Ultraviolet/ Visible Absorption Spectroscopy

CHEM*3440 Ultraviolet/ Visible Absorption Spectroscopy Widely used in Chemistry. Perhaps the most widely used in Biological Chemistry. Easy to do. Very easy to do wrong. Understand your experiment. CHEM 3440 F08 1 UV/Visible Electronic Transitions This is mainly a study of molecules and their electronic transitions. Molar Absorptivity (ε) ranges from 0 to 10 5 for use in absorbance measurements. Transitions with ε < 10 3 are considered to be of low intensity. In organic molecules, most bonding electrons are excited by λ < 185 nm (VUV). Recall that E(eV) = 1239 / λ (nm) Most functional groups have lone pairs whose energies place them in the near UV and visible range. These groups are called chromophores, although this is a bit odd, since all molecules are chromophores under the right conditions. CHEM 3440 F08 2 1

Valence Electronic Structure The valence electrons are the only ones whose energies permit them to be excited by near UV/ visible radiation. Energies are in 2-6 ev regime. σ (anti-bonding) π (anti-bonding) n (non-bonding) π (bonding) Four types of valence transitions σ σ* π π* n σ* n π* σ (bonding) * * * this classification does not include Rydberg transitions, which are observed at higher energies (> 7 ev). Rydberg excitations promote a valence electron into a highly energetic hydrogenic orbital, characterized by s,p,d,f type symmetries rather than the complex molecular orbitals associated with the valence transitions. The lowest transitions of H 2 O and the alkanes, are good examples of transitions involving Rydberg states CHEM 3440 F08 3 n σ* Transitions Still rather high in energy, with λ between 150 and 250 nm. These tend to be relatively weak absorbers. Not many molecules with n σ* transitions in UV/ vis region λ max ε max H 2 O (partial Rydberg) 167 1480 CH 3 OH 184 150 CH 3 Cl 173 200 CH 3 I 258 365 (CH 3 ) 2 S 229 140 (CH 3 ) 2 O 184 2520 CH 3 NH 2 215 600 (CH 3 ) 3 N 227 900 CHEM 3440 F08 4 2

n π* and π π* Transitions Most UV/vis spectra involve these transitions. π π* are generally more intense than n π*. λ max ε max type C 6 H 13 CH=CH 2 177 13000 π π* C 5 H 11 C C CH 3 178 10000 π π* O CH 3 CCH 3 186 1000 n σ* O CH 3 COH 204 41 n π* CH 3 NO 2 280 22 n π* CH 3 N=NCH 3 339 5 n π* CHEM 3440 F08 5 Solvent Effects - Shifts Solvents can interact with the analyte molecules and shift absorbance peaks and intensities. Red Shift ( Bathochromic) Peaks shift to longer w avelength. Blue Shift ( Hypsochromic) Peaks shift to shorter w avelength. n π* generally blue shifted by solvent; solvation of and hydrogen bonding to the lone pair. Large shifts (up to 30 nm). Both n π* and π π* red shifted; attractive polarization forces, increase with increasing solvent polarity. Small shifts (less than 5 nm). CHEM 3440 F08 6 3

Solvent Effects - Intensity Solvents can also induce significant changes in the intensity of peaks. Hyperchromic Increase in absorption intensity. Hypochromic Decrease in absorption intensity. Absorption characteristics of 2-methylpyridine Solvent λ max ε max Hexane 260 2000 Chloroform 263 4500 Ethanol 260 4000 Water 260 4000 Ethanol - HCl (1:1) 262 5200 Note that the wavelength shifts are small (2-3 nm), while the intensities can vary by factors of 2-3. What are the implications for the Analytical Chemist? CHEM 3440 F08 7 Auxochrome Auxochrome : Substitutent groups which are not themselves optically active in this energy range, but which do interact with other chromophores to shift both intensity and wavelength. Associated with redistribution of internal electronic configurations and charge densities Absorption Characteristics of Pyridine Derivatives Derivative λ max ε max Pyridine 257 2750 2-CH 3 262 3560 3-CH 3 263 3110 4-CH 3 255 2100 2-F 257 3350 2-Cl 263 3650 2-I 272 400 2-OH 230 10000 CHEM 3440 F08 8 4

