Lecture 11: Surface Characterization of Biomaterials in Vacuum

3.051J/20.340J Lecture 11 Surface Characterization of Biomaterials in Vacuum The structure and chemistry of a biomaterial surface greatly dictates the degree of biocompatibility of an implant. Surface characterization is thus a central aspect of biomaterials research. Surface chemistry can be investigated directly using high vacuum methods: • Electron spectroscopy for Chemical Analysis (ESCA)/X-ray Photoelectron Spectroscopy (XPS) • Auger Electron Spectroscopy (AES) • Secondary Ion Mass Spectroscopy (SIMS) 1. XPS/ESCA Theoretical Basis: ¾ Secondary electrons ejected by x-ray bombardment from the sample near surface (0.5-10 nm) with characteristic energies ¾ Analysis of the photoelectron energies yields a quantitative measure of the surface composition 1 3.051J/20.340J 2 Electron energy analyzer (variable retardation voltage) X-ray source (Ε = hν) Detector Lens e P ≈ 10-10 Torr θ e e Ekin Evac EF Photoelectron binding energy is characteristic of the element and bonding environment Chemical analysis! EB LIII LII LI K Binding energy = incident x-ray energy − photoelectron kinetic energy EB = hν - Ekin 3 3.051J/20.340J Quantitative Elemental Analysis Low-resolution spectrum C1s Intensity O1s N1s 500 300 Binding energy (eV) ¾ Area under peak Ii ∝ number of electrons ejected (& atoms present) ¾ Only electrons in the near surface region escape without losing energy by inelastic collision ¾ Sensitivity: depends on element. Elements present in concentrations >0.1 atom% are generally detectable (H & He undetected) ¾ Quantification of atomic fraction Ci (of elements detected) Ci = I i / Si ∑Ij / Sj j sum over detected elements Si is the sensitivity factor: - f(instrument & atomic parameters) - can be calculated 4 3.051J/20.340J High-resolution spectrum C1s PMMA Intensity 290 285 Binding energy (eV) ¾ Ratio of peak areas gives a ratio of photoelectrons ejected from atoms in a particular bonding configuration (Si = constant) 5 carbons in total Ex. PMMA −C−C− H CH3 3 −C−C− H C H CH3 (a) Lowest EB C1s EB ≈ 285.0 eV H C=O O CH3 1 O CH3 1 C=O O (b) Intermediate EB C1s EB ≈ 286.8 eV (c) Highest EB C1s EB ≈ 289.0 eV Why does core electron EB vary with valence shell configuration? 5 3.051J/20.340J Electronegative oxygen “robs” valence electrons from carbon (electron density higher toward O atoms) Carbon core electrons held “tighter” to the + nucleus (less screening of + charge) Slight shift to higher C1s binding energy Similarly, different oxidation states of metals can be distinguished. Ex. Fe FeO Fe3O4 Fe2O3 Fe 2p binding energy XPS signal comes from first ~10 nm of sample surface. What if the sample has a concentration gradient within this depth? Surface-segregating species 10nm Multivalent oxide layer Adsorbed species 6 3.051J/20.340J Depth-Resolved ESCA/XPS ¾ The probability of a photoelectron escaping the sample without undergoing inelastic collision is inversely related to its depth t within the sample: ⎛ −t ⎞ P(t ) ~ exp ⎜ ⎟ ⎝ λe ⎠ where λe (typically ~ 5-30 Å) is the electron inelastic mean-free path, which depends on the electron kinetic