Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
Proc. R. Soc. A (2009) 465, 3839–3858
doi:10.1098/rspa.2009.0144
Published online 25 September 2009
Using solvents to remove a toner print so that
office paper might be reused
BY THOMAS A. M. COUNSELL
AND
JULIAN M. ALLWOOD*
Department of Engineering, University of Cambridge, Trumpington Street,
Cambridge CB2 1PZ, UK
The environmental impact of office paper recycling might be avoided if toner print could
be removed from paper in a way that left the paper immediately reusable. This article
reports on experiments that investigate the use of solvents to allow black toner print to
be removed from white cut-size office paper. Hansen solubility parameters are estimated
for detaching the pigment in toner print from paper and compared to the parameters for
paper. These imply that solvents can be found that would allow toner print to be removed
without harming the underlying paper. However, immersion in solvents alone detaches
just 10 per cent of the toner pigment. Rubbing combined with solvents increases removal
to 50 per cent. Adding ultrasonic agitation increases removal to 80 per cent. Mixing
different solvents and increasing the volume of solvent can further improve removal to
the point where the cleaned paper is useful, although still distinguishable from new. For
instance, soaking the printed paper in a mixture of 60 per cent dimethylsulphoxide and
40 per cent chloroform for 4 min, while applying ultrasonic agitation, results in paper that
can be readily reprinted. The results need to be validated on other toner and paper types.
Further work is needed to investigate the influence of temperature, of adding surfactants
and to consider the economic, safety and environmental implications.
Keywords: paper recycling; toner print; Hansen solubility parameters
1. Reuse of office paper by removing toner
Medieval monks removed the print from unwanted parchment to allow it to be
reused (Reed 1972). Could the same be done for a sheet of modern office paper?
Counsell & Allwood (2007) estimate that, if it could be done, the climate change
gases emitted over the life of a sheet of paper might be cut by between 45 and
95 per cent.
To reuse a sheet of modern office paper, any tears, holes and folds would
need to be repaired and any unwanted print neutralized. The print might be
a dye, a pigment or a toner. Counsell & Allwood (2006b) surveyed research into
processes that neutralize print, classified according to whether the process tried
*Author for correspondence ([email protected]).
Electronic supplementary material is available at http://dx.doi.org/10.1098/rspa.2009.0144 or via
http://rspa.royalsocietypublishing.org.
Received 19 March 2009
Accepted 25 August 2009
3839
This journal is © 2009 The Royal Society
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3840
T. A. M. Counsell and J. M. Allwood
to obscure, de-colour or remove the print and whether removal processes used
adhesion, ablation, abrasion or solvents. Almost the only evidence of previous
research is patent filings and these provide little data on whether and how the
techniques work.
This article reports on a small part of the overall problem of office paper
reuse: it explores one kind of solution (removal), of one print type (toner) using
one approach (solvents). It is part of a wider programme of research into toner
removal outlined earlier by Counsell & Allwood (2006a) that includes toner
print removal by abrasion (Counsell & Allwood 2009) and by laser ablation
(Counsell 2007).
Seventy to ninety per cent of a sheet of office paper is typically fibrous bundles
of cellulose, a type of glucose polymer. These bundles are usually flattened
cylinders, 10–50 μm in width and 1–5 mm in length, stacked randomly in up to 10
layers across the thickness of the paper. The remainder of a sheet of office paper is
usually a combination of fillers (often clays), whiteners (often titanium dioxide),
traces of lignin and additives that adjust the strength and water absorption
properties of the paper. Toner is the type of print used by most photocopiers
and laser printers, but not by most pens or ink-jet printers. It consists of 5–
25 μm diameter blobs of polymer embedded with coloured pigments. The exact
composition of a toner varies by manufacturer and is usually a trade secret. They
are usually classified according to the type of polymer (an acrylic or a polyester)
and whether they include a magnetic component (typically iron ferrite). During
printing, the resin is given an electrostatic charge to allow it to be positioned
on the paper surface. Once positioned, the toner is usually fixed to the paper
by a combination of heat and pressure that cause the polymer to soften, sinter
and wet against the surface paper fibres. After printing, the polymer remains
attached to the paper surface by intermolecular adhesion. A scanning electron
microscopic image of toner print on office paper is contained in the electronic
supplementary material.
This article addresses two questions:
(i) Are there solvents that will allow toner print to be removed without
harming the paper?
(ii) Will removing the toner enable office paper reuse?
The article then goes on to investigate the effect of varying three parameters:
(i) What is the effect of agitation on toner removal?
(ii) What is the effect of mixing solvents on toner removal?
(iii) What is the effect of the volume of solvent on toner removal?
The following section of this article briefly reviews the Hansen approach to
selecting solvents. Section 3 appraises the evidence of previous attempts to remove
toner from paper using a solvent. Section 4 describes the experimental method
used in this research. Section 5 reports and §6 evaluates the results of these
experiments. There are two important limitations to this research: only a single
toner paper combination has been tested and not all the factors that influence
solubility have been investigated. These limitations are discussed in the final
section of this article.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
Un-printing toner with solvents
3841
2. The Hansen approach to selecting solvents
Hansen (1967a,b,c) has proposed a method of selecting solvents that will dissolve
a polymer based on matching three ‘Hansen solubility parameters’ for the solvent
and for the polymer. Each of the parameters represents the strength of a different
type of bond that would typically form between molecules of the solvent or
polymer: δd for dispersion bonds, δp for dipole–dipole or polar bonds and δh for
hydrogen bonds. The three parameters can be used as axes in a three-dimensional
space, with the solvent and the polymer plotted as points in the space. The closer
the two points are in this space, the more likely the polymer is to be soluble in
the solvent. Hansen suggested a measure Ra of the distance between solvent and
polymer in this space:
Ra 2 = 4(δdA − δdB )2 + (δpA − δpB )2 + (δhA − δhB )2
(2.1)
If the distance Ra is greater than a polymer specific number Ro, then the polymer
is unlikely to be soluble in the solvent. From this, Hansen defined a measure of
whether a polymer will be soluble in a solvent RED = Ra/Ro. If RED is < 1, then
the polymer will be soluble. The Hansen parameters of many common solvents
and some polymers have been published (Barton 1991; Hansen 1999). If solvents
are mixed, then Hansen suggests that mixtures tend to have Hansen parameters
approximately equal to the volumetric average of those of the solvents.
