Whenever concentrations are measured with atomic spectroscopy, they are calculated from a calibration curve. Therefore, attaining accurate quantitative results requires proper calibration, especially when measuring low-level concentrations near detection limits. This brief tutorial explains the effect of calibration on the accuracy and detection limits in atomic spectroscopy analyses. A thorough discussion of statistics in analytical chemistry is available in a series of 52 articles (1), with six articles (2–7) being the most relevant to this discussion.

Linearity must be considered when performing analyses by any atomic spectroscopy technique. In general, the linearity of atomic absorption (AA) is ≈3 orders of magnitude, inductively coupled plasma–optical emission spectrometry (ICP-OES) is ≈6 orders of magnitude, and inductively coupled plasma–mass spectrometry (ICP-MS) is ≈10–11 orders of magnitude. Although this is technically true, it does not mean that calibrating over wide ranges will produce accurate results. If a calibration curve in ICP-MS is constructed from calibration standards over 9 orders of magnitude (that is, 1 ppt to 1000 ppm) and has a correlation coefficient of 0.999992, this would be considered a linear calibration curve with an excellent correlation coefficient, implying that any concentration between 1 ppt and 1000 ppm would be read accurately against this curve, which is not true.

Consider a theoretical example of a calibration curve made from standards ranging from 0.1 to 100 ppb—four orders of magnitude. The only way to measure accurate concentrations over a wide range is if every measurement is perfect, which is not possible: every standard on the curve has an error associated with it. Consider that if each calibration standard in Figure 1 has a 2% error on the intensity, the mean intensity error increases as the concentration and intensity of each standard increases: a 2% error on the 0.1 ppb standard (100 counts per second [cps]) is only ±2 cps, but a 2% error on the 100 ppb standard (95,000 cps) is ±1,900 cps. If these data are plotted and a regression (or best-fit) line drawn through the standards (Figure 1), the error of the higher concentration standards dominates the curve. The best-fit line passes almost directly through the 100 ppb standard, but the lower standards are increasingly farther away from the line, demonstrating that the highest absolute error contributes most to the overall curve fit.

Since the error of high concentration standards dominates the calibration curve, both accuracy at lower concentrations and detection limit calculations are affected. If accuracy at low concentrations is the most important criteria, then the calibration curve should be constructed without the use of high-concentration calibration standards: the calibration curve should only contain low-level standards. For example, if selenium (Se) is measured by ICP-MS and is expected to be below 10 ppb in most samples with a reporting limit of 0.1 ppb, a calibration curve consisting of a blank and three standards at 0.5, 2.0, and 10.0 ppb will provide much better accuracy at levels of 0.1 ppb than a calibration curve with standards at 0.1, 10, and 100 ppb. This same principle also applies to ICP-OES and AA, even though their linear ranges are narrower than ICP-MS.

Now consider an example from ICP-MS: a calibration curve for Zn constructed from 11 standards (0.01, 0.05. 0.1, 0.5, 1 5, 10 50, 100, 500, and 1000 ppb) over six orders of magnitude. Figure 2a shows the resulting calibration curve, which has a correlation coefficient (

) of 0.999905, indicating excellent linearity. However, when the 0.1 ppb standard was analyzed as a sample against this curve, it read 4.002 ppb – a huge error.

The cause of this high readback is apparent in Figure 2b, which shows an expanded view of the low end of the curve (the 0.01–10 ppb standards): Zn contamination in the seven lowest standards which causes them to read higher than their nominal concentrations. However, this issue is not apparent from the correlation coefficient since the lowest standards contribute almost nothing statistically to the curve, compared to the four highest standards. This example clearly demonstrates the problem of calibrating at concentrations much higher than expected in samples.

The previous example demonstrates the problem with contamination in calibration standards, but another common issue is contamination in the calibration blank. Contamination can originate from a number of sources, including reagents (acid, water), deposition in the sample introduction system, or deposition in the instrument itself (interface cones, for example), to name a few. Since the blank is assumed to have analyte concentrations of zero, the measured signal from the blank is subtracted from all subsequent measurements. If the blank signal intensity is higher than that of a standard (or sample), the blank-subtracted concentrations will be negative, resulting in poor calibration curves. However, this may not be apparent with calibration curves which include high-concentration standards because the contamination will be small relative to the signal from the high standards. The result: the correlation coefficient will indicate a linear regression, implying that accurate results can be obtained at low concentrations.

