In this month’s column, I review the band assignments of a protein spectrum, pointing out why it can be useful to know the band assignments when attempting to use the Raman spectra to understand the functionality of proteins. In fact, the American Chemical Society (ACS) has just issued a virtual issue of 

 entitled “Protein Crowding and Stability,” discussing how the protein conformation impacts its biological functionality. I address how the Raman spectra can help with the determination of protein structure.

Because of the successes of Raman spectroscopy in recent years in many fields, including biomedical applications, its use is attracting interest. Although successful applications require the use of sophisticated software algorithms to tease out the information from complex data sets, it still can be useful to understand the origin of some of the characteristic spectral features. In the case of proteins, the CH–NH region does not show huge variations, but the fingerprint region is quite variable, and spectroscopic assignment of the chemical origin of a significant number of those variations can be made. With this type of information, it then becomes possible to begin to infer biochemical changes or function from the spectra.

In this column, I review what is known about the Raman spectra of proteins. What is discussed is not new information; it is available in numerous sources (1–4). My hope is that individuals thinking about using Raman spectroscopy in their research can benefit from my explanation.

Figure 1 shows the spectrum of a fingernail, with the major bands of interest indicated. We treat each region individually. The nails are composed of α-keratin that has a high content of alanine, leucine, arginine, and cysteine, assembled into a helix similar to the α-helix, and with a coiled arrangement that provides the strength of a fibrous structure.

In forming a protein, there is a condensation of two amino acids to form each peptide bond that is shown in Figure 2. 

 represents the side groups that provide the identification of each particular amino acid. The amide I and amide III bands represent motion of atoms of the peptide backbone. Both the amide I and amide III bands reflect coupled vibrations of the backbone, which depend on the secondary structure of the protein. Note that there is one NH group and one CH group associated with every amino acid in the backbone. But there are additional NH and CH groups from particular amino acid sidegroups that are tabulated in Table II.

When the spectra of individual regions are shown, they are shown with bandfitting, which is used to analyze a protein’s structure. In reviewing a few articles that make assignment of amide bands, there is not always agreement in the band assignments. Table I lists the frequencies of the amide I and III bands, with reference to two publications indicating the variations. Despite this lack of absolute assignment, it has still been found to be useful to bandfit and determine rough estimates of relative amounts of structure.

In addition to these amide bands, which involve coupled motions of the carbonyl stretch, the backbone C–N stretch, and the NH bend, there are also NH stretches. Because the NH bond can readily H-bond, the peak frequency moves around and broadens. There are also CH bands, all of which fall below 3000 cm

. First, we look at the CH–NH region and then the carbonyl region.

Because I am going to try to make sense of the structure in the CH and NH region, I found it useful to tabulate the additional NH and CH bands that arise from the side group structures and note where their peak frequencies are believed to fall, which is shown in Table II.

The fitting has revealed three relatively broad bands in the NH region for this protein (Figure 3). One would expect these bands to reflect changes in conformation and hydrogen bonding interactions in the molecule in the vicinity of the NH groups.

In the CH region, the highest frequency band is the aromatic CH that appears at 3061 cm

, the value that is most often quoted for an aromatic CH. Note that this band can move 10–15 cm

In the saturated CH region, we fit five bands and can match these bands to predictions from assignments of classical organic molecules. The methylene group often has symmetric and asymmetric stretches respectively near 2850–2900 cm

, whereas a methyl group has symmetric and asymmetric stretches respectively at approximately 2875–2920 cm

 is assigned to the methine CH. Generally, the envelope of CH bands looks quite similar from protein to protein, but a careful evaluation of the spectra of proteins with a small number of known differences might enable a careful interpretation in regard to the integrated intensities of these bands.

Figure 4 shows the spectrum in the amide I region. In addition to the bands of the amide backbone, there are aromatic bands that have also been assigned (4). For anyone interested to see how much the bands in the Amide I region change in different chemical environments, reference 6 is a good place to start (6).

Disulfide Stretch and Other Bands that Reflect the Presence of S

There is one more region where the spectra are not complex, and that is the region around 500 cm

, where the -S-S- stretch appears (Figure 5). The two bands that appear in the fit above 500 cm

 are labeled GGG and TGG, where the G and T refers to the gauche or trans dihedral angles of the following unit:

There is a third possibility that can also be observed—TGT, with the peak at approximately 540–545 cm-1 (its position is labeled in gray in Figure 5 because it is missing in this protein spectrum). There are also C-S bands of the cysteine or methionine unit containing the sulfur; the frequency falls between 630–760 cm

, depending on sidegroups and other considerations. For anyone who is more interested, refer to the literature (2).

In addition, if the sulfur atoms on cysteine or methionine are not bound up with sulfur atoms on other amino acids, there can be free -SH groups (unless the environment is quite basic), whose stretches appear in the region between 2500–2600 cm

Finally, we examine the bulk of the fingerprint region not yet considered, which is 800–1500 cm

 in Figure 6. The bandfitting has identified many more bands than those whose origins have been labeled. The user needs to realize that in this region of the spectrum there is extensive mixing of motion of the atoms so mathematical normal modes analysis is often necessary. However, the bands that are labeled are well established. The aromatic ring stretch for phenylalanine stands out and is easily recognizable. Note that tyrosine, which also has a ring, does not exhibit a band here because of its parasubstitution. However, the Fermi resonance doublet at 1580 and 1600 cm

 represents these two amino acids that have aromatic rings. In addition, the band at 1616 cm

 is assigned to both the tyrosine and the tryptophan.

The amide III band also falls in this region, and the fitting reveals two bands that identify the presence of the α-helix and β-sheet.

The spectral region of Figure 6 also includes two Fermi resonance doublets that have been used to characterize these proteins. The doublet near 860 and 830 cm

 is assigned to tyrosine, and its relative intensity, which varies between 0.3 and 6.7, has been shown to be an indicator of H-bonding (4).

The spectrum of tryptophan also exhibits a Fermi resonance doublet, with lines at approximately 1360–1340 cm

. When the relative intensity is greater than 1.1, the sidegroup is sitting in a hydrophobic environment, but when it is less than 0.9, it is in a hydrophilic environment. The location of this doublet is indicated in Figure 6, but the label is gray rather than black because the absence of the higher frequency component makes its assignment in this spectrum questionable.