A solvent may not completely dissolve a polymer. Depending on the nature and
accessibility of the bonds in the polymer, the polymer may only swell. In the case
of a multi-material structure such as toner print or paper, different components
may do nothing, swell or dissolve completely leading to a complex effect. Instead
of a single polymer, this article therefore calculates Hansen solubility parameters
based on which solvents create the desired or undesired overall effect in a material.
In the case of the toner print, the effect that is of interest is whether the pigment
contained in the print is separated from the paper surface. In the case of the
paper, the effect that is of interest is whether the paper is visibly harmed in a
way that inhibits reprinting. Once the solubility parameters for each effect have
been estimated, a set of possible solvents and solvent mixtures can be identified
that are likely to allow toner print to be removed without damaging print. Once
these solvent and solvent combinations are identified, the potential for improving
performance by varying operating parameters such as volume, agitation and
temperature can be explored. From this, the practical potential of the process
can be gauged.
3. Previous work on the use of solvents for toner removal
One academic paper and eight patents provide evidence of previous work in this
area. Their implication for each of this paper’s research questions is discussed.
(a) Reports on the choice of solvent
The solubility parameters of different components of paper have been tabled
in, for instance, Hansen & Björkman (1998). This work suggests that cellulose,
the main component of the paper, is unlikely to dissolve completely in organic
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3842
T. A. M. Counsell and J. M. Allwood
solvents, however Robertson (1970) established that the tensile strength of the
paper varies when the paper is immersed in different solvents and Hansen &
Björkman (1998) have used this to create a broad estimate of the solubility
parameters of paper as a whole which should give a sense of what solvents are
likely to harm the paper and therefore limit its reuse.
The solubility parameters of typical toners have not been reported.
Consequently, the quantitative difference in solubility, Ra, between the paper
and the toner has not been estimated. However, there is qualitative evidence that
the solubility of the toner is distinct from the solubility of the paper. The evidence
of this is reviewed below in chronological order.
The earliest report was in a case study on the hazards of using photocopiers
in a conservation environment: Nicholson (1989) mentioned that she was able
to use toluene and dimethyl-ethane to remove the toner from some valuable
historic documents that had accidentally been fed into a photocopier instead
of fresh paper.
Eight subsequent patents (Orita & Chikui 1989; Higuchi & Takahashi 1992;
Yasumasa 1992; Machida et al. 1996, 1997; Yamamoto et al. 1996; Bhatia et al.
1999; Bhatia 2000) have been filed on using solvents to remove toner in order to
reuse the paper, each suggesting a different set of solvents. Yasumasa (1992) uses
chloroform alone, while the others use a mixture of chemicals. Yamamoto et al.
(1996) use water, a surfactant and one of a large list of chemicals ‘includ(ing)
dihydric organic monoesters, glycol ether and the like’ together with a standard
set of laboratory solvents. Machida et al. (1996) again use water, a surfactant
and one of a large list of solvents, with the most effective discussed below.
Bhatia et al. (1999) suggest a wide range of fluids, but their preferred embodiment
has four or five components: water; an aliphatic solvent such as ‘diethylene
glycol n-butyl ether’; an aromatic solvent such as ‘ethyl 4-meth-oxy-benzoate’; a
surfactant and optionally a proprietary enzyme.
The majority of the patents therefore propose mixtures of solvents and other
chemicals. None indicate how the mixing changes removal performance.
(b) Reports on overall effectiveness
Two reports claim that solvents can remove toner sufficiently for paper to be
reused, the one that presents quantitative results will be considered first.
Machida et al. (1996) report the results of soaking the printed paper in six
organic solvent–water–surfactant mixtures for a minute, then ‘lightly rubbing
copy paper surface with a web’ to remove toner print. They define a ‘cleaning
efficiency’ measure that is the percentage increase in the amount of light reflected
from the sheet before and after the toner removal process. They claim that a
cleaning efficiency of above 70 per cent results in usable paper and that four of
their six results surpass this level. The six reported results use three different
solvents (dichloromethane, toluene and xylene) and three different surfactants.
The highest performance mixture that they report is 50 per cent toluene, 49
per cent water and 1 per cent C18 H29 NaO3 S. The table does not allow the separate
effects of mixing solvents or of adding surfactants to be understood.
Two Federal Bureau of Investigation agents, Frost & Dwyer (2001), tested
the process that was developed by DeCopier Technologies Ltd from patents
by Bhatia et al. (1999) and by Bhatia (2000). Frost & Dwyer (2001) printed
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
Un-printing toner with solvents
3843
sample paper and overhead transparency with a range of combinations of toner,
ink-jet, type-writer, pen and pencil print. A DeCopier employee manually applied
a water–solvent–surfactant fluid, brushed, heated and dried the samples. The
details of this process were not reported, and the employee was permitted to
vary the composition of the deinking solution and the time and pressure of the
brushing as they saw fit. The level of paper damage and quantity of readable
text were then tabulated qualitatively. They reported complete removal of toner
that had been printed on overhead transparency, but less success with toner that
had been printed on the paper: text was sometimes readable; there were always
traces of toner visible under a microscope; there was always a visible change in the
texture of the paper. A reason for the discrepancy between this performance and
the claims made by DeCopier Technologies of 100 per cent removal (Bhatia et al.
