Resonance Enhancement of Raman Spectroscopy: Friend Or Foe?

www.spectroscopyonline.com

Electronically reprinted from June 2013 Volume 28 Number 6

Molecular Spectroscopy Workbench Resonance Enhancement of Raman Spectroscopy: Friend or Foe?

The presence of electronic transitions in the visible part of the spectrum can provide enor- mous enhancement of the Raman signals, if these electronic states are not luminescent. In some cases, the signals can increase by as much as six orders of magnitude. How much of an enhancement is possible depends on several factors, such as the width of the excited state, the proximity of the laser to that state, and the enhancement mechanism. The good part of this phenomenon is the increased sensitivity, but the downside is the nonlinearity of the signal, making it difficult to exploit for analytical purposes. Several systems exhibiting enhancement, such as carotenoids and hemeproteins, are discussed here.

Fran Adar

he physical basis for the Raman effect is the vibra- bound will be more easily modulated. So, because  tional modulation of the electronic polarizability. electrons are more loosely bound than  electrons, the T In a given molecule, the electronic distribution is polarizability of any unsaturated chemical functional determined by the atoms of the molecule and the electrons group will be larger than that of a chemically saturated that bind them together. When the molecule is exposed to group. Figure 1 shows the spectra of stearic acid (18:0) and electromagnetic radiation in the visible part of the spec- oleic acid (18:1). These two free fatty acids are both con- trum (in our case, the laser photons), its electronic dis- structed from a chain of 18 carbon atoms, in one case fully tribution will respond to the electric field of the photons. saturated and, in the other, with one carbon double bond Then, as the atoms move during a vibration, the electronic between the ninth and 10th carbon atom from the methyl polarizability will be modulated. This modulation of the end. In addition to this chemical difference, the saturated electronic polarizability is what gives rise to the Stokes fatty acid is solid, whereas oleic acid is liquid. The liquid and anti-Stokes Raman spectrum. form accounts for the relative broader lines and some dif- If you think about it, electrons that are more loosely ferences in the band ratios. However, the most important Spectroscopy 28(6) June 2013 www.spectroscopyonline.com

400 1298 1439 350 1063 1129 300 1461 1466 1423 250 1104 1413 1370 200

150 1656 1440

100 1302 1459 Intensity (arbitrary units) 1086 1264 1064

50 1119

0 600 800 1000 1200 1400 1600 Raman shift (cm-1)

600 2880

500 284 3 400 2899 2923 300 2935 285 1 2962 289 7 287 2 200 2930

Intensity (arbitrary units) 100 3009

2600 2700 2800 2900 3000 3000 3000 3300 3400 Raman shift (cm-1)

Figure 1: Raman spectra of stearic acid (red) and oleic acid (blue).