I am going to display a series of figures with protein spectra of related origins. See if you can discern some of the spectral differences from protein to protein. For easy comparison, all spectra will be scaled to the CH stretch at 2930 cm

The cuticle, whose spectrum is shown in Figure 7, was a piece of skin that I removed from next to my fingernail. It is composed of dead skin cells, and the protein composition of skin cells is keratin. The skin cells that I collected from my epidermis are also composed of keratin but with lipids as well. What are the outstanding differences between the spectra in Figure 7? I see that the fingernail has much more disulfide. I also see that the structure in the amide I region is different. And I also see that the skin has features of lipids that are on the skin. You may remember the characteristics of the lipid spectra from a previous column.

Figure 8 overlays the spectra of fingernail, a liquid egg white, a dried egg white, and the membrane inside of an egg shell. Can you see the presence of water in the liquid egg white in Figure 8? That’s easy, but can you discern the presence of the NH in the water spectrum? What other differences can you see between the spectra?

Figure 9 overlays the spectra of fingernail, chicken fascia, chicken tendon, pink muscle excited at 532 nm, and pink muscle excited at 638 nm. You can see that the chicken tendon has much more detail in the fingerprint region. Can that be used to infer something about the structure of the collagen with its important mechanical properties? This is a fibrous protein that is composed of 65% collagen. On each chain, every third amino acid is a glycine, and many others are proline and hydroxyproline and they assemble into a left-handed helix. Three of these left-handed helices assemble into a right-handed helix with 30 residues per turn. There is cross-linking between chains via H-bonds and intermolecular ester linkages; with aging, the H-bonds convert to covalent cross links. All of this provides strength to the tendon. On the other hand, muscle also contains collagen, but is dominated more with actin and myosin for contractility, and troposin and tropomyosin for regulatory function. What differences do you see in these spectra that can be related to their composition?

Figure 9 also has a trick for which I did not prepare you. Why are the two spectra of the pink muscle different when excited at different wavelengths? Why is the muscle pink? Animal muscles can be pink because of the presence of myoglobin, mitochondria, or both. They contain hemes whose absorption in the visible results in resonance enhancement, especially when exciting with a green laser which overlaps with the absorption of the heme. The extra sharp bands in the spectrum excited in the green are resonance-enhanced bands of the heme (7).

In reviewing these spectra and the information on interpretation, I do not expect that you will be an expert. In fact, I spent much more time writing this article than I expected, because I had a general idea of what was present in the spectra, but did not understand it enough to put it in words with confidence that I would get the information correct. But if you spend some time going over this, and then need to work with Raman spectra of proteins, you will have a heads up on what you are looking at. Once you are aware of the possibilities, you can choose the tool that is most appropriate for you to get the information that you need—band-fitting, chemometrics, or 2D-COS. Good luck and have fun!

Raman Spectroscopy in Biology: Principles and Applications

Handbook of Vibrational Spectroscopy – Applications in Life, Pharmaceutical and Natural Sciences

, vol. 5, John M. Chalmers and Peter R. Griffiths, Eds. (Wiley & Sons, Hoboken, NJ, 2002), pp. 3426–3446.

(4) D. Nemecek, J. Stepanek, and G.J. Thomas, Jr., 

Proteins and Nucleoproteins Curr. Protocols in Protein Sci.

Infrared and Raman Characteristic Group Frequencies Tables and Charts

(6) N.C. Maiti, M.M. Apetri, M.G. Zagorski, P.R. Carey, and V.E. Anderson, 

(7) F. Adar, M. Gouterman, and S. Aronowitz, 

Articles

Column: Molecular Spectroscopy Workbench

Raman spectra can help determine protein structure. Here’s how.

Interpretation of Raman Spectrum of Proteins

An instrument generally measures the spectrum of a sample, and produces that spectrum, in a particular way. It also saves (or stores) that spectrum according to prespecified instructions designated by the manufacturer. But the data themselves, their format, or the way they are presented, may not be optimal from the user’s point of view, as the decisions a manufacturer makes are inevitably done in the absence of user input, despite efforts to include user or third-party evaluations. Furthermore, different users have different requirements, so it’s rare when all users want to avail themselves of the same type of data processed the same way. To satisfy the need for users to convert raw data produced by an instrument into a form more suitable for their particular needs, instrument manufacturers, as well as independent third-party software developers, include software packages that convert the raw data into other forms to address, it is hoped, many users’ needs.

This column (and perhaps the next one or two) is based on the talk given at the Gold Medal Award symposium held during the 2020 Eastern Analytical Symposium.

In the 1970s, we became acquainted with near-infrared (NIR) analysis; at that time, questions surrounding data transformations abounded. Back then, the issues with data transformations were more or less tied to other issues we were dealing with, such as technology issues—use of interference filters versus diffraction gratings versus Fourier transform (FT)–based instruments—and how to deal with what was variously called “particle-size” or “repack” effect (and which data transformation was “best”?). Ever since Karl Norris announced the technology we now call modern NIR analysis, he and almost all scientists working in this field used transformations of the data to achieve several goals. Karl himself was a superb empiricist, and often achieved impressive results using this approach. Other scientists tried to emulate his techniques, but few had Karl’s insights regarding what to do when a difficulty was encountered, and could not replace Karl’s “magic,” although some could occasionally achieve improvements in their calibration results by following his recipes.

Other scientists observed Karl’s methods and results, and learned to appreciate his insights. On the other hand, some also felt that one of the basic tenets of calling a field of study a “science” is the ability for every scientist in the given field to be able to reproduce the results that any one of them can achieve. Therefore, a good amount of effort was expended to analyze Karl’s methods and to reproduce and even expand on them. One of the key operations that Karl used when he analyzed NIR data with the goal to produce accurate, precise, and stable calibration results and models was the use of data transformations. A good deal of effort was put into generating and testing various transformations of the spectral data to improve calibration results.

