RU2809556C2 - Method for determining the timing of handwritten inscriptions on documents - Google Patents

Abstract

FIELD: forensic technical expertise.

SUBSTANCE: invention relates to the field of forensic technical expertise and relates to a method for determining the timing of handwritten inscriptions on documents. The method includes document analysis using chromatography, stroke mass measurement and Raman spectroscopy of dye samples when exposed to radiation with wavelengths of 532 and 785 nm. The timing of the inscriptions is determined by mathematical processing of the obtained data using the analysis of multidimensional data by the method of principal components.

EFFECT: extending the deadline for determining the time for applying the inscriptions and reducing the likelihood of error.

1 cl, 9 dwg, 3 tbl

Description

The invention relates to the field of forensic technical examinations, namely to methods for determining the range of possible dates for applying handwritten and stamped details on paper documents.

A known system for checking documents using a Raman spectrometer, hereinafter referred to as Raman, which contains a chamber for sample exposure, a laser source, detectors and a database with means for comparing the spectral characteristics of samples (US 6008888, G01J 3/44; G01N 21/65 , publ. 1999.12.28).

The disadvantage of using a single method of Raman spectroscopy is the fluorescence signal, which interferes with obtaining a high-quality spectrum. Fluorescence occurs due to irradiation of a sample consisting of a mixture of organic compounds. This phenomenon is caused by the creation of its own broadband radiation, which can be much more intense than Raman scattering.

There is a known hardware-information complex for dating written documents, which uses for the examination of documents the analysis of Raman spectra of samples of handwritten strokes based on a database for comparing the spectral characteristics of samples and a means for comparing them, as well as the analysis of the spectral properties of paper and ink materials, in which differential thermal analysis of samples of paper documents and thin-layer chromatography of extracted samples of inscriptions and seals (RU 107586, G01J 3/00, publ. 08/20/2011). The disadvantage of this method is the use of thin layer chromatography, during which volatile ink components are extracted along with non-volatile ones, which distorts the results of determining the timing of inscription application, and also creates an additional background signal, which, in some cases, increases the error and reduces the sensitivity of the determination.

Known patent RU2549548, G01J 3/44, publ. 04/27/2015. A method for determining the timing of handwritten inscriptions on documents using chromatography and Raman spectroscopy of dye samples when exposed to radiation with a wavelength of 785 or 532 nm. With this method, writing compositions are divided into two types A with characteristic intensity maxima of 1264 cm ^-1 and 1334 cm ^-1 and B with characteristic intensity maxima of 730 cm ^-1 and 1580 cm ^-1 , then the optimal irradiation power for writing compositions of type A is selected, equal to 10 mW at an irradiation wavelength of 785 nm, and the optimal irradiation power for type B writing compositions at an irradiation wavelength of 532 nm, equal to 2 mW; then the time dependences of the intensity ratios of the characteristic peaks for writing compositions of types A and B are determined; then a gas chromatographic analysis is carried out for the presence of 2-ethylhexanol, 2-phenoxyethanol and benzyl alcohol in the writing composition; then the ratio of the intensities of the characteristic peaks for the sample under study is determined and, based on the data of the gas chromatographic study, the actual production time of the inscription is determined from the calibration curve of the Raman spectra. The disadvantage of this method is the lack of reproducibility of the results for different types (brands) of writing compositions. Two document details that are close in production time may fall on significantly different calibration curves, which are not always technically possible to establish due to the lack of reference samples for comparison.

Known patent RU2604142, G01N 33/32, publ. 12/10/2016. A method for determining how old a document was drawn up based on letter materials, which consists in copying fragments of strokes onto PVC film, using camphor alcohol and dimethylformamide (DMF) as reagents, determining the X, Y and Z times for the object under study and for sample fragments strokes for the tested and further period of time, where X is the contact time during which no copying on camphor alcohol occurs, Y is the copying time on DMF, at which the copied fragments have an average color, and Z is the copying time on DMF, at which the copied fragments have intense coloring; if the values of X, Y and Z for strokes in the document under study coincide with the values of X, Y and Z for samples for the period of time being checked, provided that the time for copying samples over a further period of time is reduced, it is concluded that the time for executing strokes in the document being examined will correspond to the time period being checked. The disadvantage of the method is the use of organoleptic comparison methods and the associated subjectivity and inaccuracy of the analysis. In addition, the method involves working with toxic reagents and does not meet occupational safety requirements.