In reality, there is no such thing as a true “blank”: contamination always exists at some level. The goal is to limit contamination so that it is much lower than the lowest calibration standard.

Calibrating for Low and High Concentrations Levels

As shown, accurate results for low-level samples can best be obtained by calibrating at concentrations close to the expected concentrations. But this leads to the question: will low-level calibrations produce accurate results for samples with high concentrations? The answer is yes, in most cases. The reason: the error at high concentrations dominates the curve. This means that errors at low concentrations will not have as large an impact on high-concentration samples. To confirm, it is good practice to perform a linear range study on any set of calibration standards: run successively higher standards against the calibration curve. The linear range is generally considered the highest concentration that recovers within 10% of its true value.

However, there are situations where high concentrations will be negatively affected by low level concentrations. Two common causes are large errors in the low calibration standards or when the analyte concentrations are so high that they cause matrix suppression. However, this latter case can usually be compensated for by proper selection of an internal standard.

The key to accurate measurements at low concentrations and to achieving meaningful detection limits is to establish calibration curves with low-level standards. Just about any calibration curve can produce correlation coefficients when using high-concentration standards, but these will not provide meaningful, accurate low-level results or detection limits.

. Parts 1–52, (2002–2013). https://www.americanlaboratory.com/841-Search/?search=Coleman%20Vanatta%20Part%201

. Part 40 (2010). https://www.americanlaboratory.com/914-Application-Notes/1114-Part-40-Blanks/

. Part 41 (2011). https://www.americanlaboratory.com/914-Application-Notes/981-Statistics-in-Analytical-Chemistry-Part-41-Blanks-Concluded/

. Part 44 (2011). https://www.americanlaboratory.com/913-Technical-Articles/35232-Part-44-R-sup-2-sup/

. Part 45 (2011). https://www.americanlaboratory.com/913-Technical-Articles/35860-Part-45-R-sup-2-sup-Continued/

. Part 46 (2012). https://www.americanlaboratory.com/913-Technical-Articles/38042-Part-46-R-Concluded/

. Part 47 (2012). https://www.americanlaboratory.com/913-Technical-Articles/39480-Part-47-Risky-Calibration-Practices/

 is a Principal Application Scientist with PerkinElmer in Shelton, Connecticut. Direct correspondence to: 

Articles

Tutorial

Attaining accurate quantitative results in atomic spectroscopy requires proper calibration, especially when measuring low-level concentrations near detection limits. This tutorial explains key concepts of calibration, illustrated with specific examples.

Calibration: Effects on Accuracy and Detection Limits in Atomic Spectroscopy

A unique feature of carbon is its ability to exist in many different forms and exhibit a variety of properties. The versatile properties of carbon-based materials (hardness, stiffness, and electrical conductivity) makes carbon appealing across many applications, including carbon fibers, lubricants, and carbon black. The ability to identify specific carbon materials is thus imperative, especially for research and development purposes. Raman spectroscopy provides a quick, nondestructive method for identifying and characterizing different types of carbon with little to no sample preparation required. The instrumentation involves a laser, which causes inelastic Raman scattering of light to determine the vibrational modes of the molecules in the material. The result is a Raman spectrum exhibiting specific peaks—commonly referred to as the G and D bands for carbon materials. This article explains the key steps of using Raman technology to investigate carbon and carbon-based materials (such as carbon nanotubes, graphene, and carbon fibers and composites), as well as the process of analyzing the spectra. Specifically, the full width at half maximum (FWHM) of the G band and the dispersions of the G and D bands are analyzed and correlated with different carbon materials. Finally, the challenges of using Raman spectroscopy are discussed.

Raman spectroscopy is an analytical tool that uses scattered light to measure the vibrational modes of a sample. It was named after an Indian physicist (C.V. Raman) who observed the Raman scattering of light in 1928 (1). To conduct Raman spectroscopy, a source of light is required. The light, usually from a laser, can be within the visible, near-infrared (NIR), or near-ultraviolet (UV) wavelengths. The laser light interacts with the molecule, resulting in an energy shift of the laser photons, thereby giving information on the vibrational modes of the molecule.