Fast forwarding to the current day, reading the literature gives the impression that the NIR “community” has taken Karl’s leadership to heart, and routinely applies a data transformation to their NIR spectral data when performing a calibration exercise. However, a closer reading of the literature reveals that those data transformations that are applied are taken from an extremely small subset of the potential transformations that might be used. Almost invariably, the transformation consists of computing the second derivative of the spectral data (with respect to wavelength, such as, for example, 

 [wavenumbers] when the spectra were measured on an interferometer-based instrument), sometimes preceded by a “smoothing” operation (usually a running average of some number [m] of successive spectral data points). Our overall impression is that of a lack of investigative curiosity. However, to be fair, we must also take note that many workers are constrained by the equipment and software available to them, which only contain algorithms for those few data transformations. Some scientists are further constrained by limitations imposed by governmental regulatory agencies.

However, at the beginning, a wide variety of data transforms were applied to the spectral data, in attempts to improve the calibration results. Those were pruned down to the modern, nearly exclusive use of second-derivative transformation. In part, that was because of the forces described in the previous paragraph, but also in part, because NIR spectroscopists lost the will (or the initiative) to program their own computers and implement algorithms beyond the packages provided by the vendors of their calibration software.

Thus, one purpose of this column, and the next few, is to resurrect some of these more novel and sophisticated methods that were developed in those days but then disappeared, in the hope that some readers may wish to become more adventurous about their current software usage, and try some of these more advanced techniques. Our discussions include some of the theory and expected benefits of the algorithms. Some of these algorithms were developed by ourselves, some by other scientists; we indicate and credit those scientists when appropriate.

The benefit of some algorithms is not always a reduced standard error of prediction (SEP). To be sure, SEP is an important indicator of the quality of a calibration algorithm, but by itself is not the be-all and end-all diagnostic to be used in evaluating calibration performance. Other important characteristics of a calibration model can include robustness, reduced sensitivity to variations of the instrument or samples that can cause drift, outliers, nonlinearity, random noise, and other undesirable behaviors that can beset any analytical method. One important effect is the “repack” effect, the sensitivity of a calibration model to the change in readings engendered by simply pouring a powdered solid sample out of the sample cup, and then back in. Indeed, reducing this sensitivity was one of the key reasons for instituting derivative data transformation in NIR analysis of powdered solids, and, indeed, it has merit. However, even better methods are available, and will be described in future columns. Another of those undesirable behaviors is difficulty in transferring calibration models from one instrument, on which it was developed, to another instrument, on which it is to be used (calibration transfer). We have discussed calibration transfer separately in previous “Chemometrics in Spectroscopy” columns (1,2), so we will not repeat it here. This discussion can also be found in chapters 122 and 123 in the second edition of our eponymous book (3). Some of these behaviors do not have nice, convenient, and easy-to-calculate statistics that are unique measures of one of those behaviors, although some do. We will indicate those, too, when appropriate.

We’ll start off our discussion not with an algorithm, but with an explanation. In the early days of NIR, for the first few years after Karl Norris invented the technology, the most commonly used wavelength-selection technology for NIR instruments was a set of fixed-wavelength filters. Again following Karl’s lead, most instruments contained six or more filters, each filter allowing some small range of wavelengths to pass through the rest of the optical system, and all filters mounted on a mechanism that could change which filter was active at any moment. Each filter was characterized by the center wavelength of its passband, and that determined the wavelength assigned to the readings taken using that filter.

Karl’s development work showed that measuring the reflectance or absorbance at multiple wavelengths for each sample is often, even usually, required. This scheme allowed a properly programmed computer to correct a reading at one spectral point for interferences caused by errors at another spectral point for any of various other error sources, based on the data measured at a different spectral point. This is known as 

, and is discussed in more detail in the literature (4). This method can work, but it is subject to interference by some subtle characteristics of the data. One of these subtleties is illustrated in Figure 1, and was a major issue in trying to develop transferable (or universal) calibrations. It was found that no matter how much care was used in setting up the instrument, or preparing and presenting samples to the instrument, it was impossible to reproduce the same set of calibration coefficients on the second try as on the first try, despite the fact that the performance of the two calibration models were equivalent and the filters (wavelengths) used were the same.

Figure 1 illustrates the difference between the situations where absorbance at one wavelength (or, strictly speaking, a small wavelength range represented by the central wavelength), and two (or more) distinct wavelengths with absorbances that are correlated are used to create a calibration model. Advanced discussions of calibration theory demonstrate analytically that intercorrelations between the variables of a regression analysis make the computed regression coefficients for those variables unstable, because of the increased sensitivity of the computed calibration coefficients to small changes in the data as the correlation of the absorbances at those wavelengths increases (5); in Figure 1, we present that situation graphically.

Figure 1 illustrates what happens when two correlated variables are both used in an attempt to “improve” the calibration. Because there are two independent variables (the absorbances at the two wavelengths), including the variable representing the concentrations, the data exist in a 3D space. Furthermore, the data are contained within a cylindrical region of that space. A calibration model, in this situation, is comprised of a plane representing the cylindrical shape of the data, rather than a line. This plane represents the least-square fit to the data. However, since the noise of the absorbances at the two wavelengths are random and of approximately equal magnitude, the exact position of the plane is very rotationally unstable, and subject to the influences of these very small fluctuations of the noise. The upper-right-hand corner of Figure 1 is a vector diagram (somewhat stylized, to be sure) showing how the projection of an extension of the plane against lines parallel to the axes of the space define the values of the two coefficients of the calibration. If the plane were to rotate around the axis shown, to a different position, then the projections would change in length, representing changes in the values of the calibration coefficients that we observed.

Figure 1 shows a representation of a two-wavelength calibration, and is one for which we can draw a diagram, which compresses a three-dimensional situation into the two dimensions of the figure. For a calibration using more than two wavelengths, an analogous figure would be required to represent four or more dimensions, which is beyond our capability to draw, and also beyond most people’s capability to visualize (including ours). However, the mathematics assures us that the same phenomena will create the instability we see in the two-wavelength, 3D case, in those higher-dimensional situations as well.

Therefore, we can explain the cause of what we observed as an apparent inability to reproduce a calibration model when using MLR technology, even though any calibration model resulting from a different orientation of the calibration plane would, in practice, perform equally well. However, this understanding does not provide us with a tool to correct the situation, or even give us a means to select, from all the possible orientations of the calibration plane, the one orientation, with its corresponding calibration coefficients that would allow other scientists to reproduce that exact model. Thus, as we discussed above in our introductory comments to this column, data transforms and modifications to the basic MLR algorithm were applied to address this issue and other problems that arose.