There is a known method for determining the time of drawing up a document by the strokes of a ballpoint pen or imprints of stamp ink (set forth in patent RU No. 2424502 dated July 20, 2011, and a method for determining the age of strokes of coloring substances of document details by the dynamics of their fading (set forth in patent RU No. 2533315 dated 20.11 .2014, A method for determining the time of drawing up a document from ballpoint pen strokes or stamp ink impressions by extracting a cutout of the sample being determined in a solvent, characterized in that cutouts of the test and control strokes are used, the date of which is known; the complete color difference is determined in accordance with the ISO 7724 standard -1:1984 of the test and control strokes with a multiplicity of at least two times; points determined by the numerical values of the color difference between samples obtained from the first and repeated control strokes are positioned on a semi-logarithmic coordinate grid; points of the total color difference of the stroke being determined are positioned on the same coordinate grid sample on the zero and repeat days of the study; Taking into account the known date of execution of the control stroke on the coordinate plane, the date of execution of the test stroke is determined.

A method for determining the age of strokes of coloring substances of document details, which consists in selecting a coloring substance for a control stroke and creating a control stroke, obtaining samples of an experimental stroke and a control stroke, the date of which is known, extracting the samples in a solvent and determining the complete color difference and difference in lightness for the experimental and control strokes in relation to the solvent, after at least one period of time, repeat at least once obtaining samples of the experimental and control strokes, their extraction in the solvent and determination of the complete color difference and the difference in lightness of the experimental and control strokes and determine the relative age of application of the test stroke, characterized in that samples of the experimental and control strokes are obtained by obtaining impressions of the experimental and control strokes on pieces of adsorbent material wetted with a solvent, while cellulose acetate is used as an adsorbent material, and methanol is used as a solvent .

Disadvantages of the methods: unreliability of the results, confirmed by numerous researchers. Reasonable doubts are raised by the fact that when performing the examination, they use a methodology in accordance with the GOST R. 52489-2005 standard (ISO 7724-1:1984). This standard applies to paints and varnishes and establishes “an instrumental method for determining the color characteristics and color differences of pigments and pigmented paints and coatings.” The study of paint coatings in accordance with ISO 7724-1:1984 requires the use of a different type of spectrophotometer, designed to determine the characteristics of complex composite materials. According to the requirements of this standard, in order to correctly measure the color characteristics of a coating, the size of the paint coating sample under study must be at least 10 by 10 mm, which is significantly larger than the area of the usual document details. As a result, numerous publications cast reasonable doubt on this method.

There is a known method for determining how long ago handwritten texts and other writing materials were completed (RF patent 2296315, G01N 21/31, published 03/27/2007). A method for determining the age of handwritten texts and other writing materials, which consists in obtaining two samples of the writing material under study, one of which is extracted with a solvent, the other is first heated, and then extracted under the same conditions with the same solvent as the first sample, then a quantitative comparison of the extraction results is made, and before the quantitative comparison, both samples are subjected to spectrophotometry, characterized in that the ratio of the optical densities in the region of the maxima of the dye peaks per unit stroke length of each sample is obtained, and the ratio of the optical densities in the region from 400 to 1100 nm is calculated , and the determination of the age of the document under study is carried out by comparing these values with similar parameters of documents with a known execution date, presented in graphical or tabular form.

The disadvantage of the method is the practical impossibility of creating a collection of comparison samples (similar strokes) with a known date of completion, since today there are several thousand different writing compositions on the market and their number increases by dozens every month, and not all of them are available for purchase or other means of replenishment full-scale collection of an expert. In addition, the daily application of several hundred or thousands of strokes on paper sheets and the storage of these sheets requires expensive infrastructure in the form of specialized premises and staff dedicated to storing and recording samples. Finding the right prototype in such a database can take years.

The invention is based on the task of creating a method for determining the timing of handwritten inscriptions on documents using chromatography, weight analysis, Raman spectroscopy of dye samples when exposed to radiation with a wavelength of 785 or 532 nm, and a chemometric method for mathematical processing of spectroscopic data using the principal component method (PCA) .

The proposed method achieves an extension of the deadline for determining the date of execution of the document details on a paper basis, as well as a reduction in the likelihood of error due to the simultaneous use of several physical and chemical methods and a modern mathematical apparatus for processing the spectroscopic signal.

The solution to the stated technical problem is ensured by the fact that in the proposed method for determining the timing of applying handwritten inscriptions on documents using chromatography, weight analysis, Raman spectroscopy of dye samples when exposed to radiation with a wavelength of 785 or 532 nm, initially micro-cuts of the props and a section of paper free from the details, determine the mass of the clipping with a fragment of the studied props by weighing on high-precision scales, chromatograph the clipping, determine the parameter M by weighing the clipping of the props stroke before taking measurements on a gas chromatography-mass spectrometer (m ₁ ) and after taking measurements (m ₂ ), after which M is calculated using the formula: M = m ₁ - m ₂ , Raman spectra are measured using lasers with wavelengths of 532, 785 nm and a fiber optic sensor for recording sample spectra in the external cell compartment, while the spectral range of spectra registration Raman: 80-4500 cm ^-1 , wavelengths of exciting lasers are 532 nm, 785 nm, for them, respectively, a maximum power of 20 mW, 100 mW is used, and they also use a highly sensitive CCD detector with Pelte cooling to -70 ° C and two diffraction gratings - 400 lines/ mm and 1200 strokes/mm; after measurement, the resulting spectrum in the range 280-1735 cm ^-1 is divided into 7 spectral ranges: 280-600 cm ^-1 , 600-890 cm ^-1 , 850-1040 cm ^-1 , 1040-1240 cm ^-1 , 1240-1500 cm ^-1 , 1500-1670 cm ^-1 , 1670-1735 cm ^-1 then to identify the relationship between the components of the spectral data matrices and the recency of the props, the calculation of the Pearson correlation coefficient is used using the formula:

Where, – sample meansx ^m Andy ^m ,, – sample variancesr [-1,1].

| r _xy | = 1 => x, y are linearly dependent,

r _xy = 0 => x, y are linearly independent;

then the correlation coefficients of each of the six spectral ranges are determined, for a preliminary assessment of the shelf life of the props, the components with the highest correlation values are used, then the shelf life of the sample is calculated by exponentially approximating the dependence of the values of the components with the highest correlation coefficients on the period plotted along the X axis, for which Find the values of the time parameter corresponding to the calculated value of the component on the X axis; in the event that one value of a component corresponds to 2 or more values on the X axis, the selection of the correct value is made based on a comparison of the obtained result with the results of gas chromatographic and gravimetric determination of the age of the document, namely if the chromatogram contains significant peaks of volatile components exceeding 2 or more times background “noise”, then the desired period lies in the range of 0-1 year from the moment of measurement; if the peaks of volatile components are 1.5 – 1.9 times higher than the background “noise”, then the desired period lies in the range of 1-2 years from the moment of measurement, if the peaks of volatile components are comparable to “noise” - the document is 2 or more years old; if the parameter M exceeds the value of 0.04 mg, then the desired period lies in the range of 0-1 year from the moment of measurement, if the parameter M lies in the range of 0.1-0.4 mg, then the desired period lies in the interval 1-2, 5 years from the date of measurement, if the parameter M is in the range of 0-0.1 mg, then the age of the document exceeds 2.5 years; then 3D graphs of the dependence of various components with the highest correlation coefficients PC3 and PC5 from each other are constructed to separate the components into clusters depending on the period, for which the calculated value of the components PC3 and PC5 is placed on a three-dimensional graph, where the age of the document is plotted along one of the axes months, and if the calculated point falls into the area of spatial overlap of two or more clusters, further calculations are carried out by searching for time markers; for this purpose, histograms of the distribution of the values of components PC3 and PC5 with the highest value of the correlation coefficient modulus are constructed for all storage periods, and the percentage of positive and negative values of the correlation coefficients is used to determine the storage period of the sample, using for this purpose calibration data on the magnitude of the movement of the maximum of the distribution from the period of production of the props from the positive to the negative area, where each position of the maximum is associated with periods of production of the document; then they plot the relationship between the components with the highest correlation coefficients PC3\RS5 and select 5 ranges of values, grouped according to the characteristic corresponding to the different shelf life of the samples, and then determine the time interval for the production of the document based on determining the time boundaries of the area in these coordinates in which the result fell computing components PC3 and PC5.

The invention is illustrated in Figs. 1 - 9. Fig. 1 shows the selected ranges of the spectrum of the writing composition used for calculations using the X-ray diffraction method. Figure 2 shows graphs of the contributions of Raman shifts to the first 6 components and the spectrum of the sample. Figure 3 shows the results of the analysis of sections of the spectra of writing compositions using the X-ray diffraction method. Figure 4 shows a diagram for selected ranges in the region from 600 to 1034 cm ^-1 and a table for contributions at 724 and 735 cm ^-1 and for contribution ratios. The contributions of the components with the highest correlation coefficient with the term are presented. Figure 5 shows a diagram for the selected ranges in the region from 1416 to 1670 cm ^-1 and a table for contributions at 1571 and 1583 cm ^-1 and for contribution ratios. The contributions of the components with the highest correlation coefficient with the term are presented. Figure 6 shows the dependences of the components with the highest correlation coefficient on the sample storage period. Figure 7 shows a 3D graph in coordinates PC4/PC5/PC6 for the system under study, the points are colored in different colors depending on the shelf life of the sample. Figure 8 shows histograms of the distribution of PC3 values for the system in the range 600-890 cm ^-1 : A) 20-50 months, B) 55-90 months; PC5 values for the 1416-1601 cm ^-1 system: C) 20-50 months, D) 55-90 months. Figure 9 shows a graph of the dependence of the values of PC5 on the values of PC3.

The method is carried out as follows.

Micro-cuts of props and a section of paper free of props, each about 1 cm long, are weighed on high-precision scales, then introduced into a multi-stage pyrolyzer model EGA/PY-3030D. Keeping the cuttings in the pyrolyzer – 3 minutes. Analysis of the clippings is carried out on a gas chromatography-mass spectrometer GCMS-QP2010 Ultra/GCMS-QP2010 SE.