When light from the laser is incident on a sample, different interactions comprising of the scattering and emitting of light (referred to as fluorescence) may occur as illustrated in Figure 1. The terms λ

 in Figure 1 refer to incident, scattered, and emission wavelengths, respectively. For scattering, three types of scattering can occur, and they are referred to as Rayleigh, Stokes Raman, or anti-Stokes Raman scattering depending on the amount of energy transfer between the photon and the molecule.

For anti-Stokes Raman scattering, some energy from the laser light is transferred to the molecule. In this case, the molecule attains a higher virtual energy state resulting in the scattered light (λ

) having less energy compared to the incident light (λin), as shown below in equation 1.

The higher virtual energy state of the molecule is unstable and short-lived, and is re-emitted immediately. One application of anti-Stokes Raman scattering is in imaging unmodified biological live cells. Other microscopic imaging techniques require the use of contrast agents to acquire an image, which may result in the modification or alteration of the cells.

In Rayleigh scattering, no energy transfer occurs between the photon and the molecule, resulting in the scattered light (λ

) having the same wavelength as the incident light (λ

. Rayleigh scattering is strongly dependent on particle size as well as wavelength.

In Stokes Raman scattering, the molecule loses some of its energy (by relaxing to a lower vibrational state) to the incident light. In this case, the scattered light (λ

) gains energy resulting in higher energy than the incident light (λ

This shift in energy provides Raman spectra with specific features used to provide chemical and structural identification of substances through their unique Raman fingerprint.

Apart from the scattering of light, a phenomenon, known as fluorescence emission, involves the absorption of energy by the molecules. In Raman scattering, the molecules attain a virtual energy state; however, in fluorescence, the molecules attain a real energy state (excited electronic state). Fluorescence light is emitted when the molecule relaxes back to a lower energy state (Figure 1). This process depends strongly on the excitation energy of the laser. Fluorescence and Raman scattering may compete with each other, especially when the excitation laser energy is close to the electronic transition energy of the material under investigation. Fluorescence interference may occur from the material analyzed or from impurities in the sample. Higher energy laser sources, such as 514 nm or 633 nm visible laser wavelengths, produce stronger fluorescence background. On the other hand, wavelengths that are within near-infrared (NIR), such as 785 nm or 1064 nm, results in lower levels of or no fluorescence because their energy may not be sufficient enough to excite the molecules to a higher energy state (2).

This article first discusses the procedure of taking Raman spectra as well as tips and instrument considerations. Following this, a section on analyzing different carbon Raman spectra is presented and discussed. Under this section, an explanation of the key Raman features (including the D and G peaks and FWHM) and how they vary depending on the type of carbon is provided.

Modern Raman spectroscopy typically involves illuminating a sample with a laser and a filter to remove Rayleigh scattered light. In dispersive Raman, a diffraction grating disperses the signal onto a detector. In Fourier transform Raman, an interference pattern is generated to construct a Raman signal. The main issues that are encountered while running a Raman sample is fluorescence interference and thermal degradation. Fluorescence is an issue in Raman analysis because it is highly efficient and gives a signal many orders of magnitude greater than the Raman signal (3).

There are several basic instrumental considerations to make before collecting Raman spectra, which are discussed in detail below.

Wavelengths of light used in Raman spectroscopy span from UV to NIR. The choice of laser is highly dependent on the likelihood of fluorescence or degradation of the sample. Longer wavelengths give weaker Raman signals but likely have little to no fluorescence. The longer wavelength excitation laser supplies less energy, which in turn means the resulting virtual state is lower and unlikely to overlap with the upper electronic state (4). A longer wavelength excitation laser eliminates fluorescence interferences but also decreases the efficiency of Raman scattering. The Raman efficiency is defined by equation (4) where λ is the excitation wavelength.

Using the UV Raman region with a wavelength below 400 nm can generate up to 100x stronger Raman scattering than visible light according to the efficacy equation. UV Raman can also eliminate background fluorescence. However, UV Raman is a complicated and expensive technique, which makes it likely to degrade a sample. For this reason, 532 nm and 780 nm lasers are most typically used to analyze carbon-based materials as they are efficient Raman scatterers less prone to fluorescence (5).

The main source of background interference in Raman is dark current (or thermal electrons). Dark current may be generated if a highly sensitive charge coupled device (CCD) detector is used. Fortunately, dark current can be significantly reduced if the detector is adequately cooled. Nonetheless, a background measurement should be performed prior to sample measurement. Some Raman instruments have automated background correction technology, which collects the background measurement during instrument downtime (6).