We will begin our examination of alternate calibration methodologies in the next column on this topic.

 (Elsevier, Amsterdam, The Netherlands, 2nd ed., 2018).

Principles and Practice of Spectroscopic Calibration

 (John Wiley & Sons, New York, NY, 1991). 

 (John Wiley & Sons; New York, NY, 3rd ed., 1998).

Column: Chemometrics in Spectroscopy

Raw data produced by an NIR instrument undergoes some sort of processing, or transformation, to make them easier to use. In this series, we explore options for that data transformation, starting with multiple linear regression (MLR).

Data Transforms in Chemometric Calibrations: Application to Discrete-Wavelength Models, Part 1: The Effect of Intercorrelation of the Spectral Data

Inductively coupled plasma–tandem mass spectrometry (ICP-MS/MS) is a relatively new technique, with the first triple quadrupole ICP-MS (ICP-QQQ) being introduced commercially in 2012. However, the technique has quickly been accepted for a wide variety of applications in industries including, but not limited to, manufacturing, contract analysis, government, research, and academia. In this article, we explain the key steps to setting up an ICP-MS/MS method and address some of the common misconceptions surrounding the technique.

Compared to conventional single quadrupole inductively coupled plasma–mass spectrometry (ICP-MS), the main benefits of ICP-MS/MS include:

the unprecedented ability to resolve spectral interferences using controlled reaction chemistry in the collision reaction cell (CRC);

significantly better separation of adjacent mass, “peak tail” overlaps because of two mass filtering steps; and

improved detection limits through higher sensitivity and lower backgrounds.

The flexibility of ICP-MS/MS has led some people to assume that method development must be more difficult on ICP-MS/MS than on single quadrupole ICP-MS. However, existing single quadrupole methods can usually be transferred to ICP-MS/MS without significant modifications. Also, reaction gas methods run more consistently on ICP-MS/MS, so methods are more universal and operation is simpler.

The following six steps help guide method setup for ICP-MS/MS (Figure 1):

When developing an ICP-MS/MS method, users may focus on a specific interference that needs to be resolved. However, the starting point for method development on ICP-MS/MS—as with conventional ICP-MS—should consider the wider analytical needs. These needs include several factors that can be easily overlooked.

First, can the instrument handle the sample matrix? Matrix tolerance is important for many ICP-MS applications. Optimizing the plasma to give low CeO

 (<1.5%) ensures that the instrument will easily tolerate typical sample matrix levels. Low CeO/Ce indicates better matrix decomposition, which reduces deposits on the interface and leads to better stability and less maintenance. Low CeO/Ce conditions also increase analyte ionization, improving sensitivity for poorly ionized elements, such as beryllium, arsenic, cadmium, and mercury. Low CeO/Ce indicates better decomposition of molecular species that would otherwise form polyatomic ion overlaps (1).

Second, is the sample composition unknown or variable? The major element levels and sample-to-sample variability can be important factors. Variable sample matrices can give rise to unexpected polyatomic ion overlaps, which a given reaction gas mode may not resolve effectively. In such cases, helium (He) collision mode may provide a simpler, more universal, and more reliable solution (2,3).

Third, which analytes must be measured and at what level? Most ICP-MS applications can be run successfully using He collision mode, but lower analyte levels may require a reaction gas method to further reduce backgrounds and improve detection limits. Similarly, analysis of high-level major elements may require custom tuning or sample dilution if the detector cannot measure the full concentration range.

Finally, how many samples will be analyzed? Many parameters affect analysis speed for large sample batches. ICP-MS/MS systems are compatible with discrete sampling devices to shorten sample analysis times. However, the number of cell gases used—and the time required to switch between cell modes—will also impact throughput.

Step #2: Identify Critical Method Needs—Why is ICP-MS/MS Needed?

Once the basic analytical requirements have been addressed, specific, critical aspects of the analysis should be considered. This stage usually focuses on the most difficult sample types, unusual analytes, or extremely low (or high) concentration levels. In the context of ICP-MS/MS method development, this step should identify analytes that are affected by spectral overlaps.

Step #3: Apply the Simplest Approach to Address Interferences 

ICP-MS/MS offers performance benefits because of the additional quadrupole, Q1, before the CRC, which controls reaction processes in the cell. Reaction gas modes can give extremely effective interference reduction, useful for ultralow level analysis in high purity reagents. However, reaction gas methods may be specific for each overlap being resolved, so multielement methods may require several sets of cell conditions. These conditions may not always be applicable to other sample types.

As mentioned above, many ICP-MS interferences can be addressed using He collision mode. Most spectral overlaps in ICP-MS are caused by polyatomic (molecular) ions. For example, 

 75. Helium mode uses kinetic energy discrimination (KED) to reduce the transmission of polyatomic ions relative to the monatomic analyte ions they overlap (4). As a result, He mode supports multi-element analysis and allows access to secondary (confirmatory) isotopes for many elements, while also practically eliminating the need for unreliable mathematical correction equations.

The major appeal of He mode is that the same cell conditions can be used for multiple analytes across a range of sample types, so method setup is simple and consistent. He mode is the default mode for single quadrupole ICP-MS methods, and can be used in the same way on ICP-MS/MS. Using He mode for most analytes makes it much simpler to develop methods for the remaining analytes that need a reaction gas approach.

Step #4: Identify Analytes or Overlaps that Cannot Be Addressed Using He Mode

Many users employ ICP-MS/MS precisely because their application cannot be performed successfully using single quadrupole ICP-MS, even with well optimized He mode. Usually, the requirements relate to specific spectral interferences or adjacent mass overlaps that can only be resolved using a reaction gas or double mass selection (MS/MS).

ICP-MS/MS can use reaction gas modes to reduce intense spectral overlaps, such as 

S, allowing these previously difficult elements to be measured at a low level. Reaction gas modes can also reduce matrix-based overlaps to lower levels, enabling ultratrace level analysis.

He mode cannot resolve isobaric overlaps, such as 

Pb, and so on. In typical, routine ICP-MS analysis, these overlaps are avoided by measuring an alternative mass (

Pb, in the examples above). However, measurement of the overlapped isotope is required for some applications and provides performance benefits, such as higher sensitivity, in others. In these cases, a reaction gas mode may be able to resolve the overlap.