Typical Organic UV/Vis Spectra Note how the same molecule displays markedly different spectral features depending upon the physical environment in which it finds itself. In the vapour and hexane environments, there are peak clusters at 510, 527 and 550 nm. Each has fine structure in the vapour phase. Q1 : What is the approximate energy difference between these clusters? Q2: What is the probable physical origin of the clusters? Q3 : what is the probable physical origin of the fine structure? Q4 : Why is the fine structure lost in the solvated species? CHEM 3440 F08 9 Inorganic - Transition Metals Spectra from transition element ions arise from the 3d and 4d electrons. These spectra are quite broad, often in the visible (solutions are brightly coloured), and are significantly affected by ligands and solvents. Ligand Field Theory is a molecular orbital approach to understand these effects. Essentially it asserts that the d- orbital energies are split in solution (or with ligands) and transitions between these split levels are at the source of their UV/ Vis spectroscopy. Fig. 14-7 in Skoog, Holler, Nieman. CHEM 3440 F08 10 5

Inorganic - Lanthanides UV/Vis spectra of lanthanide ions arise from 4f levels (5f for the actinides). These are rather well-screened from outside influences and as such the spectra are comparatively sharp, and only weakly influenced by ligands and solvents. What are the implications for an Analytical Chemist? Fig. 14-6 in Skoog, Holler, Nieman. CHEM 3440 F08 11 Inorganic - Charge Transfer Inorganic complexes metal ions with surrounding ligands can undergo absorption processes where the electron jumps from an orbital mostly centred on the ligand to an orbital mostly centered on the metal ion (the opposite can occur, but less frequently). These transitions are intense. Fig. 14-10 in Skoog, Holler, Nieman. CHEM 3440 F08 12 6

Instrumentation Spectrometric instruments have a common set of general features. Often, one technique is distinguished from another by differences in these features. Here we look at specific features for the UV/Visible experiment. Sources: D 2 lamp, W filament (halogen lamp), and Xe arc lamp. Wavelength Selectors: Filters and Monochromators. Sample Containers: Fused silica, quartz, and glass. Detectors: Phototube, PMT, photodiode, photodiode array, CCD array CHEM 3440 F08 13 Sources - Deuterium Lamp Note the absence of sharp resonance bands Strike a low voltage DC arc in a lamp filled with D 2. Gives continuum emission from 160 to 400 nm. D 2 + e ( ) D * 2 + e E kinetic D * ' 2 D + D + E kinetic D + D + M D 2 + M + hν Deuterium lamps from Photron CHEM 3440 F08 14 7

Sources - Tungsten Filament Reflecting focusing assembly Lamp Condenser Lens A heated W filament, gives off blackbody radiation. Add a small amount of a halogen gas (usually Br 2 ). Sublimated W reacts with halogen to form tungsten halide; does not deposit on quartz cover (no blackening) but does redeposit on filament (extends life). Relative Intensity100 0 3400 K 3000 K 2600 K 2200 K 500 1500 2500 Wavelength (nm) CHEM 3440 F08 15 Sources - Xe lamp Tube filled with Xe (or sometimes a mixture of Hg and Xe), invented in 1940, commercialized in 1961 by Osram. Pass a low voltage DC current to excite Xe. The broad spectral output closely resembles natural daylight, and is often used in projection systems (e.g. 15 kw IMAX systems) UV region Visible region 150 Watt Xe lamp by Alpha source Note the sharp resonance structures superimposed on continuum background. This requires greater stability in spectroscopic applications. CHEM 3440 F08 16 8

QuickTime and a TIFF (Uncompressed) decompressor are needed to see this picture. Optical Filters 100 Transmittance Band-reject interference filter 0 400 Wavelength (nm) 750 100 Band pass interference filter Transmittance 0 A filter set from Andover Corp. Filters can absorb light with dye molecules incorporated into the glass or gel. They can also pass or reject bands of light because of interference effects with multiple layers of materials. Dye or interference filters can select a more narrow region of light to allow through to a detector. Useful in non-scanned situations and when a specific, known band of radiation needs to be monitored. 100 Transmittance Two absorption filters BG-23 0 300 Wavelength (nm) 800 BG-38 CHEM 3440 F08 17 Monochromator-1 A monochromator disperses the light in order to select a narrow bandwidth. Both gratings and prisms can be used for this dispersion. Numerous instrumental designs available to account for various optical aberrations. Entrance Slit First (Collimating) Mirror Reflection Diffraction Grating Exit Slit Second (Focusing) Mirror A couple of informative sites regarding monochromators. http://www.shsu.edu/~chemistry/primers/mono.html http://www.monochromator.com/jy/oos/oos_ch1.htm#1.1 CHEM 3440 F08 18 9

