (s-polarized) beams went through a polarizer, a band pass filter, a lens, and a monochromator, and were recorded by a charge-couple device (CCD) camera. Figure 2.5. (A) Noncollinear geometry. (B) Collinear geometry. PM is a phase modulator. P, F, and L denote a polarizer, a filter and, a lens, respectively. 13 Chapter 3 Interactions of Polyelectrolytes with Water and Ions at Air/Water Interfaces 3.1 Introduction Polyelectrolytes are polymers with ionizable groups that make the polymers charged in aqueous solutions.13,14 The interactions between polyelectrolytes, water, and ions have attracted great interest because of the applications of polyelectrolytes in oil recovery,15 water purification,16-19 food additives,20,21 drug delivery,22,23 and layer-by-layer deposition techniques.24-29 Because of the electrostatic interaction, solutions with polyelectrolytes are known to have different rheological properties, compared to their neutral counterparts.30-32 For example, the viscosity of a polyelectrolyte solution is proportional to the square root of the polymer concentration, while the viscosity for uncharged polymers is proportional to the concentration.33 Significant experimental and theoretical efforts have been made to study the rheological and conformational properties of polyelectrolyte solutions, such as the viscosity and radii of gyration of polyelectrolytes.34 Nevertheless, their interactions with water and ions are not completely understood.35 The interfacial properties of polyelectrolytes are poorly understood as only a limited number of effective probes are available. Surface tension has been the main tool for earlier studies at air/solution interfaces.36-38 From the studies, polyelectrolytes were found to show no surface activity at air/water interfaces at low concentrations, and the addition 14 of surfactants of an opposite charge can promote adsorption at the surface.37 Nevertheless, the surface composition cannot be directly calculated from the bulk concentration, and the macroscopic measurements have not provided much molecular level information. Recent studies using neutron reflectometry, which determines the composition and thickness of the surface layer, has allowed the surface properties to be related to the bulk properties,39-41 but the surface structural information and the interactions between water and polyelectrolytes remain mostly unknown. In this chapter, the interactions of polyelectrolytes, water, and ions at air/solution interfaces were studied using phase-sensitive sum frequency generation (SFG) vibrational spectroscopy. SFG provides vibrational spectra of an interface and has been used to study polymers at air/liquid interfaces.42-45 The technique has been shown to be monolayer-sensitive. Conventionally, SFG vibrational spectroscopy only measures the amplitude of the second-order susceptibility |χ(2)| where the phase of χ(2) is missing.9 Recent developments in phase-sensitive SFG have allowed the sign of χ(2) to be measured, which determines the orientation of dipoles at an interface.46-48 The phase-sensitive technique provides an opportunity to gain new insight into the behaviors of polyelectrolytes. Partially hydrolyzed polyacrylamide (HPAM) (Figure 3.1), was used in the current study as it is a widely used polyelectrolyte.15 In this chapter, the behavior of HPAM and water at the air/liquid interface was investigated in the CH and OH vibration regions. In addition, the monovalent cation, Na+, and the divalent cation, Ca2+, were 15 introduced to the system, to evaluate the effect of cations on the structures of water and the polyelectrolyte.    Figure 3.1. The structural formula of partially hydrolyzed polyacrylamide. 3.2 Experimental Section Materials and Sample Preparation. HPAM (average molecular weight ~520 000 g/mol with 80% acrylamide by weight), NaCl (> 99%), and CaCl2 (> 99%) were purchased from Sigma-Aldrich. Water with resistivity >18.2 MΩ·cm was obtained from a Millipore system. All solutions were freshly prepared before the spectroscopic measurements. Polyelectrolyte solutions with concentrations of 10-5 M and 10-6 M were made by dissolving HPAM into pure water. Solutions with concentrations of 10-7 M to 10-8 M were obtained by diluting a 10-6 M solution successively. For the polymer-salt mixtures, the polymer solution was added into an equal volume of salt solution under vigorous stirring. The SFG spectra at 0 and 2 hours after the preparation were compared, and no time dependence was observed. All SFG spectra in the current study were collected within 2 hours after the solutions were prepared and have been repeated at least two times for accuracy and reliability. 16 Phase-Sensitive SFG Setup. The optical configuration mostly adopted the design by Tahara and his coworkers.48,49 The detailed setup can be found in the references.  Briefly, a femto-second Ti:sapphire laser (800 nm, 1 kHz, and 2 mJ/pulse) was used to pump an optical parametric amplifier to generate an IR beam and to produce a narrow-band ps 800 nm beam. The ps 800 nm beam (s-polarized) and IR beam (p-polarized) were temporally and spatially overlapped on the sample with incident angles of 50º and 60º, respectively. The SFG, 800 nm, and IR beams were collected by a concave mirror (f = 150 mm) and focused on a GaAs wafer. To obtain an interferometric spectrum, the SFG from the sample was time-delayed by passing through a silica window with a thickness of 1 mm. The SFG from the sample and the GaAs then went through a polarizer (s-polarized), a band-pass filter, and a monochromator, and the interference spectrum was then recorded by a charge-coupled device (CCD) camera. The interference spectra of samples were normalized against that of a quartz crystal to obtain the phase-sensitive spectra of χ(2). Only the imaginary part of χ(2), Im(χ(2)) , is shown in the spectra as Im(χ(2)) is a measure of the net polar orientation of water molecules.46,49 3.3 Results and Discussion Figure 3.2A shows the phase-sensitive SFG spectra of the pure water surface. The spectrum is similar to those observed by Tian et al.50 and by Nihonyanagi et al.49 The Im(χ(2)) can be positive or negative, depending on the sign of the projection of the OH transition: a positive peak indicates OHs with the hydrogen pointing to the air (upward), 17 and a negative peak indicates OHs pointing toward the bulk (downward).46,49 The assignment for the individual peak at the neat air/water interface has not been completely explained.   51-53 Recently, Nihonyanagi et al. using phase-sensitive SFG along with molecular dynamics (MD) simulation, showed that the H-bonded OH groups near the surface, on average, point down to the bulk, causing the negative OH band to be at a frequency similar to that of bulk water (∼3410 cm-1).54   Tian et al. proposed that the “ice-like” tetrahedrally bonded water molecules have a dominating contribution to the positive band near 3100 cm-1.50 Nevertheless, in Nihonyanagi’s MD simulation, the positive peak was attributed to water dimers with strong hydrogen bonds, and the water molecules that are responsible for the positive 3100 cm-1 peak are not tetrahedrally coordinated. More recently, MD simulations carried out by Pieniazek et al. showed that the positive peak could emerge using a three-body potential, but the tetrahedrally bonded water molecules at the surface are similar to those in the bulk, instead of having an “ice-like” structure.55 Overall, the negative peak at 3450 cm-1 is generally accepted to be from H-bonded interfacial water with OH groups pointing down to the bulk, and the positive peak near 3100 cm-1 from water molecules involved in stronger interactions. 18  Figure 3.2. Im(χ(2)) spectra of air/water interfaces in the OH region with various concentrations of HPAM. (A) pure water, (B) 10-8 M, (C) 10-7 M, (D) 5×10-7 M, (E) 10-6 M, and (F)10-5 M. 19 Figure 3.2B-F show the SFG spectra of aqueous HPAM solutions with HPAM concentrations of 10-8 M, 10-7 M, 5×10-7 M, 10-6 M, and 10-5 M, respectively. With 10-8 M of HPAM in the solution, the water spectrum was similar to that of pure water: a relatively small positive peak at 3100 cm-1 and a negative broad peak near 3450 cm-1. As shown in Figure 3.2C, the spectrum of water changed dramatically with 10-7 M HPAM: the negative peak near 3400 cm-1 totally disappeared and the positive peak near 3100 cm-1 increased significantly. The spectrum suggested that a significant amount of HPAM was present at the surface with a bulk concentration of 10-7 M (~ 50 ppm). This is qualitatively consistent with previous surface tension studies showing the copolymer of acrylamide (AM) and acrylamidomethylpropanesulfonate (AMPS) have no surface activity at air/water interfaces at low concentrations and produced a steep decrease in the surface tension at higher concentration (~2000 ppm).37 However, the onset concentration could be polymer-dependent. Although the SFG spectrum could not quantify the surface coverage of HPAM, the amount of HPAM present at the surface was enough to flip most of the OH bonds to an upward orientation indicated by the enhancement of the positive peak (3100 cm-1) and the disappearance of the negative peak (3450 cm-1). Since the polymer contains a significant number of hydrophobic functional groups, the polymers present at the interface are assumed to be located at the top layer. The charges of HPAM also promoted the ordering of water at the surface and enhanced the overall intensity of SFG (Figure 3.2C). The disappearance of the 3450 cm-1 peak and the enhancement of the 3100 cm-1 peak suggest that stronger hydrogen bonds are formed in the presence of the 20 polyelectrolyte.56,57 Although the MD simulation carried out by Nihonyanagi et al. suggested that the positive peak at 3100 cm-1 comes from the formation of water dimers with three OH bonds parallel to the surface and one OH bond pointing into the bulk water,58 the existence of such dimers is questionable when negatively-charged polymers are present on the surface.54 Based on the electrostatic interaction, the energy should be lower when OH bonds are pointing toward the negatively charged polymers on the surface, as illustrated in Figure 3.5A. For concentrations above 10-6 M (Figure 3.2D-F), a noticeable blue shift is seen in the OH peaks with increasing HPAM concentration. Two positive peaks, near ~3200 cm-1 and ~3400 cm-1, indicate that the OHs maintain the upward orientation up to 10-5 M. (Concentrations above 10-5 M are not of interest because the solutions become viscous.) Overall, the SFG intensity of water peaks reached a maximum at 10-7 M of HPAM, and higher concentrations of HPAM decreased the SFG intensity, indicating that more HPAM will disorder the water molecules.    Figure 3.3. Im(χ(2)) spectra of air/water interfaces in the CH region with HPAM concentrations at 10-8 M (□), 10-7 M (▽), 5×10-7 M (◊), 10-6 M (Δ), and 10-5 M (○). 21       SFG spectra in the CH region were measured to study the behaviors of HPAM at the interface. Figure 3.3 shows the SFG spectra of HPAM for concentrations of 10-8 M, 10-7 M, 5×10-7 M, 10-6 M, and 10-5 M. No CH peak was observed at 10-8 M HPAM, which is consistent with the SFG spectra of water, indicating that the polyelectrolyte was not surface active at 10-8 M. Surprisingly, no peak was observed at 10-7 M of HPAM, though the water spectrum of 10-7 M HPAM showed that water molecules were mostly flipped at this concentration. It is possible that less than a monolayer of HPAM was enough to flip most water molecules, or the HPAM was not ordered at this concentration. For HPAM concentrations ≥ 10-6 M, three major peaks were observed in the SFG spectra. The assignment of the CH stretching modes has been discussed in previous IR and Raman studies.59 The peak at ~2875 cm-1 was assigned to the C-H stretch, and peaks at ~2925 cm-1 and ~2975 cm-1 were assigned to the symmetric and asymmetric stretch of CH2 groups, respectively. The SFG intensity of CH peaks was nearly saturated at 10-6 M, indicating that the polyelectrolyte reached the maximum surface density with a bulk concentration of ~10-5 M. This is consistent with the trend in the corresponding vibrational spectra of water (Figure 3.2), where no significant water structure changes were observed for polyelectrolyte concentrations above 10-6 M. The negative bands in the Im(χ(2)) spectra indicate that CH are pointed upward (hydrogen pointing toward the air) and the NH2 and [COOH-] are buried in water.60-62 The ion effects were studied with both monovalent and divalent cations. Na+ and 22 Ca2+ were chosen since they are among the most abundant salts available, naturally and industrially. Figure 3.4A and 3.4B show the SFG spectra of HPAM (5×10-7 M) in the CH and OH regions, respectively, with 0.01 M NaCl, and 0.01 M CaCl2. Previous studies of the air/water interfaces have shown that cations mainly reside in the sub-layer near the surface.63,64 In a polyelectrolyte aqueous solution, more cations are attracted to the surface as the polymer bearing negative charges are present at the surface. In the case of monovalent cations, the Na+ ions functioned as counterions and shielded the negative charges on the polyelectrolyte, which weakened the ability of HPAM to flip the surface water molecules. Therefore, the water spectrum with 5×10-7 M HPAM and 0.01 M NaCl (Figure 3.4A) resembles that of pure water (Figure 3.2A). The presence of Na+ ions caused no significant changes in the HPAM conformation (Figure 3.5B), where the spectra with and without Na+ are highly similar. In this case, the water molecules are in a more neutral environment and exhibited an orientation similar to that of pure water (Figures 3.5A and 3.5B). 23  Figure 3.4. (A) Im(χ(2)) spectra in the OH region: 5×10-7 M HPAM solution (Δ), 5×10-7 M HPAM with 0.01 M NaCl solution (○), and 5×10-7 M HPAM with 0.01 M CaCl2 solution (◊). (B) Im(χ(2)) spectra in the CH region: 5×10-7 M HPAM solution (Δ), 5×10-7 M HPAM with 0.01 M NaCl solution (○), and 5×10-7 M HPAM with 0.01 M CaCl2 solution (◊). 24  Figure 3.5. Illustration of the polyelectrolyte and water structures at air/liquid interfaces for (A) HPAM solution, (B) HPAM in the presence of Na+, and (C) HPAM in the presence of Ca2+.   