Doubly charged ion interferences form when an atom loses two electrons in the plasma and forms a doubly charged (M

) ion. A quadrupole mass filter separates ions based on mass to charge ratio (

 ions appear at half their true mass. Most elements form insignificant levels of M

 ions, but some, such as the rare earth elements (REEs), do form a small percentage of M

. Single quadrupole ICP-MS can correct for normal levels of M

 overlap, but this approach is not effective at high levels of interference. For analyses such as trace arsenic and selenium in REE minerals, a reaction gas method is more effective.

Peak tail overlaps affect trace analytes measured next to intense major element peaks. An example is trace manganese (

 55) analysis in an iron or steel matrix (the major isotopes of iron are at 

 54 and 56). Peak tail overlaps are resolved using the double mass selection of MS/MS, as the abundance sensitivity (AS) of a tandem MS is the product of the AS of the two mass filters, so Q1 AS x Q2 AS in the case of ICP-QQQ.

Step #5: Select the Preferred Reaction Gas Mode

For overlaps that cannot be addressed using He mode, an appropriate reaction gas mode is selected, which is not as daunting as new users often assume, because ICP-MS/MS ion–molecule reaction chemistry is well understood and documented. Applications can often be run using the predefined method settings provided with the instrument software. Manufacturers’ application notes (5) and users’ publications (6) document proven methods for many more applications.

ICP-MS/MS users also benefit from unique method development tools, including product ion and precursor ion scans. These are useful for identifying suitable analyte product ions and potential matrix-based product ion overlaps (7).

Step #6: Ensure Control of Cell-Formed Reaction Product Ions 

When reaction gases are used in ICP-MS, new product ions can be formed in the CRC. The creation of potentially interfering product ions is a major downside of reaction gases compared to helium KED mode, but control of product ion formation is a key benefit of a tandem MS configuration. The Q1 mass filter, before the CRC, ensures that only ions at the target mass—that is, the analyte and any on-mass interferences—enter the CRC.

Q1 therefore prevents the formation of product ions from analytes or matrix elements present at other masses. However, product ions can still be created from the reaction gas itself. For example, when 

 is measured in organic matrices, Q1 is set to 

 product ions could react further with the NH

The sequential reactions that form the NH

 product ions can be prevented using either mass discrimination or ion energy discrimination, depending on the ICP-MS/MS cell configuration and operating conditions. The energy discrimination approach is effective because product ions created from the cell gas only have the kinetic energy they derive from the collision. These low-energy product ions have insufficient energy to pass through the gas-filled cell, so they do not appear in the mass spectrum (Figure 2).

(2) N. Yamada, J. Takahashi, and K. Sakata, 

(4) E. McCurdy, N. Sugiyama, and S.M. Wilbur, 

(5) “Handbook of ICP-QQQ Applications” (Agilent Technologies, publication 5991–2802EN, 2020).

(6) Agilent ICP-QQQ bibliography https://www.agilent.com/en/promotions/8900icp-qqq-trust.

(7) E. McCurdy, G. Woods, and N. Sugiyama, 

 are with Agilent Technologies Inc. Direct correspondence to: 

Columns

Method development for ICP-MS/MS should not be difficult. The six steps here will guide you to success.

Key Steps in Method Development for Inductively Coupled Plasma–Tandem Mass Spectrometry (ICP-MS/MS)

The role and performance of handheld laser-induced breakdown spectroscopy (hLIBS) devices is discussed in the context of meteorite research exploration while targeting specific questions of geologic interest in the field. The fast data acquisition and minimal sample preparation features of these devices make them very attractive as candidate tools for future astronauts in geochemical mapping, resource prospecting, sample selection, and hazard identification. Although hLIBS instruments have recently become commercially available, they have not yet been widely applied to meteoritical field campaign. This article reviews how the very promising advanced analytical tool hLIBS may be helpful to classify and discriminate meteorites in the field.

The mineralogy, chemistry, and isotopic composition of meteorites can provide unique information on the materials that constitute the solar system and allow deciphering its cosmic history because the composition of many belt asteroids have never changed drastically. The composition of these asteroids has not changed because of gravitational high temperatures and pressures as larger bodies do. Meteorites do not often differ much from terrestrial rocks, a similarity that often compromises their recognition and preservation, leading to their easy disappearance in the landscape should they not be seen to fall and then promptly collected. In particular, iron meteorites can be easily confused with man-made iron objects simulating typical features of a true meteorite. After a careful analysis, a specimen visually thought or suspected to be an iron meteorite may be revealed to be a pseudometeorite, or meteor-wrong. These materials are attracted by a magnet and include terrestrial iron oxides such as magnetite along with many different types of man-made iron byproducts, such as slags and castoff iron with very rough, burned and melted glassy (vitreous) surfaces. There may occasionally be the presence of vesicles (gas bubbles) because corrosion occurred over time.

Although meteorites sometimes appear different from ordinary terrestrial rocks lying around them, their preliminary identification in the field needs a practiced eye. In particular, many meteorites do not contain the common earth mineral quartz, and generally do not present vesicles. Most meteorites, especially iron meteorites, are much heavier in the hand than an ordinary earth rock, and contain a significant amount of iron (Fe) and nickel (Ni). Thus, the first step to identify them is whether they stick to a magnet. However, many terrestrial rocks are also attracted by a magnet, so this test cannot be considered definitive, but only represents a preliminary proof needing of further analysis. Visual and magnetic methods are not exhaustive for a verifiable identification that requires an adequate laboratory analysis to ascertain the extraterrestrial origin of suspected, especially iron, meteorites. In particular, Ni is an element rare in terrestrial rocks, but is present very often in meteorites. Thus, if visual inspection and the magnet test are positive, the presence and content of Ni should be verified, together with that of some other diagnostic trace elements such as gallium (Ga), germanium (Ge), and iridium (Ir).

