8th International DAAAM Baltic Conference
"INDUSTRIAL ENGINEERING”
19-21 April 2012, Tallinn, Estonia
INK-JET PRINTING OF PHARMACEUTICALS
Takala, M.; Helkiö, H.; Sundholm, J.; Genina, N.; Kiviluoma, P.; Widmaier, T.;
Sandler, N. & Kuosmanen, P.
Abstract:
Traditional
tablet
manufacturing process has multiple
powder handling and mixing phases. It is
time consuming and limits individualized
dosing. Furthermore, uniformity of the
dose is difficult to achieve. Modern inkjet printing technology has been
successfully
applied
to
polymer
electronics and biomaterial applications.
It has many advantages that could be
utilized in drug manufacturing process,
such as faster production cycle with
fewer processing steps, precise dosing,
on-demand individualized dosing and
possibility for on-line quality control.
This study examines how well current
printer technologies function in printing
different types of drug solutions on
paper-like substrate. Preliminary tests
were carried out with paracetamol and
riboflavin-water solutions. The study
denoted that printing medicines and biopolymer coating was possible, the
accuracy of dose was very satisfying and
personalized dosing was easy to achieve.
Further
studies
with
specially
manufactured printers are still needed to
determine the suitability of the ink-jet
printing process in commercial drug
manufacturing.
Key words: thermal, piezoelectric, print
head,
paracetamol,
riboflavin,
personalized dosing.
1
INTRODUCTION
Manufacturing of pills is currently time
consuming and often inflexible. That is
because the manufacturing process
consists of multiple phases, which are
necessary with the technology and
methods applied in pharmaceutical
industry. Furthermore, the control of the
correct dose is done in post-production
where random pills are taken to be
examined to verify the correct content
through
statistical
methods
[1].
Sometimes the amount of the active
ingredient in a pill can be very small,
which means that the size of the pill has
to be increased with added ingredients in
order to make it large enough for even
picking up from the table. For example, a
nitro heart and artery medicine pill has a
0.5 mg active dose of glyceryl trinitrate,
whereas the pill itself weights
approximately 110 mg. By introducing
technology used in printers it could be
possible to make the process faster, more
accurate, more dynamic and easier to
control. Printing technology makes it
possible to directly inject active
ingredients onto the surface of an edible
substrate thus creating drugs that react
faster. Compared to a medicine droplet,
ink-jet droplet can be up to 20 times
smaller [2]. On top of that, new
possibilities arise when multiple active
ingredients could be injected onto the
same surface. MIT has done research
concerning the use of a 3D printer in pill
manufacturing. Their method included
both creating the actual pill and injecting
the dose into it [3].
The goal of the present study was to
determine if ink-jet techniques can be
used in pill manufacturing and how
accurate they would actually be. This was
done by selecting different manufacturers’ low-end printers with different
ink-jet technologies and performing a
series of repeated tests and measuring the
results.
2
METHODS
2.1 Printers
The tests were initially intended to
perform with two different commercial
printers: HP Photosmart B010 (Fig. 1)
and Epson Stylus SX 425W. The purpose
of using two different printers was to test
which technology is more suitable for
pharmaceutical printing – thermal ink-jet
or piezoelectric ink-jet. HP utilizes
thermal ink-jet print head and Epson
piezoelectric ink-jet print head. However,
from the beginning the piezoelectric
printer was found to be unable to print
any substance. Due to that all of the
analysed results are from the thermal inkjet printer.
The difference in these two technologies
is in the way the ink-droplet is formed.
Thermal ink-jet uses a heat element – a
small resistor – to create heat in order to
cause the ink to vaporize and to create a
bubble. When the bubble expands it
pushes ink out of the nozzle. As the
bubble collapses it creates a vacuum,
which causes more ink to be pulled from
the cartridge into the print head [4] (Fig.
2). Piezoelectric printer uses a small
crystal at the back of the ink reservoir in
each nozzle. Small electrical current is
led to the crystal, which causes the crystal
to vibrate. When the crystal expands, an
ink droplet is forced out of the nozzle and
when the crystal contracts, it pulls more
ink from the reservoir [5] (Fig. 3).
Fig. 2. The principle of bubble jet nozzle [6]
Fig. 3. The principle of piezoelectric nozzle [6]
Tests were made by replacing original ink
cartridges with cartridges filled with
riboflavin-water-glycerol solution and
paracetamol-water-glycerol
solution.
Small squares were then printed on a
standard paper, photo paper and
Fig. 1. HP Photosmart B010
transparent film to determine how well
the printers were able to print other liquid
than ink. The squares were analysed with
a microscope and an UV/VIS-spectrophotometer in order to resolve the amount
of pharmaceutical in one printed square
and also to find out whether the squares
pervaded on the paper or not.
Same test procedure was designed for
both types of printers.
substrate, which can then be analysed.
Furthermore, it makes it easier to produce
different kinds of test samples without
the need of writing down every change
made to the test. Every parameter of the
test can be automatically printed onto the
test substrate.
2.2 Software
A printer firmware is a control software
inside the printer. Its task is to interpret
the image sent to the printer into a series
of signals sent to different actuators such
as print head, paper feeders and stepper
motors [7].