The conditions for obtaining chromatograms are given in Table 1.

Table 1 – Conditions for obtaining chromatograms

No. Element Options 1 Pyrolyzer Temperature – 200 °C, 3 min 2 Column thermostat Temperature: 1. 60 °C, 3 min
2. heating to 180 °C at a rate of 10 °C/min
3. heating to 310 °C at a rate of 20 °C/min
4. 310 °C, 3.5 min 3 Gas chamber Carrier gas is helium;
Pressure 74.5 Kpa
Total flow 15.5 ml/min
Column flow 1.23 mL/min
Linear speed 40.4 cm/sec

Identification of volatile components included in the composition of the studied components is carried out by comparison according to retention parameters with individual (pure) components (2-phenoxyethanol, 1-phenoxypropan-2-ol, 2-(2-phenoxyethoxy)ethanol, glycerin, benzyl alcohol, di -, tri-, tetra-, pentaethylene glycol, phthalic anhydride, propylene glycol, hexylene glycol, etc.). The samples are then reweighed on a high-precision balance.

The area (S) of the corresponding peaks in the chromatogram of a cut from the streak is taken as a characteristic of the total content of the volatile component in the streak of the studied props. The value of the difference in mass of the cutting before obtaining the chromatogram and after the measurement is taken as a characteristic of the amount of solvents in the streak (M). Weighing is performed on Sartoris cpa2p Microbalance scales.

Parameter M is determined by weighing a cut of a props stroke before taking measurements on a gas chromatography-mass spectrometer (m ₁ ) and after taking measurements (m ₂ ), after which M is calculated using the formula: M = m ₁ - m ₂ .

Raman spectra are then measured using a Bruker Senterra rapid Raman spectrometer using an Olimpus optical sample imaging system. Measurements are made using lasers with wavelengths of 488, 532, 785 nm and a fiber optic sensor to record sample spectra in an external cuvette compartment. Spectral range for recording Raman spectra: 80-4500cm ^-1 . Wavelengths of exciting lasers (the maximum power used is indicated in parentheses): 532 nm (20 mW), 785 nm (100 mW), 488 nm (20 mW). A highly sensitive CCD detector with Pelte cooling to -70°C is used. and two diffraction gratings - 400 lines/mm and 1200 lines/mm.

After measurement, the resulting spectrum is in the range 280-1735 cm^-1 divided into 7 spectral ranges: 280-600 cm^-1, 600-890 cm^-1, 850-1040 cm^-1, 1040-1240 cm^-1, 1240-1500 cm^-1, 1500-1670 cm^-1., 1670-1735 cm^-1 The result of division into spectral ranges is illustrated in Fig. 1.

Analysis of selected spectral ranges using the PCA method is implemented in the R Studio program in the R programming language. To identify the relationship between the components of the spectral data matrices and the recency of the props, the calculation of the Pearson correlation coefficient is used using the formula:

Where, – sample meansx ^m Andy ^m ,, – sample variancesr [-1,1].

| r _xy | = 1 => x, y are linearly dependent,

r _xy = 0 => x, y are linearly independent.

After calculating the value of the correlation coefficient, the statistical significance of the resulting coefficient is assessed by testing the hypothesis that it is equal to zero.

Peaks in the region of 1500-1650 cm ^-1 contribute to all components. The main contribution to the first component is made by the baseline (or luminescent background), to the second - peaks at 915 cm ^-1 (0.642), 1620 cm ^-1 (0.575), 1371 cm ^-1 (0.284) and 800 cm ^-1 (0.235) , responsible for C-Ccenter-C and C-C vibrations in the ring, smaller contributions from peaks at 1586 cm ^-1 (0.168), 1178 cm ^-1 (0.131) and 442 cm ^-1 (0.123) (also responsible for C- vibrations C in the ring, C-Ccenter-C and C-H in the ring). The third component is predominantly contributed by the baseline (or luminescent background). The second and third components actually have no connection with the shelf life, therefore, the corresponding connections do not participate in the processes of temporary degradation of writing compositions. The fourth component has the highest correlation coefficient with a term of 0.286; the main contribution to this component is made by the features of the luminescent background, and the following series of peaks can be distinguished: 726 cm ^-1 (0.169), 915 cm ^-1 (0.136), 1344 cm ^-1 (0.116), 800 cm ^-1 (0.110), 1620 cm ^-1 (0.094), 1204 cm ^-1 (0.089), The largest contribution is made by the peak responsible for C-N vibrations (726 cm ^-1 ), probably this peak has the closest relationship with the shelf life of the sample. The fifth component is contributed by peaks at 1620 cm ^-1 (0.395), 915 cm ^-1 (0.325), 1175 cm ^-1 (0.270), 1570 cm ^-1 (0.150), 443 cm ^-1 (0.142) which are responsible for the most stable in binding time, as well as peaks at 683 cm ^-1 (0.137), 657 cm ^-1 (0.125), 943 cm ^-1 (0.105), 598 cm ^-1 (0.084) and 424 cm ^-1 (0.084) which contribute both C-Ccenter-C vibrations and C-NC vibrations. The contribution of vibrations at 730 cm ^-1 is 0.070. The main contribution to the sixth component comes from vibrations at 1615 cm ^-1 (1.117), 919 cm ^-1 (0.978), as well as a series of vibrations at 1591 (0.258), 1580 (0.212), 446 (0.181), 798 (0.135), 1307 (0.130), 1394 (0.129), 1184 (0.124). The contribution of vibrations at 735 cm ^-1 is 0.103, and at 723 cm ^-1 0.078.