difference for this discussion is the clear. There is no electronic transition represent resonant conditions. The presence of the analytical bands for in the visible part of the spectrum resonance condition for Raman scat- the -C=C- stretch at 1655 cm-1, and where the Raman effect is being ex- tering occurs when there is a transi- the CH stretch on the –C=C- bond cited and transmitted. When there tion in the vicinity of the energy at 3009 cm-1. Assuming the signal are such transitions things become of the laser photon. Because most strengths reflect the number of func- much more complicated. Aside from work, especially analytical work, is tional groups, one would expect the needing to worry about the sample done with visible lasers, this means –C=C- band to be weaker than the self-absorption (similar to the issue in a visible transition, which will be an C-C bands. In fact, the integrated the near infrared [NIR]), there is the electronic transition because typical intensity between 1600 and 1700 cm-1 enhancement issue that we want to visible photon energies (1.5–3 eV) are (C=C) is twice as strong as the inte- explore in this column. in the range of electronic transitions. grated intensity between 1030 and For the purposes of this article we 1140 cm-1 (C-C) even though there Resonant Raman Scattering will discuss molecular systems with  are 16 single bonds versus 1 double The previous example illustrates electrons, although there are certainly bond! The spectrum of an unsatu- how the presence of  electrons can other classes of electronic transitions rated fatty acid is dominated by the increase the polarizability of a mol- in the visible part of the spectrum. -C=C- stretch. ecule, thus increasing the Raman For unsaturated molecular systems These compounds are still optically cross-section. But this does not yet to produce visible optical transitions, www.spectroscopyonline.com June 2013 Spectroscopy 28(6) the  electrons must be “conjugated,” at about 200 nm (~6 eV). As double ment of alternating single and double which means the molecule must con- bonds are added to a conjugated bonds. The result is that degraded tain multiple carbon double bonds chain, the energy of the lowest energy PVC has polyenes with a population separated by single bonds. These mol- absorption occurs at decreasing ener- of sequence lengths. If you examine ecules can be linear (called polyenes) gies (or increasing wavelengths). The such materials visibly they may be or they can be represented by closed free electron model has been used to tinged orange or there may be no rings that are called “aromatic,” cer- describe these states: The electrons visible coloration. But if you record a tainly because of the characteristics of are approximated as particles in Raman spectrum the polyene can be some of the earliest known examples a box defined by the length of the observed with very high sensitivity, such as benzene. Toward the end of chain (1). As the box becomes longer, even when there is no evidence for its the 19th century and into the begin- the energy of the states become lower, presence optically. What is addition- ning of the 20th century, there was with an asymptotic long wavelength ally interesting is that the observed development of synthetic textile dyes value of about 600 nm. Raman frequency depends on the based on conjugated molecules, often There is an analogous phenom- excitation wavelength because each containing N atoms; this industry enon in describing the vibrational wavelength excites a different part actually represents the beginnings states (2). An isolated double bond of the population with its own reso- of organic chemistry. Very often the has a frequency of about 1650 cm-1 nance values. excited electronic states of dye mol- and a single bond has a frequency The second surprising instance ecules are long-lived and result in near 1000 cm-1. But in a conjugated where I observed a carotenoid oc- decay via radiative emission, which system the double bond frequency curred while I was working with is known as fluorescence and phos- is lowered, and the single bond Professor Gene Hall of Rutgers Uni- phorescence. Fluorescence tends to be frequency is raised (because of the versity. Professor Hall was studying a an intense process (quantum yields averaging of the bond order). As the series of fish oil supplements. When of the order of 0.5). Because Raman chain is lengthened, these shifts will he examined a krill oil, he found that quantum yields are of the order of increase until the frequencies hit the spectrum was overwhelmed by 10-6, any molecule that fluoresces will “limiting” values of about 1495 and the spectrum of a carotenoid, even probably not be a good candidate for 1150 cm-1. though the sample was colorless. Raman measurements; the Raman All of this means that there is a A more careful examination of the signal will be overwhelmed by its correlation between the conjuga- spectrum, while referring to the Mer- “fluorescence background.” tion length, the optical absorption lin publication that we cited earlier Examples that exhibit beautiful res- wavelength, and the Raman vibra- (2), indicated the presence of astax- onance-enhanced spectra are taken tional frequencies. However, there anthin, a well known natural carot- from biomolecules — carotenoids and is actually more than a correlation. enoid that differs from -carotene hemes. They illustrate the sensitivity The presence of the optical absorp- in that there are carbonyl and OH of the technique, as well as the infor- tion enables the resonance Raman functionalities on each of the ring mation that can be derived. They also phenomenon, and the resonance is end groups. illustrate that the symmetry of the quite sensitive to the excitation wave- It is useful to review briefly the molecule as well as the electronic and length; that is, the important param- mechanism by which the spectra of vibrational states determine the mode eter in the resonance phenomenon is polyenes is enhanced. The  to * of enhancement and the types of vi- the proximity of the laser wavelength electronic transition is an “allowed” brations that are enhanced. to the absorption maximum. So, if transition that is polarized along the you are able to tune the laser wave- axis of the molecule, so vibrations Carotenoids length, you will see a variation in the that couple to it most effectively have Carotenoids are chains of alternating resonance enhancement that follows symmetric character. But not all vi- double and single bonds. -Carotene, the visible absorption spectrum. brations will be resonance enhanced. the orange pigment in carrot root, The resonance enhancement of A vibration will be resonance Raman is the quintessential carotenoid, al- the Raman signals of polyenes shows active if there is a change in bond though there are many others, such up in surprising circumstances. The length or angle in the excited elec- as lycopene, the red pigment in to- first example that comes to mind is tronic state compared to the elec- matoes and red peppers. The visual that of degraded polyvinyl chloride tronic ground state (Franck Condon pigments are also carotenoids. In (PVC) (3). It turns out that during effect). In fact, because the  to * fact, retinal can be conceptualized as the degradation of PVC there is an excitation has the effect of shortening -carotene cleaved in half and then extraction of H and Cl across one of the single bonds and lengthening the terminated with an aldehyde group. the polymer backbone bonds result- double bonds, it is the stretching vi- Figure 2 illustrates the structures of ing in a  bond. Interestingly, the brations of the carbon-carbon bonds -carotene and retinal. extraction tends to occur on adjacent in the chain that will have the great- An isolated double bond absorbs bonds so that one ends up with a seg- est resonance Raman intensity (4). Spectroscopy 28(6) June 2013 www.spectroscopyonline.com