In contrast to 0.01 M NaCl, no ordered OH peaks were observed for the HPAM solution with 0.01 M CaCl2 (Figure 3.4A). In addition, the HPAM spectrum (Figure 3.4B) shows that the conformation of HPAM changed significantly when Ca2+ ions were present in the solution. The observed phenomena support the formation of previously proposed polymer-ion complex in the presence of divalent cations.65,66 In this model, one Ca2+ cation could interact with two [COO-] groups, from either intra-chain or inter-chain, and form a Ca[COO]2 jointing point. As many jointing points would exist in the chain, the originally extended backbone would aggregate (Figure 3.5C). The amorphous nature 25 of the polymer-ion complex explained the disappearance of the ordered water structure. The current study shows that the formation of the HPAM/Ca2+ complex completely destroys the ordered water structure at the interface. Experiments with higher Ca2+ concentrations (up to 0.1 M) were also carried out (data not shown) and they produced similar spectra. Experiments with lower Ca2+ concentrations (up to 0.001 M) were also carried out (data not shown) and the SFG spectra did not show significant changes induced by Ca2+. 3.4 Conclusions The interactions of HPAM, water, and ions at air/liquid interfaces were studied using phase-sensitive SFG vibrational spectroscopy. The presence of the polyelectrolyte caused the flipping of water molecules with the hydrogen being oriented toward the air. With 10-7 M HPAM, water had a highly ordered structure with a single peak near 3100 cm-1; higher concentrations resulted in a less ordered structure of the water molecules. The addition of monovalent cations, Na+, did not affect the conformation of the polyelectrolyte, though it flipped the water molecules back to a structure like that of an air/water interface. The addition of divalent cations, Ca2+, changed the conformation of the polyelectrolyte and completely destroyed the ordering of water molecules. The observed phenomena are consistent with the formation of the polymer-ion complex, in which a divalent cation can interact with two intra-chain or inter-chain charges and form a jointing point. 26 Chapter 4 Re-evaluating the Surface Tension Analysis of Polyelectrolyte-Surfactant Mixtures 4.1 Introduction Surfactants are amphiphilic compounds that contain both hydrophobic and hydrophilic groups. They adsorb at interfaces of hydrophobic and hydrophilic media, such as air/water or oil/water interfaces extending the hydrophobic group out of the bulk water into the air (oil) while the hydrophilic group stays in the water.67 The presence of surfactants at an interface often significantly changes the properties of the interface. For example, the wetting and penetration effect, emulsification, dispersion, foaming, and detergency, are some of the important properties of surfactants. Because of their broad applications, understanding surface activities of surfactants is an ongoing effort in both academic and industrial research.68,69 Surface tension (ST) analysis is the most popular technique for studying surface activities of surfactants.70-72 While surfactants are generally described as "substances that reduce surface tension", the situation is not so straightforward when other molecules, such as polyelectrolytes, are present in the solution.39,41,73 The interaction between the surfactant and the polyelectrolyte produces a complex ST behavior, and a clear interpretation has been difficult. For example, one of the most studied systems is the mixture of the anionic surfactant sodium dodecyl sulfate (SDS) and the cationic 27 polyelectrolyte poly-(diallyldimethylammonium chloride) (PDADMAC). (Structures shown in the inset of Figure 4.1.) The SDS-PDADMAC mixture has been extensively studied partly because the interaction becomes particularly pronounced for oppositely charged surfactants and polyelectrolytes.74-76 Figure 1 shows the ST of aqueous PDADMAC solutions with various SDS concentrations. PDADMAC is believed to be surface-inactive as it does not induce a ST change without a surfactant.41,77 Based on the ST measurements, both the SDS and PDADMAC remain surface-inactive until the SDS approaches 10-4 M. Above 10-4 M, ST shows a sharp decrease followed by a steep increase and a subsequent decrease. Previous studies by neutron reflectivity suggested that the adsorbed amount decreased by a factor of two in the 10-4 − 10-3 M region.74 However, a molecular-level understanding of the complex ST behavior is non-existent because there has been no good analytical tool to obtain molecular-level information at the liquid surface. Figure 4.1. ST of aqueous PDADMAC solution (50 ppm) with various SDS concentrations. The solid line is a guide to the eye. The colored data points indicate the 28 corresponding colored SFG spectra in Figure 3. The insets are the molecular structures of SDS and PDADMAC. Recent developments in phase-sensitive sum-frequency generation (SFG) spectroscopy have provided a new opportunity to study the surface activity of surfactants. While the traditional SFG vibrational spectroscopy measures only the amplitude of the 2nd-order nonlinear optical susceptibility |χ(2)|,43,78 the phase-sensitive SFG spectroscopy also measures the phase of χ(2). The advantage of the phase-sensitive SFG is that the spectrum of the imaginary χ(2), Im(χ(2)), characterizes surface vibrational resonances and their averaged orientations.47,79,80 In the current study, we re-evaluate the ST analysis with the new information obtained from the phase-sensitive SFG spectra and present a molecular-level understanding of the ST. We found that ST is a misleading indicator for the surface activity of the surfactant. The complex behavior of the ST in Figure 4.1 is related to a surface charge reversal at the water surface. 4.2 Experimental section Sample Preparation. Poly-(diallyldimethylammonium chloride) (PDADMAC), (sodium styrene sulfonate) (PSS), cetyltrimethylammonium bromide (CTAB), and sodium dodecyl sulfate (SDS) were purchased from Sigma-Aldrich. Polyelectrolyte-surfactant solutions were prepared using a standard mixing method.81 Briefly, both polyelectrolyte and surfactant solutions were prepared with the concentration twice the desired 29 concentration of the polyelectrolyte-surfactant mixture, then an equal volume of polyelectrolyte and surfactant solutions were mixed. All the mixtures were freshly made before SFG measurements. No significant time dependence was observed in the SFG spectra during the experiments (~3 hours). Surface Tension Measurements. The surface tension of water was measured by the Wilhelmy plate method. The surface tension was determined by measuring the force acting on a Wilhelmy plate when the plate was slowly lifted out of the water surface. The surface tension measurement was repeated at least three times for each solution. Data in Figure 4.1 and Figure 4.2(I) were measured at 293 K. Phase-Sensitive SFG Setup. A femtosecond Ti-sapphire laser (120 fs, 800 nm, 1 kHz and 2 mJ/pulse) was used to produce a picosecond 800 nm beam and to pump an optical parametric amplifier to generate a femtosecond IR beam. The IR and picosecond 800 nm beams were aligned in a collinear optical path with an incident angle of 60°.11 A reference SFG was obtained by focusing the IR and picosecond 800 nm beams into a quartz crystal with a thickness of 50 µm. The IR, picosecond 800 nm, and the reference SFG beams were then focused again on the sample. The reference SFG and the SFG generated at the sample went through a time-delay, a polarizer, a band pass filter, a lens, and a monochromator, then the interference pattern was recorded by a charge-couple device (CCD) camera. The polarization combination used in this study is SSP (s-polarized SFG, s-polarized 800 nm, and p-polarized IR). The energies of the 800 nm 30 and IR beam were ~10 µJ/pulse and ~5 µJ/pulse, respectively. The spot size on the sample was ~ 300 µm. Each spectrum presented in this chapter was an average over 1200 second (1.2 million laser pulses). All SFG spectra were measured at 293K. 4.3 Results and discussion Before looking into the ST of SDS-PDADMAC mixtures, it is useful to understand the ST of pure SDS solutions, as shown in Figure 4.2(I). Figure 4.2(II) and 4.2(III) show the SFG spectra of SDS solutions in the CH and OH regions, respectively. Figure 4.2(I) suggests that SDS is nearly surface-inactive until its concentration reaches ~5×10-4 M. However, Figure 4.2(II) shows that a significant amount of SDS has been observed on the water surface at 1×10-4 M (orange), and Figure 4.2(III) shows that the spectrum of water has significantly changed at 2×10-5 M (green). These SFG spectra suggest that ST is a lagging indicator for the surfactant’s surface activity. Figure 4.2. (I) ST of water with various SDS concentrations. (II) Im(χ(2)) spectra of air/water interfaces in the CH stretch region. Deuterated water was used to avoid 31 interference with the OH stretch. The peaks near 2875 cm-1 and 2930 cm-1 were assigned to the CH3 symmetric stretch and Fermi resonance. The CH peaks appear negative when the CH3 pointing up.82 (III) Im(χ(2)) spectra in the OH stretch region. Data of the same color indicates the same SDS concentration: (a) 0 M (blue), (b) 2×10-5 M (green), (c) 1×10-4 M (orange), (d) 3.2×10-4 M (red), (e) 2×10-3 M (magenta), (f) 1×10-2 M (purple). A big puzzle was how the ST of SDS solutions near 3.2×10-4 M (red color) could remain similar to that of pure water when the SFG spectra indicated that a significant amount of SDS was present on the surface and the structure of water had dramatically changed. Answering this question requires a better understanding of the Im(χ(2)) spectra. The pure water spectrum in Figure 4.2(III)(a) has two distinguishable OH bands: a positive OH band near 3100 cm-1 and a negative OH band near 3450 cm-1. For the OH symmetric stretch, the Im(χ(2)) can be positive or negative, depending on the sign of the OH projection with respect to the surface normal: a positive peak indicates water molecules with the hydrogen pointing toward the air (up), and a negative peak indicates the OHs pointing toward the bulk (down).12,49,83 Nihonyanagi et al., using phase-sensitive SFG along with molecular dynamics (MD) simulation, showed that the H-bonded OH groups near the surface, on average, point down toward the bulk, causing the negative OH band to be at a frequency similar to that of bulk water (3410 cm-1).54 The origin of the low-frequency peak near 3100 cm-1 has been controversial. Tian et al. proposed that “ice-like” tetrahedrally bonded water molecules have a dominating contribution to the 32 positive band near 3100 cm-1.84 Nevertheless, in Nihonyanagi’s MD simulation, the positive peak was attributed to water dimers, which generates a vertical induced dipole pointing toward the air,54 rather than tetrahedrally coordinated water molecules. On the other hand, the MD simulations carried out by Pieniazek et al. using a three-body-interaction model showed that the positive peak at the lower frequency is a result of cancellation between the positive contributions from four-hydrogen-bonded molecules and the negative contribution from those molecules with one or two broken hydrogen bonds.85 In the current study, the origin of the low frequency band is not of a particular concern because flipping the 3450 cm-1 band from negative to positive has to involve flipping the averaged water orientation from pointing down to pointing up.80 It has been experimentally shown that an all-negative spectrum was observed when the water surface was occupied by positively-charged surfactants, and an all-positive spectrum was observed when the water surface was occupied by negatively-charged surfactants.60,61,86 Therefore, the broad positive bands in Figure 4.2(III)(d)-(f) indicate that the negatively charged SDS induced an upward averaged OH orientation. When the SDS concentration increased, a more ordered water structure formed as indicated by the enhanced OH peaks. The origin of the OH frequency shift is not clearly known, but it could be related to a change in the strength of H-bonds or vibrational coupling.43 The formation of a more ordered water structure induced by SDS provides an explanation of how the ST can remain nearly constant at a lower SDS concentration. ST 33 is the Gibbs free energy per unit surface area: ( )T,PAGγ ∂∂≡ where A is the surface area, T is the temperature, and P is the pressure. Since ss TSHγ −= , where HS is the surface enthalpy and SS is the surface entropy, the change of ST, γΔ , is a combined effect of ΔHS and ΔSS. The ΔHS is negative for the adsorption process. On the other hand, the formation of more ordered water and SDS structures suggests a decrease in the surface entropy (ΔSS<0 or -TSS>0), which counteracts the negative ΔHS and can potentially keeps the ST stable. The ST would then drop after SS reaches its minimum. To test this model, the surface entropy and surface enthalpy were measured using ( )Ps TγS ∂∂−= and ss TSγH += , respectively. Table 4.1 summarizes the measured SS and HS. It is clear that the ST shows little decrease at SDS concentration below 2×10-3 M because the decrease in HS is compensated by the decrease in SS (or the increase of -TSS). Since the surface entropy is an excess property relative to the entropy of bulk water, SS can become negative. The surface entropy reached its minimum near 2×10-3M which is consistent with the onset of the ST drop in Figure 4.2(I). Theoretically, the adsorbed SDS molecules also contribute to the ΔSS, but its contribution cannot be separated from that of water in the measurements. The increase of HS from 48 mN/m at 2×10-3 M to 52 mN/m at 1×10-2 M can be explained by the electrostatic repulsive force between SDS as a closely packed SDS layer is enthalpically unfavorable. The SS increases from -0.05 mN/m·K at 2×10-3 M to 0.04 mN/m·K at 1×10-2 M cannot be explained by the SFG spectra of water and SDS as both SDS and water show the highest SFG intensities at 1×10-2 M. It is possible that the SS increase at 1×10-2 M SDS is related to the relative geometry between water and 34 SDS because the relative position of water and SDS cannot be seen from the separate spectra of SDS and water. This assumption is consistent with fact that the critical micelle concentration (CMC) of SDS is around 8 ×10-3 M.87 Table 4.1. Surface tension, γ, surface entropy, SS, -TSS, and surface enthalpy HS of SDS solutions measured at T=293 K. -TSS and HS are calculated values using γ and SS. [SDS] (M) γ (mN/m) SS (mN/m·K) -TSS (mN/m) HS (mN/m) 0 72.7 ± 0.2 0.15 ± 0.04 -44 ± 12 117 ± 12 2×10-5 72.5 ± 0.2 0.14 ± 0.04 -41 ± 12 114 ± 12 3.2×10-4 71.1 ± 0.2 0.10 ± 0.04 -29 ± 12 100 ± 12 2×10-3 62.5 ± 0.2 -0.05 ± 0.04 15 ± 12 48 ± 12 1×10-2 40.6 ± 0.2 0.04 ± 0.04 -12 ± 12 52 ± 12 With the addition of PDADMAC, the behavior of water is significantly different. Figure 4.3 shows the SFG spectra of the aqueous PDADMAC solutions (50 ppm) with various SDS concentrations in the OH and CH regions. The CH spectra in Figure 4.3(II) were obtained with deuterated SDS in the solutions, therefore the spectra represent the CH peaks from PDADMAC. Because deuterated PDADMAC was unavailable, Figure 