Monochromator-2 White light enters It is collimated to fill the grating The grating disperses the light into its separate colours. Only a narrow band of wavelengths are dispersed into the precise angle to fall upon the exit slit. Only a limited colour range passes through the exit slit. CHEM 3440 F08 19 Echellette Grating Has been most common grating. Ruling engine cuts sawtooth grooves in a master. Replicas are cast from the master. Surface normal vector Θ Θ Vector normal to groove face. Blaze angle Two main defining parameters are grooves/ mm and blaze angle. CHEM 3440 F08 20 10

How A Grating Works Collimated (light rays all traveling parallel to each other) white light irradiates the grating surface. This shows monochromatic light (one wavelength) for simplicity. Also, it is shown coming in normal to the grating plane, but it does not have to. CHEM 3440 F08 21 How A Grating Works - 1 Each facet of the grating becomes a source of spherical re-emission of all of the scattered light. These spherical wavelets look like plane w aves over small arcs when far enough away from surface. CHEM 3440 F08 22 11

How A Grating Works - 2 Zero-order First order Second order Third order In the far field (many wavelength s distance from the surface) the light only propogates in directions where there is constructive interference between wavelets. Different directions are called diffraction orders. This depends upon the spacing between grooves (groove density). CHEM 3440 F08 23 How A Grating Works - 3 Previous first order direction Different first order direction When light of a different wavelength impinges upon the grating, its diffraction direction changes for all non-zero order diffraction processes. It is this difference that leads to the dispersion of light of different wavelengths. CHEM 3440 F08 24 12

How A Grating Works - 4 When the incident light contains two wavelengths, one exactly half the wavelength of the other, their orders overlap scattering into the same angle. Second order red Fourth order green First order red Second order green Third order green First order green CHEM 3440 F08 25 Grating Equation Most fundamental equation for gratings: sin α + sin β = k N λ α β k = diffraction order N = groove density λ = radiation wavelength Note : For those who will take CHEM 4400, there is an alternative way to consider this (apparently) simple equation that provides more physical insight CHEM 3440 F08 26 13

Angular and Linear Dispersion A grating s effectiveness is in its ability to have two different wavelengths leave its surface at different angles. A direct measure of that property is its angular dispersion, the range of angles into which a range of wavelengths are distributed. The larger the angular dispersion, the better is the grating. Within reason dβ dλ = k N cosβ Most monochromators scan a grating s output across a fixed slit at some fixed distance from the grating. In this case, the more direct measure of grating effectiveness is linear dispersion what is the distance at the slit over which a range of wavelengths are distributed. knl exit Linear dispersion can be improved by -increasing the groove density (N) -increasing the diffraction order (k) -increasing the focal length (L) dx = dλ cos β Distance to the exit slit (focal point). Note : pay special attention to the definition of the linear and angular dispersions. As written here, THESE definitions of dispersions are per nm, and we want to have it as LARGE as possible (i.e. disperse the wavelengths across as large a given physical width on the optical slit as possible). This way, the light that gets through the small slits will be highly monochromatic. CHEM 3440 F08 27 Resolving Power The resolving power further shows how the size of a grating can impact on the resolution of a grating. Why would the size of the grating affect its ability to separate wavelengths? R = λ dλ = knw grating Illuminated width of the grating. As an example, for a grating with N = 1200 lines/mm and a grating width of 110 mm, and working in 1st order: R = 1200 x 110 = 132,000 Therefore, at 500 nm, dλ = 500/132,000 = 0.0038 nm for the bandpass of the grating. CHEM 3440 F08 28 14