A shift in the position of the contributions relative to the position of the peak maximum in the Raman spectrum of the sample may indicate the composite nature of the peak, and the identification of the contributions of certain specific vibrations. The contributions of Raman shifts to the SAR components are illustrated in Fig. 2.

The largest absolute values of the Pearson correlation coefficients were obtained for the ranges 600 – 890 cm ^-1 and 1500 – 1670 cm ^-1 : -0.484 and -0.333, respectively. The above ranges include the peaks used to build the degradation model of writing compositions - 729 and 1580 cm ^-1 . Similar calculations were carried out for narrower sections of the spectrum, including peaks at 729 and 1580 cm ^-1 to select the optimal range (with the maximum value of the correlation coefficient) (Figure 3).

Based on the calculation results, it was established that the largest modulus of the correlation coefficient was found for section No. 2 in the range of 600-890 cm ^-1 , the Pearson correlation coefficient was -0.484. The peak in the region of 729 cm ^-1 when analyzed by XRD is decomposed into two peaks: at 724 cm ^-1 and at 735 cm ^-1 . In this case, there is an increase in the modulus of the correlation coefficient with an increase in the contribution ratio 724/735 (Figure 4). In the ranges 2B 692-1034 cm ^-1 and 2G 600-1034 cm ^-1 the maximum correlation coefficient with the period corresponds to PC5. In other ranges, the maximum correlation coefficient corresponds to PC3. At the same time, in the ranges 2B and 2G (in Fig. 4 there are no contributions at 735 cm ^-1 , the large value of the correlation coefficient can be explained by the presence of a contribution at 943 cm ^-1 (Fig. 4), since this peak in the Raman spectrum of writing compositions is the contribution of fluctuations in CN bonds, which are most susceptible to temporal changes.

The peak in the region of 1580 cm ^-1 when analyzed by XRD is decomposed into two peaks: at 1571 cm ^-1 and at 1583 cm ^-1 . In this case, one can note an increase in the modulus of the correlation coefficient with an increase in the ratio of contributions 1571/1583 (Fig. 5).

Based on the results of calculations for narrower intervals in the region from 1416 to 1670 cm-1, it was established that the absolute value of the Pearson correlation coefficient for section No. 6B in the range 1416-1601 cm ^-1 is 0.348.

For a preliminary assessment of the statute of limitations of the details, the components with the highest correlation values are used.

After carrying out the basic calculations to solve the assigned problems of determining the dating of handwritten details, the following is done:

1. Chromatography and weighing of a micro-cut from document paper containing a fragment of the studied props

2. plotting the dependence of one of the components on the period of application of the writing composition to the paper,

3. division into clusters depending on the sample storage period.

4. identifying time markers that make it possible to make a rough estimate of the timing of applying the writing composition.

5. plotting the dependence of the components with the highest values of the correlation coefficient with the shelf life in PC3/RS5 coordinates

6. Weighing microcuts on high-precision balances before and after chromatography

The results obtained can be effectively used in forensic technical examination of documents to determine the timing of applying details to the document.

Results of calculations of the production time of details

The document is illustrated with examples.

Example 1

Analysis of the results obtained from chromatographic and gravimetric examination of documents

is done as follows:

In the chromatograms of paper clippings in areas of the document free from details, no peaks of the characteristic components of props 1 were detected (Fig. 6). In the chromatogram of a clipping from detail 1 of document 1 during the initial study, there is a peak of the characteristic component, namely 2 - phenoxyethanol (retention time 11.1 ± 0.1 min), which is used for further calculations (Table 2.).

Table 2. – Results obtained from the study of details.