4.3(III) contained signals from both PDADMAC and SDS. However, it was dominated by SDS because the signal of SDS was one order-of-magnitude larger than that of PDADMAC. A higher SFG intensity from SDS is expected because SDS forms a much more ordered surface layer in comparison to PDADMAC. The peaks at 2875, 2925, and 2970 cm-1 were assigned to the CH3 symmetric stretch, Fermi resonance, and CH3 35 asymmetric stretch, respectively. Figure 4.3. (I) Im(χ(2)) spectra of (a) pure water (blue), and PDADMAC solutions (50 ppm) with (b) 0 M (cyan), (c) 7×10-5 M (green), (d) 2.5×10-4 M (orange), (e) 8×10-4 M (red), (f) 1.6 ×10-3 M (magenta), and (g) 10-2 M (purple) of SDS. (II) Spectra of PDADMAC. (III) Spectra of SDS. Spectra of the same color have the same SDS concentration. The SFG spectra in Figure 4.3(I)(b) and 4.3(II) (cyan) confirmed that the PDADMAC is not surface-active without SDS. Figure 4.3(I)(b) is nearly identical to the spectrum of pure water, and Figure 4.3(II) (cyan) shows PDADMAC was undetectable without SDS. Surprisingly, adding 7×10-5 M of the negatively charged SDS makes the surface positively charged as indicated by an all-negative water spectrum in Figure 4.3(I)(c) 36 (green). A reasonable explanation is that SDS at the surface attracted the positively-charged PDADMAC to the surface. Since SDS is mono-valent, and the polyelectrolyte is multi-valent, a single SDS molecule at the water surface can attract multiple positive charges carried by a PDADMAC molecule, making the surface overall positive, as illustrated in Figure 4.4(a). This is consistent with previous ST studies suggesting that the SDS and PDADMAC complex formed at this low concentration is between an isolated SDS and PDADMAC.73,88 Figure 4.3(II) and 4.3(III) also confirmed that both PDADMAC and SDS were observed at the water surface with 7×10-5 M of SDS (green curves). A comparison between Figure 4.2(II) and Figure 4.3(III) indicates that at the same SDS concentration significantly more SDS was attracted to the surface in the presence of PDADMAC. Figure 4.4. Illustration of the proposed models. (a) A small amount of SDS attracts the PDADMAC to the surface, making the surface overall positively-charged and the net orientation of water's OHs points down. (b) A larger amount of SDS on the surface makes the surface negatively-charged, and the net orientation of OHs points up. The dramatic decrease and recovery of the ST between 10-4 M and 10-3 M of SDS, as 37 shown in Figure 4.1, has been difficult to explain with hypotheses developed based on ST and neutron reflection studies.39,88-90 The SFG spectra in Figure 4.3 suggest that the rapid ST decrease and recovery originates from a surface charge reversal. From 2.5×10-4 M to 8×10-4 M of SDS, the spectra of water changed from an all-negative OH spectrum in Figure 4.3(I)(d) (orange) to an all-positive spectrum in Figure 4.3(I)(e) (red), which indicated that the surface charge changed from positive to negative. Above 8×10-4 M SDS, the ordering of water increased with increasing SDS concentration and eventually reached its maximum where the ST drops again above 1×10-3 M, similar to that of pure SDS solution in Figure 4.2(I). Figure 4.3(II) shows that the increased adsorption of SDS at the surface depleted PDADMAC from the surface. At 1×10-2 M SDS (purple curves), PDADMAC became undetectable, whereas the SDS signal reached maximum. It has been observed that the strong interaction between polyelectrolytes and surfactants leads to formation of aggregates and phase separation in the bulk at a higher concentration. The detailed structure of the SDS-PDADMAC complex at the surface is not well understood.91 However, the current study suggests that PDADMAC is located below the top SDS layer as illustrated in Figure 4.4(b). Previous neutron reflectivity studies have shown that the thickness of the surface layer stayed within 17-19 Å when SDS was varied from 10-6 to 10-1 M.74 Because the length of a SDS molecule is about 18 Å, the study suggested that surface aggregation or phase separation did not occur at the water surface. The SFG spectra suggest that during the process of the surface charge reversal, the surface charge density decreases, which causes an increase in the surface entropy and a 38 decrease in the ST. As the surface accumulates more SDS and becomes negatively-charged, the entropy decreases again and the ST recovers. To correlate the SFG spectra with thermodynamic functions, measurements of the SS and HS were attempted. However, the measurements were unreliable because of a well-known slow decrease of ST over time in surfactant-polyelectrolyte mixtures.92 Although the time dependence is only a few percents, it is larger than the temperature dependence. Since the SFG spectra had a background fluctuation of 5-10%, the time-dependence showed little effect on the SFG spectra. To test the generality of the observed surface charge reversal, mixtures of cationic surfactant cetyltrimethylammonium bromide (CTAB) and anionic polyelectrolyte poly(sodium styrene sulfonate) (PSS) were studied. In this case, flipping of the SFG spectrum from positive to negative was observed between 10-4 M and 10-3 M of CTAB. (Figure 4.5) Therefore, the surface charge reversal is a common phenomenon of surfactant-polyelectrolyte mixtures, not linked to the specific properties of SDS and PDADMAC. 39 Figure 4.5. (I) Surface tension of poly(sodium styrene sulfonate) (PSS) solution (50 ppm) with various concentrations of cetyltrimethylammonium bromide (CTAB). The Insets are the molecular structures of CTAB and PSS. (II) Im(χ(2)) spectrum of the PSS solution with 1×10-4 M of CTAB. (III) Im(χ(2)) spectrum of the PSS solution with 6×10-4 M of CTAB. 4.4 Conclusions In conclusion, ST was found to be an inaccurate indicator for the surface activities of surfactants because the decrease of surface entropy counteracted the decrease of surface enthalpy and kept the ST nearly unchanged at a lower surfactant concentration. The phase-sensitive SFG spectra showed that the origin of the dramatic changes in the ST of ionic surfactant-polyelectrolyte solutions is a result of a surface charge reversal. 40 Chapter 5 Revisiting the Thermodynamics of Water Surfaces and the Effects of Surfactant Head Group 5.1 Introduction Besides mercury, water has the highest surface tension among all common liquids: water's surface tension is 72 mN/m compared to ethanol's 22 mN/m and heptane's 20 mN/m.  93 The high surface tension is due to the rather strong hydrogen bonding between water molecules. A typical textbook description for the origin of surface tension is illustrated in Figure 5.1.94 While a molecule in the bulk liquid experiences attractive force from all directions, a molecule at the surface experiences a net inward force. This force always tends to minimize the surface area of the liquid. By definition, the energy required to increase the liquid surface is the surface tension. In the presence of surfactants (green colored in Figure 5.1), the surface tension of water decreases because the surfactant-water and surfactant-surfactant interactions are weaker than the water-water interactions. Although this description has offered a simple picture to understand surface tension and the effect of surfactants, it has ignored an important thermodynamic function of water: entropy. 41 Figure 5.1. Scheme of the attractive forces commonly used to explain the origin of surface tension. While the forces (red arrows) on an inner molecule are balanced, a molecule at the surface experiences a net inward force. In the presence of surfactants (green spheres with tails), the intermolecular attractive forces at the surface are weakened, and the surface tension consequently decreases. Water has the lowest entropy among all common liquids. The entropy of liquid water is 70 J/(mol·K), compared to ethanol's 161 J/(mol·K) and heptane's 328 J/(mol·K).93 The low entropy of water is also due to the strong hydrogen bonding between water molecules. In Boltzmann's statistical formulation, entropy is related to the freedom of motion: entropy S = k·ln(Ω), where k is the Boltzmann's constant, and Ω is the number of possible microstates in which a system can be found. The more freedom the molecules have to occupy different microstates, the higher the entropy. On average, each liquid water molecules has 3 – 3.5 hydrogen bonds.95-97 Therefore the possible motions of water molecules in the bulk liquid are relatively constrained, resulting in a low number of possible microstates, i.e. a low entropy state. On the other hand, water molecules at the surface have significantly more freedom for rotational and translational movements. 42 Therefore, the entropy of the surface is higher than that of the bulk. While surface tension is the most popular measure to study water surface and the effect of surfactants, the surface entropy of water has been largely ignored. The thermodynamic definition of surface tension (γ) is the Gibbs free energy per unit area, which is a function of surface enthalpy (HS) and surface entropy (SS): 𝛾 =  𝐻! + 𝑇 ∙ 𝑆!. In the current study, we measured 𝐻! and 𝑆! separately to gain a better understanding of water surface. Additionally, surface vibrational spectra obtained by phase-sensitive SFG vibrational spectroscopy were correlated to the measured macroscopic thermal dynamics functions, and MD simulations were carried out to gain deeper insight into the properties of water surfaces. 5.2 Experimental section Surface Tension Measurements. The surface tension ( γ ) of water was measured at 20 °C by the Wilhelmy plate method. Each measurement was repeated at least three times for every solution. The temperature dependence of the surface tension was obtained by changing the system temperature (T) between 15 °C and 25 °C. The surface entropy of water presented in Table 5.1 was calculated using the slope: ( )TγS s ΔΔ−= . Since the γΔ was very small, a larger ΔT was needed to measure an accurate γΔ . Since sS was not strongly dependent on the temperature, a ΔT of 10 °C was not the dominating error of the measured sS presented in Table 5.1. The uncertainty in the measured surface tension (±0.2 mN/m) was the dominating error. 43 Phase-Sensitive SFG. A femtosecond Ti-sapphire laser (2 mJ/pulse at 1 kHz) was used to pump an optical parametric amplifier (TOPAS, Coherent, USA) to generate a femtosecond IR beam. The IR beam and a narrowband picosecond 800 nm beam were aligned in a collinear optical path previously described by Shen and cowrokers.11 A reference SFG beam was obtained by focusing the IR and picosecond 800 nm beams into a quartz crystal with a thickness of 50 µm. The IR, picosecond 800 nm, and reference SFG beams were then focused again on the sample. The reference SFG and the SFG generated at the sample went through a time-delay, a polarizer, a bandpass filter, a lens, and a monochromator, then the interference pattern of SFG was recorded by a charge-couple device (CCD) camera. The Im(χ(2)) spectra were obtained following the methodology described previously by Tahara and his coworkers.80 The polarization combination used in this study was SSP (s-polarized SFG, s-polarized 800 nm, and p-polarized IR). The incident angle was 60°. The energies of the 800 nm and IR beam were ~10 µJ/pulse and ~5 µJ/pulse, respectively. The spot size on the sample was ~ 300 µm. Each spectrum presented in this chapter was an average over 20 min. All SFG spectra were measured at 20 °C. Sample Preparation. Sodium dodecyl sulfate (>99%), dodecyltrimethylammonium bromide (>99%), 3-(N, N-Dimethyldodecylammonio)propanesulfonate (>99.5%), and tetraethylene glycol monododecyl ether (>98%) were purchased from Sigma-Aldrich. Water used for solution preparation was obtained from a Millipore system (resistivity > 18.2 MΩ·cm). 44 MD Simulation. We implemented a molecular dynamics simulation using GROMACS 5.0.2 in the canonical ensemble, in which the number of particles (N), the box volume (V) and the temperature (T) are kept constant.98-100 To describe the behavior of the surfactant molecules (SDS), we used the all-atom force field parameters as provided by Shen and Sun for AMBER type potential equations.101 The SPC/E model was applied to describe the behavior of the water molecules.102 The simulation box had dimensions of 4 × 4 × 40 nm3. Initially a slab of 4 nm × 4 nm × 16 nm was filled with 8567 water molecules. This slab of water was placed at the center of the simulation box, and a vacuum of approximately 12 nm depth was at both sides of the water slab. Then, 36 SDS surfactants were randomly distributed on each side of the water interface using the PACKMOL package such that the head group pointed toward the water.103 A surface coverage of 0.44 nm2/molecule was created, corresponding to the surface coverage of SDS at its CMC. To maintain electronic neutrality of the system, 72 water molecules were randomly replaced with the 72 Na+ counterions that come with the SDS molecules. The same system configuration without surfactant molecules was used to simulate the neat water/air interface for comparison with the surfactant/water interface. The steepest descent energy minimization was conducted to prepare the system for dynamic simulation. The system temperature was maintained at 293 K using V-rescale thermostat with temperature constant, 𝜏! , equal to 0.1 ps.104 All bonds, including water OH bond, were constrained by SHAKE algorithm with a tolerance of 10-4.105 The Lennard-Jones interactions were truncated at a cut-off radius of 1.2 nm. Unlike-atom interactions were computed using 45 standard Lorentz-Berthelot combination rules.106,107 Periodic boundary conditions were applied to all three directions. Particle mesh Ewald (PME) algorithm with real cut-off radius of 1.2 nm and grid spacing of 0.12 nm was used to consider the long-range columbic interactions.108 The simulation was carried out for 120 ns with time steps of 2 fs for integration of the equations of motion. The system took 100 ns to reach its equilibrium state, therefore the last 20 ns were used for analysis. The visualizations were made by VMD 1.9.1.109 5.3 Results and discussion The decrease of water's surface tension is highly dependent on the concentration and type of the surfactant. Typically, surfactants are classified according to their head groups: nonionic, zwitterionic, anionic and cationic. The insets in Figure 5.2 show the molecular structures of four model surfactants used in the current study: tetraethylene glycol monododecyl ether (C12E4) (nonionic), 3-(N, N-Dimethyldodecylammonio) propanesulfonate (DDAPS) (zwitterionic), sodium dodecyl sulfate (SDS) (anionic), and dodecyltrimethylammonium chloride (DTAC) (cationic). To eliminate possible tail-length effects, all surfactants used in the current study have the same tail length of 12 carbons. As shown in Figure 5.2, the surface tension of water typically has little change at a low surfactant concentration (< 10-6 M). Once the surface tension starts to decrease, it quickly reaches the critical micelle concentration (CMC), at which micelles form in the bulk water, and increasing the surfactant concentration does not further decrease the 46 surface tension of water. It has been shown that these surfactants have a similar surface coverage at their CMCs, falling in the range of 42-50 Ǻ2/molecule with their surface tension in the range of 30-40 mN/m.68,110-113 Since they all have similar surface coverage and surface tension, the decrease of surface tension would appear to be well explained by the decrease of attractive force illustrated in Figure 5.1. However, the following surface entropy measurements show that surface entropy plays a significant role in determining the surface tension of water, and its effect is highly dependent on the head group. Figure 5.2. Surface tension of water surface with various concentrations of C12E4 (◆), DDAPS (▲), SDS (■), and DTAC (●).The insets are the molecular structures of the surfactants. Measurements separating the surface entropy and surface enthalpy are necessary to fully understand how surfactants affect the surface tension of water. As surface tension 47 𝛾 = 𝐻! + 𝑇 ∙ 𝑆!, the change of surface tension is described by Δγ = ΔHS -T·ΔSS at a constant temperature T. While the presence of surfactants weakens the interactions between surface molecules (ΔHS < 0), the decrease in surface entropy (ΔSS < 0 or -T·ΔSS > 0) produces a positive contribution to Δγ , which counteracts the enthalpy term. Experimentally, the surface entropy of water can be measured using the thermodynamic identity SS = − !"