Analytical Instrumentation for Field Analysis

The availability of analytical instrumentation for routine chemical analysis in the field has been one of the most long-standing needs of the scientific community. The requirements of such a field analyzer include: (i) lightweight and portability, to allow use by one individual; (ii) ruggedized and robust structure for use in a variety of environments; (iii) simplicity of use; and (iv) capability of identifying a large number of elements over a short acquisition time. Many different analytical techniques have been tested in developing a portable instrumentation appropriate for chemical analysis in the field, including mass spectrometry, ion mobility spectrometry, gas chromatography, nuclear magnetic resonance, electron spin resonance, millimeter wave, and terahertz spectrometry, as well as many different forms of optical spectroscopy, such as X-ray fluorescence (XRF), Raman, near-infrared, mid-infrared, ultraviolet–visible, and laser-induced breakdown spectroscopy (LIBS). In particular, XRF and LIBS appear to meet the criteria for performing elemental analysis in the field under ambient environmental conditions.

Instrumentation, Analytical Procedures, and Features of LIBS 

The main components of a LIBS instrument are: (a) a laser, the most used being the short-pulsed Q-switched laser that releases a huge amount of energy in a short period, which is used to ablate the sample; (b) a time control system, allowing for the precise control of the beginning of laser pulse, the number of pulses, the interval between pulses, the beginning of spectrum capture, and the acquisition time; (c) a set of mirrors and lenses focusing the laser light onto the sample and collecting the light emitted from the plasma; and (d) a spectrometer–detector system for collecting and spectral discriminating of plasma light emissions that contain all the information on the sample composition and physical conditions of the plasma.

In a LIBS experiment, the high-energy laser pulses are focused on the surface of the sample that is placed in air or an inert gas at ambient or controlled pressure, so to ablate a small portion of it, which is then vaporized, generating the plasma that contains thermally excited atoms, ions, and molecules of the analyte. As the plasma cools down, the excited species decay to lower energy levels emitting electromagnetic radiations typical of each of them, which are collected by a spectrometer that resolves the specific emissions. The acquired signals are then processed, visualized, and analyzed using a dedicated software. An optically thin plasma in the condition of local thermodynamic equilibrium (LTE) is expected to yield a LIBS spectrum whose pattern corresponds to the elemental composition of the sample (1).

Quantitative elemental analysis by LIBS is based on the proportionality between the spectral emission line intensity of the element and its concentration, and can be performed by constructing a classical calibration curve that relates the measured elemental line intensity to the elemental concentration of standards of known composition, or by the calibration free (CF) approach that avoids the use of calibration standards and can address sample matrix issues (2). Furthermore, multivariate statistical methods are used, which can process all spectral information contained in a LIBS spectrum and allow relating them to the composition of the sample, and identify and discriminate different materials (3).

With respect to various destructive and expensive analytical techniques, LIBS features significant advantages, including sensitivity to light elements, versatility, minimal destructivity, rapidity, relatively low operating costs, and the possibility of in-field operations (4). Furthermore, several handheld (h) LIBS instruments, which consist of compact, lightweight, self-contained units that can be used comfortably while being held in the hand of an operator, have been developed and commercialized (4). Recently, hLIBS instruments were used for the rapid in situ measurement of most major and trace elements in a variety of rocks including meteorites, which cannot be achieved by using other field portable instruments.

A Few Representative Examples of hLIBS Application

In the last decade, LIBS has shown to be a very promising technique for meteorite analysis and identification (5–7). In two companion works (8,9), a commercial hLIBS instrument was used for the first time to achieve a preliminary chemical identification and classification of meteorites. The objective of one study (8) was to discriminate between a classified and certified meteorite named Agoudal, a suspect meteorite fragment termed meteor-wrong and a pig iron product (Figure 1a) on the basis of their major and trace element composition. A calibration-free (CF)-LIBS approach was used to quantify the main elements Fe, Ni, and Co in the iron meteorite and Si, Fe, Mn, and Ti in the other two samples (Figure 1b). Although the quantitative data obtained were only partially satisfactory, requiring further improvements of hLIBS analyzer components to increase analytical sensitivity, the technique was able to recognize and discriminate between a true meteorite specimen and mislabeled or unlabeled dubious meteorite samples, which allowed a decision to be made about whether to undertake further laboratory analysis or not.

The same hLIBS instrument was used in a companion effort (9) to discriminate between iron, stony-iron, and stone meteorites and meteor-wrongs (Figure 2). In this work, an innovative strategy was developed allowed the identification and classification of meteorites according to their nature and origin. This novel approach was based on the integration of results of automated feature selection from broadband LIBS spectra and supervised learning by a novel fuzzy-rule-based inference algorithm that could be easily interpreted by human users. The algorithm developed was able to assign correctly 25 out of 26 unclassified meteorite samples to their class with a high total accuracy (96.15%), and identify the most relevant intragroup variations, while only 18 of the samples were properly recognized by partial least square–discriminant analysis (PLS-DA). These results confirm that the classification of LIBS spectra of meteorites is a nonlinear process, and that a high accuracy cannot be achieved using prevalent traditional models, such as principal component analysis (PCA) and PLS-DA, which can only evaluate linear combinations of input variables. Meteorite identification, especially for chemically similar samples, is a challenging decision-making problem that requires a compromise among human experience, visual evidence, and efficient analytical data processing that can be addressed by a fuzzy logic approach. Thus, this approach based on the identification of elements using the six wavelengths selected is expected to be extended to subsequent meteorite investigations, and suggests that such a methodology may also be suitable for other types of chemometric studies.

The examples described above confirm that hLIBS is particularly promising for performing minimally destructive (Figure 2 a,b,c) in-situ measurements aiming at a preliminary identification, discrimination and classification of meteorites, and recognition of mislabeled or unlabeled and dubious meteorite specimens in the field, museums, and private collections, as well as deciding about any further laboratory analysis. More widespread applications for real-time analysis in the field are expected to be possible if hLIBS analyzers are used in combination with an on-board global positioning system (GPS) and wireless data transmission devices. Furthermore, it would also be helpful to develop robust algorithms that allow hLIBS data to provide reliable chemical fingerprints for a variety of meteorites, as well as construct spectral libraries to permit sample discrimination and provenance determination in real time.