When printing a coloured image the
firmware is given a CMYK-value for
every pixel that needs colouring from the
printer driver program. CMYK-values
define how the firmware uses cyan,
magenta, yellow and black inks to
formulate every coloured pixel.
Printer driver is the software that
communicates with the printer itself. Its
main task is to transform images and
texts into a form that the printer
understands. Commonly images are
presented in RGB colour space when
viewed from a computer. By default the
printer driver does some colour
management to define the best possible
CMYK values for the printing device.
This leads to situations where, for
example, pure RGB yellow is not given
as pure CMYK yellow to the printer and
thus the printed yellow colour is a
composition of ink from multiple ink
cartridges.
To be able to print different drug
solutions poured into the ink cartridges,
the printer is needed to be controlled in a
way that guarantees the printed solution
contains
pure ‘cyan’, ‘magenta’,
‘yellow’ or ‘black’ drug. For this purpose
drug printing software was created (Fig.
4). Its main objective is to produce
identical sets of drug doses on a paper
Fig. 4. The user interface of drug printing program.
Another reason for building software is to
demonstrate the whole process of printing
pharmaceuticals to a substrate in an
understandable way. Printing different
colours with a common graphic program
is not very informative. The ability to
change the actual effective drug and drug
dosing instead of different colours and
colour hues gives way better impression
of the process.
The concentration of the dose is
proportional to the colour intensity. If the
printer is set to use only yellow cartridge,
the colour intensity is the Y-value of the
CMYK-values. When low Y-value is set
for a certain area the printer jets droplets
in large intervals but when high Y-value
is set the jet interval is much lower and
the surface is visually more constant.
There are few parameters that are needed
to resolve, before it is possible to print
the needed amounts of riboflavin or
paracetamol on a substrate. A common
ink droplet ranges from few to ten
picoliters. But what is unknown is the
amount of solution the printer jets in one
square and how linearly the solution
consumption rise when printing test
pages with identical squares of the drug
solution but different drug concentration.
The exact amount of drug solution
printed on the substrate will be calculated
with
the
help
of
UV/VISspectrophotometer. When multiple test
sheets are analysed the results can be
parameterized for the printing software
and the used printer model. Then known
amounts of active pharmaceutical
ingredient can be printed on a substrate
from the software.
2.3 Testing method
A series of test samples were printed with
the software and HP Photosmart B010
printer. The yellow ink cartridge was
filled with the drug solution and other
cartridges were left empty. Riboflavin
sodium phosphate (RSP) (riboflavin 5’monophosphate sodium salt, Ph. Eur.,
Fluka Analytical, Sigma-Aldrich, France)
was used as water soluble active
pharmaceutical ingredient (API) of
orange colour. Glycerol (85%) was used
as moisturizer and viscosity modifier.
Purified water was used as a solvent.
Printing was done on two different
substrates: usual copy paper and
photocopy paper (Fig. 5). Different
printing quality settings were used to
analyse visually the linearity of the colour
hue when increasing the colour intensity
from 10 to 100.
Fig. 5. Printed RSP on photocopy paper.
Concentrations printed: 10, 25, 50, 75 and 100.
3
RESULTS
The visual analysis of the prints with the
Olympus BH-2 microscope showed a
clear difference in colour between 10, 25
and 50 colour intensities but no visual
difference in colour between 50, 75 and
100 colour intensities. Fig. 6 shows the
print quality with two different quality
settings and two different dose
intensities. The doses are printed on a
transparent film and the pictures are taken
from one corner of the printed squares.
2.4 Content analysis
Visual analysis was performed with
Olympus BH-2 microscope. Further
analysis was performed in water by using
UV-VIS
spectrophotometer
(Perkin
Elmer, Lambda 25, Germany). Each
printed area was cut out, put into 1.5 ml
of purified water, mixed and incubated
for 1 to 3 h. The absorbance values of the
solutions were measured at 267 nm.
Seven areas were used for each
concentration printed. In addition, blank
copy and photocopy paper samples were
used as reference material.
Fig. 6. Microscope images of different samples of
riboflavin-water-glycerol printed on an overhead film.
(A): 100 per cent dose with normal print quality. (B):
100 per cent dose with best printing quality. (C): 25 per
cent dose with normal print quality. (D): 25 per cent
dose with best printing quality.
The content of API in the copy and
photocopy paper was similar (Fig. 7).
The content analysis confirmed the
results obtained by visual observation of
the intensity of colour in the printed
areas. There was no linear correlation
between the expected printed amount and
actual content.
0,035
Content [mg]
0,030
0,025
0,020
0,015
0,010
Copy paper
0,005
Photo paper
0,000
0
25
50
75
100
Colour intensity [%]
Fig. 7. The content analysis of RSP printed areas. Data
are presented as mean ± SD (n = 7).
The results
with
the UV/VISspectrophotometer support initial visual
analysis. The printer does not increase the
amount of printed solution linearly when
using ‘normal print quality’-setting.
The low standard deviation of the tests
clearly showed that the solution
concentration was constant through the
different tests. This makes the dosing
very accurate after the printer and the
software are calibrated.