!" !,!,114 and subsequently the surface enthalpy can be calculated using 𝐻! =  𝛾 + 𝑇 ∙ 𝑆!. The measured values of water’s 𝐻! and 𝑆! for various surfactants at their CMCs are summarized in table 5.1. While the surfactants decreased the surface enthalpy by ~50-70%, water’s surface entropy could drop to near zero or even negative values for the ionic surfactants. Because surface entropy is a surface excess property, zero surface entropy indicates that the motions of molecules near the surface of water have become as constrained as those in the bulk water. Table 5.1. Surface tension γ, surface entropy SS, -TSS, and surface enthalpy HS of water at T=293 K with various surfactants at their CMC concentrations. surfactantγ (mN/m)S S(mN/m•K)-TS S (mN/m)H S (mN/m)pure water 72.7 ± 0.2 0.15 ± 0.02 -44 ± 6 117 ± 6PEO (non-ionic) 30.2 ± 0.2 0.10 ± 0.02 -29 ± 6 59 ± 6DDAPS (zwitterionic) 41.6 ± 0.2 0.07 ± 0.02 -20 ± 6 62 ± 6SDS (anionic) 40.6 ± 0.2 0.04 ± 0.02 -12 ± 6 52 ± 6DTAB (cationic) 41.6 ± 0.2 -0.02 ± 0.02 6 ± 6 36 ± 6 48 To study the nature of this extremely low entropy state of water surface, phase-sensitive SFG was used to probe the molecular ordering and orientation at the water surface in the presence of these four surfactants. In contrast to the conventional SFG spectroscopy, which only measure the amplitude of the second-order nonlinear susceptibility |χ(2)|, phase-sensitive SFG measures the spectra of the imaginary χ(2), Im(χ(2)), which reveals both the ordering (the amplitude of peaks) and orientations (the sign of peaks) of surface molecules.47,79,80 Figure 5.3 shows the Im(χ(2)) spectra of the water surface in the CH and OH regions near the CMCs of the four surfactants. As expected, the CH spectra of the surfactants in Figure 5.3a are similar, which agree with measurements using neutron reflection showing that the surface coverage of these surfactants are similar (42-50 Ǻ2/molecule).68,110-113 In contrast, the OH spectra of water shown in Figure 5.3b are dramatically different. In a OH spectrum, the Im(χ(2)) can be positive or negative, depending on the sign of the averaged OH projection with respect to the surface normal: a positive peak indicates water molecules with the hydrogen pointing toward the air (up), and a negative peak indicates the hydrogen pointing down.47 The ionic surfactants SDS and DTAC induced a much more ordered water structure at the surface, indicated by the magnitudes of the SFG peaks. The opposite charges carried by SDS and DTAC flip the water molecules such that their dipoles point in opposite directions, as indicated by the sign of the peaks. In thermodynamics, an ordered state has lower entropy. Therefore, the extremely low entropy state of water surface in the presence of SDS or DTAC is associated with the ordered water structure induced by the 49 charged surfactant. Figure 5.3. (a) Phase-sensitive SFG spectra of various surfactants on water at their CMCs. The peaks near 2875 cm-1 and 2930 cm-1 are the CH3 symmetric stretch and Fermi resonance. The CH peaks appear negative when the CH3 pointing up.60 (b) The corresponding SFG spectra of water. The OH peaks appear negative when the OH bonds pointing down and positive when the OH bonds pointing up. MD simulations were carried out to gain a further molecular-level understanding of the low-entropy water surface. SDS was chosen for MD simulations because it is the most studied surfactant with well-tested force fields. Figure 5.4a shows the final configuration of SDS and water molecules after 100 ns of energy optimization. Figure 5.4b shows the average orientation of water’s dipoles cos𝜃 as a function of depth (z), where the surface normal of water is defined as θ = 0. Figure 5.4b shows that, in the 50 absence of SDS, surface water molecules are preferentially ordered over a depth of only 1 nm (blue curve). In the presence of SDS, the ordered water structure propagates over 5 nm into water surface (red curve). Additionally, the MD simulation showed that adsorption of SDS on the water surface promotes the formation of hydrogen bonds (HBs) near the water surface. Figure 5.4c shows that the averaged number of HBs for each water molecule is ~3.2 in the bulk water. At the pure water surface, the averaged number of HBs per water molecule can be as low as 2 because water molecules at the top layer have a limited number of neighbors available for HB formation. In the presence of SDS, the number of HBs increased to ~2.5. The increased number of HBs also contributes to the reduction of entropy as the degrees of freedom decreases. Overall, both the increase in the ordering of water molecules and the formation of additional HBs lead to lower surface entropy. 51 Figure 5.4. (a) Structures of water and SDS in the MD simulation. (b) Water dipole order parameter where water’s surface normal is defined as θ = 0. (c) The averaged number of hydrogen bonds per water molecule vs. the depth. 5.4 Conclusions We studied the surface entropy of water and the effects of four different surfactants by separately measuring the enthalpy and entropy of water surfaces. All surfactants induced a significant surface entropy decrease that counteracts the enthalpic effect. Interestingly, both the anionic and cationic surfactants lowered the surface entropy of water to near zero. SFG vibrational spectroscopy showed that this extremely low entropy state of water surface was associated with the surfactant-induced ordering of surface water molecules. Additionally, MD simulation indicated that ionic surfactants promoted 52 the formation of hydrogen bonds near water surfaces. Both effects lead to the reduction of water’s surface entropy, which play a critical role in determining the surface tension of water. 53 Chapter 6 Surface Charge at Bitumen/Water Interface - Effects of pH, Ions, and Surfactants 6.1 Introduction Bitumen chemistry has been gaining a great deal of interest because there are large bitumen deposits available for development to supply the future energy needs. Bitumen is a complex mixture of hundreds of organic molecules. Its molecular structure and composition are not well understood but studies have suggested that bitumen is composed primarily of polyaromatic hydrocarbons with heteroatoms such as oxygen, nitrogen, and sulfur.115,116 Properties of bitumen/water interface affect many processes in bitumen production, including crude oil extraction, transportation, treatment, and tailing management.117 A better understanding of bitumen/water interface chemistry is needed for developing more environmental-friendly and cost-effective industrial processes. Water characteristics, such as its alkalinity, ion concentrations, and surfactants contents, affect the properties of water/bitumen interfaces.118 Early studies, which measured the electrophoretic mobilities of bitumen particles in aqueous solutions, suggested that the acid groups at the bitumen surface deionized in water and gave rise to negative surface charges.119 Acevedo et al. showed that salts, such as sodium carbonate, lowered the interfacial tension at bitumen/water interfaces.120 Studies using interfacial rheometry showed that surfactants interacted with bitumen.121,122 Surfactants such as 54 sodium dodecyl sulfate (SDS) and dodecylamine hydrochloride were reported to adsorb onto the interface through the hydrophobic effect, which also affect the zeta potential of the bitumen/water interface.123,124 However, these macroscopic studies provided little microscopic information about bitumen/water interfaces.125-128 In this chapter, we investigate the structures of water on bitumen surfaces in responses to pH, ions, and surfactants using a new approach. Sum frequency generation (SFG) is known for its high surface sensitivity because under the electric-dipole approximation the second-order optical process is forbidden in a centrosymmetric medium, such as bulk water. Vibrational spectroscopic studies at bitumen/water interfaces are difficult because both water and bitumen are highly absorptive in the IR region. We overcame this difficulty by placing a bitumen film on the water surface, which caused minimal IR absorption and allowed the IR laser to access the buried interface. Previous SFG studies on bitumen showed that the vibrational peaks in the CH region consist of three peaks at 2850 cm-1, 2870 cm-1 and 2935 cm-1, corresponding to the CH2 symmetric stretch, CH3 symmetrical stretch, and CH3 Fermi resonance, respectively.129 However, the CH peaks did not provide much information on the bitumen surface. We found that vibrational peaks of water in the OH region are highly sensitive to the pH, ion, and surfactant concentrations, which provided a better understanding of the interaction between water and bitumen. Additionally, recent developments in phase-sensitive SFG have enabled measurements of the imaginary part of the 55 second-order nonlinear optical susceptibility, i.e., Im(𝜒!""(!) ), which characterizes both the surface’s vibrational resonances and the averaged orientation of the functional groups.9,46 This new optical technique provides an opportunity to gain molecular-level information at the buried bitumen/water interface. 6.2 Experimental Section Material and Sample Preparation. NaCl (>99%), CaCl2 (>99%), sodium dodecyl sulphate (98.5%) and dodecyl trimethyl ammonium chloride (DTAC) (99%) were purchased from Sigma Aldrich. Water used in all experiments was obtained from a Millipore system (18.2 MΩ·cm, pH ≈ 6). Water of pH = 9 was prepared by adding NaOH (Sigma-Aldrich, 99%) into pure water. Bitumen films on water were prepared using 1 g bitumen (Athabasca, Alberta, density 1.012 g/mL at 20 ˚C) dissolved in 15 mL of toluene (Fischer Scientific). Approximately 0.015 mL of the bitumen-toluene solution was dropped on the surface of water (in a Petri dish with a diameter of 8 cm) to form a bitumen film. The film was then kept unperturbed for 20 min allowing the toluene to evaporate.130 Based on the volume of the bitumen and the surface area of the film, the film thickness was estimated to be ~ 0.2 µm. At this thickness, the absorption of SFG is negligible (~3%). We have compared the spectra of the sample 20 min and 1 hour after the preparation, and no significant difference was observed. All experiments were performed at room temperature (20±1°C). Phase-sensitive SFG setup. A femtosecond Ti-sapphire laser (120 fs, 800 nm, 1 kHz and 56 2 mJ/pulse) was used to pump an optical parametric amplifier to generate a femtosecond IR beam. The IR beam and a narrow-band picosecond 800 nm beam were aligned in a collinear optical path with an incident angle of 60°.11,131 A reference SFG was obtained by focusing the IR and picosecond 800 nm beams into a 50 µm-thick quartz crystal. The IR, picosecond 800 nm, and time-delayed reference SFG beams were again focused on the sample. The reference SFG and the SFG generated at the sample went through a polarizer, a band pass filter, a lens, and a monochromator. The SFG interference pattern was recorded by a charge-couple device (CCD) camera. In this setup, the spectral bandwidth of the SFG spectra was limited by the spectral bandwidth of the femtosecond IR laser. The polarization combination used in this study is SSP (s-polarized SFG, s-polarized 800 nm and p-polarized IR). The energies of the 800 nm and IR beams were ~10 µJ/pulse and ~3 µJ/pulse, respectively. Each spectra presented in this chapter was acquired over a period of 10 min and averaged over four different spots at the surface. 6.3 Results and Discussion Figure 6.1 shows the SFG spectra of bitumen/water interfaces at pH = 6 (pure water) and 9. The surface properties of bitumen at pH = 9 are particularly important because large-scale industrial bitumen extraction has been carried out at pH ~ 9.132,133 The spectra in Figure 6.1 show two major features: a broad positive peak centered near 3150 cm-1 and a relatively small negative peak near 3500 cm-1. It has been proposed that the low frequency band comes from a strongly hydrogen-bonded water structure while the high 57 frequency band originates from a weakly hydrogen-bonded water structure.5,134,135 In phase-sensitive SFG vibrational spectroscopy, the sign of a peak indicates the projection of the averaged OH vibrational transition moment with respect to the surface normal.46 For the OH symmetric stretch, the Im(χ(2)) can be either positive or negative. A positive peak indicates water molecules with the hydrogen pointing up (toward the bitumen surface), and a negative peak indicates the OHs pointing down (toward the bulk water).48 Since SFG is forbidden in a centrosymmetric medium, the amplitude of the SFG peak measures the ordering of water molecules. In Figure 6.1, the spectra of water at the bitumen/water interface are dominated by the positive peak near 3150 cm-1, indicating the bitumen surfaces are overall negatively charged and induce an ordered water structure near the bitumen surface with the averaged OH orientation pointing toward the bitumen. At pH = 6, the spectrum (Figure 1a) is very similar to the spectrum of a negatively charged palmitic acid layer on water.86 Therefore, our results agree with previous studies suggesting that the hydrolysis of acidic groups in the bitumen, such as naphthenic acids, create the negative charges at the interface.136 It has also been proposed that the negative charges might originate from the dissociation of heteroatom-containing groups, such as -SOH.137,138 A higher pH value, which facilitates the dissociation of more acidic groups, increases the negative surface charge of bitumen and enhances the ordering of water, as indicated by the higher SFG intensity in Figure 6.1b. Additionally, hydroxide adsorption may also increase the negative charge in the bitumen surface.139 Because mineral surfaces also carry negative surface charges, it partially explains why a higher pH promotes the 58 detachment of bitumen from mineral surfaces.140 Since the industrial bitumen extraction operates at pH = 9 to improve bitumen recovery, the following studies will focus on the interfacial properties of bitumen at pH = 9. Figure 6.1. Im(𝜒!""