Review: Analysis of Writing Inks on Paper Using Direct Analysis in Real Time Mass Spectrometry

Emily C. Lennert





pen, ink, gel, ballpoint, fluid, direct analysis in real time, DART, mass spectrometry, MS, DART-MS, questioned, documents, writing

Article Reviewed

  1. Jones, R. W.; McClelland, J. F. Analysis of writing inks on paper using direct analysis in real time mass spectrometry. Forensic Science International. 2013, 231, 73-81.


The opinions expressed in this review are an interpretation of the research presented in the article. These opinions are those of the summation author and do not necessarily represent the position of the University of Central Florida or of the authors of the original article.


Examination of questioned documents is a broad field that includes the analysis of writing inks, i.e. pen inks. Writing inks are of particular importance for signatures and handwritten records. Ink analysis may be used to determine whether a document has been altered, back dated, or to determine the approximate age of a document.

The authors of this study present direct analysis in real time – mass spectrometry (DART-MS) as a non-destructive technique for the analysis of writing inks on paper. The technique allows for sampling directly from the document without affecting the appearance of the document, e.g. sample does not need to be cut out or extracted and no visible effect is observed. Other MS techniques often produce spectra that are dominated by the ink dyes, which are common between dyes and offer little value in differentiating inks from one another. DART-MS allows for the detection of many components, other than the dyes, in the ink sample which may allow for samples to be differentiated.

Samples were introduced into the DART ion stream using a custom sample mount to ensure reproducibility. The DART was operated in positive ion mode. Samples were created by drawing lines freehand on paper.

First the authors studied the effect of paper on the resulting mass spectra obtained from DART-MS analysis. To do so, three inks were studied on 16 different papers. Inks included a ballpoint, gel, and fluid ink. Lines were drawn on the paper and the paper was stored in a file folder, in the dark, for 8-9 months prior to testing. Each sample was tested in triplicate, for a total of 48 spectra per ink type. After DART-MS analysis, a library of mass spectra was created for each ink, individually. The spectra for each ink were searched against its respective library to determine whether the library would return a match for the proper ink-paper combination, e.g. the sample replicates. A match factor of 999 was considered a perfect match. For ballpoint inks, high-quality matches were observed with the exception of one paper, Hammermill Color Laser Gloss. This paper produced considerable background that interfered with the ink spectra, even upon background subtraction. Overall, for ballpoint inks, a median match factor of 861 was observed, indicating very good match quality, according to the authors. For gel ink, a median match factor of 829 was observed and was considered good quality matching. The glossy paper did not interfere with the gel inks. The authors attribute the lower match quality to the tendency of gel inks to produce weaker spectra with fewer peaks, compared to ballpoint inks. For fluid inks, a median match quality of 765 was observed. This is due to two interfering papers. One interfering paper was the same glossy paper that interfered with the ballpoint inks. The other paper was Crane’s Resume Paper, a thick 100% cotton paper. The authors attribute the low match quality in this paper to the interaction of the fluid ink with the paper. Fluid inks soak into the paper and are therefore harder to analyze by DART-MS; the authors theorize that this particular paper may be especially effective at absorbing the ink.

Ink age was examined next to determine how the mass spectra change over time. Immediately upon deposition, volatile components of the inks will begin to evaporate, resulting in a change in mass spectra. Once again, a ballpoint, gel, and fluid ink were examined. Samples were prepared over the course of 11 months by drawing lines freehand on paper, and storing in file folders in the dark. Comparing the 4 day old and 332 day old ballpoint ink spectra, considerable changes are observed in the lower mass region where the more volatile components are. The higher mass peaks appear unaffected by aging, while the lower mass peaks, i.e. 100 m/z, 139 m/z, and 199 m/z, tend to show a decrease in intensity over time, as seen in Figure 3 within the study. In comparing the spectra at 9 and 336 days for the gel ink, the loss of a peak at 301 m/z is observed. This component, which was not identified, drastically decreased within the first few months of aging, as seen in Figure 5 within the study. Pentaethylene glycol, 239 m/z, decreased in intensity by approximately half over the course of the aging study. Other components appeared relatively stable, with some appearing to increase over time. However, the authors attribute this apparent increase in intensity to a decrease in the intensity of the peak used for scaling. Fluid ink analysis revealed three homologous series in the spectra. A homologous series is a series of peaks representing compounds with the same structure, with one varying parameter. For example, one of the observed series was for polyethelene glycol (PEG), [HO(C2H4O)nH + H]+, where n represents a number that, in this case, varied from 3 to 9. The next two series consisted of PEG derivatives. As the ink aged, several peaks were observed to decrease over time, as seen in Figure 7 within the study. The peak at 239 m/z appeared stable over time and was used to scale the other peaks for relative intensity determinations.

Finally, ink sample spectra were searched against libraries to determine whether inks could be identified based on mass spectra. Six separate libraries were created: black ballpoint, blue ballpoint, black fluid, blue fluid, black gel, and blue gel. Each library contained two spectra for each ink. The “unknown” ink sample was searched in the appropriate library, and results were considered a match if the search returned the ink’s replicate spectra, i.e. the other spectra of the pair, as the number one hit. For all inks, the collective success rate observed was 92% correct. The percent correct classification rate for each library can be observed in Table 1 within the study. Overall, the greatest matching success was observed in ballpoint pen inks, with 99% and 100% correct classification for black and blue inks, respectively. Fluid inks had the lowest correct classification rate, with 81% correct for black and 77% correct for blue inks. However, the authors note that all “incorrect” matches may not be, in fact, incorrect, as a manufacturer may use the same ink in multiple pen models. The authors note that several incorrect matches were of pens from the same manufacturer, but different model. Another method of determining the quality of the match for inks was to examine the match factor of the number one hit compared to the next, incorrect, hit. Black ballpoint pens had high match factors for the number one hit, but also returned high match factors for the number 2 hit. This indicates similarity between samples, or that black ballpoint pens may be more uniform and therefore more difficult to differentiate.

Scientific Highlights

  • DART mass spectra may be obtained easily and rapidly from ink samples on paper without damaging the document.
  • Highly processed or coated papers, such as the glossy paper reported in the study, may cause background interference
  • Papers that are thick and absorbent may reduce the accuracy of analysis for fluid inks.
  • The DART spectra of ink changes most within the first few months of storage, as volatile components are lost. Changes generally stabilize after 11 months.
  • Libraries for reference ink spectra may be searched with high success, particularly for ballpoint and gel inks.


Document analysis may require the analysis of writing inks, however some current methods are destructive to the document. DART-MS offers a non-destructive technique for ink analysis on documents.

Potential Conclusions

  • DART-MS may offer a noon-destructive option for ink analysis on documents.
  • Inks may be characterized by DART-MS, and searchable libraries may be created to aid in the identification of an unknown ink sample.