Optical object detection in paper improved by refractive index matching and mechanical treatment

J. Saarela1, S. Heikkinen2, T. Fabritius1,3, and R. Myllylä1

1Department of Electrical and Information Engineering and Infotech Oulu, University of Oulu, P.O. Box 4500, 90014 University of Oulu, Finland

2Department of Process and Environmental Engineering, University of Oulu, P.O. Box 4300, 90014 University of Oulu, Finland

3Computational Optics Group, University of Tsukuba, Tennodai, Tsukuba, Ibaraki 305-8573, Japan

ABSTRACT

Two different paper grades were tested with a clearing agent to measure how much mechanical smoothening can improve transparency inside paper. The paper grades were newsprint and supercalendered paper. The paper furnishes of both papers were alike, but the supercalendered paper was mechanically smoothened. Anise oil was used as the clearing agent, but similar measurements were also done with air and water. Black lines 8.5 µm to 281.1 µm wide were placed behind layers of cleared paper and transparency was measured with a microscope. When anise oil was the clearing agent, supercalendering improved transparent paper grammage from 139 g/m2 to 164 g/m2. With water the improvement was from 40 g/m2 to 51 g/m2. With air the improvement was not determinable. As a conclusion, it is recommended that paper is smoothened if it needs to be studied optically. Optical coherence tomography, for example, would benefit from this treatment.

Keywords: Transparent paper grammage, transparency slope, refractive index matching, paper testing, optical clearing, clearing agent, multiple scattering, light transmission, optical testing

1. INTRODUCTION From the optical point of view, the three-dimensional, stochastic network structure of tissue consisting of cellulose fibers can be categorized as a highly scattering anisotropic medium. This makes the structure of paper hard to study optically. Refractive index matching decreases scattering and thus improves transparency.

A review on transparentizing paper was written in 1960 by Vaurio [1]. As Vaurio studied the phenomenon, Radvan et.al. used it to study paper structure [2]. Since those days new interests and techniques have developed. For example, recycled paper has been a focus of interest for only a few years. Alarousu et. al. used refractive index matched paper to improve the visibility of a new technique called optical coherence tomography in 2005 [3]. In 2006 further studies on refractive index matching in paper and optical tomography were introduced: Fabritius and Myllylä investigated swelling behavior [4], Fabritius et. al. characterized optically cleared paper [5] and Kirillin et. al. simulated measurements [6]. A method of defining the refractive index of paper using refractive index matching was presented in 2006 by Fabritius et. al. [7].

1.1 Refractive index matching, i.e. optical clearing

The refractive index mismatch between components within a highly scattering material has a strong effect on its optical properties such as transmittance and reflectance. The optical properties of a turbid medium like paper can be effectively changed by matching the refractive indices of scatterers and the base material. This so-called optical clearing method is well known in dispersive optics of physical systems and it has been used successively in many biomedical applications, including microscopy studies [8]. In this study the optical clearing method is used in paper examination.

The pores and cavities of dry paper are filled with air, and the refractive index mismatch between the pores and the raw material is significant. Thus, many scattering events happen within dry paper, and the paths of separate photons rapidly become chaotic. When refractive index matching liquids are used, the medium of the pores is replaced by a liquid with a refractive index that is larger than that of air (na > 1), thereby changing the scattering properties of the paper. The lower the refractive index mismatch, the shorter the trajectories of the photons. The medium becomes more transparent and

Advanced Laser Technologies 2007, edited by Risto Myllylä, Alexander V. Priezzhev,Matti Kinnunen, Vladimir I. Pustovoy, Mikhail Yu. Kirillin, Alexey P. Popov,

Proc. of SPIE Vol. 7022, 70221A, (2008) · 0277-786X/08/$18 · doi: 10.1117/12.804109

Proc. of SPIE Vol. 7022 70221A-12008 SPIE Digital Library -- Subscriber Archive Copy

1mm

also improves the probing depth of a microscope. The principle of refractive index matching in terms of photon migration is demonstrated in figure 1 and its effect in paper is shown in figure 2.

Fig. 1. The refractive indexes of wood fiber and air differ considerably. This causes a high refraction angle and in the case of

multiple scattering, blurs the image quickly. Therefore, an absorbent object in dry paper cannot be detected, as seen on the left. But, if air is replaced with a clearing agent with a refractive index close to that of wood fiber, the refractive angle is smaller and the object can be detected, as seen on the right.

Fig. 2. A clear difference in transparency can be seen with a sample without a clearing agent on the left and with benzyl

alcohol as a clearing agent on the right. The scale bar is 1 mm long.

Proc. of SPIE Vol. 7022 70221A-2

1.2 Aim of the study

Air and cellulose fibers form a strongly scattering medium. Usually, this is a desired property called high opacity, which enables printing on both sides of a paper without the two prints showing through to the other side. However, to study the structure of paper it would be useful to see through the paper. Refractive index matching can be used for this purpose. Surface roughness is expected to increase opacity. Therefore we studied if the effect of refractive index matching is improved if the paper surface is smooth.