Document being examined Researchable props Peak area in chromatograms of line clippings (S), mA*s Peak height in chromatograms of line clippings (h), mA Primary Research 2-phenoxyethanol Document 1 Props 1 0.225 0.052 Repeated study 2-phenoxyethanol Document 1 Props 1 - -

Practical research carried out at the Russian Federal Center for Economic Examination under the Ministry of Justice of Russia [“Theory and practice of forensic examination”, scientific and practical journal, No. 2 (30) 2013. E.A. Trosman, G.S. Bezhanishvili, N.A. Batygina, N.M. Arkhangelskaya, R.A. Yurova. Methodology “Determining how long ago details were made in documents based on the relative content of volatile solvents in strokes”, pp. 80-88], show that in strokes made with pastes of ballpoint and gel pens of different compositions and having an absolute old age of over 24 months, volatile components as a rule, are no longer contained, or are contained in “trace” quantities (slightly exceeding the “background peaks” of the paper).

Based on the results of measuring chromatograms and preliminary and subsequent weighing during the initial and repeated examination for the studied details, it can be concluded that the residual amount of volatile components in the writing compositions indicates that handwritten details 1 were completed within a period of no more than 2.5 years from the date of examination. According to the results of weighing during the primary study: the M parameter for clippings from document 1 averages 0.024 mg; with repeated testing - 0.009 mg. According to the results of weighing during the primary study: the M parameter for clippings from document 1 averages 0.024 mg; with repeated testing - 0.009 mg. This indicates that at the start of the analysis, a significant amount of volatile components was present in the test streak, which decreased during the analysis due to a long shelf life of 1 month). It has been experimentally established that if the chromatogram contains significant peaks of volatile components that are 2 or more times higher than the background “noise”, then the desired period lies in the range of 0-1 year from the moment of measurement, if the peaks of volatile components are 1.5 - 1. 9 times higher than the background “noise”, then the required period lies in the interval of 1-2 years from the moment of measurement, if the peaks of volatile components are comparable to the “noise” - the document is 2 or more years old; if the parameter M exceeds the value of 0.04 mg, then the desired period lies in the range of 0-1 year from the moment of measurement, if the parameter M lies in the range of 0.1-0.4 mg, then the desired period lies in the interval 1-2, 5 years from the date of measurement, if the parameter M is in the range of 0-0.1 mg, then the age of the document exceeds 2.5 years. Similar data obtained by methods of chromatography, gas chromatography, IR spectroscopy are contained in the literature, for example [-Juanuan Xu, Jinghan Wang, Lijuan Yao. Dating the writing age of black roller and gel inks by gas chromatography and UV-vis spectrometer. Forensic science international, 162 (2006), 140-143; -Agnes Koenig, Celine Weyermann, Ink Dating, part 1: Statistical distribution of selected aging parameters in a ballpoint ink reference population. Science and Justice, 58 (2018), 17-30. ]

To verify the results of the study using GCMS and gravimetric analysis, spectral studies are carried out, described in example 2.

The calculation results are shown in Table 2.5.3.

Table 2.5.3 – Results obtained from the study of details.

Document being examined Researchable props Weight of the cutting BEFORE obtaining the chromatogram (m ₁ ), mg Weight of the cutting AFTER obtaining the chromatogram (m ₂ ), mg Change in mass (M), mg Primary Research Document 1 Paper 0.772 0.748 0.024 Props 1 0.855 0.830 0.025 Repeated study Document 1 Paper 0.878 0.868 0.010 Props 1 0.321 0.314 0.007

Results of document examination using GCMS method

Document 1 was produced within a period not exceeding 30 months from the date of the examination.

Example 2.

Spectral measurements of dye temporal degradation.

Raman spectra were measured, in accordance with the procedure, using a Senterra express Raman spectrometer from Bruker using an Olimpus optical sample imaging system. Measurements are made using lasers with wavelengths of 488, 532, 785 nm and a fiber optic sensor to record sample spectra in an external cuvette compartment. Spectral range for recording Raman spectra: 80-4500cm ^-1 .

Based on the measurement results, the correlation coefficients of the components in coordinates PC4, PC5 and PC6 versus time (months) were calculated, their correlation coefficients with storage time were calculated using the formulas, and the Pearson correlation coefficient was calculated using the formula:

Where, – sample meansx ^m Andy ^m ,, – sample variancesr [-1,1].

| r _xy | = 1 => x, y are linearly dependent,

r _xy = 0 => x, y are linearly independent.

Calculation of the shelf life of a sample is carried out by exponentially approximating the dependence of the values of the fourth and fifth components on the period plotted along the X-axis. Figure 6 shows graphs of the dependence of the components with the highest correlation coefficients PC4, PC5 and PC6 on the shelf life of the sample. The timing is calculated by finding on the X-axis the values of the time parameter corresponding to the calculated value of the component. If one component value corresponds to 2 or more values on the X axis, the selection of the correct value is made based on studying the materials of the Case and comparing the obtained result with the results of gas chromatographic and weight determination of the document’s age, described in example 1.

To verify the research results obtained by GCMS, gravimetric and spectral analysis methods, calculations are carried out using X-ray diffraction analysis methods described in examples 1 and 2.

Example 3.

XRD analysis of document age.