1999) was not given.
(c) Reports on other operating parameters
The patents do not specify potentially important operating parameters, such
as the volume of the solvent used. The patents all suggest agitation but they do
not indicate how agitation influences removal performance. Higuchi & Takahashi
(1992), Orita & Chikui (1989) and Yasumasa (1992) use ultrasound; Bhatia et al.
(1999), Machida et al. (1996, 1997), Yamamoto et al. (1996) and Bhatia (2000)
brush or rub the paper surface and Nicholson (1989) places the paper on a
suction table.
In summary, existing reports give some confidence that solvents can allow toner
print to be removed but doubts remain as to how useful the results are, the
operating parameters, how the parameters relate to performance and whether
the processes have been optimized.
4. Methods
Six sets of experiments were carried out on a single toner–paper combination.
The first set was used to estimate the solubility parameters for the toner, the
second was to investigate mixing solvents to vary the solubility parameter and
the third was to investigate the reusability of the paper once the toner was
removed. The final three sets of experiments explored the influence of mechanical
agitation, ultrasonic agitation and the volume of solvent. This section describes
the experiments. The next section reports their results.
(a) The sample toner and paper
There is significant variation in both the formulation of toners and in the
structure of the paper and the additives within it. The tests reported in this
article were carried out on a single toner–paper combination. Both the print and
the paper are typical of an office environment. The paper was white, uncoated,
wood-free, 80 g m−2 Canon copy paper. The printer was an HP 4200 dtn blackand-white laser printer. The exact toner formulation is proprietary to HP and has
been reported as 40–50% polyester–resin, 40–50% ferrous iron-oxide, 1 per cent
amorphous silica (HP 2004).
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3844
T. A. M. Counsell and J. M. Allwood
(b) Measuring toner removal performance
Three performance measures were used: a relative and a quantitative measure
of toner removal, and a relative measure of reusability.
(1) During the early tests, the visible effect of the solvent was slight and
uncertain, so a relative measure of removal was used. A 1 cm2 of paper, printed
100 per cent black was used as the test sample. After carrying out the test, the
sample and the solvent were poured onto a piece of filter paper. The solvent was
allowed to soak and pass through the filter paper, before both the filter paper and
the sample were air-dried. The sample was visually inspected for traces of white,
and the filter for traces of black. The results were then qualitatively compared to
each other rather than placed on an absolute scale.
(2) The quantitative measure of removal was based on the resulting whiteness
of a black sample. A 1 cm2 the paper, printed 100 per cent black was used as the
test sample. After carrying out the test, the sample was air-dried and a 600 dpi
optical scan of the black surface taken on a conventional flatbed scanner. The
level of toner removal was quantified by calculating the mean whiteness of the
sample (on a scale of 0–255) of the resulting image. On this measure, it was found
that an untouched black printed sample has a mean whiteness of 42, and a fresh
sheet of paper has a mean whiteness of 252. The whiteness results were then
linearly converted into a percentage, with a black printed sample given 0 per cent
and a white sheet of paper 100 per cent.
(3) Reusability is closely tied to readability and was therefore measured
qualitatively using a larger 9 cm2 of a paper that was printed with black text.
After the test, the sample was air-dried and a 600 dpi optical scan of the printed
surface taken as for the quantitative method above. The sample was then attached
to a backing sheet of paper and reprinted with a different text. A scan of the
reprinted sample was then taken. The unprinted and reprinted text were then
qualitatively compared to freshly printed sheets of the original and new text.
The reusability performance is measured using the paper printed with black
text, representing the typical state of used paper. The tests of the effects of
varying operational parameters are done with a paper that is completely covered
on one side with black print because this makes differences between performance
levels more visible and quantifiable. In this way, both the typical and extreme
states of the paper are explored.
(c) Test 1: Hansen solubility parameters
The exact formulation of the toner resin is not known, so its solubility
parameters were estimated using a variation of the experimental method proposed
by Hansen (1999). This interpolates a polymer’s solubility parameters from the
solvents in which the polymer dissolves. In this case, a 1 cm2 sample of 100 per cent
black printed paper was placed in 50 ml of each of the 16 solvents listed in table 1
for 24 h at room temperature. The effectiveness of the solvent was then evaluated
using the qualitative measure 1 explained in the previous section. The solubility
parameters of the solvents and their ability to allow the pigment in toner print to
become separated from the paper surface were fed into a custom program.1 The
program followed the method described in Hansen (2000) and made iterative
1A
listing of the programme is available from the author on request.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
Un-printing toner with solvents
3845
Table 1. Solubility of toner in 16 solvents.
white flecks visible on the
sample and black flecks
on the filter paper
acetone
dimethyl formamide
tetrahydrofuran
black flecks on the
filter paper
traces of black on the
filter paper
dichloromethane
diethyl-ether
ethyl-acetate
toluene
1-butanol
acetonitrile
chloroform
no visible effect
dimethylsulphoxide
ethanol
hexane
isopropanol
methanol
water
estimates of the solubility parameters and solubility radius (Ra) of the toner,
seeking to maximize the number of tested solvents whose performance would be
correctly classified using those values.
The approach used varies from that of Hansen (1999) in that a direct measure
of solubility is not being made. Hansen recommends that the interaction of
the solvent and the polymer be examined under a microscope to spot signs of
dissolution or swelling. This method looks for an indirect result of the dissolution:
black pigment that has become detached from the paper. It will, therefore, ignore
occasions when a solvent does swell or dissolve one or more of the materials in
toner print but that do not result in the pigment becoming suspended in the
solvent away from the paper.
(d) Test 2: mixing solvents
Once the Hansen parameters of the toner pigment removal were estimated, a
series of tests were carried out with mixed pairs of solvents in varying propotions
to explore the effect of varying the effective Hansen parameters of the solvent.