aromatic molecule even though it is not based on aromatic six-membered β-Carotene rings. Porphine, on which porphyrin is based, is a tetrapyrrole shown in Figure 3. Molecular orbital theory argues that the inner ring of 16 atoms has 14  electrons plus four out-of- All-trans-retinal plane N electrons, with the outer  electrons being essentially ethylenic CH CH H3CCH3 3 3 O (6). H The electronic absorption spectrum of porphine consists of a strong CH 3 band near 400 nm (termed the B or Soret band) and four weaker bands Figure 2: Structures of -carotene and all-trans retinal. between 500 and 900 nm (the Q bands). In the metal complex, the four weaker bands in the visible part of the spectrum collapse into two bands. This has been interpreted to be evidence for twofold symmetry in the porphine, and fourfold symmetry NN NH N in the metal–porphine complex. This spectral behavior has been explained M in terms of a cyclic polyene model (6). Note that the hemes that occur in he- N HN N N moglobin, myoglobin, and the cytochromes all contain side groups that, in principle, lower the symmetry of the porphine core. But the electronic spectra have still been successfully Figure 3: Structure of porphine (left) and porphine complexed with a metal as in heme or explained by the fourfold symmetry chlorophyll (right); note that metal complex with porphine has two tautomeric equivalent forms model (except for chlorophyll, the (second not shown) resulting in fourfold symmetry. Mg complex of a substituted chlorin, in which the  system’s symmetry is This becomes interesting in the and then reisomerization back to the lowered by some saturation). The Q study of the biological polyenes be- light-adapted form (4). Isomeriza- and Soret absorptions were respec- cause there are important biological tion can be followed in the resonance tively assigned to transitions between processes that depend on isomeriza- Raman spectra because isomeriza- the two highest filled orbitals and the tion (4). There are two light-sensing tion changes the conjugation of the lowest empty orbital. Martin Gouter- systems that function because of  electrons, which will shift the elec- man then included the evolving tech- isomerization. In vertebrate eyes, the tronic absorption maximum and the niques of molecular orbital theory to pigment for black and white vision resonance Raman observed frequen- include configuration interactions has an 11-cis retinal chromophore cies. What is even more interesting is to correct for the neglect of electron bound to lysine on a particular pro- that dynamic Raman and hole burn- correlations. Using group theory tein by a protonated Schiff base bond. ing absorption studies (on the pico- he found that the excited B and Q In photosynthetic bacteria there is second time scale!) have been used to states have the same symmetry, so a bacteriorhodopsin containing an follow cis–trans photoisomerization they mixed. In addition, there can be all-trans retinal protonated Schiff (5). vibronic mixing between the B and base. During the visual process, the Q states, but because these states are vertebrate rhodopsin undergoes a se- Hemeproteins already orthonormal, only asymmet- ries of bleaching intermediates until Hemeproteins form another class ric vibrations can mix the states. In the chromophore is separated from of biomolecule that experiences a this Hertzberg-Teller coupling, the Q the protein in the all-trans isomer. In strong resonance enhancement. The band acquires intensity from the B or photosynthetic bacteria light absorp- heme, which is the part that is inter- Soret band via vibronic mixing. And tion results in conversion to the 13- acting strongly with visible light, is this same Hertzberg-Teller coupling cis isomer, driving the translocation an iron complex of porphyrin. The accounts for the resonance in the of protons across the cell membrane, porphyrin molecule is classified as an Raman scattering (7). The predic- www.spectroscopyonline.com June 2013 Spectroscopy 28(6) tions for the resonance Raman active Table I: Inverse polarized resonance Raman lines of cytochrome c reductase as observed with excitation lines at 568, 531, and 521 nm vibrations have A2g, B1g, and B2g sym- 568 nm 1299 cm-1 Shoulder 1338 cm-1 metry in the D4h point group used to describe the heme. Unlike A1g vibra- 531 nm 1298 cm-1 1314 cm-1 1336 (sh) 1341 cm-1 tions, which are usually the most in- -1 tense features in the spectrum, these 521 nm 1298 cm 1316 1335 1340 asymmetric vibrations will not have high intensity in parallel polarized scattering and, in fact, the highest intensity was observed for the antisym- 1585 metric A2g modes (7). 1579 Some of the hemeproteins, espe- 1129 cially reduced cytochrome c (cyt c+2) 1534 and to some extent oxyhemoglobin 1625