(!) ) spectra of bitumen/water interfaces in the OH vibrational region at (a) pH = 6 and (b) pH = 9. The interactions between the negatively charged bitumen surface and cations were studied at pH = 9 using solutions of NaCl and CaCl2. While Na+ is an indifferent ion, Ca2+ is known to be a potential-determining ion.141 Figure 6.2a shows the spectra of water at the bitumen/water interface with various concentrations of NaCl. Overall, an increase in the salt concentration decreases the ordering of the interfacial water, which can be explained by the charge screening of bitumen’s surface charge. This results in agreement with previous surface tension studies suggesting that Na+ cations interact with the dissociated acid groups and form the corresponding salt, such as –COONa;120 however, 59 the SFG spectra could not provide direct evidence for the formation of the complex. The OH peak decreased significantly with 0.2 mM NaCl, and the ordered interfacial water structure nearly disappeared with 100 mM of NaCl. Figure 6.2b shows that Ca2+ also results in a decrease of the SFG intensity. A comparison between Figures 6.2a and 6.2b shows that Ca2+ is more effective than Na+ for the surface charge screening because of its higher charge. With 50 mM of CaCl2, the ordered water structure at the bitumen/water interface nearly disappeared. Figure 6.2. Im(𝜒!""(!) ) spectra of bitumen/water interface in the OH vibrational region with (a) NaCl concentrations at 0 (1), 0.2 mM (2), 10 mM (3) , and 100 mM (4) and (b) CaCl2 concentrations at 0 (1), 0.1 mM (2), 5 mM (3), and 50 mM (4). All spectra were taken at 60 pH = 9. SDS and DTAC were used as model anionic and cationic surfactants, respectively, to study their interactions with the bitumen surface. Figure 6.3 shows the SFG spectra of water at the bitumen/water interfaces with no surfactants (Figure 6.3a), 1 mM of SDS (Figure 6.3b), and 1 mM of DTAC (Figure 6.3c). Because both the bitumen and SDS have a negative net charge, it would be expected that the SDS experienced a repulsive force from the bitumen. In contrast to this expectation, Figure 6.3b shows that an enhancement of water’s positive peak was observed in the presence of SDS, indicating the bitumen surface became more negatively charged. The SFG spectra are consistent with previous oil-drop-detachment studies proposing that SDS adsorbed onto bitumen surfaces via its hydrophobic tail.123 However, it should not be ruled out that local positive charges might exist on the surface of bitumen. For example, it has been proposed that ionization of groups containing either nitrogen (e.g. amines, pyrroles, amides, pyridines, etc.) or sulfur (e.g. thiophene, sulfoxide, etc.) are possible origins of the surface charge on bitumen.137  The pKa of protonated amines is ~10, and they may remain positively charged at pH 9. These local positive charges would attract the anionic surfactants. If they exist, the small amount of local positive charges would not be observable using macroscopic techniques, such as zeta potential and electrophoresis measurements because these techniques measure the net charge of the surface. On the other hand, the adsorption of the cationic surfactant DTAC was favorable by both the electrostatic attraction force 61 and the hydrophobic interaction. A surface charge reversal was observed, indicated by the sign change of Figure 6.3c compared to Figure 6.3a and 6.3b. Figure 6.3. Im(𝜒!""(!) ) spectra of bitumen/water interface in OH vibrational region with (a) water, (b) 1 mM SDS, (c) 1 mM DTAC in the aqueous subphase. All spectra were taken at pH = 9. 6.4 Conclusions The interfacial properties of bitumen/water interfaces were studied using phase-sensitive SFG vibrational spectroscopy. Bitumen surfaces in water were overall negatively charged, aligning the water molecules in a more ordered structure with the OH on average pointing toward the bitumen. The surface charge of bitumen increased when the pH of water increased. The presence of salt in water neutralized the surface charge of bitumen and nearly destroyed the ordered structure of water on bitumen. Ca2+ was found to be more effective than Na+ in screening the bitumen surface charge. Both the anionic 62 surfactant SDS and the cationic surfactant DTAC interacted with the bitumen. While anionic surfactant slightly increased the negative charge of bitumen, the cationic surfactant strongly interacted with the bitumen and produced a surface charge reversal. 63 Chapter 7 Surface Charge Reversal of Ionic Liquid at the Air/Water Interface 7.1 Introduction Room temperature ionic liquids (ILs) are salts in liquid phase at room temperature. They usually consist of organic cations and inorganic or organic anions. Most ILs are thermally stable, nonvolatile, environmentally friendly, and recyclable. These fascinating properties have encouraged scientists to use ILs as alternative solvents or media in fields such as organic synthesis, catalysis, solar cell system research, and gas capture processes.142-145 Because many applications take place at surfaces or interfaces, it is important to gain fundamental understandings of their surface or interface behaviors. Ionic liquids, such as 1-butyl-3-methylimidazolium tetrafluoroborate ([BMIM][BF4]), are composed of an inorganic anion and an amphiphilic cation as illustrated in Figure 7.1. Pure ILs have cations and anions distributed with charge neutrality at the surface.146 When they are mixed with water, their surface properties are similar to surfactants. For example, Sung et al showed that imidazolium cations formed a monolayer on the water surface, and the surface tension of water decreased dramatically at an IL concentration as low as ~0.001 M.147 The surface propensity of imidazolium ions was explained as the cation’s hydrophobic butyl chain extended into the air.148-150 On the other hand, the distribution of inorganic ions, such as BF4-, was interpreted 64 thermodynamically as being repelled from the water surface and staying in the bulk or sublayer.151 However, molecular dynamic simulations by Jungwirth et al. proposed that some anions, for instance Br- and I-, were more populated at the water surface relative to the bulk.152 Complex ions, such as SO42-, NH4+, and NO3-, have been experimentally demonstrated to be of surface propensity.63 Thus, the behaviors of imidazolium cations and counter anions at the air/water interface need careful investigation. However, previous research has focused on the surface behaviors of imidazolium cations whereas the anion behaviors remain unclear.153,154 Here we use phase-sensitive SFG vibrational spectroscopy to observe the water molecules at the air/ionic liquid solution interface in order to investigate the ions’ distribution at the interface. When ions are present at the air/water interface, the electrostatic field generated by the accumulated ions induces the OH dipoles of water molecules to change orientation and bonding, consequently affecting the interfacial water network. Phase-sensitive SFG vibrational spectroscopy is able to measure the imaginary part of second order susceptibility, Im(χ(2)), that contains the resonance information of water molecules. The sign of Im(χ(2)) indicates the average orientation of OH dipoles. For water, a negative sign indicates OH dipoles with hydrogen pointing toward the bulk (OH down) while a positive sign indicates OH dipoles with hydrogen pointing toward the air (OH up). In this study, we will show that, unlike traditional salts or surfactants, the imidazolium cation [BMIM+] and its counter-ion [BF4-] exhibit a surface charge reversal 65 as the concentration increases: [BMIM+] starts to accumulated at water surface at a low concentration, but the amount of [BF4-] exceeded that of the cation with increasing bulk concentration, resulting in a negatively charged surface. This result indicates that both [BMIM+] and [BF4-] have surface propensities, which are influenced by the concentration. Figure 7.1. Molecular structure of 1-butyl-3-methylimidazolium tetrafluoroborate ([BMIM][BF4]). 7.2 Experimental section Sample preparation. 1-butyl-3-methylimidazolium tetrafluoroborate (98%) was purchased from Sigma-Aldrich and used as received. IL/water binary solutions were prepared by mixing the desired mass of IL and water in a volumetric flask. Phase-sensitive SFG setup. A femtosecond Ti-sapphire laser (120 fs, 800 nm, 1 kHz and 2 mJ/pulse) was used to pump an optical parametric amplifier to generate a femtosecond IR beam. The IR beam and a narrow-band picosecond 800 nm beam were aligned in a collinear optical path with an incident angle of 60°.11,131 A reference SFG was obtained by focusing the IR and picosecond 800 nm beams into a 50 µm-thick quartz crystal. The 66 IR, picosecond 800 nm, and the time-delayed reference SFG beams were again focused on the sample. The reference SFG and the SFG generated at the sample went through a polarizer, a band pass filter, a lens, and a monochromator. The SFG interference pattern was recorded by a charge-couple device (CCD) camera. The polarization combination used in this study is SSP (s-polarized SFG, s-polarized 800 nm and p-polarized IR). The energies of the 800 nm and IR beams were ~10 µJ/pulse and ~3 µJ/pulse, respectively. Each spectra presented in this chapter was acquired over a period of 20 min. 7.3 Results and discussion Figure 7.2 shows the water spectra of the air/IL solution interface with the IL mole fraction (XIL) ranging from low to high. At XIL = 0.00018 ( ≈ 1×10-5 M ), the spectrum displays two negative broad bands at ~3200 cm-1 and ~3400 cm-1 that originated from the OH vibrations of water. There is currently no consensus about the origin of these two bands so far. Shen et al. assigned the 3200 cm-1 band to an “ice-like” water structure and the 3400 cm-1 ban to a “liquid-like” water structure because the wavenumbers overlap with those of ice and liquid water, respectively. Nihonyanagi et al. reported that at a highly charged interface, the two broad OH bands result from the splitting of intramolecular and/or intermolecular couplings.155 However, it is questionable whether the water molecules at the surface of IL aqueous solution behave similarly. It is commonly accepted that the lower frequency band stems from strongly bonded water molecules while the higher frequency band from weakly bonded water molecules. The 67 negative sign indicates that the surface water OH dipoles are pointing down (hydrogens pointing toward the bulk water); that is, the imidazolium cations are present on the water surface, making the surface positively charged. The imidazolium cation carrying a butyl chain makes it less energetically favorable to stay in the bulk compared with the anions; therefore, they move to the surface at low concentration. With an increase in the IL mole fraction to XIL = 0.0005 ( ≈ 0.027 M), the two OH vibrational bands become more negative. This suggests the water surface is populated with more cations, creating more water molecules oriented with OH dipoles pointing downward. In the CH spectra shown in Figure 7.3, weak CH vibrational peaks become detectable at XIL = 0.0005. The peaks at 2875 cm-1, 2930 cm-1, and 2970 cm-1 are assigned to the symmetric, Fermi, and asymmetric vibrations of methyl groups on the butyl chain. The stretches of the methyl group attached to the aromatic nitrogen atom as well as the C-H bonds on the aromatic ring are not observable, which might be due to a disorder of the aromatic rings at the air/water interface.153 The trend of CH the spectra is consistent with the OH spectra in that the amount of imidazolium cations increased from XIL = 0.00018 to 0.0005. 68 Figure 7.2. Water spectra of air/IL aqueous solution interface at IL mole fraction = 0.00018 (magenta), 0.0005 (purple), 0.004 (orange), 0.02 (cyan), 0.04 (blue), 0.2 (green). The color of each concentration corresponds to the same concentration as in Figure 7.3. However, a further increase of the IL mole fraction in the bulk to XIL = 0.004 results in a significant decrease in the amplitudes of OH vibrational bands. This decrease may be due to the fact that the imidazolium cations depleted from the surface or that the [BF4-] anions were present at the surface, which counteracted the positive charges. Both cases can lead to a decrease in the net positive charges. Figure 7.3 shows a significant increase in the cation population, as indicated by the enhancement in CH peaks at XIL = 0.004. It is therefore reasonable to claim that the anions aggregate at the water surface at this concentration. Because of the decrease in net positive charges, less water molecules 69 oriented with OH pointing downward. Our results are not consistent with those reported by Sung et al. Their work concluded that the anions started to appear on the surface at a much higher mole fraction, XIL ≈ 0.016, which corresponds to the breakpoint from decrease to increase in the surface tension plot of the IL aqueous solution.147 Nevertheless, the surface tension can only provide macroscopic information about the surface property. The SFG results show that a significant amount of the anions accumulated on the surface at a mole fraction as low as XIL = 0.004. The combination of the CH and OH spectra suggests that the [BF4-] anion population increases faster than that of the imidazolium cation at a higher concentration. Figure 7.3. CH vibrational spectra of air/IL aqueous solution interface at IL mole fraction = 0.00018 (magenta), 0.0005 (purple), 0.004 (orange), 0.02 (cyan), 0.04 (blue), 0.2 (green). 