To divide components into clusters depending on the period, 3D graphs of the dependence of various components on each other are constructed. In this case, it is possible to achieve partial separation, which has a “diffuse” appearance. Figure 7 presents an example of such an attempt to cluster samples based on the relationship PC5-PC6-T (time in months). The calculated value of components 5 and 6 is placed on a three-dimensional graph, where time is plotted along one of the axes (the age of the document in months). The results obtained are verified by the results obtained using GCMS, gravimetric analysis and spectral studies. If the calculated point falls into the area of spatial overlap of two or more clusters, further calculation is performed, described in example 4, by searching for time markers.

Example 4: Identifying time markers.

To identify time markers, for example, the ranges 600-890 cm ^-1 and 1416-1601 cm ^-1 are selected. For the selected systems, histograms of the distribution of the value of the components with the highest value of the module of the correlation coefficient are constructed for all storage periods (Fig. 8).

From Fig. 8 it is clear that after fifty months of storage of the samples, the PC3 values shift to the negative region: when stored for up to 50 months, 67% of the PC3 values are positive, after 50 months of storage their statistical weight decreases to 13%. And for PC5 values, on the contrary, after fifty months of sample storage, the values of the 5th component shift from negative to positive: up to 50 months of sample storage, 43% of PC5 values are positive, after 50 months, 76%. The values become positive.

In the process of practical work, researchers may discover other time markers, the use of which is allowed based on the results of calculations using the X-ray diffraction method.

Thus, the percentage of positive and negative values of PC3 and PC5 is used to determine the storage period of the sample (more/less than 50 months). The obtained result is verified by the results obtained using GCMS methods, gravimetric analysis, spectral studies and calculations using X-ray diffraction methods. If the result obtained using the method of searching for time markers does not allow determining the age of the document with the required accuracy, then further calculations are carried out, described in example 5.

Example 5. Two-dimensional clustering of calculation results.

Illustrated in FIG. 9.

On the dependence graph in PC3\RS5 coordinates, 5 areas are distinguished, corresponding to different storage periods of samples: 20-25, 30-50, 55-70, 80-85 and 90 months. The determination of the time interval for document production is made based on the determination of the time boundaries of the area in coordinates PC3\RS5, shown in Fig. 9, which contains the result of the calculations.

Thus, using the method of principal component analysis, a more accurate clustering of the studied samples by shelf life is achieved. The developed clustering method is the most objective, since its implementation uses all points in the Raman spectra, and the error due to the “human factor” is minimized. All calculations are carried out by the program using a specially generated script.

A method for determining the timing of handwritten inscriptions on documents using chromatography and Raman spectroscopy of dye samples when exposed to radiation with a wavelength of 785 or 532 nm, in the presence of a database of strokes of writing compositions of pens, with different periods of application on paper, characterized in that it is first produced weighing of the micro-clipping on high-precision Sartorius cpa2p Microbalance scales, then chromatography of the micro-clipping is performed. To do this, micro-cuts of the props and a section of paper free of props, each about 1 cm in length, are introduced into a multi-stage pyrolyzer model EGA/PY-3030D. Keeping the cuttings in the pyrolyzer – 3 minutes. Analysis of the clippings is carried out on a gas chromatography-mass spectrometer GCMS-QP2010 Ultra/GCMS-QP2010 SE. At the end of chromatography, the microcut is re-weighed on the same scales. The area (S) of the corresponding peaks in the chromatogram of a cut from the streak is taken as a characteristic of the total content of the volatile component in the streak of the studied props. The value of the difference in mass of the cutting before obtaining the chromatogram and after the measurement is taken as a characteristic of the amount of solvents in the streak (M). Weighing is performed on Sartorius cpa2p Microbalance scales.

The M parameter is determined by weighing a cut of a props stroke before taking measurements on a gas chromatography-mass spectrometer (m1) and after taking measurements (m2), after which M is calculated using the formula: M = m1- m2.

Raman spectra are then measured using a Bruker Senterra rapid Raman spectrometer using an Olimpus optical sample imaging system. Measurements are made using lasers with wavelengths of 488, 532, 785 nm and a fiber optic sensor to record sample spectra in an external cuvette compartment. Spectral range for recording Raman spectra: 80-4500 cm-1. Wavelengths of exciting lasers (the maximum power used is indicated in parentheses): 532nm (20mW), 785nm (100mW). A highly sensitive CCD detector with Pelte cooling to -70°C is used. and two diffraction gratings - 400 lines/mm and 1200 lines/mm.

After measurement, the resulting spectrum in the range: 280-1735 cm ^-1 is divided into 7 spectral ranges, 280-600 cm ^-1 , 600-890 cm ^-1 , 850-1040 cm ^-1 , 1040-1240 cm ^-1 , 1240-1500 cm ^-1 , 1500-1670 cm ^-1 , 1670-1730 cm ^-1 . To identify the relationship between the components of the spectral data matrices and the recency of the props, the calculation of the Pearson correlation coefficient is used.