4
DISCUSSION
Initially piezoelectric based printers were
though to work better due to not-socomplicated liquid dynamics. They
enable the viscosity to be higher and
accommodate other than water-based
solutions [5]. However, Epson failed to
print anything. Sometimes the solution
did not even eject the cartridge into the
print head assembly. There are multiple
possibilities why they do not function but
one major issue could be liquid viscosity.
If the viscosity is too high, the
pharmaceutical ink is not able to flow
through the printer head. Then, if the
viscosity is too low, the pharmaceutical
ink will go through the print head even
when it is not supposed to do so. There is
no undisputed information why it is so
difficult to print with piezoelectric
printers. Therefore, more studies about
liquid viscosity and surface tension needs
to be made.
Even though thermal ink-jet proved to be
applicable technique
for printing
pharmaceuticals, commercial ink-jet
printers are not applicable straight out of
the box due to their fault sensitivity. Test
printer tended to malfunction, because the
drugs dried up so easily on the print head.
It would also be vital to get to the
firmware of the printer because it is
important to know that the printer prints
just
the
required
amount
of
pharmaceuticals. Also the reactions in the
solution caused by the thermal element
should be considered.
The tests proved that the amount of
printed
pharmaceutical
did
not
correspond to the amount it was supposed
to print with ‘normal print quality’setting. There are few possible reasons
for this: the printer still does some colour
corrections for the CMYK values or the
viscosity of the solution makes the
solution either jet in an uncontrollable
way or prevents the solution to jet
altogether. Also the used ‘normal print
quality’ prints a row of squares in one
sweep whereas using best quality results
in multiple sweeps per row. Visual
analysis indicated clearly that a single
sweep gave poor linearity of the different
doses. Using ‘best printing quality’
results in a significant improvement of
the linearity between the different dose
consentrations.
The tests proved that there is a lot of
potential in pharmaceutical printing and it
can change the way drugs are
manufactured and delivered in the future.
One possible scenario could, for example,
be that instead of making pills, the drug
companies could deliver different readyto-use formulations of their drug, which
could then be printed on demand in the
local pharmacy.
Although more research has to be made
in order to fully determine the
possibilities of drug printing, this
research gave insight of some of the
challenges that need to be faced before
making printable drugs as an alternative
solution to pill manufacturing.
5
REFERENCES
[1] Sandler, N., Määttänen, A., Ihalainen,
P., Kronberg, L., Meierjohann, A.,
Viitala, T., Peltonen, J. 2011. Inkjet
Printing of Drug Substances and Use of
Porous
Substrates-Towards
Individualized Dosing. Journal of
Pharmaceutical Sciences Volume 100,
Issue 8 (2011-08-01), pp. 3386-3395.
ISSN 1520-6017
[2] University Of Leeds News. 2010.
New research into safer drugs puts pills
through
the
printer.
Available: http://www.engineering.leeds.
ac.uk/faculty/news/2010/safer-drugsputs-pills-through-the-printer.shtml . (10.
3. 2012)
[3] MIT Tech Talk. 1997. Medicines are
‘printed’
into
pills.
Available: http://web.mit.edu/press/1997/
pills-0416.html . (10.3. 2012)
[4] Magdassi, S. 2010. Chemistry of
inkjet inks. World Scientific. ISBN: 978981-281-821-8.
[5] Wijshoff, H. 2010. The dynamics of
the piezo inkjet printhead operation.
Physics Reports vol. 49 pp. 77-177.
[6] Tyson, J. 2001. How inkjet printer
works.
Available: http://computer.howstuffwork
s.com/inkjet-printer.htm. (1.3.2012).
[7] Sen, A.K., Darabi, J. 2007. Droplet
ejection performance of a monolithic
thermal inkjet print head. Journal of
micromechanics and microengineering
Vol.
17.
DOI:10.1088/09601317/17/8/002
6
CORRESPONDING ADDRESS
Panu Kiviluoma, D.Sc. (Tech.), Post-doc
researcher
Aalto University School of Engineering
Department of Engineering Design and
Production
P.O.Box 14100, 00076 Aalto, Finland
Phone: +358 9 470 23558,
E-mail: panu.kiviluoma@aalto.fi
http://edp.aalto.fi/en/
7
ADDITIONAL
AUTHORS
DATA
ABOUT
Takala, Marko, B.Sc (Tech)
E-mail: marko.takala@aalto.fi
Helkiö, Henri, B.Sc (Tech)
E-mail: henri.helkio@aalto.fi
Sundholm, Joni, B.Sc (Tech)
E-mail: joni.sundholm@aalto.fi
Widmaier, Thomas, M.Sc (Tech)
E-mail: thomas.widmaier@aalto.fi
Kuosmanen, Petri, D.Sc. (Tech.), Prof.
E-mail: petri.kuosmanen@aalto.fi
Niklas Sandler, PhD (Pharm.), Prof.
niklas.sandler@abo.fi *
Natalja Genina, PhD (Pharm.)
natalja.genina@abo.com *
* Pharmaceutical Sciences
Department of Biosciences
Åbo Akademi University
Tykistökatu 6A
FI-20520 Turku, Finland