70 Interestingly, a positive band around ~ 3150 cm-1 is detected at X = 0.02, as illustrated in Figure 7.2, meaning that the interfacial water molecules were realigned with OH dipoles up toward the air. This indicates that the surface number density of [BF4-] is higher than that of [BMIM+], thus the water surface is negatively charged. The vibrational peak at ~3150 cm-1 is attributed to the strongly hydrogen-bonded water structure. The low vibrational intensity might be due to the fact that fewer water molecules are present at the liquid surface at a low water mole fraction. The increased amount of ions on the surface might disrupt the ordered water structure, which may also lead to a low OH peak intensity. As shown in Figure 7.2, three OH peaks are present in the spectrum at XIL = 0.04 (green). The peaks at 3200 cm-1 and 3350 cm-1 are assigned to the strongly bonded and loosely bonded water molecules at the solution surface populated with BF4-. The broadening of the positive OH bands compared with the XIL = 0.02 spectrum demonstrates that more water molecules are realigned with OH upward orientation and that more BF4- anions are populated on the surface. The weak peak at ~3500 cm-1 has been proposed to stem from the “free” water molecule that coordinates between two anions, forming a complex BF4-···HOH···BF4- through hydrogen bonds.156 This model also supports our finding that a significant amount of anions aggregate at the liquid surface. In the spectrum of XIL = 0.2, the water vibration is nearly undetectable due to the low water number density at the surface. The surface is mostly covered by IL ions. 71 Figure 7.4. Illustrative scheme on the water molecules orientation and population of imidazolium cations and BF4- anions at (a) low and (b) high IL mole fraction. Figure 7.3 shows that the surface number density of imidazolium cations increases significantly at X = 0.02 and is close to saturation at X = 0.04. This might be because that the bulky aromatic ring limits the imidazolium ions’ access to the surface. The presence of CH asymmetric vibration of the methyl groups on imidazolium cations might indicate that the methyl groups tilted away from surface normal. Previous work by Rivera-Rubero et al. demonstrated that the average angle between surface normal and methyl groups was about 40° to 50°.149 72 Given that the imidazolium cations surface number density exhibits an increasing trend, the negatively charged surface occurs because the BF4- anion surface number density is higher than that of the cation at a higher IL mole fraction (X > 0.04). In other words, surface populations of cation and anion reverse as IL mole fraction increases. It might be surprising that the inorganic anion is more surface-active than the cation bearing a hydrophobic butyl group. For example, the surfactants, which consist of similar parts, are known to present the amphiphilic part (containing a hydrophobic tail and a hydrophilic head) at the water surface for all concentrations, while the counter-ions stay in the bulk. However, this rule does not apply for the IL aqueous solution. A reasonable explanation is that, at a low IL mole fraction, the cations bearing butyl groups are more hydrophobic and have easier access to the surface in comparison with BF4-; at a high IL mole fraction, when the cations’ access is limited by the bulky rings, the small BF4- anions can stay in the space between imidazolium cations, making the overall surface negatively charged (see Figure 7.4). 7.4 Conclusions In summary, phase sensitive SFG vibrational spectroscopy has provided informative pictures of the IL distribution at the air/water interface. Our results show that both [BMIM] and [BF4-] have surface propensity. At diluted concentration, the imidazolium cations tend to approach the water surface with the butyl chain extending to the air, making the water surface positively charged. As the [BMIM][BF4-] mole fraction is 73 increased, the cations accumulate at the surface. On the other hand, the [BF4-] anions start to distribute at the water surface but the surface number increases faster than that of cations, making the water surface less positive. At a higher [BMIM][BF4-] mole fraction, more anions reside at the water surface, and the water surface is negatively charged. This population reversal of cation/anion might occur because the small anions are able to accommodate between the cations and aggregate at the water surface. 74 Chapter 8 Conclusion This dissertation presents studies of air/water interfaces in the presence of charged molecules using phase-sensitive SFG; studies include the interaction of water and polyelectrolyte, the water structure at the surface with co-adsorption of a polyelectrolyte/surfactant mixture, the effect of water’s ordering and orientation on surface tension, interfacial charge at the bitumen/water interface, and the ionic liquid distribution at the air/water interface. At a neat air/water interface, the OH groups of water primarily pointed downward toward the bulk. However, the presence of the polyelectrolyte caused the water molecules to flip, with the hydrogen being oriented upward (Chapter 3). The addition of the monovalent cation Na+ to the polyelectrolyte solution did not affect the conformation of the polyelectrolyte, though the water molecules flipped back to a structure similar to that of a neat air/water interface, with the hydrogen pointing down. On the other hand, the addition of divalent cation Ca2+ not only changed the conformation of the polyelectrolyte but also completely destroyed the ordered structure of interfacial water molecules. The observed phenomena are consistent with the formation of the polymer-ion complex, in which a divalent cation can interact with two intra-chain or inter-chain charges and form a jointing point. Surfactant and polyelectrolyte mixtures can cause complex surface tension 75 behaviors that are challenging to interpret. For example, an aqueous solution of sodium dodecyl sulfate (SDS) and poly-(diallyldimethylammonium chloride) (PDADMAC) shows dramatic changes in surface tension when the concentration of SDS varies (Chapter 4). The dramatic surface tension decrease and recovery of the SDS-PDADMAC mixtures were discovered to be the results of a surface charge reversal. Similar surface charge reversal was also observed in cationic surfactant and anionic polyelectrolyte mixtures. The effect of surfactants on lowering the water surface tension has been studied through separate measurement of the surface entropy and surface enthalpy in the presence of four types of surfactant head groups: nonionic,  zwitterionic,  anionic,  and  cationic (Chapter 5). While  all  these  surfactants  decreased  the  surface  enthalpy  of  water  by  50-­‐70%,  ionic  surfactants  make  the  surface  entropy  of  water  drop  to  near  zero  or  even  negative  values.  Phase-­‐sensitive  SFG  vibrational  spectroscopy  and  molecular  dynamics  (MD)  simulations  suggested  that  the  zero-­‐entropy  state  of  the  water  surface  was  associated  with  the  enhanced  water  ordering  induced  by  surfactants  and  an  increased  number  of  hydrogen  bonds.  Both  effects  reduce  water  molecules'  degrees  of  freedom  for  motion  and  consequently  lower  the  surface  entropy  of  water.  The  ability  of  a  surfactant  to  decrease  the  surface  entropy  of  water  is  in  the  following  order:  ionic  >  zwitterionic  >  nonionic.  Phase-sensitive SFG was applied to study the bitumen/water interface. Water 76 molecules at the bitumen/water interface form a well-ordered structure due to net negative charges at the bitumen surface, which aligned water molecules with hydrogen pointing toward the bitumen (Chapter 6). The ordering of water was enhanced at the industry-relevant pH value of 9. The presence of salt in water neutralized the surface charge of the bitumen and nearly destroyed the ordered structure of water on the bitumen. Ca2+ was found to be more effective than Na+ in screening the bitumen surface charge. Both the anionic surfactant SDS and the cationic surfactant DTAC interacted with the bitumen. While the anionic surfactant slightly increased the negative charge of bitumen, the cationic surfactant strongly interacted with the bitumen and produced a surface charge reversal. The water structure and orientation at the surface of the ionic liquid-water mixture were studied in various ionic liquid mole fractions (Chapter 7). At the air/1-butyl-3-methylimidazolium tetrafluoroborate ([BMIM][BF4]) aqueous solution interface, we observed surface charge reversal as the IL mole fraction gradually increased. The imidazolium cations resided at water surface at a low IL mole fraction, which realigned the water molecules with hydrogen toward the bulk. However, with an increased IL mole fraction, the surface number of [BF4-] anions increased, making the surface negatively charged, with water molecules orienting the hydrogen toward the air. The significant increase in the [BF4-] surface population might be due to the fact that the small anions could fit in the space between imidazolium rings in the higher IL mole 77 fraction, whereas the cations’ access to the surface was limited by the bulky imidazolium ring. 78 References   (1)   Du,  Q.;  Superfine,  R.;  Freysz,  E.;  Shen,  Y.  R.  Vibrational  Spectroscopy  of  Water  at  the  Vapor  Water  Interface.  Phys.  Rev.  Lett.  1993,  70,  2313-­‐2316.     (2)   Hiemenz,  P.  C.;  Rajagopalan,  R.  Principles  of  colloid  and  surface  chemistry,  3rd  ed.;  Marcel  Dekker:  New  York,  1997.     (3)   Thomas,  R.  K.  Neutron  reflection  from  liquid  interfaces.  Annu.  Rev.  Phys.  Chem.  2004,  55,  391-­‐426.     (4)   Mendelsohn,  R.;  Mao,  G.  R.;  Flach,  C.  R.  Infrared  reflection-­‐absorption  spectroscopy:  Principles  and  applications  to  lipid-­‐protein  interaction  in  Langmuir  films.  BBA-­‐Biomembranes  2010,  1798,  788-­‐800.     (5)   Shen,  Y.  R.;  Ostroverhkov,  V.;  Chen,  E.  C.  Y.;  Ji,  N.;  Waychunas,  G.  Phase-­‐sensitive  sum-­‐frequency  vibrational  spectroscopic  studies  of  water  interfaces.  Abstr.  Pap.  Am.  Chem.  Soc.  2006,  231.     (6)   Hunt,  J.  H.;  Guyotsionnest,  P.;  Shen,  Y.  R.  Observation  of  C-­‐H  Stretch  Vibrations  of  Monolayers  of  Molecules  Optical  Sum-­‐Frequency  Generation.  Chem.  Phys.  Lett.  1987,  133,  189-­‐192.     (7)   Stiopkin,  I.  V.;  Jayathilake,  H.  D.;  Bordenyuk,  A.  N.;  Benderskii,  A.  V.  Heterodyne-­‐detected  vibrational  sum  frequency  generation  spectroscopy.  J.  Am.  Chem.  Soc.  2008,  130,  2271-­‐2275.     (8)   Zhu,  X.  D.;  Suhr,  H.;  Shen,  Y.  R.  Surface  Vibrational  Spectroscopy  by  Infrared-­‐Visible  Sum  Frequency  Generation.  Phys.  Rev.  B  1987,  35,  3047-­‐3050.     (9)  Shen,  Y.  R.  Basic  Theory  of  Surface  Sum-­‐Frequency  Generation.  J.  Phys.  Chem.  C  2013,  117,  11884-­‐11884.     (10)   Shen,  Y.  R.  The  principles  of  nonlinear  optics,  Wiley  classics  library  ed.;  Wiley-­‐Interscience:  Hoboken,  N.J.,  2003.     (11)   Ji,  N.;  Ostroverkhov,  V.;  Chen,  C.  Y.;  Shen,  Y.  R.  Phase-­‐sensitive  sum-­‐frequency  vibrational  spectroscopy  and  its  application  to  studies  of  interfacial  alkyl  chains.  J.  Am.  Chem.  Soc.  2007,  129,  10056-­‐+.     (12)   Shen,  Y.  R.  Phase-­‐Sensitive  Sum-­‐Frequency  Spectroscopy.  Annu.  Rev.  Phys.  Chem.  2013,  64,  129-­‐150.     (13)   Forster,  S.;  Schmidt,  M.  Polyelectrolytes  in  Solution.  Adv.  Polym.  Sci.  1995,  120,  51-­‐133.     (14)   Barrat,  J.  L.;  Joanny,  J.  F.  Theory  of  polyelectrolyte  solutions.  Adv.  Chem.  Phys.  1996,  94,  1-­‐66.     (15)   Wever,  D.  A.  Z.;  Picchioni,  F.;  Broekhuis,  A.  A.  Polymers  for  enhanced  oil  recovery:  A  paradigm  for  structure-­‐property  relationship  in  aqueous  solution.  Prog.  Polym.  Sci.  2011,  36,  1558-­‐1628.     (16)   Bolto,  B.;  Gregory,  J.  Organic  polyelectrolytes  in  water  treatment.  Water  Res.  2007,  41,  2301-­‐2324.   79   (17)   Yang,  S.  Y.;  Rubner,  M.  F.  Micropatterning  of  polymer  thin  films  with  pH-­‐sensitive  and  cross-­‐linkable  hydrogen-­‐bonded  polyelectrolyte  multilayers.  J.  Am.  Chem.  Soc.  2002,  124,  2100-­‐2101.     (18)   Bolto,  B.;  Gregory,  J.  Organic  polyelectrolytes  in  water  treatment.  Water  Res.  2007,  41,  2301-­‐2324.     (19)   Yang,  S.  Y.;  Rubner,  M.  F.  Micropatterning  of  polymer  thin  films  with  pH-­‐sensitive  and  cross-­‐linkable  hydrogen-­‐bonded  polyelectrolyte  multilayers.  J.  Am.  Chem.  Soc.  2002,  124,  2100-­‐2101.     (20)   Fang,  Y.  P.;  Al-­‐Assaf,  S.;  Phillips,  G.  O.;  Nishinari,  K.;  Funami,  T.;  Williams,  P.  A.  Binding  behavior  of  calcium  to  polyuronates:  Comparison  of  pectin  with  alginate.  Carbohyd.  Polym.  2008,  72,  334-­‐341.     (21)   Mattison,  K.  W.;  Wang,  Y.  F.;  Grymonpre,  K.;  Dubin,  P.  L.  Micro-­‐  and  macro-­‐phase  behavior  in  protein-­‐polyelectrolyte  complexes.  Macromol.  Symp.  1999,  140,  53-­‐76.     (22)   Hamman,  J.  H.  Chitosan  Based  Polyelectrolyte  Complexes  as  Potential  Carrier  Materials  in  Drug  Delivery  Systems.  Mar.  Drugs  2010,  8,  1305-­‐1322.     (23)   Peyratout,  C.  S.;  Dahne,  L.  Tailor-­‐made  polyelectrolyte  microcapsules:  From  multilayers  to  smart  containers.  Angew.  Chem.  Int.  Ed.  2004,  43,  3762-­‐3783.     (24)   Decher,  G.;  Eckle,  M.;  Schmitt,  J.;  Struth,  B.  Layer-­‐by-­‐layer  assembled  multicomposite  films.  Curr.  Opin.  Colloid  In.  1998,  3,  32-­‐39.     (25)   Hammond,  P.  T.  Recent  explorations  in  electrostatic  multilayer  thin  film  assembly.  Curr.  Opin.  Colloid  In.  1999,  4,  430-­‐442.     (26)   Schonhoff,  M.  Self-­‐assembled  polyelectrolyte  multilayers.  Curr.  Opin.  Colloid  In.  2003,  8,  86-­‐95.     (27)   Lee,  G.  S.;  Lee,  Y.  J.;  Yoon,  K.  B.  Layer-­‐by-­‐layer  assembly  of  zeolite  crystals  on  glass  with  polyelectrolytes  as  ionic  linkers.  J.  Am.  Chem.  Soc.  2001,  123,  9769-­‐9779.     (28)   Chapel,  J.  P.;  Berret,  J.  F.  Versatile  electrostatic  assembly  of  nanoparticles  and  polyelectrolytes:  Coating,  clustering  and  layer-­‐by-­‐layer  processes.  Curr.  Opin.  Colloid  In.  2012,  17,  97-­‐105.     (29)   Ali,  M.;  Yameen,  B.;  Cervera,  J.;  Ramirez,  P.;  Neumann,  R.;  Ensinger,  W.;  Knoll,  W.;  Azzaroni,  O.  Layer-­‐by-­‐Layer  Assembly  of  Polyelectrolytes  into  Ionic  Current  Rectifying  Solid-­‐State  Nanopores:  Insights  from  Theory  and  Experiment.  J.  Am.  Chem.  