After carrying out the basic calculations, in order to solve the assigned problems of determining the dating of handwritten details, the following is carried out: constructing the dependence of one of the components on the period of application of the writing composition to the paper, dividing into clusters depending on the shelf life of the sample, identifying time markers that allow making a rough estimate of the timing of applying the writing composition composition, plotting the dependence of the components with the highest values of the correlation coefficient with the shelf life in PC3/RS5 coordinates.

The results of determining the possible timing of document production are produced at each stage of measurements and calculations. The results obtained are compared with each other and thus verified.

The results obtained can be effectively used in forensic technical examination of documents to determine the timing of applying details to the document.

Claims (6)

A method for determining the timing of handwritten inscriptions on documents using chromatography, weight analysis, Raman spectroscopy of dye samples when exposed to radiation with a wavelength of 785 or 532 nm, characterized in that initially micro-cuts of the props and a section of paper free from details are made, the mass of the cut is determined with a fragment of the props under study by weighing on high-precision scales, the clipping is chromatographed, the parameter M is determined by weighing the clipping of the props stroke before taking measurements on a gas chromatography-mass spectrometer (m ₁ ) and after taking measurements (m ₂ ), after which M is calculated using the formula: M = m ₁ - m ₂ , Raman spectra are measured using lasers with wavelengths of 532, 785 nm and a fiber optic sensor for recording sample spectra in the external cuvette compartment, while the spectral range for recording Raman spectra: 80-4500 cm ^{- 1} , the wavelengths of the exciting lasers are 532 nm, 785 nm, for them, respectively, maximum powers of 20 mW, 100 mW are used, and they also use a highly sensitive CCD detector with Peltier cooling to -70°C and two diffraction gratings - 400 lines/mm and 1200 strokes/mm; after measurement, the resulting spectrum in the range of 280-1735 cm ^-1 is divided into 7 spectral ranges: 280-600 cm ^-1 , 600-890 cm ^-1 , 850-1040 cm ^-1 , 1040-1240 cm ^-1 , 1240-1500 cm ^-1 , 1500-1670 cm ^-1 , 1670 – 1735 cm ^-1 , then to identify the relationship between the components of the spectral data matrices and the duration of the props, the Pearson correlation coefficient is calculated using the formula: Where – sample means x ^m and y ^m , – sample variances | r _xy | = 1 => x, y are linearly dependent, r _xy = 0 =>x, y linearly independent; then the correlation coefficients of each of the seven spectral ranges are determined, for a preliminary assessment of the shelf life of the props, the components with the highest correlation values are used, then the shelf life of the sample is calculated by exponentially approximating the dependence of the values of the components with the highest correlation coefficients on the period plotted along the X axis, for which on the X axis the values of the time parameter corresponding to the calculated value of the component are found; in the event that one value of a component corresponds to 2 or more values on the X axis, the selection of the correct value is made based on a comparison of the obtained result with the results of gas chromatographic and gravimetric determination of the age of the document, namely if the chromatogram contains significant peaks of volatile components exceeding 2 or more times background “noise”, then the desired period lies in the range of 0-1 year from the moment of measurement; if the peaks of volatile components are 1.5–1.9 times higher than the background “noise”, then the desired period lies in the range of 1-2 years from the moment of measurement, if the peaks of volatile components are comparable to “noise” - the document is 2 or more years old; if the parameter M exceeds the value of 0.04 mg, then the desired period lies in the range of 0-1 year from the moment of measurement, if the parameter M lies in the range of 0.1-0.4 mg, then the desired period lies in the interval 1-2, 5 years from the date of measurement, if the parameter M lies in the range of 0-0.1 mg, then the age of the document exceeds 2.5 years; then 3D graphs of the dependence of various components with the highest correlation coefficients PC3 and PC5 from each other are constructed to separate the components into clusters depending on the period, for which the calculated value of the components PC3 and PC5 is placed on a three-dimensional graph, where the age of the document is plotted along one of the axes in months, and if the calculated point falls into the area of spatial overlap of two or more clusters, further calculations are performed by searching for time markers; for this purpose, histograms of the distribution of values of components PC3 and PC5 with the highest value of the correlation coefficient modulus are constructed for all storage periods , and the percentage of positive and negative values of the correlation coefficients is used to determine the storage period of the sample, using for this purpose calibration data on the magnitude of the movement of the maximum of the distribution from the period of production of the props from the positive to the negative area, where each position of the maximum is associated with periods of production of the document; then plot the relationship between the components with the highest correlation coefficients PC3\RS5 and select 5 ranges of values, grouped according to the characteristic corresponding to the different shelf life of the samples, and then determine the time interval for the production of the document based on determining the time boundaries of the area in these coordinates in which the result of the calculations fell component RS3 and RS 5.