The first pair was chloroform and acetone. The second pair was chloroform and
dimethylsulphoxide. In both cases, six tests were carried out on 1 cm2 samples
of 100 per cent black printed paper that were placed in 5 ml of solvent, of which
between 0 and 100 per cent was chloroform, and then agitated for 10 s with
ultrasound.
The level of toner pigment removal was analysed with the quantitative method
2 outlined in the previous section. The results were compared to the estimated
Hansen solubility parameters of the solvent mixture. These parameters were
calculated as the volumetric average of the parameters of the component solvents,
as suggested by Hansen.
(e) Test 3: overall performance
The overall performance was measured using larger 9 cm2 samples of the paper
printed with black text. These were placed in 10 ml of solvent and agitated
for 240 s. The solvents tested were water, dimethylsulphoxide, chloroform,
dichloromethane, acetone and mixtures of each. The results were scanned and
visually compared to the original sample. The samples were then reprinted with
a second layer of black text and compared.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3846
T. A. M. Counsell and J. M. Allwood
(f ) Tests 4–6: the effects of agitation and volume
Once the overall performance had been established, three further sets of tests
were carried out to explore the effect of ultrasonic and mechanical agitation and
of varying the volume of the solvent used. In each case, the solvents were chosen
from those tested above, using one that did not remove toner, two that had little
effect and one that was successful. In each case, performance was measured using
the quantitative method 2 outlined in the previous section.
To measure the effect of ultrasound, 1 cm2 samples of 100 per cent black printed
paper were placed in 4 ml of solvent. The jar containing the solvent and the sample
was placed in an ultrasound bath for 240 s, operating at 40 kHz and 80 W for a
period of between 1 and 240 s at the end of the soaking period.
To explore mechanical agitation, 1 cm2 samples of 100 per cent black printed
paper were placed in 4 ml of solvent for 60 s. The sample was then removed and
rapidly placed face down on a piece of fresh white tissue paper. It was then
dragged across the tissue paper until the solvent had dried. This process was
repeated on the same sample between one and five times.
To measure the effect of varying the volume of solvent on toner removal,
1 cm2 samples of 100 per cent black printed paper were placed in acetone for
60 s and agitated with ultrasound. The volume of the solvent was varied from
0.5 to 9 ml.
5. Results
This section presents the results of each test. The following section uses the results
to answer the five research questions. The final section of this article discusses
the limitations and opportunities of the method and results.
(a) Test 1: Hansen solubility parameters
The results of the tests to estimate the Hansen solubility parameters of the
toner pigment removal are shown in table 1. In this table, the solvents have
been grouped according to the visible effect they had, and then organized
alphabetically. Of the 16 solvents tested, three caused the toner pigment to be
suspended in the solvent and six did not. The remaining seven solvents appeared
to have some influence on the toner.
From these results, the estimated Hansen solubility parameters for the toner
that lead to √
pigment removal are δd = 15.3–16.9, δp = 8.5–10.1, δh = 3.1–8.8 and
Ro = 5–10.5 MPa. The range is based on whether the best 3, 7 or 10 solvents
are counted as being solvents for the toner and is detailed in table 2.
No solvents had a significant impact on the paper, and therefore the Hansen
solubility parameters for cellulose are taken from Hansen & Björkman (1998)
which were calculated from the results of the experiments carried out by
Robertson (1970) on how the tensile strength of the paper varies when the paper
is immersed in a solvent. From this Ra, the difference
in solubility between the
√
paper and the toner was estimated as 14.0–18.1 MPa.
The Hansen parameters based on 10 solvents are used in the remainder of
this article.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3847
Un-printing toner with solvents
Table 2. Best estimates for solubility parameters.
fit
papera
toner, based on
the 10 working solvents in table 1
excluding the three least effective solvents
excluding the seven least effective solvents
1
0.9
1
δ√
d
( MPa)
δ√
p
( MPa)
δ√
h
( MPa)
Ro
√
( MPa)
20.3
16.3
18.7
11.7
15.3
16.9
16.6
8.5
10.1
9.9
6.2
3.1
8.8
10.5
9
5
a Estimate
from Hansen & Björkman (1998) that is based on the tensile strengths of paper in
different solvents, as reported by Robertson (1964).
Table 3. Eleven illustrative two-solvent mixtures with estimated solubility parameters close to those
for toner print removal.
mixture
√
δd ( MPa)
δp ( MPa)
√
δh ( MPa)
18%
36%
57%
24%
88%
41%
30%
52%
44%
21%
67%
15.9
15.9
16.3
15.8
15.8
15.7
16.4
16.6
16.7
16.1
18.0
9.1
7.8
7.8
9.3
9.9
7.3
9.4
7.9
9.7
7.1
7.5
6.8
6.9
6.4
7.2
6.9
7.6
7.4
4.9
5.9
8.1
7.2
chloroform + acetone
dimethylsulphoxide + diethyl ether
dimethyl formamide + hexane
tetrahydrofuran + acetone
acetone + dichloromethane
dimethyl formamide + diethyl ether
acetonitrile + tetrahydrofuran
hexane + dimethylsulphoxide
acetonitrile + chloroform
dimethyl formamide + ethyl acetate
chloroform + dimethylsulphoxide
√
REDe
toner
paper
0.1
0.2
0.2
0.2
0.2
0.2
0.3
0.3
0.3
0.3
0.5
1.4
1.5
1.4
1.4
1.4
1.5
1.3
1.5
1.4
1.4
1.3
(b) Test 2: mixing solvents
Table 3 illustrates some two-component mixtures that are calculated to have
the Hansen solubility parameters close to those estimated for the toner pigment
removal above in test 1. The left-hand column contains the composition, the
middle columns contains the estimated solubility parameters and the right-hand
columns contain the estimated solubility of the mixture with respect to the toner
and to the paper.
