Full text of DOI:10.1557/JMR.2002.0231

RAPID COMMUNICATIONS
The purpose of this Rapid Communications section is to provide accelerated publication of important new results in the
fields regularly covered by Journal of Materials Research. Rapid Communications cannot exceed four printed pages in
length, including space allowed for title, figures, tables, references, and an abstract limited to about 100 words.
Rapid prototyping of micropatterned substrates using
conventional laser printers
Michael L. Branham
Department of Aerospace Engineering, Mechanics, and Engineering Science, University of Florida,
Gainesville, Florida 32611-6250, and Department of Chemistry and Biochemistry, Florida State
University, Tallahassee, Florida 32306-4390
Roger Tran-Son-Tay
Department of Aerospace Engineering, Mechanics, and Engineering Science, University of Florida,
Gainesville, Florida 32611-6250
Christopher Schoonover, Patrick S. Davis, and Susan D. Allen
Department of Chemistry and Biochemistry, Florida State University,
Tallahassee, Florida 32306-4390
Wei Shyy
Department of Aerospace Engineering, Mechanics, and Engineering Science, University of Florida,
Gainesville, Florida 32611-6250
(Received 15 November 2001; accepted 20 March 2002)
We demonstrated rapid prototyping of templates for replica molding using a
conventional laser printer. A polymer, polydimethylsiloxane, was cast directly on the
transparency templates to make the replicas. The templates and replicas were
characterized by scanning electron microscopy, profilometry, and optical microscopy.
Four patterns, including an Electronic Industries Association resolution test pattern,
were printed on transparencies at 600 dots per inch on a HP LaserJet 4M printer
(Hewlett-Packard, Palo Alto, CA). Optimal precision and clarity occurred between
intensity settings of 50–100. Mean pattern height/depth ranged from 8–13 ␮m, and
width was as small as a few tenths of a millimeter. Mean surface roughness of the
template patterns ranged from 1 to 4 ␮m on the top surface and from 5 to 10 nm on
the bare transparency surface. This method provides access to microfabricated patterns
for the broader research community without the need for sophisticated micromachining
facilities.
The interest in miniaturized systems for biochemical
analysis and biomedical research continues to grow rap-
idly. Devices commonly referred to as micro total ana-
lytical systems (␮TAS) have been reported for a diverse
range of applications, such as the determination of mono-
clonal antibodies,1 the determination of phosphate2 and
nitrite,3 and for high-speed DNA sequencing.4 The ap-
plication of similar devices to cell culture and cell me-
chanics5 is also expanding to include the geometric
control of cell shape,6 orientation, and gene expression,7
as well as the adhesion of neurons8 and growth cone
guidance.9 The devices themselves usually require pho-
tolithographic microfabrication techniques to produce
them since conventional machining will not provide suf-
ficiently small devices (colloquially referred to as chips).
A review by Madou provides information relating to all
aspects of microfabrication.10 Chip-based devices manu-
factured from glass are perhaps the most widely used in
part because of the versatility and chemical resistivity of
glass, the relatively straightforward fabrication, and the
ease of optical detection in glass devices. However, a
major problem frequently encountered by researchers
without specialized facilities is how to produce prototype
devices rapidly and at a realistic cost.
This paper describes a practical method for the fabri-
cation of templates, stamps, or molds using a laser printer
with resolution of 600 dots per inch (dpi) or greater and
solvent-free transparency film as a substrate (Fig. 1). The
overall time from a computer design to a micropatterned
surface is 14–24 h, most of which is for curing the poly-
mer. No access to spin coaters, ultraviolet radiation, or
tools normally found in the semiconductor process is
J. Mater. Res., Vol. 17, No. 7, Jul 2002
© 2002 Materials Research Society
1559
http://journals.cambridge.org
Downloaded: 30 Mar 2015
IP address: 35.8.191.249

Rapid Communications
Pattern height and aspect ratio is determined by the in-
tensity setting on the printer, which controls the pattern
charge on the drum. Pattern roughness is determined by
toner particle size and fusing.
Several micropatterned polydimethylsiloxane (PDMS)
substrates were produced and analyzed. Test patterns
were designed with computer graphics software (Claris
Draw 5.0, Claris Corp., Sunnyvale, CA) as shown in
Figs. 2(a)–2(c). An EIA resolution test pattern [Fig. 2(d)]
was also used to test pattern blur and continuity.16 The
pattern designs were then printed onto 3M (solvent free)
laser color printer film using a HP LaserJet 4M printer
(St. Paul, MN). PDMS was cast directly onto the trans-
parency template after being blown dry under nitrogen
gas for 5 min. The replicas were prepared from a mixture
9:1 vol% of Sylgard 184 elastomer and curing agent
(Dow Chemical, Midland, MI). The mixture was then
poured over the transparency templates in a glass dish
and cured at room temperature for 14 h under reduced
pressure to remove any air bubbles generated during
mixing. The stamps were carefully removed from the
master mold using a scalpel. The stamps were cleaned by
vortexing in distilled water and then ethanol (95%) for
10 min each and dried under a stream of nitrogen gas.