energy and the material. (Physically, λe = avg. distance traveled between inelastic collisions.) ⇒ For t = 3 λe P(t) = 0.05 e θ =90° 95% of signal from t ≤ 3 λe ¾ By varying the take-off angle (θ), the sampling depth can be decreased, increasing surface sensitivity e e θ θ t = 3λe sin θ 7 3.051J/20.340J Ci 5 θ (degrees) 90 ¾ Variation of composition with angle may indicate: - Preferential orientation at surface Surface segregation Adsorbed species (e.g., hydrocarbons) etc. ¾ Quantifying composition as a function of depth The area under the jth peak of element i is the integral of attenuated contributions from all sample depths z: ⎛ −z I ij = CinstT ( Ekin )Lijσ ij ∫ ni (z) exp ⎜ ⎝ λe sin θ ⎞ ⎟dz ⎠ Cinst = instrument constant T(Ekin) = analyzer transmission function Lij = angular asymmetry factor for orbital j of element i σij is the photoionization cross-section ni(z) is the atomic concen. of i at a depth z (atoms/vol) 8 3.051J/20.340J For a semi-infinite sample of homogeneous composition: ∞ ⎛ −z ⎞ I ij = −I ij ,o ni λe sin θ exp ⎜ ⎟ = I ij ,o ni λe sin θ = Si ni = I ij ,∞ λ sin θ ⎝ e ⎠0 where I ij ,o = CinstT ( Ekin )Lijσ ij Relative concentrations of elements (or atoms with a particular bond configuration) are obtained from ratios of Iij (peak area): • Lij depends on electronic shell (ex. 1s or 2p); obtained from tables; cancels if taking a peak ratio from same orbitals, ex. I C1s / I O1s • Cinst and T(Ekin) are known for most instruments; cancel if taking a peak ratio with Ekin ≈ constant, ex. I C1s (C −C −O) / I C1s (C −CH 3 ) • σij obtained from tables; cancels if taking a peak ratio from same atom in different bonding config., ex. I C1s (C −C −O ) / I C1s (C −CH 3 ) • λe values can be measured or estimated from empirically-derived expressions For polymers: For elements: −2 0.5 λe (nm) = ρ −1 ( 49Ekin + 0.11Ekin ) 0.5 −2 + 0.41( Ekin a ) ⎤ λe (nm) = a ⎡538Ekin ⎣ ⎦ For inorganic compounds (ex. oxides): 0.5 −2 + 0.72 ( Ekin a ) ⎤ λe (nm) = a ⎡ 2170Ekin ⎣ ⎦ 9 3.051J/20.340J where: 1/ 3 a = monolayer thickness (nm) ⎛ MW ⎞ a = 10 ⎜ ⎟ ⎝ ρ N Av ⎠ 7 MW = molar mass (g/mol) ρ = density (g/cm3) Ekin = electron kinetic energy (eV) Ex: λe for C1s using a Mg Kα x-ray source: EB = hν - Ekin For Mg Kα x-rays: hν = 1254 eV For C1s : EB = 284 eV −2 0.5 λe (nm) = ρ −1 ( 49Ekin + 0.11Ekin ) λe = 3.1 nm Ekin = 970 eV Assume ρ = 1.1 g/cm3 10 3.051J/20.340J For non-uniform samples, signal intensity must be deconvoluted to obtain a quantitative analysis of concentration vs. depth. Case Example: a sample comprising two layers (layer 2 semi-infinite): d 1 2 I ij = Cins T ( ⎛ −z )L σ (z) exp n ⎜ kin ij ij ∫ i ⎝ λe sin θ ⎞ ⎟dz ⎠ ∞ d ⎛ −z ⎞ ⎛ −z ⎞ (2) − I ij = −I ij(1),o ni,1 λe sin θ exp ⎜ I n λ sin θ exp ⎟ ⎜ ⎟ ij , o i,2 e ⎝ λe sin θ ⎠ 0 ⎝ λe sin θ ⎠ d ⎛ ⎛ −d I ij = I n λ sin θ ⎜1− exp ⎜ ⎜ λ sin θ ⎜ ⎝ e,1 ⎝ (1) ij ,o i,1 e,1 or I ij = I (1) ij ,∞ ⎛ ⎛ −d ⎜1 − exp ⎜⎜ ⎜ ⎝ λe,1 sin θ ⎝ ⎞ ⎞ (2) ⎛ −d ⎟⎟ ⎟⎟ + I ij ,o ni,2λe,2 sin θ exp ⎜⎜ ⎠⎠ ⎝ λe,1 sin θ ⎞ ⎞ (2) ⎛ −d ⎟⎟ ⎟⎟ + I ij ,∞ exp ⎜⎜ ⎠⎠ ⎝ λe,1 sin θ ⎞ ⎟⎟ ⎠ ⎞ ⎟⎟ ⎠ Why λe,1? Electrons originating in semi-infinite layer 2 are attenuated by overlayer 1 (i ) where I ij ,∞ = measured peak area from a uniform, semi-infinite sample of material i. 11 3.051J/20.340J Methods to solve for d Scenario 1: ni,2=0 (ex., C1s peak of a polymer adsorbed on an oxide): I ij = I (1) ij ,∞ ⎛ ⎛ −d ⎜1 − exp ⎜⎜ ⎜ ⎝ λe,1 sin θ ⎝ ⎞⎞ ⎟⎟ ⎟⎟ ⎠⎠ d 1 2 (1) ¾ measure a bulk sample of the upper layer material ⇒ I ij ,∞ ⎛ I ij ⎞ −d ln ⎜1− (1) ⎟ = ⎟ ⎜ ⎝ I ij ,∞ ⎠ λe,1 sin θ ⎛ I ij ln 1 − ⎜ ¾ obtain slope of ⎜ I (1) ij,∞ ⎝ ⎞ ⎟⎟ vs. cscθ ⇒ –d/λe,1 ⎠ ¾ for a fixed θ: ⎡ I ij ⎤ d = −λe,1 sin θ ln ⎢1 − (1) ⎥ ⎢⎣ I ij,∞ ⎥⎦ ¾ substitute a calculated or measured λe,1 to obtain d 12 3.051J/20.340J Scenario 2: ni,1=0 (ex., M2p peak from underlying metal oxide (MOx): I ij = I (2) ij,∞ ⎛ −d exp ⎜ ⎜ λ sin θ ⎝ e,1 ⎞ ⎟⎟ ⎠ 1 2 ¾ measure Iij for same peak at different take-off angles (θ1, θ2) I ij ,θ1 I ij ,θ 2 I ij ,θ1 I ij ,θ 2 ⎛ −d ⎞ I ij(2),o ni,2 λe,2 sin θ1 exp ⎜⎜ ⎟⎟ sin λ θ 1 ⎠ ⎝ e,1 = ⎛ −d ⎞ I ij(2),o ni,2 λe,2 sin θ 2 exp ⎜⎜ ⎟⎟ sin λ θ 2 ⎠ ⎝ e,1 = ⎛ −d ⎞ sin θ1 csc exp ⎜ csc θ θ − 1 2 )⎟ ⎜λ ( ⎟ sin θ 2 ⎝ e,1 ⎠ d = λe,1 ( cscθ 2 − cscθ1 ) −1 ⎛ I ij ,θ1 sin θ 2 ⎞ ln ⎜ ⎜ I ij,θ sin θ1 ⎟⎟ ⎝ 2 ⎠ ¾ substitute a calculated or measured λe,1 to obtain d d 13 3.051J/20.340J Scenario 3: element present in distinguishable bonding configurations in layers 1 & 2 (ex., O1s peak from -C-O-C- and MOx): ⎛ ⎛ −d I ij = I ij(1),∞ ⎜1 − exp ⎜ ⎜ λ sin θ ⎜ ⎝ e,1 ⎝ I ij(2) I ij(1) ⎞ ⎞ (2) ⎛ −d I exp + ⎟ ⎟⎟ ⎟ ij,∞ ⎜⎜ ⎠⎠ ⎝ λe,1 sin θ ⎛ −d ⎞ I ij(2),o ni,2 λe,2 sin θ exp ⎜⎜ ⎟ λe,1 sin θ ⎟⎠ ⎝ = ⎛ ⎛ −d ⎞ ⎞ I ij(1),o ni,1λe,1 sin θ ⎜1− exp ⎜⎜ ⎟⎟ ⎟⎟ ⎜ λ θ sin ⎝ e,1 ⎠⎠ ⎝ ⎞ ⎟⎟ ⎠ 1 2 (2) (1) ¾ measure element peak areas I ij and I ij (1) (2) ¾ for same element and orbital: I ij ,o = I ij ,o ni,2 ¾ for same element and orbital: n i,1 I ij(2) I ij(1) = Ci,2 Ci,1 ⎛ −d ⎞ Ci,2 λe,2 exp ⎜⎜ ⎟ λe,1 sin θ ⎟⎠ ⎝ = ⎛ ⎛ −d ⎞ ⎞ Ci,1λe,1 ⎜1 − exp ⎜⎜ ⎟⎟ ⎟⎟ ⎜ λ θ sin ⎝ e,1 ⎠⎠ ⎝ ¾ solve numerically for d, substituting calculated values of λe,2 & λe,1 d 14 3.051J/20.340J ¾ if d << λe,1 sinθ : ( ax ) exp(−ax) ≈ 1 − ax + 2 I ij(2) I ij(1) ⎛ d Ci,2 λe,2 ⎜⎜1− ⎝ λe,1 sin θ = ⎛ ⎞ d Ci,1λe,1 ⎜⎜ ⎟⎟ ⎝ λe,1 sin θ ⎠ ⎞ ⎟⎟ ⎠ ⎡ I ij(2)Ci,1λe,1 ⎤ +1⎥ d = λe,1 sin θ ⎢ (1) I C λ ⎢⎣ ij i,2 e,2 ⎥⎦ −1 2 − ... 15 3.051J/20.340J Ion Etching Depth profiling for depths > 10 nm (100 nm – 1 µm) Ar+, Xe+ or He+ ions etch surface layer XPS spectra re-recorded Signal Intensity Ti2p P2p Hydroxyapatite (Ca10(PO4)6(OH)2) coating on Ti implant sputter time Calibration of sputter rates: time ⇒ depth 16 3.051J/20.340J 2. Auger Electron Spectroscopy Theoretical Basis: ¾ Auger electrons created by electron bombardment of sample are ejected from near surface (1-3 nm) with characteristic energies ¾ Analysis of the Auger electron energies yields a quantitative measure of the surface composition ejected core electron Evac EK – EM = EN + Ekin Auger electron Ekin = EK – EM – EN N M Exyz (ECVV or ECCV) LIII LII LI • ejection from x shell • electronic transition y→x • release of z-level Auger with Ekin K INFORMATION: Exyz is characteristic to element & bonding AES vs. XPS Advantages Disadvantages - focused e-beam gives high x,y spatial resolution (5 nm vs. ~1 µm) - charging effects on nonconductive