The results of the tests to investigate the influence of mixing pairs of solvents
are shown in figure 1 for chloroform–dimethylsulphoxide mixtures and in the
electronic supplementary information for chloroform–acetone mixtures. The left
column lists the proportion of chloroform used in the mixture, with the balance
made up of the other solvent. The middle columns list the predicted solubility
parameters and solubilities with respect to the toner and cellulose. The right
columns contain a scanned image of the sample and a corresponding measure of
the whiteness of the sample.
Mixtures of the two solvents generally had an increased ability to remove toner
pigment compared to the solvents alone, but this did not appear to correlate with
the increase in solubility predicted by the Hansen solubility parameters. Nor did
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3848
T. A. M. Counsell and J. M. Allwood
% chloroform
HSP (√MPa)
δpe
δde
δhe
REDe
toner
paper
%whiteness after
60 s ultrasound
100
17.8
3.1
5.7
0.7
1.6
86
90
17.9
4.4
6.1
0.6
1.5
85
80
17.9
5.8
6.6
0.6
1.4
76
70
18.0
7.1
7.0
0.5
1.3
90
60
18.0
8.4
7.5
0.5
1.2
86
50
18.1
9.8
7.9
0.6
1.1
89
40
18.2
11.1
8.4
0.6
1.1
88
30
18.2
12.4
8.8
0.7
1.0
94
20
18.3
13.7
9.3
0.8
0.9
79
10
18.3
15.1
9.8
0.9
0.8
85
0
18.4
16.4
10.2
1.0
0.8
85
Figure 1. Effect on toner removal of mixing chloroform and dimethylsulphoxide.
the removal vary smoothly as the proportion of each component varied. Mixtures
of chloroform and acetone resulted in sample whiteness of around 80 per cent,
while some mixtures of chloroform and dimethylsulphoxide resulted in sample
whiteness of above 90 per cent.
(c) Test 3: overall effectiveness
Figure 2 shows the effectiveness of solvents for removal of toner print and
figure 3 shows the usability of subsequent reprints on the cleaned paper.
The best performance of a single solvent appeared to be with chloroform.
However, this leaves the original text faintly visible. Better performance was
achieved by the two mixtures of solvents, with a 40 per cent chloroform 60 per cent
dimethlysulphoxide mixture particularly effective.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
Un-printing toner with solvents
(a)
(b)
(c)
(d)
(e)
(f)
(g)
3849
(h)
Figure 2. Unprinted office paper. (a) For comparison: original sample, (b) water, (c) dimethylsulphoxide, (d) chloroform, (e) dichloromethane, (f ) acetone, (g) 80% acetone + 20% chloroform,
(h) 40% chloroform + 60% dimethylsulphoxide.
(d) Tests 4–6: agitation and volume
The results of the tests to investigate the influence of ultrasonic agitation are
shown in figure 4. This table has the four solvents tested across the top organized
in increasing order of solubility as estimated in the test 1. Vertically the table is
organized according to the duration of ultrasound, increasing downwards. Each
cell in the table contains an image of the scanned sample and a number indicating
the average whiteness of the sample as specified in the previous section.
The use of ultrasound improved whiteness. The best removal performance
without ultrasound was 12 per cent. The best performance with ultrasound was
more than 80 per cent, achieved by using chloroform and 240 s of agitation.
The results of the tests to investigate the influence of mechanical rubbing
are shown in figure 5. Similar to the ultrasound results table, this table has the
four solvents tested across the top organized in increasing order of solubility as
estimated in test 1. Vertically the table is organized according to the number
of times the sample was soaked in the solvent and then rubbed, increasing
downwards. Each cell in the table contains an image of the scanned sample and a
number indicating the average whiteness of the sample as specified in the method
section above.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3850
T. A. M. Counsell and J. M. Allwood
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
Figure 3. Reprinted office paper. (a) For comparison: text on fresh paper, (b) water, (c) dimethylsulphoxide, (d) chloroform, (e) dichloromethane, (f ) acetone, (g) 80% acetone + 20% chloroform
(h) 40% chloroform + 60% dimethylsulphoxide.
One mechanical rub improved whiteness, but subsequent rubbing did not
necessarily result in improvement. The best performance, with a whiteness of
68 per cent, was achieved by rubbing an acetone-soaked sample once. Rubbing
samples that had been soaked in chloroform or dichloromethane was much less
effective, possibly because the solvent evaporates from the paper more quickly.
Close inspection of the sample after rubbing suggests that the paper may have
been damaged—individual paper fibres were much more visible than when not
using agitation or when agitated with ultrasound. The extreme case was water,
where one rub was sufficient to remove a layer of paper fibres and subsequent
rubs tore the sample.
The results of the tests to investigate the influence of solvent volume are
shown in figure 6. The volume of the solvent used increases down the table. The
cells of the table contain scanned images of the sample and a corresponding
whiteness measure.
For chloroform, whiteness of 80 per cent was reached using 3 ml of solvent,
for dichloromethane 5 ml was required. A peak whiteness of 78 per cent was also
reached at 5 ml for acetone.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3851
Un-printing toner with solvents
ultrasound
water (%)
chloroform (%) dichloromethane (%) acetone (%)
none
3
3
5
12
1s
12
33
43
40
3s
4
40
64
44
5s
4
64
43
37
7s
4
58
62
56
10 s
2
77
52
64
15 s
7
80
53
56
30 s
10
83
66
71
60 s
4
86
83
74
120 s
4
83
75
74
240 s
2
82
83
78
for comparison:
black sample
0 (%)
white paper
100 (%)
Figure 4. Effect on toner removal of ultrasound duration.
The result was checked by repeating some of the tests with a half-size sample.