Photobleaching is an excellent method to use to reduce fluorescence interferences. This method involves exposing the sample to the laser prior to measurement. This irradiation can induce photolytic decomposition, which breaks down fluorescent impurities and reduces background fluorescence (7).

High-energy cosmic rays can cause random unavoidable spikes in Raman spectra. Algorithms have been developed to reject these spikes. Many Raman instruments have built-in cosmic ray rejection capabilities (8).

The aperture determines how much signal is detected (otherwise known as spectral resolution). An aperture is typically a slit or pinhole, and the larger the aperture, the greater the signal. A smaller aperture would be required for higher resolution information. Larger apertures improve the signal but may also increase noise (9).

In some instances, sufficient Raman spectra can be obtained in seconds. It is good practice to start off with the maximum exposure time before adjusting the number of exposure scans to improve spectral quality. However, this process will only reduce noise. Raman signal is primarily improved by laser power and aperture adjustments (9).

Raman scattering is proportional to the power of the laser (mW). However, it is advisable to begin scanning with lower laser power and increase the power gradually. Both fluorescence emission and thermal degradation of the sample are more likely to occur with increased laser power. Carbon-based materials are typically dark in color and absorb much energy from the laser. Care must be taken when interpreting the spectra of carbon-based materials as a high-energy Raman laser can actually induce the graphitization or other structural change in some samples (10–12).

For measuring small details on samples in the order of micrometer and sub-micrometer scale (that is, defects of carbon nanotubes), confocal Raman microscopy may be required. Confocal Raman spectroscopy involves excitation laser focus in conjunction with Raman collection optics. Confocal Raman spectroscopy can also nondestructively probe depths of the sample without the need for cross sectioning. Any X, Y, or Z location can be probed, and the spatial resolution is maintained in the micrometer range. Care must be taken when focusing on certain areas of the sample to ensure that the area of interest is in focus and is representative of the sample. In addition, precise laser alignment (focal point and spectrograph match) followed by calibration (accuracy of Raman shift and Raman intensity) of the microscope is essential and should be performed regularly (10–12).

All carbon-based materials irrespective of whether they are amorphous or crystalline will exhibit characteristic first order peaks at approximately 1380 and 1580 cm

 (see Figure 2), which are commonly referred to as the D and G bands, respectively. The G band is because of the stretching of a bond connecting two 

 sites arranged in either aromatic rings or as olefinic chains; meanwhile, the D mode corresponds to a breathing vibration of sixfold aromatic rings (14).

Typically, the visible Raman spectra of the crystalline and amorphous carbons will display clearly resolved peaks and peak overlap, respectively. The positions, widths, and ratio of the intensities of the peaks can be indicative of 

 content of materials, degree of disorder and crystallite size. Specifically, the parameters most extensively investigated are full width at half maximum (FWHM), dispersion, and position of the G band as denoted by FWHM(G), Disp(G), Pos(G), respectively as well as the ratio of the intensities of the D and G bands [I(D)/I(G)].

However, to obtain the values of these aforementioned parameters, the peaks need to be resolved and this is often achieved through curve fitting techniques. The two most commonly used models for the fitting of the two first order peaks are: (a) two Gaussians and (b) Lorentzian and Breit-Wigner-Fano (BWF). In the latter, Lorentzian and BWF curves model the D and G peaks, respectively. Gaussian and Lorentzian functions are symmetrical in shape whereas BWF exhibits asymmetry (13). Importantly, the choice of the model dictates the calculation of the I(D)/I(G) quantity; the two Gaussian model uses peak areas whereas the Lorentzian and BWF model requires the intensity heights (14). Displayed in Figure 2 are visible Raman spectra of a single-walled carbon nanotube and diamond-like carbon coatings which are fitted with the Lorentzian and BWF and two Gaussians models, respectively. The shapes of these spectra are characteristic of typical crystalline and amorphous carbon materials.