Soc.  2010,  132,  8338-­‐8348.     (30)   Aitkadi,  A.;  Carreau,  P.  J.;  Chauveteau,  G.  Rheological  Properties  of  Partially  Hydrolyzed  Polyacrylamide  Solutions.  J.  Rheol.  1987,  31,  537-­‐561.     (31)   Lewandowska,  K.  Comparative  studies  of  rheological  properties  of  polyacrylamide  and  partially  hydrolyzed  polyacrylamide  solutions.  J.  Appl.  Polym.  Sci.  2007,  103,  2235-­‐2241.     (32)   Hu,  Y.;  Wang,  S.  Q.;  Jamieson,  A.  M.  Rheological  and  Rheooptical  Studies  of  Shear-­‐Thickening  Polyacrylamide  Solutions.  Macromolecules  1995,  28,   80 1847-­‐1853.     (33)   Fuoss,  R.  M.  Viscosity  Function  for  Polyelectrolytes.  J.  Polym.  Sci.  1948,  3,  603-­‐604.     (34)   Dobrynin,  A.  V.;  Rubinstein,  M.  Theory  of  polyelectrolytes  in  solutions  and  at  surfaces.  Prog.  Polym.  Sci.  2005,  30,  1049-­‐1118.     (35)   Collins,  K.  D.  Why  continuum  electrostatics  theories  cannot  explain  biological  structure,  polyelectrolytes  or  ionic  strength  effects  in  ion-­‐protein  interactions.  Biophys.  Chem.  2012,  167,  43-­‐59.     (36)   Jones,  M.  N.  Interaction  of  Sodium  Dodecyl  Sulfate  with  Polyethylene  Oxide.  J.  Colloid  Interface  Sci.  1967,  23,  36-­‐&.     (37)   Asnacios,  A.;  Langevin,  D.;  Argillier,  J.  F.  Mixed  monolayers  of  cationic  surfactants  and  anionic  polymers  at  the  air-­‐water  interface:  Surface  tension  and  ellipsometry  studies.  Eur.  Phys.  J.  B  1998,  5,  905-­‐911.     (38)   Ritacco,  H.;  Kurlat,  D.;  Langevin,  D.  Properties  of  aqueous  solutions  of  polyelectrolytes  and  surfactants  of  opposite  charge:  Surface  tension,  surface  rheology,  and  electrical  birefringence  studies.  J.  Phys.  Chem.  B  2003,  107,  9146-­‐9158.     (39)   Taylor,  D.  J.  F.;  Thomas,  R.  K.;  Penfold,  J.  Polymer/surfactant  interactions  at  the  air/water  interface.  Adv.  Colloid  Interfac.  2007,  132,  69-­‐110.     (40)   Campbell,  R.  A.;  Arteta,  M.  Y.;  Angus-­‐Smyth,  A.;  Nylander,  T.;  Varga,  I.  Multilayers  at  Interfaces  of  an  Oppositely  Charged  Polyelectrolyte/Surfactant  System  Resulting  from  the  Transport  of  Bulk  Aggregates  under  Gravity.  J.  Phys.  Chem.  B  2012,  116,  7981-­‐7990.     (41)   Bain,  C.  D.;  Claesson,  P.  M.;  Langevin,  D.;  Meszaros,  R.;  Nylander,  T.;  Stubenrauch,  C.;  Titmuss,  S.;  von  Klitzing,  R.  Complexes  of  surfactants  with  oppositely  charged  polymers  at  surfaces  and  in  bulk.  Adv.  Colloid  Interface  Sci.  2010,  155,  32-­‐49.     (42)   Backus,  E.  H.  G.;  Abrakhi,  S.;  Peralta,  S.;  Teyssie,  D.;  Fichet,  O.;  Cantin,  S.  Sum-­‐Frequency  Generation  Spectroscopy  of  Cinnamate  Modified  Cellulosic  Polymer  at  the  Air-­‐Water  Interface.  J.  Phys.  Chem.  B  2012,  116,  6041-­‐6049.     (43)   Tyrode,  E.;  Johnson,  C.  M.;  Kumpulainen,  A.;  Rutland,  M.  W.;  Claesson,  P.  M.  Hydration  state  of  nonionic  surfactant  monolayers  at  the  liquid/vapor  interface:  Structure  determination  by  vibrational  sum  frequency  spectroscopy.  J.  Am.  Chem.  Soc.  2005,  127,  16848-­‐16859.     (44)   Lu,  R.;  Gan,  W.;  Wu,  B.  H.;  Chen,  H.;  Wang,  H.  F.  Vibrational  polarization  spectroscopy  of  CH  stretching  modes  of  the  methylene  goup  at  the  vapor/liquid  interfaces  with  sum  frequency  generation.  J.  Phys.  Chem.  B  2004,  108,  7297-­‐7306.     (45)   Ohe,  C.;  Kamijo,  H.;  Arai,  M.;  Adachi,  M.;  Miyazawa,  H.;  Itoh,  K.;  Seki,  T.  Sum  frequency  generation  spectroscopic  study  on  photoinduced  isomerization  of  poly(vinyl  alcohol)  containing  azobenzene  side  chain  at  the  air-­‐water  interface.  J.  Phys.  Chem.  C  2008,  112,  172-­‐181.   81   (46)   Ostroverkhov,  V.;  Waychunas,  G.  A.;  Shen,  Y.  R.  New  information  on  water  interfacial  structure  revealed  by  phase-­‐sensitive  surface  spectroscopy.  Phys.  Rev.  Lett.  2005,  94,  046102.     (47)   Ji,  N.;  Ostroverkhov,  V.;  Tian,  C.  S.;  Shen,  Y.  R.  Characterization  of  vibrational  resonances  of  water-­‐vapor  interfaces  by  phase-­‐sensitive  sum-­‐frequency  spectroscopy.  Phys.  Rev.  Lett.  2008,  100, 096102.     (48)   Yamaguchi,  S.;  Tahara,  T.  Heterodyne-­‐detected  electronic  sum  frequency  generation:  "Up"  versus  "down"  alignment  of  interfacial  molecules.  J.  Chem.  Phys.  2008,  129,  101102.     (49)   Nihonyanagi,  S.;  Yamaguchi,  S.;  Tahara,  T.  Direct  evidence  for  orientational  flip-­‐flop  of  water  molecules  at  charged  interfaces:  A  heterodyne-­‐detected  vibrational  sum  frequency  generation  study.  J.  Chem.  Phys.  2009,  130,  204704.     (50)   Tian,  C.  S.;  Shen,  Y.  R.  Isotopic  Dilution  Study  of  the  Water/Vapor  Interface  by  Phase-­‐Sensitive  Sum-­‐Frequency  Vibrational  Spectroscopy.  J.  Am.  Chem.  Soc.  2009,  131,  2790-­‐+.     (51)   Sovago,  M.;  Campen,  R.  K.;  Wurpel,  G.  W.  H.;  Muller,  M.;  Bakker,  H.  J.;  Bonn,  M.  Vibrational  response  of  hydrogen-­‐bonded  interfacial  water  is  dominated  by  intramolecular  coupling.  Phys.  Rev.  Lett.  2008,  100,  173901.     (52)   Sovago,  M.;  Campen,  R.  K.;  Wurpel,  G.  W.  H.;  Muller,  M.;  Bakker,  H.  J.;  Bonn,  M.  Comment  on  "Vibrational  Response  of  Hydrogen-­‐Bonded  Interfacial  Water  is  Dominated  by  Intramolecular  Coupling''  -­‐  Reply.  Phys.  Rev.  Lett.  2008,  101, 173901.     (53)   Tian,  C.  S.;  Shen,  Y.  R.  Comment  on  "Vibrational  Response  of  Hydrogen-­‐Bonded  Interfacial  Water  is  Dominated  by  Intramolecular  Coupling''.  Phys.  Rev.  Lett.  2008,  101.     (54)   Nihonyanagi,  S.;  Ishiyama,  T.;  Lee,  T.;  Yamaguchi,  S.;  Bonn,  M.;  Morita,  A.;  Tahara,  T.  Unified  Molecular  View  of  the  Air/Water  Interface  Based  on  Experimental  and  Theoretical  chi((2))  Spectra  of  an  Isotopically  Diluted  Water  Surface.  J.  Am.  Chem.  Soc.  2011,  133,  16875-­‐16880.     (55)   Pieniazek,  P.  A.;  Tainter,  C.  J.;  Skinner,  J.  L.  Surface  of  Liquid  Water:  Three-­‐Body  Interactions  and  Vibrational  Sum-­‐Frequency  Spectroscopy.  J.  Am.  Chem.  Soc.  2011,  133,  10360-­‐10363.     (56)   Wei,  X.;  Miranda,  P.  B.;  Zhang,  C.;  Shen,  Y.  R.  Sum-­‐frequency  spectroscopic  studies  of  ice  interfaces.  Phys.  Rev.  B  2002,  66, 085401.     (57)   Tyrode,  E.;  Johnson,  C.  M.;  Rutland,  M.  W.;  Claesson,  P.  M.  Structure  and  hydration  of  poly(ethylene  oxide)  surfactants  at  the  air/liquid  interface.  A  vibrational  sum  frequency  spectroscopy  study.  J.  Phys.  Chem.  C  2007,  111,  11642-­‐11652.     (58)   Nihonyanagi, S.; Ishiyama, T.; Lee, T.; Yamaguchi, S.; Bonn, M.; Morita, A.; Tahara, T. Unified Molecular View of the Air/Water Interface Based on Experimental 82 and Theoretical Chi((2)) Spectra of an Isotopically Diluted Water Surface. J. Am. Chem. Soc. 2011, 133, 16875−16880.   (59)   Murugan,  R.;  Mohan,  S.;  Bigotto,  A.  FTIR  and  polarised  Raman  spectra  of  acrylamide  and  polyacrylamide.  J.  Korean  Phys.  Soc.  1998,  32,  505-­‐512.     (60)   Mondal,  J.  A.;  Nihonyanagi,  S.;  Yamaguchi,  S.;  Tahara,  T.  Structure  and  Orientation  of  Water  at  Charged  Lipid  Monolayer/Water  Interfaces  Probed  by  Heterodyne-­‐Detected  Vibrational  Sum  Frequency  Generation  Spectroscopy.  J.  Am.  Chem.  Soc.  2010,  132,  10656-­‐10657.     (61)   Mondal,  J.  A.;  Nihonyanagi,  S.;  Yamaguchi,  S.;  Tahara,  T.  Three  Distinct  Water  Structures  at  a  Zwitterionic  Lipid/Water  Interface  Revealed  by  Heterodyne-­‐Detected  Vibrational  Sum  Frequency  Generation.  J.  Am.  Chem.  Soc.  2012,  134,  7842-­‐7850.     (62)   Guyotsionnest,  P.;  Hunt,  J.  H.;  Shen,  Y.  R.  Sum-­‐Frequency  Vibrational  Spectroscopy  of  a  Langmuir  Film  -­‐  Study  of  Molecular-­‐Orientation  of  a  Two-­‐Dimensional  System.  Phys.  Rev.  Lett.  1987,  59,  1597-­‐1600.     (63)   Tian,  C.  S.;  Byrnes,  S.  J.;  Han,  H.  L.;  Shen,  Y.  R.  Surface  Propensities  of  Atmospherically  Relevant  Ions  in  Salt  Solutions  Revealed  by  Phase-­‐Sensitive  Sum  Frequency  Vibrational  Spectroscopy.  J.  Phys.  Chem.  Lett.  2011,  2,  1946-­‐1949.     (64)   Hua,  W.;  Jubb,  A.  M.;  Allen,  H.  C.  Electric  Field  Reversal  of  Na2SO4,  (NH4)(2)SO4,  and  Na2CO3  Relative  to  CaCl2  and  NaCl  at  the  Air/Aqueous  Interface  Revealed  by  Heterodyne  Detected  Phase-­‐Sensitive  Sum  Frequency.  J.  Phys.  Chem.  Lett.  2011,  2,  2515-­‐2520.     (65)   Peng,  S.  F.;  Wu,  C.  Light  scattering  study  of  the  formation  and  structure  of  partially  hydrolyzed  poly(acrylamide)/calcium(II)  complexes.  Macromolecules  1999,  32,  585-­‐589.     (66)   Pai,  R.  K.;  Pillai,  S.  Divalent  cation-­‐induced  variations  in  polyelectrolyte  conformation  and  controlling  calcite  morphologies:  Direct  observation  of  the  phase  transition  by  atomic  force  microscopy.  J.  Am.  Chem.  Soc.  2008,  130,  13074-­‐13078.     (67)   Nagarajan,  R.;  Ruckenstein,  E.  Theory  of  Surfactant  Self-­‐Assembly  -­‐  a  Predictive  Molecular  Thermodynamic  Approach.  Langmuir  1991,  7,  2934-­‐2969.     (68)   Lu,  J.  R.;  Thomas,  R.  K.;  Penfold,  J.  Surfactant  layers  at  the  air/water  interface:  structure  and  composition.  Adv.  Colloid  Interface  Sci.  2000,  84,  143-­‐304.     (69)   Alexandridis,  P.;  Hatton,  T.  A.  Poly(Ethylene  Oxide)-­‐Poly(Propylene  Oxide)-­‐Poly(Ethylene  Oxide)  Block-­‐Copolymer  Surfactants  in  Aqueous-­‐Solutions  and  at  Interfaces  -­‐  Thermodynamics,  Structure,  Dynamics,  and  Modeling.  Colloid  Surface  A  1995,  96,  1-­‐46.     (70)   Eastoe,  J.;  Dalton,  J.  S.  Dynamic  surface  tension  and  adsorption  mechanisms  of  surfactants  at  the  air-­‐water  interface.  Adv.  Colloid  Interface  Sci.  2000,  85,  103-­‐144.     (71)   Claussen,  W.  F.  Surface  Tension  and  Surface  Structure  of  Water.  Science  1967,  156,  1226-­‐&.   83   (72)   Li,  D.  C.;  Wagner,  N.  J.  Universal  Binding  Behavior  for  Ionic  Alkyl  Surfactants  with  Oppositely  Charged  Polyelectrolytes.  J.  Am.  Chem.  Soc.  2013,  135,  17547-­‐17555.     (73)   Bahramian,  A.;  Thomas,  R.  K.;  Penfold,  J.  The  Adsorption  Behavior  of  Ionic  Surfactants  and  Their  Mixtures  with  Nonionic  Polymers  and  with  Polyelectrolytes  of  Opposite  Charge  at  the  Air-­‐Water  Interface.  J.  Phys.  Chem.  B  2014,  118,  2769-­‐2783.     (74)   Penfold,  J.;  Tucker,  I.;  Thomas,  R.  K.;  Taylor,  D.  J.  F.;  Zhang,  X.  L.;  Bell,  C.;  Breward,  C.;  Howell,  P.  The  interaction  between  sodium  alkyl  sulfate  surfactants  and  the  oppositely  charged  polyelectrolyte,  polyDMDAAC,  at  the  air-­‐water  interface:  The  role  of  alkyl  chain  length  and  electrolyte  and  comparison  with  theoretical  predictions.  Langmuir  2007,  23,  3128-­‐3136.     (75)   Campbell,  R.  A.;  Angus-­‐Smyth,  A.;  Arteta,  M.  Y.;  Tonigold,  K.;  Nylander,  T.;  Varga,  I.  New  Perspective  on  the  Cliff  Edge  Peak  in  the  Surface  Tension  of  Oppositely  Charged  Polyelectrolyte/Surfactant  Mixtures.  J.  Phys.  Chem.  Lett.  2010,  1,  3021-­‐3026.     (76)   Nizri,  G.;  Lagerge,  S.;  Kamyshny,  A.;  Major,  D.  T.;  Magdassi,  S.  Polymer-­‐surfactant  interactions:  Binding  mechanism  of  sodium  dodecyl  sulfate  to  poly(diallyldimethylammonium  chloride).  J.  Colloid  Interface  Sci.  2008,  320,  74-­‐81.     (77)   Lieske,  A.;  Jaeger,  W.  Synthesis  and  characterization  of  block  copolymers  containing  cationic  blocks.  Macromol.  Chem.  Phys.  1998,  199,  255-­‐260.     (78)   Rao,  Y.;  Li,  X.;  Lei,  X.  G.;  Jockusch,  S.;  George,  M.  W.;  Turro,  N.  J.;  Eisenthal,  K.  B.  Observations  of  Interfacial  Population  and  Organization  of  Surfactants  with  Sum  Frequency  Generation  and  Surface  Tension.  J.  Phys.  Chem.  C  2011,  115,  12064-­‐12067.     (79)   Stiopkin,  I.  V.;  Weeraman,  C.;  Pieniazek,  P.  A.;  Shalhout,  F.  Y.;  Skinner,  J.  L.;  Benderskii,  A.  V.  Hydrogen  bonding  at  the  water  surface  revealed  by  isotopic  dilution  spectroscopy.  Nature  2011,  474,  192-­‐195.     (80)   Nihonyanagi,  S.;  Mondal,  J.  A.;  Yamaguchi,  S.;  Tahara,  T.  Structure  and  Dynamics  of  Interfacial  Water  Studied  by  Heterodyne-­‐Detected  Vibrational  Sum-­‐Frequency  Generation.  Annu.  Rev.  Phys.  Chem.  2013,  64,  579-­‐603.     (81)   Tonigold,  K.;  Varga,  I.;  Nylander,  T.;  Campbell,  R.  A.  Effects  of  Aggregates  on  Mixed  Adsorption  Layers  of  Poly(ethylene  imine)  and  Sodium  Dodecyl  Sulfate  at  the  Air/Liquid  Interface.  Langmuir  2009,  25,  4036-­‐4046.     (82)   Mondal,  J.  A.;  Nihonyanagi,  S.;  Yamaguchi,  S.;  Tahara,  T.  Structure  and  Orientation  of  Water  at  Charged  Lipid  Monolayer/Water  Interfaces  Probed  by  Heterodyne-­‐Detected  Vibrational  Sum  Frequency  Generation  Spectroscopy.  J.  Am.  Chem.  Soc.  2010,  132,  10656-­‐10657.     (83)   Hu,  D.;  Yang,  Z.;  Chou,  K.  C.  Interactions  of  Polyelectrolytes  with  Water  and  Ions  at  Air/Water  Interfaces  Studied  by  Phase-­‐Sensitive  Sum  Frequency  Generation  Vibrational  Spectroscopy.  J.  Phys.  Chem.  C  2013,  117,  15698-­‐15703.   84   (84)   Tian,  C.  S.;  Shen,  Y.  R.  Sum-­‐frequency  vibrational  spectroscopic  studies  of  water/vapor  interfaces.  Chem.  Phys.  Lett.  2009,  470,  1-­‐6.     (85)   Pieniazek,  P.  A.;  Tainter,  C.  J.;  Skinner,  J.  L.  Interpretation  of  the  water  surface  vibrational  sum-­‐frequency  spectrum.  J.  Chem.  Phys.  2011,  135.     (86)   Chen,  X.  K.;  Hua,  W.;  Huang,  Z.  S.;  Allen,  H.  C.  Interfacial  Water  Structure  Associated  with  Phospholipid  Membranes  Studied  by  Phase-­‐Sensitive  Vibrational  Sum  Frequency  Generation  Spectroscopy.  J.  Am.  Chem.  Soc.  2010,  132,  11336-­‐11342.     (87)   Fuguet,  E.;  Rafols,  C.;  Roses,  M.;  Bosch,  E.  Critical  micelle  concentration  of  surfactants  in  aqueous  buffered  and  unbuffered  systems.  Anal.  Chim.  Acta.  2005,  548,  95-­‐100.     (88)   Staples,  E.;  Tucker,  I.;  Penfold,  J.;  Warren,  N.;  Thomas,  R.  K.;  Taylor,  D.  J.  F.  Organization  of  polymer-­‐surfactant  mixtures  at  the  air-­‐water  interface:  Sodium  dodecyl  sulfate  and  poly(dimethyldiallylammonium  chloride).  Langmuir  2002,  18,  5147-­‐5153.     (89)   Taylor,  D.  J.  F.;  Thomas,  R.  K.;  Li,  P.  X.  Adsorption  of  oppositely  charged  polyelectrolyte/surfactant  mixtures.  Neutron  reflection  from  alkyl  trimethylammonium  bromides  and  sodium  poly(styrenesulfonate)  at  the  air/water  interface:  The  effect  of  surfactant  chain  length.  