The results of these checks implied that the results above were due to solvent
volume rather than influence of a change in volume on the way that the the
ultrasonic agitation was transmitted into the sample.
6. Evaluation: useful, but not perfect, removal
This section will use the results to answer the five questions set out §1, starting
with the big questions of whether it is a useful process and then discussing the
effect of varying the operating parameters. The next and final section will discuss
the overall limitations of this research.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3852
number of rubs
T. A. M. Counsell and J. M. Allwood
water (%)
chloroform (%)
dichloromethane (%) acetone (%)
0
3
3
5
12
1
58
20
24
68
2
.
a
32
34
65
3
.
.
31
34
55
4
.
.
37
40
52
5
.
.
42
49
50
for comparison:
black sample
0 (%)
white paper
100 (%)
Figure 5. Effect on toner removal of rubbing with tissue. a, All subsequent rubs destroyed
the paper.
To answer the first two questions posed in §1: (i) there are solvents that appear
to dissolve toner print sufficiently for the pigment to no longer be attached to
the paper; (ii) some of those solvents do not appear to harm the paper; therefore,
useful but not perfect removal can be achieved.
Toner pigment removal and paper damage are triggered by solvents that
fall in distinct but overlapping ranges of Hansen solubility parameters. Three
projections of the three-dimensional Hansen solubility space are shown in figure 7.
The large dotted circle is the radius of solubility (Ro) of the paper and the large
solid circle is that of the toner. The small circles are tested solvents. As before,
filled circles are those that have been found to detach or dissolve the toner
pigment. As can be seen by the relative positions of the two circles on the three
axes, it is likely that forces between components of the toner are created by having
more dispersive bonds and fewer hydrogen bonds than those between components
of the paper. This difference is also illustrated in an alternative Teas plot of the
same data contained in the electronic supplementary material. The difference in
bond types leads to an√
estimated Ra solubility difference between the paper and
the toner of 14.0–18.1 MPa. The overlap between the solubility ranges is also
visible√in the figure. The sum of √
the estimated radii of solubility of the paper
MPa) is greater than the distance between
(11.7 MPa) and the toner (10.5 √
their Hansen parameters (14.0–18.1 MPa).
Taken together, this means that the toner can be removed without also harming
the paper. A database of Hansen parameters for 896 solvents was used to classify
solvents according to whether they should result in toner removal, harm to the
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3853
Un-printing toner with solvents
volume of solvent (ml)
chloroform (%)
dichloromethane (%) acetone (%)
1
2
47
40
40
1
56
31
53
1 12
63
58
62
2
65
40
62
2 12
69
58
59
3
80
54
69
4
80
76
66
5
86
79
78
6
85
80
70
7
88
79
77
8
85
80
73
9
84
82
75
for comparison: black sample
0 (%)
white paper
100 (%)
Figure 6. Effect on toner removal of the volume of solvent used.
paper, both or neither. Approximately 80 per cent of the solvents checked should
allow the toner pigment to be removed without harming the paper. Based on
the estimated solubility parameters, propylene oxide should be one of the best
solvents for removing the toner, with a RED of 0.1 with the toner and 1.5 with the
paper. However, this solvent is highly flammable and toxic, and its RED value is
not a significant improvement on the values achievable by mixing more common
solvents as described below.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3854
T. A. M. Counsell and J. M. Allwood
(a)
50
δp
0
50
δh
(b)
50
(c)
50
δh
δp
0
25
δd
0
25
δd
Figure 7. Plots of Hansen solubility parameters of the toner (solid line), paper (dotted line) and
solvents tested (circle). The plus symbols represent the central estimate of the parameters and the
circle indicates the boundary where RED√
< 1. Filled black circles indicate that the solvent was
found to dissolve the toner. The units are MPa. (a) δh δp (b) δd δh (c) δd δp .
There are sources of doubt in this result: only one toner and paper combination
has been tested; there is quite a wide range in the estimates for the toner
solubility parameters; the analysis does not explore the effects of the solvent
on the individual materials that make up toner print or paper.
These experiments have found solvents that allow one type of toner print to
be removed from one type of paper. It is not known how applicable the result
is to other toners and papers. The test should be repeated to estimate the
solubility parameters of other toners and papers. The volume of intersection
between different toners could then be established. If there is a volume of
intersection and that is distinct from any of the paper Hansen parameters,
then it is likely that a solvent could be found to operate on a broad set
of toners.
The range of estimates for the solubility parameters of the toner has arisen
because of the range in volumes of detached pigment when using solvents to
remove toner in the test, and the small number of solvents used in testing. The
range might be narrowed by testing with further solvents and potentially by
carrying out solubility tests under a microscope so that any swelling or dissolution
of the toner or paper might be seen directly without relying on the indirect
measure of pigment movement. In particular, the solubility parameter for the
toner has been based on the solubility of whatever must be dissolved in the toner
for the pigment to become detached and then separated from the paper surface.
It is presumed that this is the toner polymer, but the nature of the pigment may
also be a factor.
It has also been presumed in these estimates that the most appropriate
solubility parameter for paper is one calculated from how the tensile strength
of paper varies when it is soaked in a solvent. There may be other ways that
the paper has been altered by the solvent that may influence long-term stability.
In some cases, the drying and wetting processes may cause the paper sample to
ripple (this can be seen in sample figure 2h) that indicates some other effect is
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
Un-printing toner with solvents
3855
taking place. This rippling has not affected the ability to reprint the samples
during the experiments. Once a practical ability to remove print has been
demonstrated, further work would be required to confirm the stability of the
paper over multiple cycles of reuse over extended periods.
Useful, but not perfect, removal can be achieved. Image h in figure 2 shows
the best level of removal achieved. This placed the sample in a mixture of 4 ml of
chloroform and 6 ml of dimethlysulphoxide for 240 s while applying ultrasound.
The removal was not complete and it is still possible to read the original text.