Holographic Grating Ruling errors with gratings cut using a ruling engine. Produces ghosts, very weak replicas of intense lines at unexpected wavelengths. Photoresist layer exposed to diffraction pattern from laser. Chemical and Ion etching to imprint grating pattern onto substrate. Rectangular (non-blazed) Sinusoidal Can be made in sawtooth (blazed) configuration. CHEM 3440 F08 29 Echelle Grating Just like an echellette grating, except irradiate on short edge of sawtooth and grating density is much lower (30-200 lines/mm). Use in very high order, however (20-120th order). Grating has much higher dispersion than echellette of same size. Needs additional dispersion element (often a prism) to help sort out the orders. Scanning through the spectrum can involve moving through 90 consecutive orders of diffracted radiation. Example with light at 300 nm: Echellette, 1200 lines/mm, first order, resolution 62,400, linear dispersion 16 Å/mm. Echelle, 79 lines/ mm, 75th order, resolution 763,000, linear dispersion 1.5 Å/mm CHEM 3440 F08 30 15

Sample Containers Successful spectroscopy requires that all materials in the beam path other than the analyte should be as transparent to the radiation as possible. Also, the geometries of all components in the system should be such as to maximize the signal and minimize the scattered light. The material from which a sample cuvette is fabricated controls the optical window that can be used. Some typical materials are: Optical Glass - 335-2500 nm Special Optical Glass 320-2500 nm Quartz (Infrared) 220-3800 nm Quartz (Far-UV) 170-2700 nm Keep the cuvette clean. Don t clean with paper products (Kim-wipe); use optical paper. Store dry. Don t get finger prints on them. Store carefully and gently. CHEM 3440 F08 31 Detectors - Phototube Almost without exception, a photodetector is needed to convert photons into electrons (figuratively speaking) which can be measured and processed electronically. CHEM 3440 F08 32 16

Phototube - 1 The basic phototube is packaged in vacuum and presents a photocathode covered with a particular photoemissive material. These electrons are accelerated by a voltage of ~ 90 V towards an anode which collects the ejected electrons and an external meter measures the current that flows. photon photocathode anode Note that the work function of most materials is 2.5-3 ev. This means that the highest signal that could be expected using UV/VIS radiation would be ~1-2 electrons/photon. This is very low, when compared to the devices to follow CHEM 3440 F08 33 Detectors - PMT PMT s are for low light level situations. Significant gain (10 5-10 8 ) in electron flow through secondary electron emission at dynodes. This is a vacuum photomultiplier tube (PMT), based on the original 1936 design. light photochathode dynodes electrons anode voltage divider network high voltage Secondary Electron Emissive Materials for Dynodes BeO(Cs) GaP (Cs) MgO(Cs) Cs 3 Sb KCl Note that while electronic amplification (gain) of 10 5-10 8 is possible using OpAmps, these always include very large feedback resistors, which have Johnson noise. PMTs have no Johnson noise. CHEM 3440 F08 34 17

Channeltron A continuous dynode chain is built into a single unit. Excellent and widely used electron multiplier. If the front end is a photoemissive surface then you have a compact PMT. Channeltrons require high vacuum to operate. Historical Note : PMTs were largely developed by RCA until the 1986, when RCA was broken up and purchased by General Electric. The PMT line was then acquired by Eric Burlefinger in 1987 to create Burle Industries. Galileo developed the Channeltron for NASA; Galileo has subsequently been purchased by Burle Industries. CHEM 3440 F08 35 Multichannel Plate (MCP) The channeltron concept has been extended to an array of micron-sized holes in a glass plate. Each tunnel functions like its own channeltron. light photoemissive coating electrons high voltage It is an excellent electron amplifier. The output goes to an anode for recording. A photoemissive front surface makes it a photon detector. Can be used as an imaging device also (this is the basis for night-vision goggles). Like the PMT and the Channeltron, there is no Johnson noise to consider. CHEM 3440 F08 36 18

Detectors - Photodiode A photodiode is formed by sandwiching an undoped layer of Si between a heavily doped p-layer and a heavily doped n-layer. Photons whose wavelength is between 400 nm and 1100 nm can be absorbed in the intrinsic layer, producing an electron-hole pair. The bias potential sweeps these carriers to the opposite regions, producing a current in the external circuit. Photodiodes are more sensitive than phototubes (yawn), but far less sensitive than PMT s, since they only generate ~ 1 electron-hole pair per photon. On the other hand they are about the size of a transistor (~ 1 mm 2 ) and require no high voltage support. CHEM 3440 F08 37 Detectors - Photodiode Array An array of photodiodes can be fabricated using large-scale integration methods. When this placed in the image plane of a spectrometer, wide ranges of the spectrum can be simultaneously acquired. This greatly increases the speed of acquisition, enabling longer exposure times at each wavelength, and hence improved S/N. The S4111 series devices from Hamamatsu. Critical parameters are the number of elements (256 possible), the size of the device, its sensitivity, spectral range, rise time. CHEM 3440 F08 38 19