The FWHM(G) and Disp(G) both serve as indicators of the degree of disorder in amorphous carbon materials. FWHM(G) corresponds to the structural disorder, which refers to the distortion in the bond length and angles (14). Specifically, the FWHM(G) values are proportional to the disorder and thus materials with unstrained or defect free 

 clusters will exhibit small FWHM(G) values. On the other hand, Disp(G) is related to the topological disorder, which involves the distribution of the size and shapes of the 

 clusters (15). By its definition in equation 5, Disp(G) is the rate of change of Pos(G) with respect to excitation wavelength, λ, of the laser. Thus, Disp(G) can only be obtained using multiwavelength Raman techniques. Disp(G) exhibits nonuniqueness behavior and therefore, care must be taken when utilizing it for analysis.

As depicted in Figure 3, one can observe how the quantities Pos(G) and I(D)/I(G) when taken together, can provide approximations on the 

 content of amorphous carbons (a-C) depending on which stage it occupies in the amorphization trajectory. Amorphous carbons (a-C) with high 

 content are referred to as tetrahedral amorphous carbon (ta-C). Further, carbon-based materials exhibiting more graphite like structure (stage 1), have their (ID)/I(G) ratios inversely proportional to the grain size. Consequently, this quantity, (ID)/I(G), can be used to approximate the diameter of the clusters for more crystalline graphite like materials (15). This size difference is the salient distinguishing feature between graphite and nanocrystalline (NC) graphite where the latter exhibits the smaller crystallite size (16).

Finally, other peaks such as the 2D band (or second-order D band) provide a measurement of the layer thickness of graphite. Single-layer graphene has a single symmetric band but with more layers the band broadens (21). This peak shares a common feature with D in that it exhibits dispersive behavior as depicted in Figure 4. However, unlike the D peak, this band is not activated by disorder (22).

The authors discuss Raman spectroscopy as a quick, nondestructive method of identifying different carbon materials. The other advantage of using Raman spectroscopy is that little sample preparation is required before taking measurements. The choice of the laser wavelength or type of Raman spectra is imperative to minimize fluorescence or sample damage. Crystalline carbon materials spectra have isolated peaks while amorphous carbon spectra will show overlapping peaks. The spectra for amorphous carbon will require peak deconvolution before the data can be analyzed. Parameters such as FWHM, the D and G peaks, and their dispersion provides useful information in identifying the different properties of the carbon material.

(2) E. Illy, A. Draksharapu, and W.R. Browne, in 

, supplement to Spectroscopy, 16 (September 2014). https://www.spectroscopyonline.com/.

Dispersive Raman Spectroscopy, Current Instrumental Designs. In: Encyclopedia of Analytical Chemistry

 (John Wiley & Sons, West Sussex, United Kingdom, 2006).

(5) R.R Jones, D.C. Hooper, L. Zhang, D. Wolverson, and V.K. Valev, 

 (Wiley Online Library, Hoboken, New Jersey, 2002).

(7) Y. Wang, D.C. Alsmeyer, and R.L. McCreery, 

(8) D. Roy, S. Kanojia, K. Mukhopadhyay, and N. Eswara Prasad, 

Raman Technology for Today’s Spectroscopists

, supplement to Spectroscopy (June 2010). https://www.spectroscopyonline.com/view/understanding-raman-spectrometer-parameters 

(10) L. Bokobza, J-L. Bruneel, and M. Couzi, 

(12) L. Bokobza, J-L Bruneel, and M. Couzi, 

(13) A. Merlen, J. Buijnsters, and C. Pardanaud, 

(15) C. Casiraghi, A.C. Ferrari, and J. Robertson, 

(18) K. Jurkiewicz, S. Duber, H.E. Fischer, and A.J. Burian, 

(19) V.L. Deringer, A.C. Miguel, R. Jana, A. Aarva, S.R. Elliott, T. Laurila, G. Csányi, and L. Pastewka, 

(22) L.M. Malard, M.A. Pimenta, G. Dresselhaus, and M.S. Dresselhaus, 

 are with the Maritime Division in the Defense Science and Technology Group, in Fishermans Bend, Australia. Direct correspondence to:

This article explains the key steps of using Raman technology to investigate carbon and carbon-based materials—such as carbon nanotubes, graphene, and carbon fibers and composites—as well as the process of analyzing the spectra.