However, the original text does not interfere with the reprinted text (image in
figure 3h) meaning the paper is reusable.
However, the cleaned sample has a slightly off-white colour, a slightly rougher
surface and some ripples when compared with a new sheet of paper. This implies
damage to some components of the paper, but this damage does not seem to
affect reprinting.
Varying the operating parameters significantly improves removal:
(i)
(ii)
(iii)
(iv)
agitation is more important than the detailed choice of solvent,
ultrasonic agitation is more effective than mechanical rubbing,
mixing solvents can improve removal,
a minimum of 100–150 ml of solvent is needed for effective removal of a
typical density of print from an A4 sheet.
Each of these points are discussed in turn.
Figure 4 shows that the variation in toner removal is much smaller between
solvents than between levels of ultrasound. Water is the only one of the four
solvents whose Hansen solubility parameters imply no effect on toner and
therefore the only solvent where ultrasonic agitation appears to make little
difference to removal levels. The other solvents have solubility parameters that
imply varying degrees of solubility, and once any level of solubility is possible
it appears that agitation is critical. This may imply that the polymers in
toner that bind the pigment to paper have some cross-linking that prevents
complete dissolution.
Comparison of the results of ultrasonic agitation (figure 4) and mechanical
rubbing (figure 5) suggests that ultrasound is more effective than this particular
approach to agitation. Rubbing visibly damaged the surface of the paper and
tended to embed some of the toner’s pigment deep into the paper structure.
No similar effect was seen in ultrasound.
Alternative approaches to mechanical rubbing may be more effective.
In particular, a few trials were made using a brush to flick the toner from the
paper surface once it had been soaked in solvent. This tended to result in less
damage to the paper and less pigment embedded into the paper but overall was
less effective than ultrasound in removing print.
Furthermore, the effect of varying ultrasound frequency, power and the way
the ultrasound is coupled to the solvent and the paper have not been investigated.
Adjusting these parameters may further improve the process.
Table 3 shows that the removal of toner can be improved by mixing solvents
so as to emulate a solvent with Hansen solubility parameters closer to that
of the toner. This is shown particularly clearly by comparing figure 2c,d,h.
Dimethylsulphoxide alone is poor at removing print, chloroform is better but
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3856
T. A. M. Counsell and J. M. Allwood
leaves readable text and a grey surface. Mixing the two results in a white surface
with barely readable text. Ultrasound remains as important with solvent mixtures
as with single solvents.
However, the table also shows that a particular mixture’s performance does
not completely match that predicted. This could be because: the estimate of
the solubility parameters of the toner are incorrect or the multi-component
nature of the toner and the paper means that the solvent mixture does not
perform as the volumetric average of its component solvents. One of the reviewers
of this paper has highlighted that dimethylsulphoxide is a good solvent for
removing lignin from milled wood. There may therefore be a more complex,
and beneficial, interaction between the mixture and different components of
the paper.
Figure 4 shows that removal of toner suffers when the volume of the solvent
drops below a particular level but does not improve substantially when above
the level. This level appears to be dependent on the solvent used. The results
imply that at least 100–150 ml of solvent would be required to clean a sheet of
paper that had been covered with the standard 5 per cent coverage of print that
is quoted by printer manufacturers. When phrased as 10 sheets of paper cleaned
per litre of solvent, this raises questions over the viability of the approach unless
the solvent can be reused—a point discussed below.
In summary, these experiments suggest that solvents could be used to remove
toner print in a way that leaves office paper in an immediately reusable state.
The following section discusses the limits of this claim and the scope for
further research.
7. Conclusions
This article is the first report of a systematic investigation of the use of solvents
to help remove toner from office paper. For the print paper combination tested,
it is possible to remove sufficient print for paper to be reused using a mixture of
dimethylsulphoxide, chloroform and ultrasound. However, this result has not been
fully validated on other print–paper combinations; there are other parameters
that could be studied and more data are required to operationalize the process.
This section will discuss how to surmount these limitations.
(a) The results must be validated on other toner–paper combinations
If all toners have solubility parameters distinct from all papers, then it will need
to be established whether the solubility parameters of all toners are sufficiently
close to each other for a single universal solvent to be found, or whether in practice
the paper should be immersed in a series of solvents to remove all types of toner.
Furthermore, the solubility parameters of the minor ingredients of toner and
paper may need to be investigated in detail. In particular, it is not known if
the solvents remove any paper additives that are critical to the reliability of
subsequent reprinting or to the life-span of the paper. This could be explored by
estimating the Hansen solubility parameters for each common ingredient of paper
and by evaluating the implication of the loss of that component with paper and
printer manufacturers.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
Un-printing toner with solvents
3857
(b) The influence of temperature and surfactants has not be explored
Removing the toner pigment from the paper is a two-stage process: first the
bond between it and the paper must be destroyed (in this case by dissolving the
toner polymer that attaches it) and second the pigment and the paper must be
separated in location.
Temperature is likely to influence solubility. This has not been explored in
this article.
Most of the patents that have been filed in this area suggest that a surfactant
is mixed with the solvent, presumably to help with the second stage of
separating the pigment from the paper once the polymer is dissolved. Improving
this separation has not been explored in this article. It seems likely that
surfactants, solvent circulation and the introduction of electrostatic or magnetic
attractors might lead to better separation, and hence a whiter appearance for the
cleaned paper.
(c) There are practical challenges
Even though a solvent mixture has been found that is effective at allowing
toner print to be removed without damage to the paper, there are two practical
challenges to using this method outside the laboratory: solvent safety and solvent
recycling.