(HbO ) have vibronic states in the Q 1539 2 1613 1358

band that are so long lived that the 1174 1335 1392 520.8 nm 1486 absorption spectra recorded close to 1298 the temperature of liquid nitrogen 1220 1243 exhibit sharp bands, from which the 1152 1340 1316 968 988 lifetimes of the states can be calcu- 1128 1229 lated (8). In reference 8 the quantum 748 yields of the Raman and the reso- 976 673 530.9 nm nance fluorescence were measured 1175 from five hemeproteins and were 1226 found to be consistent with the life- 751 1088 1298 times as determined by the optical 1157 1336 1005 970 923 1360

spectra. 1390 997 1587 603 940 The optical spectra of hemes (the 1314 1580 1532 675 iron complexes of porphyrin) depend 1341 691 on the oxidation and spin state of 568.2 nm 1583 the iron, the protein environment 1338 (including the axial ligands), and the 747 1567 1221 1299 1126 1385 side groups on the porphyrin. Many 1622 1527 1245 1556 1112 hemeproteins (including hemoglobin 1175 988 1087 1003 947 1356 and myoglobin) incorporate proto- 673 1612 1140

heme in which two of the side groups 995 are vinyl (C=C) groups. The  elec- 967 trons of the vinyl groups are conjugated to the porphine ring, which causes a red shift in the absorption Figure 4: Resonance Raman spectra of cytochrome c reductase (cytochrome b-c1 complex) in the spectra. Mitochondria, the energy- reduced state, excited by 520.8, 530.9, and 568.2 nm at a concentration of 66 μM cytochrome c1. These releasing organelles in cells, contain spectra were recorded on a single channel detector on a scanning double monochromator using about a series of cytochromes that includes 350–400 mw of laser power and scanning at a rate of 50 cm-1/min. This figure has been adapted with “b-” and “c-“ types. The b-type permission from reference 9. cytochromes are also based on pro- toheme, but the c-type cytochrome (also known as cytochrome c reduc- structure) and the controlled capture is covalently attached to its protein tase) solubilized in a detergent solu- of energy that enables the production via a reaction of the vinyl sidegroups tion and in whole mitochrondria (9). of adenosine triphosphate (ATP). We with the protein resulting in a loss of These mitochrondrial membranes determined that the b-type hemes their  electron character. Because contain one c-type heme and two have extra resonance Raman bands the absorption spectra of the c-type b-type hemes. The goal was to iden- resulting from the extra  electrons cytochromes are blue-shifted relative tify spectral features of the different on the vinyl side groups, which pro- to that of the b-type cytochromes, heme types and then follow them as a vided a marker to differentiate the the enhancement of their vibrational function of redox potential, a param- two cytochrome types. For instance, spectra will be different. In my pre- eter controlling the oxidation-reduc- the 1315 cm-1 inverse polarized band vious life I measured the resonance tion state of the hemes (which would in cytochrome c is replaced by two -1 Raman spectra of the b-c1 complex be determined by the membrane bands near 1305 and 1340 cm in Spectroscopy 28(6) June 2013 www.spectroscopyonline.com cytochrome b. Figure 4 shows the tion wavelength. I remember one of Fran Adar is the resonance Raman spectra of the b-c1 the last things that I did was to mea- Worldwide Raman complex excited at 568.2, 530.9, and sure a sample of reductase at 10 μM. I Applications Manager 520.8 nm (lines of a krypton laser). had previously standardized my mea- for Horiba Jobin Yvon (Edison, New Jersey). Table I lists the marker lines that ap- surements at 100 μM and then was She can be reached by pear in the spectra. surprised to find that a lower concen- e-mail at fran.adar@ Because of the red shift in the tration gave me a stronger signal. For horiba.com electronic spectrum for the b-type sure, this