A promising relevant future application of LIBS is expected in the study and analysis of the chondrite class of meteorites, as they are known to preserve the chemistry of the earliest solar system and, to some extent, the pre-solar molecular cloud from which it originated. In particular, carbonaceous chondrites consist in a primitive subgroup of stony meteorites featuring an elemental composition very similar to that of the sun and the overall universe. These meteorites are aggregate rocks, meaning they consist mainly of a matrix of hydrous and anhydrous silicates packed together, which do not show any sign of metamorphism or alteration by high heating. The most intriguing feature of these chondrites is their abundance in carbon (about 1.5–4%), which is mostly present in the organic form. Thus, the investigation of carbon forms and distribution in chondrites can contribute better than most other elements to understanding the early stages of solar system formation, due to the sensitivity of this element to chemical and physical conditions that prevailed in that situation. Recently, an important increase was noticed in the collection of carbonaceous chondrites and achondrites in Africa, although these types of meteorites are very difficult to spot on the ground.

The detection and quantification of free carbon in meteorites have been commonly performed by either destructive techniques, including total organic carbon (TOC) and inductively coupled plasma–mass spectrometry (ICP-MS), or by techniques implying sample preparation, such as scanning electron microscopy (SEM), which uses Au to coat the sample surface, whereas vibrational spectroscopies such as infrared (IR) and Raman are generally used to analyze carbon compounds. Unlike the above techniques, LIBS is able to detect both elemental and combined carbon, even at ppm levels on the whole fragment or its petrographic thin sections, and reveal differences in the spectral emission line of C in different meteorites also in comparison to a meteor-wrong sample (10). This application may be considered a great advance in the context of human space exploration, where hLIBS instruments have shown to be very promising candidate tools in planetary geochemical mapping, resource prospecting, sample selection, and hazard identification.

Handbook of Laser-Induced Breakdown Spectroscopy

 (John Wiley & Sons, Chichester, United Kingdom, 2nd ed., 2013).

(2) A. Ciucci, M. Corsi, V. Palleschi, S. Rastelli, A. Salvetti, and E. Tognoni, 

Handbook of Laser‐Induced Breakdown Spectroscopy

, D.A. Cremers and L.J. Radziemski, Eds., (Wiley, Chichester, United Kingdom, 2nd ed., 2013), pp. 223–255.

(4) G.S. Senesi, R.S. Harmon, and R.R. Hark, 

, 106013 (2021). https://doi.org/10.1016/j.sab.2020.106013.

, 231–267 (2014). https://doi.org/10.1016/j.earscirev.2014.09.008.

, G. Colonna, M. Capitelli, and A. Larucchiuta, Eds., (IOP Series in Plasma Physics, Bristol, United Kingdom, 2019), pp. 1–39.

, 104929 (2021). https://doi.org/10.1016/j.apgeochem.2021.104929.

(8) G.S. Senesi, P. Manzari, G. Tempesta, G. Agrosì, A.A. Touchnt, A. Ibhi, and O. De Pascale, 

, 607–614 (2018). https://doi.org/10.1111/ggr.12220.

(9) G.S. Senesi, P. Manzari, A. Consiglio, and O. De Pascale, 

, 1664–1675 (2018). https://doi.org/10.1039/C8JA00224J. 

(10) O. De Pascale, P. Manzari, B. Marangoni, G. Nicolodelli, and G.S. Senesi, 

Reports of the National Academy of Sciences: Memoirs of Physical and Natural Sciences

 is a researcher at the Institute for Plasma Science and Technology, in Bari, Italy, which is part of the National Research Council (CNR) of Italy. Direct correspondence to: 

Special Feature

The fast data acquisition and minimal sample preparation of handheld LIBS devices make them very useful for geochemical mapping, resource prospecting, sample selection, and hazard identification.

Handheld Laser-Induced Breakdown Spectroscopy (hLIBS): A Valuable Tool for Terrestrial and Extraterrestrial Recognition of Meteorites in the Field

Nickel-based alloys have excellent corrosion resistance. They are also well-known for their mechanical properties, high temperature durability, and heat transfer performance. Because nickel-based alloys have these properties, they are widely used in creating nuclear safety level pressure equipment. It is the first-choice material for the heat exchanger tube plate and U-tube plate of the nuclear island steam generator, which has a profound impact on the safe operation of the reactor. Therefore, it is necessary to calibrate the main elements in nickel-based alloys. Accurate detection of the main elements in nickel-based alloys is important in ensuring the advancement of process research. A method for determining chromium, iron, and nickel in a nickel-based alloy by X-ray fluorescence spectroscopy is presented here. The detection range of chromium, iron, and nickel is 0.2–1.6 g/L, 0.2–2.3 g/L, and 0.5–3.3 g/L, respectively. After the three elements are corrected respectively, good correlation is obtained, 

 (nickel)= 0.999114. The relative standard deviations for chromium, iron, and nickel are 0.75%, 0.65%, and 0.65% respectively. The standard sample is tested to confirm the measured value is in agreement with the certified value.

Nickel-based alloys have high strength and oxidation resistance. At high temperatures, nickel-based alloys also have corrosion resistance. Nickel-based alloy materials are found regularly in nuclear reactor pressure vessels and steam generators. They are widely used in manufacturing reactor equipment and for building new equipment. There have been many reports on the structure and properties of chromium–iron–nickel (1–5) to understand their properties for better business production development (6–10). Accurate determination of the main elements in nickel-base alloys can provide strong data and technical support for chemical processes.

Currently, detection methods for nickel-based alloys include inductively coupled plasma–mass spectrometry (ICP-MS), graphite furnace atomic absorption spectroscopy (GFAAS), and high frequency combustion infrared (IR) absorption spectroscopy (AS). To achieve rapid and nondestructive detection, X-ray fluorescence (XRF) spectroscopy is used in a unique role to characterize the elemental composition of substances (11–16). XRF spectroscopy has become an important means of qualitative and quantitative analysis because of its simple sample preparation and its ability to analyze a wide range of elements. XRF has been widely used in many industries, such as the ore, chemistry, biology, environmental, and nuclear power fields (17–21).

The sample preparation methods of XRF mainly include three methods: powder, solid, and solution. The solution method is not limited by the shape of the sample, and can be used to determine the samples with the shapes of filaments and chips. Because the solution method is not limited by the shape of the sample, this method was used for detection in this study.

