Encyclopedia of Biophysics

Living Edition
| Editors: Gordon Roberts, Anthony Watts, European Biophysical Societies

Absorption Spectroscopy: Practical Aspects

  • Alison RodgerEmail author
  • Paul Wormell
Living reference work entry
DOI: https://doi.org/10.1007/978-3-642-35943-9_779-1


UV-visible absorbance is most commonly used to determine the concentration of a sample and also to give an indication of its purity. It is very easy to collect absorbance spectra of biomolecules; however, they are not always useful because of some of the issues outlined below (Nordén et al. 2010; Rodger and Nordén 1997). Nucleic acids (DNA and RNA), proteins, and peptides absorb very little light at wavelengths greater than 300 nm in the absence of ligands or prosthetic groups with chromophores (absorbing units). However, it is usually wise to collect absorbance data from 200 nm to about 350 nm. For a biomolecule that does not absorb between 350 and 310 nm, if the spectrum is not flat in this region, then the sample includes particles whose size is of the order of the wavelength of light; therefore, what is being measured includes scattering of the incident light rather than simply absorption. An extreme example of this is when an absorbance spectrometer is used to measure cell density in a culture using say 450 nm light – the measured spectrum is only attributable to light scattering despite usually being referred to as absorbance.

Basic Characteristics

The protocol outlined below is one way of acquiring a simple wavelength scan of the absorbance of a solution of biological macromolecules; this covers spectroscopic parameters, solvents, and baselines. Consideration should be given to artifacts that may arise including the abovementioned scattering and also absorbance flattening due to inhomogeneous samples and local high concentrations. The protocol assumes the instrument being used is a double-beam instrument with sample and reference beams. Some modification is required for a single-beam instrument.

The procedure to perform a wavelength scan is as follows:
  1. 1.



The wavelength range will usually be 350–200 nm, unless the solvent or buffer cuts out the low wavelength end (see below) or ligands are included in the sample. Only go below 200 nm if the instrument is nitrogen purged.

A data interval of 0.2–1 nm is adequate as the bands in the spectrum of a solution are all broad. Shorter data intervals will produce an unnecessarily large data file, though a shorter data interval may help any noise reduction analysis. If data are being collected to aid interpretation of linear dichroism (LD) or circular dichroism (CD) spectra, ensure that the data interval is the same for the absorbance and LD or CD spectra.

The (spectral) bandwidth is the wavelength range of the radiation at any specified wavelength. If data are collected every 0.5 nm, then a bandwidth of 0.5 nm is appropriate. A larger value will lead to a greater averaging of data than the data interval suggests. Do not choose a much smaller bandwidth than the data interval as the incident light intensity will be reduced and the spectroscopic noise will increase without improving the quality of your data.

The data averaging time determines the signal to noise ratio and affects the scan speed required to produce undistorted spectra. Between 0.033 and 0.1 s is usually a good choice. Longer times are required if the absorbance is very small (less than 0.1 absorbance units), or small differences in absorbance are being determined (as for DNA melting curves, see below). There is no point in using a long averaging time if the noise level is not apparent in the spectrum at a shorter averaging time.
  1. 2.

    Solvent or buffer


Choose the solvent or buffer for your experiment and then run a spectrum of it with the reference holder empty. Run a spectrum over the desired wavelength range to check that the solvent/buffer absorbance is not significant over the wavelength range of interest. If it is, then you need to change the solvent/buffer.

Note that phosphate buffers are essentially spectroscopically invisible over the wavelength range usually used; however, phosphate does interact with some samples, especially membranes. Cacodylate and Tris buffers are also transparent down to ∼190 nm. Ammonium acetate cuts off at ∼210 nm. Chloride ions begin absorbing at longer wavelengths than 200 nm, so high salt spectra cannot be collected at the lower wavelength end of the spectrum. Phosphate-buffered saline (PBS) is often described as “phosphate” but typically includes 50 mM NaCl. Higher buffer concentrations may be able to be accommodated in shorter path length cuvettes or cells (cf. Beer–Lambert law), so if it is not possible to dilute or change the buffer, try a smaller path length cuvette – though this will also reduce the number of analyte molecules in the light beam. 5-mm cuvettes will stand on their own in a normal 1-cm cuvette holder; 1-mm cuvettes will need spacers to hold them vertical – if they are not held vertical, then you will be working with a variable path length. Smaller path length cuvettes will need special holders.
  1. 3.