2. MATERIALS AND METHODS 2.1 Paper samples

Two different kinds of paper samples were used: newsprint (News) and supercalendered paper (SC). In rough figures, newsprint is made of recycled fibers in Europe and mechanical pulp in other countries. The raw material for SC paper is mainly mechanical pulp and up to 35 % fillers. The smooth surface of SC paper is produced mechanically. [9]

The main properties of the papers are given in table 1.

Table 1. Initial paper sample properties.

Paper News SC

Grammage (g/m2) 42 51

Dry content (%) 94 96

Dry mass of fillers (%) (= Ash 525°C) 10 30

Dry mass of carbonate (%) 3.0 1.3

Brightness (ISO2470) 58.3 71.6

Opacity C C/2 (ISO 2471) 93.3 88.5

s C C/2 56.95 50.77

k C C/2 5.86 1.93

Table 2. Refractive indexes of the materials used in this work. [7]

Material Wood fiber Air Water Anise oil

Refractive index 1.557 1.000 1.329 1.555

2.2 Clearing agents

In paper and paper furnish the normal media between wood fibers are air and water. In addition to these, anise oil was used as a clearing agent. Table 2 gives the refractive indexes of the ingredients used.

2.3 Measurement setup and measurements

The measurement setup is shown in figure 3. The light applied to the sample was white and came from the bottom. First the light encountered a glass plate with a test pattern. Then the light entered the paper sample. The light was detected with a microscope (Leica MZ FLIII). A CCD camera (Leica DFC320) was attached to the microscope and used to document the images.

The test pattern was etched on glass. It consisted of 18 black lines. The thickest lines were single with widths of 281.11 µm, 200.37 µm, 140.74 µm and 100.00 µm. The thinnest lines were double with widths of 8.52 µm, 12.59 µm, 17.77 µm, 25.19 µm, 35.18 µm, 50.37 µm and 70.37 µm. Using this measurement setup, pictures were taken with different amounts of paper on top. Figure 4 shows diagrammatically how the image blurs. The grey area between black lines

Proc. of SPIE Vol. 7022 70221A-3

grows due to scattering. With one paper all the lines are distinguishable, with three papers only the four thickest lines are distinguishable, with five only the thickest two, and with seven, none are distinguishable.

Image detection with a microscope and

a CCD camera

Glass plate with an test paternetched

Layers of paper

Microscope’s light source

Fig. 3. Measurement setup. The light came from beneath. First the light encountered a glass plate with a test pattern. Then the light entered the paper sample. The image on top of the paper was detected with a microscope and a CCD camera.

Fig. 4. The image blurs as the light propagates in paper. On the left a test pattern is illuminated and then light enters the

sample. On the right is seen what kind of image could be detected at a selected level.

2.4 Determining transparency

Paper is not a uniform material. The grammage, thickness and filler content vary within a sample. Therefore, in borderline cases the border between distinguishable and non-distinguishable lines is not clear. For this reason computer image analysis was not used, but instead a person decided if a line was more distinguishable than not distinguishable. So, if a line was distinguishable over 50 % of its length, it was assumed that in the case of average grammage it would be distinguishable. As an example, in figure 2 without refractive index matching only the two thickest lines were distinguishable, but with refractive index matching all 18 lines could be seen.

3. RESULTS 3.1 Transparency improvement

The graphs in figure 5 to 7 were plotted from the measurement results. They all show that transparency in paper was better with SC paper than with newsprint. Air and water are not good clearing agents, as seen from figures 5 and 6. The paper is opaque, which normally is a desired property. Anise oil has a refractive index close to that of paper, but nevertheless transparency improves if the surface is mechanically smoothened, as seen from figure 7.

Proc. of SPIE Vol. 7022 70221A-4

0

50

100

150

200

250

300

0 50 100 150 200 250

Grammage (g/m2)

Vis

ible

line

( µm

)

News

SC

Fig. 5. Mechanical treatment improves transparency in dry paper.

Fig. 6. Mechanical treatment improves transparency in wet paper.

Proc. of SPIE Vol. 7022 70221A-5

0

50

100

150

200

250

300

0 100 200 300 400 500 600 700

Grammage (g/m2)

Vis

ible

line

( µm

)

News

SC

Fig. 7. Mechanical treatment improves transparency in paper optically cleared with anise oil.

3.2 Transparency slope and transparent paper grammage

In order to determine what size of object can be detected behind a certain paper grammage with a clearing agent, transparency slopes were calculated. As a rough measure, the smaller the slope value, the slower transparency decreases. Three values closest to zero were used to plot a linear regression line. The slope of this line is the transparency slope. Table 3 shows the calculated slopes. The slope doesn’t cross the grammage axis at zero, but at some other value. This value is the grammage when the paper is so thin that it doesn’t hide anything, and it can be called the transparent paper grammage. Table 4 shows the calculated values for the papers and clearing agents used. One method of comparing transparency is to compare the transparency slope and the transparent paper grammage.

Table 3. Transparency slopes for Newsprint and SC paper and clearing agents.