Raman Spectroscopy: A Key Technique in Investigating Carbon-Based Materials

The ability to enhance the Raman signal from low concentrations of molecules using nanoparticles has made surface-enhanced Raman scattering (SERS) an attractive approach for trace detection in diverse applications in medicine, energy and the environment (1). Understanding SERS has improved significantly since the initial discovery in the 1970s (2,3). The detection and imaging, via tip-enhanced Raman scattering (TERS), of individual molecules suggests the ultimate in detection sensitivity (4–7). Particularly, recognizing how the energy from the excitation laser is concentrated by the nanoparticles through localized surface plasmon resonances has helped foster understanding of the appropriate experimental conditions to use for SERS measurements. Hot spots are important for generating the largest enhancements (8). In addition to enhancing the local electric field, other effects can occur on the surface of nanoparticles and impact the observed signals. Considering how the excitation laser intensity is concentrated and confined into small regions, it is important to consider the incident laser power. When performing SERS experiments, the excitation laser power needs to be kept at a level that enables detection but avoids damaging the sample.

Generally, surface-enhanced Raman spectroscopy (SERS) practitioners often recommend the laser power be kept below 1 mW in a diffraction limited laser spot. The spot size in a microscope objective can be approximated by the diffraction limit below

where changes with the numerical aperture (NA) of the objective and laser wavelength (λ) will alter the spot size. 1 mW of power in a 1 μm spot diameter corresponds to an energy density of approximately 1 mW/μm

), which is a common experimental condition for many SERS experiments. High NA objectives and shorter wavelength lasers require lower powers. Similarly, other factors that lower the energy density can also prevent sample damage in SERS experiments.

Figure 1 gives an illustrative example of how too much laser excitation can significantly damage a sample. In this example, individual nanoparticles and nanoparticle dimers were confined in an array of lithographically fabricated wells on the surface. Single nanoparticles are evident as green spots in the dark field image on the left. Coupling between nanoparticles to form a dimer causes the scattering to appear red from some of the wells in the darkfield image. After exciting with a few milliwatts (1–2 mW) of 633 nm laser excitation, the energy released destroyed the sample as shown on the right. Figure 1 is relatively dramatic, and the exact mechanism of the damage is unclear; however, it illustrates that small amounts of laser power can detrimentally impact your experiment.

There are several ways increased laser power is detrimental to SERS. For example, sustained heating at 250 °C converts nanorods into nanospheres in less than an hour (9). The strong dependence of the SERS signal on the nanostructure shape and spacing means melting can easily diminish the observed SERS intensity. In my own laboratory, we have seen that the temperature increase associated with increased laser power has a detrimental effect on the SERS signal (10). The energy absorbed and heat produced by the nanoparticles can easily exceed 250 °C for aggregated nanoparticles in air. Keeping the nanoparticles in a medium with higher heat dissipation (water as opposed to air) during the SERS measurement can minimize the heat-related damage. Generally, photo-induced heating is minimized by providing a way for heat to dissipate. Laser power has also been reported to damage the molecules present on the nanoparticle surfaces. This is also not too surprising, because the same plasmonic nanoparticles that are used for signal enhancement are also used for chemical catalysis, where the excited electrons accelerate chemical reactions at the nanoparticle surface (11). When molecules capture these energetic electrons, bonds can break or the vibrations of the molecule change, which alters, or even destroys, the SERS signal (12,13). This effect may explain changes between the spontaneous Raman spectrum and the observed SERS spectrum for any given molecules. In many cases, using low laser power for excitation can minimize these effects.

Figure 2 shows the spectra obtained from a protein deposited and dried on a commercial SERS substrate (Silmeco, Au SERStrate) at different laser excitation powers. The spectra recorded with increased laser power do show more signal intensity; however, much of the fine detail that is used for chemical analysis is lost at the higher powers. The distinct vibrational bands near 900 and 1500 cm

 at low power are lost at higher laser power excitation. It is worth noting that the specifications provided by Silmeco recommend using <10 W/cm

, which corresponds to a power of ~0.1 μW in a diffraction limited laser spot. It can be difficult to observe signals at such low powers, but damage will be avoided. The spectra shown in Figure 2 were acquired with a 785 nm laser and a 0.5 NA microscope objective, which corresponds to an energy density ranging from 2.0 x 10

. In the normalized spectra, the lowest power acquisition clearly shows bands that disappear with increased laser power. These lost bands may result from changes in conformation, thermal damage, or even chemical changes to the protein on the surface. Additionally, the lost features are particularly valuable when using SERS spectra from chemometric, or machine learning, models to decipher important properties of mixtures and complex samples.