is because at 100 μM the cytochrome, the 568 nm line only sample had a definite pink tinge to describe my own work (which was excites those cytochromes. However, it and was absorbing the green light. done more than 35 years ago). I used the other lines excite both types. In principle, one can calculate what the example of the hemes because I We did make measurements of this this absorption would be, but the cal- know it well, and it really is a spec- preparation as a function of redox culation would require knowing the tacular example of resonance Raman. potential to follow the disappear- pathlength into the sample, and then ance of the various species (10). What back out, as well as all the other usual References we expected to see was the bands of parameters. This means that using (1) M. Orchin and H.H. Jaffe, Symmetry, each species disappearing separately these measurements for analytical Orbitals, and Spectra (Wiley Inter- as that species became oxidized (the purposes in which concentration science, New York, 1971). resonance Raman spectrum is much information is the goal may not be (2) J.C. Merlin, Pure & Appl. Chem. 57(5), stronger for the reduced rather than realistic. 785–592 (1985). the oxidized species). What we found (3) D.L. Gerrard and W.F. Maddams, Mac- was that the spectra seemed to disap- Summary romolecules 14, 1356–1362 (1981). pear in parallel, as if the hemes are In this installment of “Molecular (4) B. Curry, I. Palings, A.D. Broek, J.A. all coupled together on the mem- Spectroscopy Workbench,” we have Pardoen, J. Lugtenburg, and R. Ma- brane! considered the phenomenon of reso- thies, in Advances in Infrared and nance enhancement of Raman sig- Raman Spectroscopy Vol.12, R.J.H. Downside nals. We have attempted to introduce Clark and R.E. Hester, Eds. (Wiley, Ok, so what is the downside of mak- the origin of the phenomenon with New York, 1985), Chapter 3. ing resonance Raman measurements? a few examples taken from biomole- (5) W.T. Pollard, C.H. Brito Cruz, C.V. With these examples, we have dis- cules that indicate the potential value Shank, and R.A. Mathies, J. Chem. cussed the extremely high sensitiv- of the information available, but we Phys. 90(1), 199–208 (1988). ity and shown how powerful the then explained that because of the (6) F. Adar, in The Porphyrins, Vol. III, information can be in characterizing electronic absorption that provides D. Dolphin, Ed. (Academic Press, samples. What we have not discussed the enhancement mechanism, self Waltham, Massachusetts, 1978). is the fact that the intensity of the absorption of the sample will make (7) T.G. Spiro and T.C. Strekas, Proc. Nat. measurement is not linear with con- extraction of concentration informa- Acad. Sci. 69(9), 2622–2626 (1972). centration. The samples are colored; tion problematic. (8) F. Adar, M. Gouterman, and S. the source of the coloration, in fact, is Please note that in citing publica- Aronowitz, J. Phys. Chem. 80(20), where the resonance is coming from. tions, it is difficult for me to make a 2184–2191 (1976). But that color will cause self-absorp- thorough search of the literature. For (9) F. Adar and M. Erecinska, tion, reducing the signal, and that the work on carotenoids, I have cited Biochemisty 17(25), 5484–5488 reduction will change with excitation articles that are available and known (1978). wavelength, and will also change to me. My apologies to anyone whose (10) F. Adar and M. Erecinska, Febs. Lett. over the range of the resonance work was not cited. In describing the 80(1), 195–200 (1977). Raman spectrum at a given excita- work on hemes, my goal was not to

Posted with permission from the June 2013 issue of Spectroscopy ® www.spectroscopyonline.com. Copyright 2013, Advanstar Communications, Inc. All rights reserved. For more information on the use of this content, contact Wright’s Media at 877-652-5295.

103249