The most effective removal was found to be a chloroform–dimethylsulphoxide
combination. If the process is carried out industrially at a central plant, then use
of these solvents is probably acceptable. However, one ambition is for a process
that could be used at an office level, ideally by integrating a removal process into
an office photocopier, printer or recycling bin. In this case, chloroform is unlikely
to be considered sufficiently safe for use. A safer alternative would need to be
found. This may be possible by analysing the solubility parameters of benign
solvents and solvent mixtures to find those that fall within the Ro of the toner
estimated in this article.
The results of experiment 4 above imply that a litre of solvent will be required
to remove print from about 10 pages, a volume that is likely to be uneconomic
and potentially too damaging to the environment. Solutions could be to alter the
process to use less solvent or to reuse the solvent. An alteration to the process
that might be researched would be to replace the solvent bath with a jet or spray
of solvent (combined with other forms of agitation). To reuse the solvent, two
steps are required: the first is to remove the pigment from the solvent. This is
probably easily done on the basis of size or density. The second is to remove
the polymer from the solvent once the solvent has become saturated. This will
presumably require the solvent to be evaporated and recondensed, leaving the
polymer behind. This evaporation and condensation may cause further safety
challenges and may require too much energy for reuse to be a viable alternative
to conventional paper recycling. Such challenges could be reduced by selecting a
solvent with an evaporation temperature as close as possible to room temperature.
In summary, there are both practical challenges and scope for further
improvement in removing toner print from paper with solvents, but the essential
principle appears plausible: print can be dissolved without damaging the paper
and, so long as ultrasonic agitation is used, the resulting paper can be readily
reprinted and reused.
Proc. R. Soc. A (2009)
Downloaded from http://rspa.royalsocietypublishing.org/ on July 31, 2017
3858
T. A. M. Counsell and J. M. Allwood
We would like to thank Wilhelm Huck and Jane Comrie of the Department of Chemistry and Len
Howlett of the Department of Engineering at the University of Cambridge for their assistance in
setting up some of the experiments reported in this article.
References
Barton, A. F. M. 1991 CRC handbook of solubility parameters and other cohesion parameters, 2nd
edn. Boca Raton, FL: CRC Press.
Bhatia, S. 2000 Method for deinking paper. US Patent US6022423.
Bhatia, S., Pichel, M. & Ashu, J. T. 1999 Method and apparatus for deinking paper. Worldwide
Patent WO9947743.
Counsell, T. A. M. 2007 Remove toner: reuse paper. PhD thesis, University of Cambridge,
Cambridge, UK.
Counsell, T. A. M. & Allwood, J. M. 2006a Un-printing toner: early results. In 13th CIRP Int.
Conf. on Life Cycle Engineering, Leuven, 31 May–2 June 2006. Paris, France: CIRP.
Counsell, T. A. M. & Allwood, J. M. 2006b Desktop paper recycling: a survey of novel technologies
that might recycle office paper within the office. J. Mater. Process. Technol. 173, 111–123.
(doi:10.1016/j.jmatprotec.2005.11.017)
Counsell, T. A. M. & Allwood, J. M. 2007 Reducing climate change gas emissions by
cutting out stages in the life cycle of office paper. Resour. Conserv. Recycl. 49, 340–352.
(doi:10.1016/j.resconrec.2006.03.018)
Counsell, T. A. M. & Allwood, J. M. 2009 Using abrasives to remove a tonerprint so that office
paper might be reused. Wear 266, 782–794. (doi:10.1016/j.wear.2008.10.003)
Frost, B. T. & Dwyer, K. T. 2001 Office paper decopier. J. Forensic Sci. 46, 631–636.
Hansen, C. M. 1967a The three-dimensional solubility parameter—key to paint component
affinities. I. Solvents plasticizers, polymers, and resins. J. Paint Technol. 39, 104–117.
Hansen, C. M. 1967b The three-dimensional solubility parameter—key to paint component
affinities. II. Dyes, emulsifiers, mutual solubility and compatibility, and pigments. J. Paint
Technol. 39, 509–510.
Hansen, C. M. 1967c The three-dimensional solubility parameter—key to paint component
affinities. III. Independent calculations of the parameter components. J. Paint Technol. 39,
511–514.
Hansen, C. M. 1999 Hansen solubility parameters: a user’s handbook. Boca Raton, FL: CRC Press.
Hansen, C. M. & Björkman, A. 1998 The ultrastructure of wood from a solubility parameter point
of view. Holzforschung 52, 335–344. (doi:10.1515/hfsg.1998.52.4.335)
Higuchi, I. & Takahashi, M. 1992 Method for recycling copying paper. Japanese Patent JP4091298.
HP 2004 Q1338A materials safety data sheet: HP color LaserJet black print cartridge. Bracknell,
UK: Hewlett-Packard, Ltd.
Machida, J., Tanaka, S. & Yoshida, M. 1996 Method of recycling a waste recording member. US
Patent US5542985.
Machida, J., Tanaka, S., Furusawa, K. & Yoshida, M. 1997 Regenerating apparatus of recording
medium. US Patent US5678157.
Nicholson, K. 1989 Photocopier hazards and a conservation case study. The Book and Paper Group
Annual, vol. 8. Washington, DC: The American Institute for Conservation of Historic and
Artistic Works.
Orita, K. & Chikui, Y. 1989 Method and device for reproducing copying paper. Japanese Patent
JP1101576.
Reed, R. 1972 Ancient skins, parchments and leathers. London, UK: Seminar Press.
Robertson, A. 1964 Cellulose–liquid interactions. Pulp Paper Mag. Can. 65, 171–178.
Robertson, A. A. 1970 Interactions of liquids with cellulose. Tappi 53, 1331–1339.
Yamamoto, T., Yamomoto, K., Nishiguchi, K., Yoshida, M. & Atarashi, H. 1996 Processing device
for sheet-like media. US Patent US5528788.
Yasumasa, H. 1992 Regeneration of copying paper and production therefore. Japanese Patent
JP4300395.
Proc. R. Soc. A (2009)