Table 4. Transparent paper grammage for Newsprint and SC paper and clearing agents.

4. DISCUSSION In principle, refractive index matching provides transparency in paper without limitation. In reality, principle limitations are set by multiple scattering, the refractive index mismatch in paper and finally absorption in the paper and clearing agent.

Slope [µm/(g/m2)] Paper media Air Water Anise oil

Paper grade News 2.8 2.9 0.32

SC 1.3 2.5 0.091

Transparent paper grammage [g/m2] Paper media Air Water Anise oil

Paper grade News 21 40 140

SC -5 51 160

Proc. of SPIE Vol. 7022 70221A-6

Of these phenomena, mechanical smoothening of paper affects multiple scattering but does not change the refractive index mismatch or absorption. Three ways that mechanical smoothening affects multiple scattering come to mind. The first is that the paper thickness changes. When the thickness decreases, the distance between scattering events decreases. Since the refractive indexes don’t change, the angles of scattering don’t change either. Therefore, the number of scattering events also decreases. These mechanisms decrease the effect of multiple scattering. The second way is that the fibers break. If part of a fiber remains attached to the fiber, it is fibrillated. If the part is detached, it becomes a so-called fine particle. These mechanisms increase the effect of multiple scattering, since more scattering points mean more scattering events. The third way affects through changes in the mechanical properties of the fibers. Clearing agents dilate fibers. Mechanical treatment affects this ability to dilate. It can affect by both increasing and decreasing scattering. If the ability to dilate increases, the change in paper thickness increases but the distances between scattering events decrease. The ability to dilate can also decrease, whereupon the effect is the opposite. The results indicate that the dominating effect is thinning of paper, since object detection improved.

One source of error is the human eye. In this study only one person made the final decision on whether a line was distinguishable. This minimized the error of different points of view, although in most cases the result was obvious and a reference person saw the same way. In borderline cases the issue was whether a line could be seen 50 % rather than which line was in question. If this method is widely used in the future, a computer would give standard results, but this would require work to standardize the environment, like lighting and focusing, and to standardize the computer’s image analysis.

Another source of uncertainty is the fact that the paper furnishes in newsprint and SC paper are not identical. Typically SC furnish has more fines, which increases multiple scattering, and more chemical pulp, which decreases multiple scattering.

The real value of this work is for persons trying to see optical objects like ink or pitch particles in a test sheet. This article shows that smoothening improves visibility.

5. CONCLUSION To detect an optical object, like an absorber, in paper is challenging. Refractive index matching improves transparent paper grammage for newsprint from 21 g/m2 with air as a media to 40 g/m2 with water and 140 g/m2 with anise oil. Mechanically smoothened newsprint, so-called supercalendered paper, has transparent paper grammage of 51 g/m2 with water and 160 g/m2 with anise oil. The improvement in transparency is also shown by the transparency slopes. Thus, mechanical treatment can improve the transparency of refractive index matched paper.

ACKNOWLEDGEMENTS

This work was supported by the Graduate School in Electronics, Telecommunications and Automation (GETA). Special thanks to Jyrki Lappalainen from the Microelectronics Laboratory.

REFERENCES

1. F. Vaurio, “Transparentizing of Paper,” Tappi J. 43(1), 18-24 (1960). 2. B. Radvan, C. Dodson and C. Skold, “Detection and cause of layered structure of paper,” Consolidation of the paper web, Trans. IIIrd Fund. Res. Symp. Cambridge, ed F. Bolem, 189-215, BPBMA, London (1966). 3. E. Alarousu, L. Krehut, T. Prykäri and R. Myllylä, “Study on the use of optical coherence tomography in measurements of paper properties,” Meas Sci Technol. 16, 1131-1137 (stacks.org/MST/16/1131) (2005). 4. T. Fabritius and R. Myllylä, “Investigation of swelling behaviour in strongly scattering porous media using optical coherence tomography,” J Phys D Appl Phys 39, 2609-2612 (2006). 5. T. Fabritius, E. Alarousu T. Prykäri J. Hast and R. Myllylä, “Characterization of optically cleared paper by optical coherence tomography,” Quantum Electron+ 36(2), 181-187 (2006). 6. M. Kirillin, A. Priezzhev, J. Hast and R. Myllylä, “Monte Carlo simulation of optical clearing of paper in optical coherence tomography,” Quantum Electron+ 36(2), 174-180 (2006). 7. T. Fabritius, J. Saarela and R. Myllylä, “Determination of the refractive index of paper with clearing agents,” Proc.

Proc. of SPIE Vol. 7022 70221A-7

SPIE 6053, 60530X-1 – 60530X-8 (2006). 8. V. Tuchin, “Optical clearing of tissues and blood using the immersion method,” J Phys D Appl Phys 38, 2497-2518 (2005). 9. A. Haarla, “Printing and writing papers”, In: Paper and Board Grades, Papermaking Science and Technology, vol. 18 ed H. Paulapuro, 14-53, Fapet Oy, Jyväskylä, 2000.

Proc. of SPIE Vol. 7022 70221A-8