To conclude, the examples provided illustrate how using the incorrect laser power can degrade, or in some cases destroy, your sample. Table I lists recommendations to help you make successful SERS measurements. Some factors that can help are acquiring spectra in an environment with higher heat dissipation, such as in water compared to in air. Other approaches that illuminate a larger number of spots, such as line focus or rapidly moving the beam around the surface, can help avoid sample damage. Overall, less power is often better when making SERS measurements.

(1) J. Langer, D. Jimenez de Aberasturi, J. Aizpurua, R.A. Alvarez-Puebla, B. Auguié, J.J. Baumberg, G.C. Bazan, S.E.J. Bell, A. Boisen, A.G. Brolo, J. Choo, D. Cialla-May, V. Deckert, L. Fabris, K. Faulds, F.J. García de Abajo, R. Goodacre, D. Graham, A.J. Haes, C.L. Haynes, C. Huck, T. Itoh, M. Käll, J. Kneipp, N.A. Kotov, H. Kuang, E.C. Le Ru, H.K. Lee, J-F. Li, X.Y. Ling, S. Maier, T. Mayerhoefer, M. Moskovits, K. Murakoshi, J-M. Nam, S. Nie, Y. Ozaki, I. Pastoriza-Santos, J. Perez-Juste, J. Popp, A. Pucci, S. Reich, B. Ren, G.C. Schatz, T. Shegai, S. Schlücker, T. Li-Lin, K.G. Thomas, Z-Q. Tian, R.P. Van Duyne, T. Vo-Dinh, Y. Wang, K.A. Willets, C. Xu, H. Xu, Y. Xu, Y.S. Yamamoto, B. Zhao, and L.M. Liz-Marzán, 

J. Electroanal. Chem. Interf. Electrochem

(1), 1–20 (1977). DOI: https://doi.org/10.1016/S0022-0728(77)80224-6.

(3) M. Fleischmann, P.J. Hendra, and A.J. McQuillan, 

(4) R. Zhang, Y. Zhang, Z.C. Dong, S. Jiang, C. Zhang, L.G. Chen, L. Zhang, Y. Liao, J. Aizpurua, Y. Luo, J.L. Yang, and J.G. Hou, 

(6) K. Kneipp, Y. Wang, H. Kneipp, L.T. Perel- man, I. Itzkan, R.R. Dasari, and M.S. Feld, 

(9), 1667–1670 (1997). DOI: 10.1103/PhysRevLett.78.1667.

(7) J. Lee, K.T. Crampton, N. Tallarida, and V.A. Apkarian, 

(7750), 78–82 (2019). DOI: 10.1038/s41586-019-1059-9.

(8) Y. Fang, N-H. Seong, and D.D. Dlott, 

(9) H. Petrova, J. Perez Juste, I. Pastoriza-Santos, G.V. Hartland, L.M. Liz-Marzán, and P. Mulvaney, 

(7), 814–821 (2006). DOI: 10.1039/B514644E.

(10) Z.C. Zeng, H. Wang, P. Johns, G.V. Hartland, and Z.D. Schultz, 

(21), 11623–11631 (2017). DOI: 10.1021/acs.jpcc.7b01220.

(11) E. Cortés, L.V. Besteiro, A. Alabastri, A. Baldi, G. Tagliabue, A. Demetriadou, and P. Narang, 

(12), 16202–16219 (2020). DOI: 10.1021/acsnano.0c08773.

(12) J. Szczerbiński, L. Gyr, J. Kaeslin, and R. Zenobi, 

(11), 6740–6749 (2018). DOI: 10.1021/acs.nanolett.8b02426.

(13) S. Sloan-Dennison, C.M. Zoltowski, P.Z. El-Khoury, and Z.D. Schultz, 

(17), 9548–9558 (2020). DOI: 10.1021/ acs.jpcc.0c01436.

 is with the Department of Chemistry and Biochemistry at The Ohio State University. Direct correspondence to: 

In SERS experiments, the excitation laser power needs to be kept at a level that enables detection but avoids damaging the sample, as these examples illustrate.

August 2021

Calibration: Effects on Accuracy and Detection Limits in Atomic Spectroscopy

Raman Spectroscopy: A Key Technique in Investigating Carbon-Based Materials

Not Too Hot: The Importance of Optimizing Laser Power for Surface-Enhanced Raman Spectroscopy (SERS) Measurements