Detectors - CCD Array Another solid state silicon device. A charged gate collects either the holes or electrons generated by the absorption of a photon. Charge accumulates in the potential well for as long as exposure is maintained. This device integrates the light levels between readouts. This is very useful for low light applications. Device is read-out by charge injection (CID) or charge transfer (CCD). Comparable sensitivity to PMT, but functions as array detector also. 1024 x 256 back illuminated UV sensitive CCD detector from Jobin Yvon. Pixel size is 26 x 26 µm. These are often used in imaging applications such as astrophotography CHEM 3440 F08 39 Quantitative Analysis with UV/Vis UV/Vis spectroscopy - and the other spectroscopic techniques we will discuss - can all be used to analyze a sample for analyte concentration. Need to know LOQ and LOL in order to define useful region of applicability. 1. Calibration Curve 2. Standard Addition (matrix effects/ complex sample) 3. Internal Standard (compensate for random and systematic errors; difficult) CHEM 3440 F08 40 20

Calibration Curves: What do they tell us? Consider the analysis of, for instance, Co 2+ (strong absorption band at ~ 500 nm). Concentration (ppm) Average Signal Replicants Standard Deviation 0.00 0.0363 25 0.0037 2.13 0.111 5 0.0091 5.63 0.268 5 0.0108 10.29 0.503 5 0.0103 15.33 0.715 5 0.0129 21.71 1.016 5 0.0141 CHEM 3440 F08 41 Calibration Curve - 1 1. Use a Least Squares approach to find the parameters for this curve and plot the curve. 2. Use Excel (or similar) spreadsheet. 1. Trendline 2. LINEST function m = 0.0456 b = 0.0231 m is slope of line and encodes how the instrument responds to sample concentration. b is the y-intercept. It is the magnitude of the blank that will be subtracted from every measurement. m = x i y i 2 x i x i N ( ) 2 x i N 1.2 1 0.8 0.6 0.4 0.2 y i 0 y = 0.0456x + 0.0231 R 2 = 0.9992 0 5 10 15 20 25 Concentration ( ppm ) y b = i N m x i N CHEM 3440 F08 42 21

Calibration Curve - 2 What is the standard deviation for b and m? Part of the output from LINEST in Excel or Use the following equations. s r = 2 y i ( ) 2 y i N 1 s m = s r 2 x i x i N ( ) 2 2 x s b = s i r 2 N x i x i ( ) 2 N 2 x i y i 2 x i x i N ( ) 2 x i N y i 2 CHEM 3440 F08 43 Calibration Curve - 3 What is the calibration sensitivity for the transfer function? This is simply m, the slope of the least squares fit: m= 0.0456 What is the analytical sensitivity? This value changes with concentration so we obtain a value of g for each data point in the plot. γ i = m s i Concentration 2.13 5.63 10.29 15.33 21.71 Analytical Sensitivity 5.0 4.2 4.4 3.5 3.2 What is the LOD (Limit of Detection)? L.O.D. = 3 s blank m What is the LOQ (Limit of Quantitation)? L.O.Q. = 10 s blank m = 3 0.0037 0.0456 = 10 0.0037 0.0456 = 0.24 ppm = 0.81 ppm CHEM 3440 F08 44 22