samples (unsuitable) - larger bonding effects - degradation of organics 17 3.051J/20.340J 3. Secondary Ion Mass Spectroscopy (SIMS) Experimental Approach: ¾ Energetic ions (1-15 keV) bombard sample surface ¾ Secondary ions/charged fragments are ejected from surface and detected ion gun energy filter mass filter ion detector sample Ion Guns + types + • Nobel gas: Ar , Xe + + • liquid metal ion: Ga , Cs (~1nm beam size ⇒ x,y mapping) • pulsed LMI (time-of-flight source) • low currents used: 10-8-10-11A/cm2 Ion beam current (A/cm2) 10-5 10-7 10-9 10-11 1 Amp = 6.2×1018 ions/sec surface monolayer lifetime (s) 16 1600 1.6×105 1.6×107 18 3.051J/20.340J Detectors • sensitive to the ratio of mass/charge (m/z) • resolution defined as m/∆m (larger = better!) • Quadrupole (RF-DC): resol. m/∆m ~ 2000; detects m < 103 amu Oscillating RF field destabilizes ions: only ions with specified m/z can pass • Magnetic sector: m > 104 amu; m/∆m ~ 10,000 Applied B-field results in circular trajectory (radius=R) of charged particles R 1/ 2 1 ⎛ 2mV ⎞ R= ⎜ ⎟ B⎝ z ⎠ accelerating voltage • Time-of-flight (TOF): m ~ 103-104 amu; m/∆m ~ 10,000 1/ 2 ⎛ m ⎞ time = ⎜ ⎟ ⎝ 2zV ⎠ L Pulsed primary beam generates secondary ion pulses detected at distance L flight tube length 19 3.051J/20.340J Modes of Operation Static SIMS ¾ low energy ions: 1-2 keV; penetration ~5-10 Å ¾ low ion doses: < 1013 ions/cm2sec 1 cm2 ≈ 1015 atoms 95% of signal from 1st atomic layer! Information: • surface composition • surface bonding chemistry (sputtered fragments) Example: SIMS of silica powder O− Negative spectrum (SiO2)2OH− SiO2− Intensity (arb. units) OH− 16 17 − SiO3 SiO3H− SiO2H− 60 61 76 77 137 m/z OH Which structure is suggested from SIMS? Si O O O OH HO Si O O 20 3.051J/20.340J Positive spectrum CH3+ OH+ Intensity (arb. units) O+ H2O+ 15 16 SIMS suggests presence of adsorbed methanol 17 18 m/z SIMS vs. XPS/AES Advantages - high sensitivity (ppm – ppb) - more sensitive to top surface - applicable to any solid Dynamic SIMS ¾ 1-20 keV primary beam ¾ rastered beam sputters a crater in sample ¾ secondary ions gives depth profiling Disadvantages - not quantitative 21 3.051J/20.340J primary beam path 10-100 µm 10-100 µm N ion-implanted Ti (for wear resistance) Signal Intensity N− sputter time References J.C. Vickerman, Ed., Surface Analysis—The Principle Techniques, J. Wiley & Sons: New York, NY, 1997, Chapters 3, 4 & 5. D. Briggs, Surface Analysis of Polymers by XPS and Static SIMS, Cambridge University Press: United Kingdom, 1998. C.-M. Chan, Polymer Surface Modification and Characterization, Hanson: Cincinnati, OH: 1994, Chapters 3 & 4. P.J. Cumpson, “Angle-resolved XPS and AES: depth-resolution limits and a general comparison of properties of depth-profile reconstruction methods”, J. Electron Spectroscopy 73 (1995) 25­ 52. B. Elsener and A. Rossi, “XPS investigation of passive films on amorphous Fe-Cr alloys”, Electrochimica Acta 37 (1992) 2269-2276. A.M. Belu, D.J. Graham and D.G. Castner, “Time-of-flight secondary ion mass spectroscopy: techniques and applications for the characterization of biomaterial surfaces”, Biomaterials 24, (2003) 3635-3653.