To collect data down to 200 nm, cuvettes (also known as cells) should be made of fused silica. This material is sometimes referred to as “quartz,” hence the “Q” label that appears on some cuvettes. Optical glass (which may be labeled “G”) and polystyrene cuvettes are not suitable for these applications, although they may perform well at visible wavelengths. Stoppers should be used whenever possible, especially if the solvent is volatile. Cleanliness is important: human perspiration contains amino acids and peptides, so fingerprint residues absorb at UV wavelengths and an apparently clean cuvette can have a significant absorbance below 250 nm. Cuvettes should therefore be handled by their frosted sides or edges or the top and bottom where the light will not be passing through to avoid fingerprints on the surface. A variety of cleaning methods is available (see, e.g., Australian Standard AS 3753–2001 (R2016) Recommended practice for chemical analysis by ultraviolet/visible spectrophotometry). These are of varying rigor. Cuvettes should be wiped with a lint-free tissue before use. For double-beam spectrometers, matched cuvettes (which may be purchased in sets from the manufacturer) are preferable if one wishes to put a reference cuvette in the reference beam. According to AS 3753–2001 (R2016), the absorbances of a pair of matched silica cells should differ by no more than 0.008 at 240 nm.

The optimal path length for a given experiment depends on the sample and buffer concentrations. If the molecular structure remains the same on dilution, then the choice can be based on convenience or by reducing the amount of buffer salts in the path length. If a demountable cell is used, it generally takes significant practice to ensure it is being assembled reproducibly. Potassium chromate in basic solution has an extinction coefficient of 4830 mol−1 dm3 cm−1 at 372 nm. This can be used to measure the path length of a cuvette.
  1. 4.

    The baseline


There are a number of ways of collecting the baseline of an absorbance spectrum. The key thing is to ensure that the absorbance of the cuvette and buffer is subtracted from that of the analyte in the final spectrum. Two options are as follows:

  1. a.

    The usual method is to place the solvent/buffer in a pair of matched cuvettes in both the reference holder (usually the rear position) and the sample holder and perform a baseline accumulation. This is often stored in the instrument as a baseline and automatically subtracted from subsequent spectra. It is prudent to check that baseline subtraction from a second scan of the solvent/buffer spectrum gives a flat spectrum with an acceptable signal-to-noise ratio. This ratio will affect the measurement uncertainty in the absorbance readings from the spectrophotometer and hence the precision of any concentrations that are calculated using the Beer–Lambert law or any other data analysis undertaken. The reference cuvette then remains filled with the solvent/buffer when sample spectra are collected.

  1. b.

    For low-absorbance samples or high-absorbance solvents/buffers, it is generally preferable to store an instrument (i.e., with empty sample holders) baseline for automatic subtraction. One then collects a spectrum of the solvent/buffer referenced against an empty sample compartment and manually subtracts this from subsequent sample spectra. This approach then lets one see where the solvent/buffer baseline is contributing to observed spectral features.

  1. 5.

    The spectrum


Place the sample in the sample cuvette in the sample holder, fill the reference beam with whatever it contained when you measured the baseline, and record the spectrum and subtract the baseline.



  1. Nordén B, Rodger A, Dafforn TR (2010) Linear dichroism and circular dichroism: a textbook on polarized spectroscopy. Royal Society of Chemistry, CambridgeGoogle Scholar
  2. Rodger A, Nordén B (1997) Circular dichroism and linear dichroism. Oxford University Press, OxfordGoogle Scholar

Copyright information

© European Biophysical Societies' Association (EBSA) 2018

Authors and Affiliations

  1. 1.Department of Molecular SciencesMacquarie UniversityMacquarie ParkAustralia
  2. 2.School of Science and HealthWestern Sydney UniversityPenrithAustralia

Section editors and affiliations

  • Alison Rodger
    • 1
  1. 1.Department of Molecular Sciences, Macquarie UniversityNWSAustralia