Calibration Curve - 4 What is the concentration of an unknown sample when the mean signal measured for 5 replicated measurments is 0.447? c x = 0.447 b m What is the standard deviation of this measurement? ( M = 5). s x = s r m 1 M + 1 N + = 0.447 0.0231 0.0456 ( c x y ) 2 m 2 2 x i x i ( ) 2 N What are the 95% confidence limits for this measurement? = 9.3 ppm recall that we made 5 measurements for this unknown. From Student s t- table, t = 2.78 for this number of measurements at this confidence level. ( )( 0.16) = 0.20 ppm 5 ts N = 2.78 C x = 9.3 ± 0.2 ppm with 95% confidence = 0.16 ppm CHEM 3440 F08 45 Standard Addition Most samples are not clean ; there are a many other components to the matrix that may interfere either chemically or physically. Recreating the exact environment to use the calibration curve method can be difficult and even impossible. Standard Addition method is a good attempt around that. 1. Take an aliquot of the sample into a volumetric flask, add any needed additional components (ph buffer, complexing agent, etc.), and dilute to the final volume. Take another volumetric flask and introduce an identical aliquot and same treatments. However, in addition, introduce a volume of a known standard solution of the analyte. Then dilute to volume. 1 2 CHEM 3440 F08 46 23

Standard Addition - 2 Measure the signal (S 1 and S 2 ) for both. The signal depends upon the concentration of the target analyte, which is simply diluted to the final volume, V t. The experiment has some response factor k that relates the concentration to the signal amplitude. We can write S 1 = k V x V t c x S 2 = k V x c V x + V s c t V s t From these two measurements, we can solve for the unknown concentration, and obtain the expression c x = S 1 ( ) S 2 S 1 V s V x c s CHEM 3440 F08 47 Standard Addition - 3 The previous scheme only required two measurements. Better results are made by spiking several samples with varying volumes of added standard. 1. Form one sample with the unknown as before. 2. Form a series of samples with unknown and increasing volumes of added standard (perhaps 5 total samples). These 5 samples are measured, each giving its own signal S i. These data can be graphed as signal against added standard volume. The slope and intercept can be determined and used to calculate the unknown concentration. CHEM 3440 F08 48 24

Standard Addition - 4 Consider the following data. The volume of the unknown and each standard increment is 5.00 ml. The standard has a concentration of 8.7 ppm. Added Volume (ml) 0.00 5.00 10.00 Signal (arbitrary) 0.251 0.422 0.617 1.2 1 0.8 0.6 Standard Addition Method y = 0.035x + 0.2548 R 2 = 0.9994 15.00 0.785 0.4 20.00 0.957 0.2 25.00 1.121 0 0 5 10 15 20 25 Volum e of Added Standard ( m L) CHEM 3440 F08 49 Standard Addition - 5 A least squares analysis gives the slope and intercept (m and b) for this straight line. The equations solve to give an expression for the unknown concentration as c x = b mv x c s = 0.2548 8.7 = 12.7 ppm ( 0.035) ( 5.00) With the equations given previously, we can find the error in m and b and assuming that those errors dominate, we can find the error in our result as 2 s s c = c m x + s 2 b = c 0.00045 m b x 0.03498 = 12.7 0.0129 = 0.16 ppm Use Student s t-table for 95% and 5 samples. We would report the final answer as c x = 12.7 ± 0.20 ppm at 95% confidence 2 + 0.00018 0.2548 2 CHEM 3440 F08 50 25

Standard Addition - 6 A graphical solution can be easily obtained. Extrapolate the curve back to the x-axis (in the negative-x region). This x-intercept represents the volume of standard solution which has the same amount of analyte as the unknow n solution. Standard Addition Method -V 0,s c s = V x c x 1.2 1 y = 0.035x + 0.2548 R 2 = 0.9994 c x = -(V 0,s /V x ) c s = -(-7.28/5.00) 8.7 ppm =12.7 ppm 0.8 0.6 V 0,s = -7.28 ml 0.4 0.2 0-10 -5 0 5 10 15 20 25-0.2 Volum e of Added Standard ( m l) CHEM 3440 F08 51 Internal Standard In this experiment, a substance different from the analyte is added in equal amounts to all unknowns and calibration standards. We measure both the analyte signal and internal standard signal for all samples and calculates the ratio of analyte signal to internal standard signal and plots this ratio against the standard analyte concentration, forming a calibration curve. The rest proceeds as with any calibration curve. This procedure can correct for many matrix interferences. The major problem is finding the right internal standard and adding it in a reproducible manner for all standards and unknowns. Best procedure is when internal standard is an isotope of the analyte. Then all interfering chemistry is identical. But must not occur naturally and the two species must be measureable and distinguishable. CHEM 3440 F08 52 26