The transparency templates and PDMS replicas were
analyzed by several techniques. Bright field images
were captured using an Olympus BH-2 Epifluorescent
Microscope (Melville, NY) with a 10× objective and a
high-resolution digital camera. Scanning electron micro-
scope (SEM) images were obtained using a Hitachi
S-4000 FE-SEM (San Jose, CA). This instrument has a
maximum resolution of 1.5 nm, a magnification range of
18–300,000×, and accelerating voltage of 0.5–30 kV.
Surface topography of the ink transparencies was char-
acterized using a Tencor Alphastep 200 profilometer
(Mountain View, CA) at room temperature and a scan
rate of 0.02 nm/s. Atomic force microscope (AFM) scans
FIG. 1. Outline of the fabrication of micropatterned PDMS substrates.
On the replicas, recesses are shown face down (not drawn to scale).
required, significantly reducing the cost and complexity
of microengineering and micropatterned polymer sur-
faces. Laser11 and ink-jet printers12 are increasingly be-
ing used in production technologies for devices such as
biosensors13 and biomaterials.14 The pattern resolution
of the output produced by laser printers is dependent on
toner particle size distribution and many other factors,
e.g., the template transfer efficiency, melting, spreading,
and of course the surface properties of the paper or trans-
parencies used. For most printers, this corresponds to the
maximum number of dots per inch.
Two laser printing processes are used commercially.
The standard for most desktop printers (including the HP
LaserJet 4M, Hewlett Packard, Palo Alto, CA) is the
“write black” process, although some printers and most
copiers use the “write white” process.15 In the “write
black” process, every spot illuminated by the laser on the
print drum will be black on the print media. At each point
the laser or light emitting diode array illuminates the
photoconductive drum, the electrostatic charge is
changed from negative to positive. As the drum contin-
ues to rotate it passes the toner cartridge that contains
negatively charged finely divided toner powder, which is
electrostatically attracted to the positive regions on the
drum. As the paper in contact with the drum rotates,
the toner is transferred to it. The paper then passes be-
tween the fusing rollers that are heated to soften low-
molecular polymers in the toner and bind it to the paper.
FIG. 2. Template designs: (a) ring, (b) low-density dots, (c) high-
density dots, and (d) EIA resolution test pattern.16
1560
J. Mater. Res., Vol. 17, No. 7, Jul 2002
http://journals.cambridge.org
Downloaded: 30 Mar 2015
IP address: 35.8.191.249

Rapid Communications
FIG. 3. Effect of decreasing print darkness on the print quality of EIA resolution pattern: (a) highest darkness (printer setting 50); (b and c) optimal
precision and image clarity occurs at printer intensity settings of 50 and 100 respectively; (d) pattern begins to loose precision (printer setting 150);
(e) lowest printer darkness setting (printer setting 200).
using a Digital Multimode 3000 (Digital Instruments,
Santa Barbara, CA) were also performed and confirmed
the profilometer surface roughness measurements. The
transparency templates were stored in a laminar flow
hood and blown clean with nitrogen gas prior to profilo-
metric analysis.
[Figs. 5(b) and 5(c)]. The edges of the ring pattern in
Fig. 5(b) were found to be irregular in shape and include
individual toner particles 3–8 ␮m in diameter.
The minimum particle size and pattern roughness is
determined from the toner particle size and/or particle
fusion during the heating cycle. In Fig. 6(a), a profile of
The effect of print darkness settings on print quality is
illustrated in Fig. 3. Micrograph 3(a) shows that at a
higher print darkness setting (higher toner concentra-
tion), pattern blur occurs as larger amounts of toner are
added per unit surface area. At intermediate print dark-
ness settings 50 [Fig. 3(b)] and 100 [Fig. 3(c)] optimal
precision and image clarity occurs. As observed in the
image [Fig. 3(d)] at a low print darkness setting (150),
the pattern begins to lose precision and break up as too
little toner is used per unit surface area. At a setting of
200 [Fig. 3(e)] the clarity of the pattern is further
degraded.
A SEM image of the ring pattern from Fig. 2(a) is
shown in Fig. 4(a). It consists of a 600-␮m-wide toner
band with irregular morphology on the surface and at the
inner and outer edges of the band. The mean height of
the pattern (11 ␮m) can be controlled only within a nar-
row range by the printer darkness settings on this printer
without affecting pattern resolution. The mean surface
roughness of the ring pattern is 1.2 ␮m, as determined by
surface profilometry [Fig. 4(b)]. The top surface Fig. 5(a)
of the ring pattern contains pockets and cavities but is, as
would be expected, more homogeneous than the edges
FIG. 4. Control ring dimensions. (a) Ring consists of a toner band
with irregular morphology on the top surface and at the inner and outer
edges of the band. (b) Ring profile shows band height (11.99 ␮m) and
width (600 ␮m.) Mean surface roughness of the band was 1.180 ␮m.