XRF spectroscopy utilizes primary X-ray photons or other excitation sources to illuminate the sample for analysis. After the inner electrons of the element in the sample are knocked out, the electrons outside transition. The transition process produces energy difference, which is released in the form of X-ray energy. The energy of electrons in each orbit is determined, so the energy difference caused by electron transition and the energy of X-ray released can also be determined. 

 is the process of converting the measured intensity of the elemental analysis line of the sample into the elemental concentration. For infinitely thick samples, the analysis intensity is only related to the concentration of the analysis elements, so when the sample composition is simple, the analysis intensity and element concentration show a simple linear relationship. For complex samples, because of the existence of matrix effects, there is still a positive correlation between the elemental emission analysis intensity and the element concentration of the sample. However, the quantitative relationship is slightly more complicated and needs to be corrected.

A S8 Tiger wavelength dispersive XRF spectrometer (Bruker) was used for this study. The instrument parameters are shown in Table I.

High purity iron (content >99.988%), chromium powder (purity: 99.95%), high purity nickel block (content: 99.999%), nitric acid (superior purity), hydrochloric acid (superior purity), nickel-based alloy standard samples BS H3C, IARM 53F, BS 690A, and B.S. 600-2; and ultrapure water (resistivity: 18.2 MΩ·cm) were used for this study. The above samples are national standard samples and analytical reagents.

Because this experiment uses the solution method for detection and analysis, the nickel-based alloy can be detected only after the alloy is dissolved. Therefore, the standard curve is established with the standard solution of chromium, iron, and nickel, and the concentration range of the working curve is determined, according to the element content in the sample.

For the standard stock solution, a specific amount of chromium, iron, and nickel was weighed to prepare a 50 g/L standard stock solution. For the standard solution concentrations, the standard solution concentration of chromium is 0.2, 0.4, 0.6, 0.8, 1.0, 1.2, 1.4, and 1.6 g/L. The standard solution concentration of iron is 0.2, 0.5, 0.8, 1.1, 1.4, 1.7, 2.0, and 2.3 g/L. The standard solution concentration of nickel is 0.5, 0.9, 1.3, 1.7, 2.1, 2.5, 2.9, and 3.3 g/L.

Uniform design is an experimental design through a set of carefully designed tables. With this method, the concentration point has good dispersion. The concentration point distribution of the standard solution in columns 1, 3, and 4 (

 = 0.2000) of uniform design test table U

) is adopted, and the concentration distribution of the solution in the working curve is shown in Table II.

A certain amount of the nickel-based alloy standard sample is weighed and dissolved in hydrogen chloride (HCl) and nitric acid (HNO

), and the volume is fixed in a 50-mL volumetric flask. The mass concentration of ferrochromium and nickel can be calculated by mass and volume.

The liquid sample was placed into the liquid cup for testing. Then, the power supply was turned on, cooling water, the P10 gas (10% CH

 and 90% argon), and helium. Next, the high-pressure generator was turned on. Once the generator was on, the instrument was measured after it was stable. The measuring conditions of the instrument are shown in Table III.

Determination of Working Curve of Chromium, Iron, and Nickel

According to the properties of each element, the least absolute deviation and the best square 

 of the correlation coefficient are taken as the target to fit the data. In the chromium–iron–nickel system, there is an absorption enhancement relationship, where nickel can stimulate the signals of iron and chromium to create an enhancement effect. Chromium can absorb the signals of iron and nickel, and an absorption effect occurs when this happens. As a result, it was integral that the fitting curve was corrected so good measurement results could be achieved. The absorption enhancement effect between elements can be corrected by a coefficient correction. To get the minimum standard deviation, the appropriate correction method was selected in the SpectraPlus software. The element chromium was corrected by the fixed theoretical α coefficient, and the elements iron and nickel by the empirical α coefficient. The square of the correlation coefficient after correction is 

 (nickel)= 0.999114, respectively. The working curves of chromium, iron, and nickel are shown in Figures 1, 2, and 3.

This method is used to carry out precision tests on nickel-based alloy standard samples, repeat the measurement 11 times, and produce statistics on the results. Table IV displays the results for the XRF precision test.

The method established by the XRF spectrometer needs to be verified for accuracy. The accuracy of the method can be verified by determining standard samples. This method analyzes the standard samples BS H3C, IARM 53F, BS 690A, and B.S. 600-2 to verify the reliability of the method. Table V shows that the measured value is similar to the certified value of the standard sample.

The sample prepared by solution method can be free from the limitation of sample shape. According to different elements, appropriate instrument parameters were selected to ensure accurate and stable measurement. The detection range of chromium, iron, and nickel is wide, which meets the requirements of detection and analysis. The method has good correlation. The square of correlation coefficient is 

 (nickel) = 0.999114, respectively. The result of precision measurement is satisfactory. The result of accuracy measurement shows that the measurement value is close to the certified value. It shows that the method has been successfully established, which lays a good foundation for future research.

(1) K. Leinartas, M. Samuleviien, A. Bagdonas, R. Juskenas, and E. Juzeliūnas, 

 70–77 (2003). http://dx.doi.org/10.1016/S0257-8972(03)00003-3

(2) K. Leinartas, M. Samuleviien, A. Bagdonas, A. Sudaviius, V. Lisauskas, and E. Juzeliūnas, 

 494–499 (2001). https://doi.org/10.1016/S1388-2481(01)00206-5

 1449–1452 (1997). https://doi.org/10.1016/S1359-6462(97)00322-9

(4) S. Jiang, D. Sun, Y. Zhang, and B. Yan, 

 1–13 (2018). https://doi.org/10.3390/met8040217

(5) S. Shi, C. Chen, Z. Guo, G. Li, Y. Wu, M. Su, and H. Wang, 

Accurate determination of the elemental composition of nickel-based alloys is essential, given their use in high-performance equipment. This XRF technique enables rapid and nondestructive detection, as an alternative to existing approaches.

February 2022

Interpretation of Raman Spectrum of Proteins

Data Transforms in Chemometric Calibrations: Application to Discrete-Wavelength Models, Part 1: The Effect of Intercorrelation of the Spectral Data

Key Steps in Method Development for Inductively Coupled Plasma–Tandem Mass Spectrometry (ICP-MS/MS)

Handheld Laser-Induced Breakdown Spectroscopy (hLIBS): A Valuable Tool for Terrestrial and Extraterrestrial Recognition of Meteorites in the Field

Determining Chromium, Iron, and Nickel in a Nickel-Based Alloy by X-ray Fluorescence Spectroscopy

Influence of Bulk Density on Near-Infrared Spectrum Detection of Cement Raw Materials and Compensation Methods