Langmuir  2003,  19,  3712-­‐3719.     (90)   Merta,  J.;  Stenius,  P.  Interactions  between  Cationic  Starch  and  Anionic  Surfactants  .1.  Phase-­‐Equilibria  and  Surface  Tensions.  Colloid  Polym.  Sci.  1995,  273,  974-­‐983.     (91)   Monteux,  C.;  Williams,  C.  E.;  Meunier,  J.;  Anthony,  O.;  Bergeron,  V.  Adsorption  of  oppositely  charged  polyelectrolyte/surfactant  complexes  at  the  air/water  interface:  Formation  of  interfacial  gels.  Langmuir  2004,  20,  57-­‐63.     (92)   Noskov,  B.  A.;  Grigoriev,  D.  O.;  Lin,  S.  Y.;  Loglio,  G.;  Miller,  R.  Dynamic  surface  properties  of  polyelectrolyte/surfactant  adsorption  films  at  the  Air/Water  interface:  Poly(dially1dimethylammonium  chloride)  and  sodium  dodecylsulfate.  Langmuir  2007,  23,  9641-­‐9651.     (93) Lide, D. R. Handbook of Chemistry and Physics, 85th ed.; CRC Press: Cleveland, OH, 2004.   (94)   Atkins,  P.  W.;  Jones,  L.  Chemical  principles  :  the  quest  for  insight,  4th  ed.;  W.H.  Freeman:  New  York,  2008.     (95)   Smith,  J.  D.;  Cappa,  C.  D.;  Wilson,  K.  R.;  Messer,  B.  M.;  Cohen,  R.  C.;  Saykally,  R.  J.  Energetics  of  hydrogen  bond  network  rearrangements  in  liquid  water.  Science  2004,  306,  851-­‐853.     (96)   Kuo,  I.  F.  W.;  Mundy,  C.  J.  An  ab  initio  molecular  dynamics  study  of  the  aqueous  liquid-­‐vapor  interface.  Science  2004,  303,  658-­‐660.     (97)   Wernet,  P.;  Nordlund,  D.;  Bergmann,  U.;  Cavalleri,  M.;  Odelius,  M.;  Ogasawara,  H.;  Naslund,  L.  A.;  Hirsch,  T.  K.;  Ojamae,  L.;  Glatzel,  P.;  Pettersson,  L.  G.  M.;  Nilsson,  A.  The  structure  of  the  first  coordination  shell  in  liquid  water.  Science  2004,  304,  995-­‐999.   85   (98)   Van  der  Spoel,  D.;  Lindahl,  E.;  Hess,  B.;  Groenhof,  G.;  Mark,  A.  E.;  Berendsen,  H.  J.  C.  GROMACS:  Fast,  flexible,  and  free.  J.  Comput.  Chem.  2005,  26,  1701-­‐1718.     (99)   Hess,  B.;  Kutzner,  C.;  van  der  Spoel,  D.;  Lindahl,  E.  GROMACS  4:  Algorithms  for  highly  efficient,  load-­‐balanced,  and  scalable  molecular  simulation.  J.  Chem.  Theory  Comput.  2008,  4,  435-­‐447.     (100)   Pronk,  S.;  Pall,  S.;  Schulz,  R.;  Larsson,  P.;  Bjelkmar,  P.;  Apostolov,  R.;  Shirts,  M.  R.;  Smith,  J.  C.;  Kasson,  P.  M.;  van  der  Spoel,  D.;  Hess,  B.;  Lindahl,  E.  GROMACS  4.5:  a  high-­‐throughput  and  highly  parallel  open  source  molecular  simulation  toolkit.  Bioinformatics  2013,  29,  845-­‐854.     (101)   Shen,  Z.;  Sun,  H.;  Liu,  X.  Y.;  Liu,  W.  T.;  Tang,  M.  Stability  of  Newton  Black  Films  Under  Mechanical  Stretch  -­‐  A  Molecular  Dynamics  Study.  Langmuir  2013,  29,  11300-­‐11309.     (102)   Berendsen,  H.  J.  C.;  Grigera,  J.  R.;  Straatsma,  T.  P.  The  Missing  Term  in  Effective  Pair  Potentials.  J.  Phys.  Chem.  1987,  91,  6269-­‐6271.     (103)   Martinez,  L.;  Andrade,  R.;  Birgin,  E.  G.;  Martinez,  J.  M.  PACKMOL:  A  Package  for  Building  Initial  Configurations  for  Molecular  Dynamics  Simulations.  J.  Comput.  Chem.  2009,  30,  2157-­‐2164.     (104)   Bussi,  G.;  Donadio,  D.;  Parrinello,  M.  Canonical  sampling  through  velocity  rescaling.  J.  Chem.  Phys.  2007,  126.     (105)   Ryckaert,  J.  P.;  Ciccotti,  G.;  Berendsen,  H.  J.  C.  Numerical-­‐Integration  of  Cartesian  Equations  of  Motion  of  a  System  with  Constraints  -­‐  Molecular-­‐Dynamics  of  N-­‐Alkanes.  J.  Comput.  Phys.  1977,  23,  327-­‐341.     (106)   Lorentz,  H.  A.  Ueber  die  Anwendung  des  Satzes  vom  Virial  in  der  kinetischen  Theorie  der  Gase.  1881,  248,  127–136.     (107)   Berthelot,  D.  Sur  le  mélange  des  gaz.  1898,  126,  1703–1855.     (108)   Essmann,  U.;  Perera,  L.;  Berkowitz,  M.  L.;  Darden,  T.;  Lee,  H.;  Pedersen,  L.  G.  A  Smooth  Particle  Mesh  Ewald  Method.  J.  Chem.  Phys.  1995,  103,  8577-­‐8593.     (109)   Humphrey,  W.;  Dalke,  A.;  Schulten,  K.  VMD:  Visual  molecular  dynamics.  J.  Mol.  Graph  Model.  1996,  14,  33-­‐38.     (110)   Yaseen,  M.;  Lu,  J.  R.;  Webster,  J.  R.  P.;  Penfold,  J.  The  structure  of  zwitterionic  phosphocholine  surfactant  monolayers.  Langmuir  2006,  22,  5825-­‐5832.     (111)   Lu,  J.  R.;  Purcell,  I.  P.;  Lee,  E.  M.;  Simister,  E.  A.;  Thomas,  R.  K.;  Rennie,  A.  R.;  Penfold,  J.  The  Composition  and  Structure  of  Sodium  Dodecyl-­‐Sulfate  Dodecanol  Mixtures  Adsorbed  at  the  Air-­‐Water-­‐Interface  -­‐  a  Neutron  Reflection  Study.  J.  Colloid  Interf.  Sci.  1995,  174,  441-­‐455.     (112)   Lyttle,  D.  J.;  Lu,  J.  R.;  Su,  T.  J.;  Thomas,  R.  K.;  Penfold,  J.  Structure  of  a  Dodecyltrimethylammonium  Bromide  Layer  at  the  Air-­‐Water-­‐Interface  Determined  by  Neutron  Reflection  -­‐  Comparison  of  the  Monolayer  Structure  of  Cationic  Surfactants  with  Different  Chain  Lengths.  Langmuir  1995,  11,  1001-­‐1008.   86   (113)   Lu,  J.  R.;  Thomas,  R.  K.  Neutron  reflection  from  wet  interfaces.  J.  Chem.  Soc.,  Faraday  Trans.  1998,  94,  995-­‐1018.     (114)   Good,  R.  J.  Surface  Entropy  and  Surface  Orientation  of  Polar  Liquids.  J.  Phys.  Chem.  1957,  61,  810-­‐813.     (115)   Czarnecka,  E.;  Gillott,  J.  E.  Formation  and  Characterization  of  Clay  Complexes  with  Bitumen  from  Athabasca  Oil  Sand.  Clays  Clay  Miner.  1980,  28,  197-­‐203.     (116)   Hansen,  J.  S.;  Lemarchand,  C.  A.;  Nielsen,  E.;  Dyre,  J.  C.;  Schroder,  T.  Four-­‐component  united-­‐atom  model  of  bitumen.  J.  Chem.  Phys.  2013,  138.     (117)   Masliyah,  J.;  Zhou,  Z.  J.;  Xu,  Z.  H.;  Czarnecki,  J.;  Hamza,  H.  Understanding  water-­‐based  bitumen  extraction  from  athabasca  oil  sands.  Can.  J.  Chem.  Eng.  2004,  82,  628-­‐654.     (118)   Zhao,  H.  Y.;  Long,  J.;  Masliyah,  J.  H.;  Xu,  Z.  H.  Effect  of  divalent  cations  and  surfactants  on  silica-­‐bitumen  interactions.  Ind.  Eng.  Chem.  Res.  2006,  45,  7482-­‐7490.     (119)   Takamura,  K.;  Chow,  R.  S.  The  Electric  Properties  of  the  Bitumen/Water  Interface  .3.  Application  of  the  Ionizable  Surface-­‐Group  Model.  Colloids  Surf.  1985,  15,  35-­‐48.     (120)   Acevedo,  S.;  Gutierrez,  X.;  Rivas,  H.  Bitumen-­‐in-­‐water  emulsions  stabilized  with  natural  surfactants.  J.  Colloid  Interface  Sci.  2001,  242,  230-­‐238.     (121)   Angle,  C.  W.;  Hua,  Y.  J.  Tailings  Pond  Surfactant  Analogues:  Effects  on  Toluene-­‐Diluted  Bitumen  Drops  in  NaHCO3/K2CO3  Solution.  Part  2:  Dilational  Interfacial  Viscoelasticity.  Energy  Fuels  2013,  27,  3613-­‐3621.     (122)   Chaverot,  P.;  Cagna,  A.;  Glita,  S.;  Rondelez,  F.  Interfacial  tension  of  bitumen-­‐water  interfaces.  Part  1:  Influence  of  endogenous  surfactants  at  acidic  pH.  Energy  Fuels  2008,  22,  790-­‐798.     (123)   Rowe,  A.  W.;  Counce,  R.  M.;  Morton,  S.  A.;  Hu,  M.  Z.  C.;  DePaoli,  D.  W.  Oil  detachment  from  solid  surfaces  in  aqueous  surfactant  solutions  as  a  function  of  pH.  Ind.  Eng.  Chem.  Res.  2002,  41,  1787-­‐1795.     (124)   Liu,  J.  J.;  Zhou,  Z.;  Xu,  Z.  H.  Electrokinetic  study  of  hexane  droplets  in  surfactant  solutions  and  process  water  of  bitumen  extraction  systems.  Ind.  Eng.  Chem.  Res.  2002,  41,  52-­‐57.     (125)   Hua,  Y.  J.;  Angle,  C.  W.  Brewster  Angle  Microscopy  of  Langmuir  Films  of  Athabasca  Bitumens,  n-­‐C5  Asphaltenes,  and  SAGD  Bitumen  during  Pressure-­‐Area  Hysteresis.  Langmuir  2013,  29,  244-­‐263.     (126)   Rudin,  J.;  Wasan,  D.  T.  Mechanisms  for  Lowering  of  Interfacial-­‐Tension  in  Alkali  Acidic  Oil  Systems  -­‐  Effect  of  Added  Surfactant.  Ind.  Eng.  Chem.  Res.  1992,  31,  1899-­‐1906.     (127)   Tchoukov,  P.;  Yang,  F.;  Xu,  Z.  H.;  Dabros,  T.;  Czarnecki,  J.;  Sjoblom,  J.  Role  of  Asphaltenes  in  Stabilizing  Thin  Liquid  Emulsion  Films.  Langmuir  2014,  30,  3024-­‐3033.     (128)   Hou,  J.;  Feng,  X.  H.;  Masliyah,  J.;  Xu,  Z.  H.  Understanding  Interfacial   87 Behavior  of  Ethylcellulose  at  the  Water-­‐Diluted  Bitumen  Interface.  Energy  Fuels  2012,  26,  1740-­‐1745.     (129)   Yang,  Z.;  Nikakhtari,  H.;  Wolf,  S.;  Hu,  D.;  Gray,  M.  R.;  Chou,  K.  C.  Binary  Solvents  with  Ethanol  for  Effective  Bitumen  Displacement  at  Solvent/Mineral  Interfaces.  Energy  Fuels  2015,  29,  4222-­‐4226.     (130)   Orbulescu,  J.;  Leblanc,  R.  M.  Importance  of  the  Spreading  Solvent  Evaporation  Time  in  Langmuir  Monolayers.  J.  Phys.  Chem.  C  2009,  113,  5313-­‐5315.     (131)   Hu,  D.;  Chou,  K.  C.  Re-­‐Evaluating  the  Surface  Tension  Analysis  of  Polyelectrolyte-­‐Surfactant  Mixtures  Using  Phase-­‐Sensitive  Sum  Frequency  Generation  Spectroscopy.  J.  Am.  Chem.  Soc.  2014,  136,  15114-­‐15117.     (132)   Clark,  K.  A.;  Pasternack,  D.  S.  Hot  water  separation  of  bitumen  from  Alberta  bituminous  sand.  Ind.  Eng.  Chem.  1932,  24,  1410-­‐1416.     (133)   Basu,  S.;  Nandakumar,  K.;  Lawrence,  S.;  Masliyah,  J.  Effect  of  calcium  ion  and  montmorillonite  clay  on  bitumen  displacement  by  water  on  a  glass  surface.  Fuel  2004,  83,  17-­‐22.     (134)   Yang,  Z.;  Li,  Q.  F.;  Chou,  K.  C.  Structures  of  Water  Molecules  at  the  Interfaces  of  Aqueous  Salt  Solutions  and  Silica:  Cation  Effects.  J.  Phys.  Chem.  C  2009,  113,  8201-­‐8205.     (135)   Leung,  B.  O.;  Yang,  Z.;  Wu,  S.  S.  H.;  Chou,  K.  C.  Role  of  Interfacial  Water  on  Protein  Adsorption  at  Cross-­‐Linked  Polyethylene  Oxide  Interfaces.  Langmuir  2012,  28,  5724-­‐5728.     (136)   He,  L.;  Lin,  F.;  Li,  X.;  Sui,  H.;  Xu,  Z.  Interfacial  sciences  in  unconventional  petroleum  production:  from  fundamentals  to  applications.  Chem.  Soc.  Rev.  2015,  44,  5446.     (137)   Das,  S.;  Thundat,  T.;  Mitra,  S.  K.  Analytical  model  for  zeta  potential  of  asphaltene.  Fuel  2013,  108,  543-­‐549.     (138)   Wang,  J.;  Lu,  Q.  Y.;  Harbottle,  D.;  Sjoblom,  J.;  Xu,  Z.  H.;  Zeng,  H.  B.  Molecular  Interactions  of  a  Polyaromatic  Surfactant  C5Pe  in  Aqueous  Solutions  Studied  by  a  Surface  Forces  Apparatus.  J.  Phys.  Chem.  B  2012,  116,  11187-­‐11196.     (139)   Beattie,  J.  K.;  Djerdjev,  A.  M.  The  pristine  oil/water  interface:  Surfactant-­‐free  hydroxide-­‐charged  emulsions.  Angew.  Chem.  Int.  Ed.  2004,  43,  3568-­‐3571.     (140)   Dai,  Q.;  Chung,  K.  H.  Hot  water  extraction  process  mechanism  using  model  oil  sands.  Fuel  1996,  75,  220-­‐226.     (141)   Schwarz,  J.  A.;  Contescu,  C.  I.;  Putyera,  K.  Dekker  encyclopedia  of  nanoscience  and  nanotechnology;  M.  Dekker:  New  York,  2004.     (142)   Olivier-­‐Bourbigou,  H.;  Magna,  L.;  Morvan,  D.  Ionic  liquids  and  catalysis:  Recent  progress  from  knowledge  to  applications.  Appl.  Catal.,  A  2010,  373,  1-­‐56.     (143)   Wang,  P.;  Zakeeruddin,  S.  M.;  Comte,  P.;  Exnar,  I.;  Gratzel,  M.  Gelation  of  ionic  liquid-­‐based  electrolytes  with  silica  nanoparticles  for  quasi-­‐solid-­‐state  dye-­‐sensitized  solar  cells.  J.  Am.  Chem.  Soc.  2003,  125,  1166-­‐1167.   88   (144)   Romanos,  G.  E.;  Zubeir,  L.  F.;  Likodimos,  V.;  Falaras,  P.;  Kroon,  M.  C.;  Iiev,  B.;  Adamova,  G.;  Schubert,  T.  J.  S.  Enhanced  CO2  Capture  in  Binary  Mixtures  of  1-­‐Alkyl-­‐3-­‐methylimidazolium  Tricyanomethanide  Ionic  Liquids  with  Water.  J.  Phys.  Chem.  B  2013,  117,  12234-­‐12251.     (145)   Welton,  T.  Room-­‐temperature  ionic  liquids.  Solvents  for  synthesis  and  catalysis.  Chem.  Rev.  1999,  99,  2071-­‐2083.     (146)   Baldelli,  S.  Interfacial  Structure  of  Room-­‐Temperature  Ionic  Liquids  at  the  Solid-­‐Liquid  Interface  as  Probed  by  Sum  Frequency  Generation  Spectroscopy.  J.  Phys.  Chem.  Lett.  2013,  4,  244-­‐252.     (147)   Sung,  J.;  Jeon,  Y.;  Kim,  D.;  Iwahashi,  T.;  Iimori,  T.;  Seki,  K.;  Ouchi,  Y.  Air-­‐liquid  interface  of  ionic  liquid  plus  H2O  binary  system  studied  by  surface  tension  measurement  and  sum-­‐frequency  generation  spectroscopy.  Chem.  Phys.  Lett.  2005,  406,  495-­‐500.     (148)   Clavero,  E.;  Rodriguez,  J.  Ionic  liquids  at  the  air/water  interface.  J.  Mol.  Liq.  2011,  163,  64-­‐69.     (149)   Rivera-­‐Rubero,  S.;  Baldelli,  S.  Surface  characterization  of  1-­‐butyl-­‐3-­‐methylimidazollum  Br-­‐,  I-­‐,  PF6-­‐,  BF4-­‐,  (CF3SO2)(2)N-­‐,SCN-­‐,  CH3SO3-­‐,  CH3SO4-­‐,  and  (CN)(2)N-­‐  ionic  liquids  by  sum  frequency  generation.  J.  Phys.  Chem.  B  2006,  110,  4756-­‐4765.     (150)   Rivera-­‐Rubero,  S.;  Baldelli,  S.  Influence  of  water  on  the  surface  of  the  water-­‐miscible  ionic  liquid  1-­‐butyl-­‐3-­‐methylimidazolium  tetrafluoroborate:  A  sum  frequency  generation  analysis.  J.  Phys.  Chem.  B  2006,  110,  15499-­‐15505.     (151)   Sung,  J.;  Jeon,  Y.;  Kim,  D.;  Iwahashi,  T.;  Seki,  K.;  Iimori,  T.;  Ouchi,  Y.  Gibbs  monolayer  of  ionic  liquid  plus  H2O  mixtures  studied  by  surface  tension  measurement  and  sum-­‐frequency  generation  spectroscopy.  Colloid  Surface  A  2006,  284,  84-­‐88.     (152)   Jungwirth,  P.;  Tobias,  D.  J.  Molecular  structure  of  salt  solutions:  A  new  view  of  the  interface  with  implications  for  heterogeneous  atmospheric  chemistry.  J.  Phys.  Chem.  B  2001,  105,  10468-­‐10472.     (153)   Baldelli,  S.  Influence  of  water  on  the  orientation  of  cations  at  the  surface  of  a  room-­‐temperature  ionic  liquid:  A  sum  frequency  generation  vibrational  spectroscopic  study.  J.  Phys.  Chem.  B  2003,  107,  6148-­‐6152.     (154)   Iimori,  T.;  Iwahashi,  T.;  Kanai,  K.;  Seki,  K.;  Sung,  J.  H.;  Kim,  D.;  Hamaguchi,  H.  O.;  Ouchi,  Y.  Local  structure  at  the  air/liquid  interface  of  room-­‐temperature  ionic  liquids  probed  by  infrared-­‐visible  sum  frequency  generation  vibrational  spectroscopy:  1-­‐alkyl-­‐3-­‐methylimidazolium  tetrafluoroborates.  J.  Phys.  Chem.  B  2007,  111,  4860-­‐4866.     (155)   Nihonyanagi,  S.;  Yamaguchi,  S.;  Tahara,  T.  Water  Hydrogen  Bond  Structure  near  Highly  Charged  Interfaces  Is  Not  Like  Ice.  J.  Am.  Chem.  Soc.  2010,  132,  6867-­‐+.     (156)   Cammarata,  L.;  Kazarian,  S.  G.;  Salter,  P.  A.;  Welton,  T.  Molecular  states   89 of  water  in  room  temperature  ionic  liquids.  Phys.  Chem.  Chem.  Phys.  2001,  3,  5192-­‐5200.