J. Mater. Res., Vol. 17, No. 7, Jul 2002
1561
http://journals.cambridge.org
Downloaded: 30 Mar 2015
IP address: 35.8.191.249

Rapid Communications
FIG. 5. SEM of ring pattern: (a) top surface; (b) edge; (c) top surface near edge; (d) pattern edge.
the “low density dots” template [pattern Fig. 2(b)] shows
culture substrata. For example, we recently investigated
the influence of substrate surface structure on epithelial
cell adhesion and spreading using the described tech-
nique.17 We were able to show that cellular responses to
surface topography depend both on cell phenotype and
microtexture of the surface in the absence of extracellular
matrix proteins. In other reports, a high-resolution printer
was shown capable of generating 20 ␮m patterns with
good quality.11,18,19 Such resolution18 is sufficient for
many applications in biotechnology and tissue engineer-
ing where the ability to fabricate confined areas of mi-
cron scale surface roughness is desired. On the other
hand, the major advantages of the method are its very
low cost, versatility, ease of use, rapid turnaround time,
and pattern resolution and roughness suitable for the pro-
duction of substrates for cell growth and bioassays.
a surface topography that consists of a lattice structure
with a average height of 7 ␮m, width of 380 ␮m, a pat-
tern surface roughness of 2.5 ␮m, and a transparency
surface roughness of 10 nm. These templates produced in
the replicas an array of microwells [Fig. 6(b)] separated
by 290 ␮m borders of unpatterned PDMS with nanoscale
surface roughness similar to the transparency films. The
“high density dot” templates [Fig. 2(c)] and replicas
were very similar except that slightly deeper micro-
wells were produced (8 ␮m) with mean bottom surface
roughness of 2 ␮m and top (unpatterned surface) rough-
ness of 5 nm.
In conclusion, we demonstrated a practical method for
rapid production of replica molding templates using laser
printer transparencies as templates for soft lithography.
This method requires no specialized microlithographic
equipment or cleanroom facilities. The PDMS replicas
directly cast on the laser printed templates accurately
reproduce the laser printer resolution and pattern surface
topography. The major disadvantages of the present
method are a poor resolution and a lack of control over
the depth of the patterns. However, this is a useful tech-
nique to generate microstructures with a resolution ap-
propriate for use in bioanalytical devices or as cell
REFERENCES
1. N. Chiem and D.J. Harrison, Anal. Chem. 69, 373 (1997).
2. G.N. Doku and S.J. Haswell, Anal. Chim. Acta 382, 1 (1999).
3. G.M. Greenway, S.J. Haswell, and P.H. Petsul, Anal. Chim. Acta
387, 1 (1999).
4. S.R. Liu, Y.N. Shi, W.W. Ja, and R.A. Mathies, Anal. Chem. 71,
566 (1999).
5. M.M. Peel and P. Dimilla, Annals of Biomed. Eng. 27, 236 (1999).
6. C.S. Chen, M. Mrksich, S. Huang, G.M. Whitesides, and
D.E. Ingber, Science 276, 1425 (1997).
7. Y. Ito, Biomaterials 20, 2333 (1999).
8. A.M. Turner, N. Dowell, S.W. Turner, L. Kam, M. Isaacson,
J.N. Turner, H.G. Craighead, and W. Shain, J. Biomed. Mater.
Res. 51, 430 (2000).
9. H.C. Tai and H.M. Buettner, Biotechnol. Prog. 14, 364 (1998).
10. M. Madou and J. Florkey, Chem. Rev. 100, 2679 (2000).
11. T. Deng, H. Wu, S.T. Brittain, and G.M. Whitesides, Anal. Chem.
72, 3176 (2000).
12. B.Y. Tay and M.J. Edirisinghe, J. Mater. Res. 16, 373 (2001).
13. M. Mott, J.H. Song, and J.R.G. Evans, J. Am. Ceram. Soc. 82,
1653 (1999).
14. J.D. Newman, A.P.F. Turner, and G. Marrazza, Anal. Chim. Acta
262, 13 (1992).
15. R. White, in How Computers Work, 6th ed. (Que Publishing, In-
dianapolis, IN, 2002), pp. 349–386.
16. http://www.bealecorner.com/trv900/respat/EIA1956.pdf.
17. M.L. Branham, S.C. Glover, R.V. Benya, S.D. Allen, W. Shyy,
and R. Tran-Son-Tay (2002, unpublished).
18. D. Qin, Y. Xia, and G.M. Whitesides, Adv. Mater. 8, 917 (1996).
19. A. Tan, K. Rodgers, J.P. Murrihy, C. O’Mathuna, and J.D. Glennon,
Lab on a Chip 7–9, (2001).
FIG. 6. Bright-field micrograph of the PDMS replica taken from the
“low-density dots” pattern template in Fig. 2(b). The resulting “mi-
crowells” resemble parallelepipeds are approximately 12 ␮m deep and
300–400 ␮m in length.
1562
J. Mater. Res., Vol. 17, No. 7, Jul 2002
http://journals.cambridge.org
Downloaded: 30 Mar 2015
IP address: 35.8.191.249