Protocol

Adhesive Micropatterns for Cells: A Microcontact Printing Protocol

¹ Laboratoire Biopuces, iRTSV, DSV, CEA, 38054, Grenoble, France
² Systems Cell Biology of Cell Division and Cell Polarity, UMR144, Institut Curie, CNRS, Paris, France

³Corresponding authors (manuel.thery{at}cea.fr; matthieu.piel{at}curie.fr)

INTRODUCTION

This protocol describes a simple, fast, and efficient method for making adhesive micropatterns that can be used to control individual cell shape and adhesion patterns. It is based on the use of an elastomeric stamp containing microfeatures to print proteins on the substrate of choice. The process can be subdivided into three parts. First, a silicon master is fabricated, which contains the microfeatures of interest. Once fabricated, the master can be used multiple times to make stamps. Masters with customized patterns can also be purchased commercially. Second, a polydimethylsiloxane (PDMS) stamp is fabricated. Unlike fabrication of the master, this step can be performed without specialized equipment. The PDMS stamp is inked with extracellular matrix proteins. Proteins are printed on a substrate (e.g., a tissue culture polystyrene dish or a glass coverslip covered with a thin layer of polystyrene). The nonprinted areas are back-filled with poly-L-lysine-polyethylene glycol, which renders them resistant to cell adhesion. The production of these micropatterned substrates can be completed in <2 h. The third and final portion of the protocol describes the deposition of cells onto the micropatterned substrate.

RELATED INFORMATION

Movie 1 illustrates many of the steps for performing microcontact printing as described in this protocol.

View as movie: (111K): [in QuickTime]

Movie 1. A step-by-step illustration of preparation of an elastomeric stamp containing microfeatures, preparation of the substrate, and cell growth on the microfeatures. (Movie created by Nicolas Carpi.)

MATERIALS

Reagents

Acetone

AP6000 adhesion promoter (Dow Chemical)

Cells of choice

Chlorotrimethylsilane

Culture medium, complete (DMEM or DMEM-F12 containing 10% fetal calf serum [FCS] and 1% penicillin/streptomycin)

Developer (Clariant AZ 400K)

Dulbecco’s phosphate-buffered saline (with calcium and magnesium; with phenol red) (PBS)

Ethanol

Fetal calf serum

Fibronectin (Sigma-Aldrich F1141) (50 µg/mL in PBS)

Fixative (optional; see Step 33)

HEPES (10 mM, pH 7.4)

Photoresist (Clariant AZ 4562)

PLL(20)-g-[3,5]-PEG(2) (PLL-g-PEG; SurfaceSolutions)

Polydimethylsiloxane (PDMS) (Sylgard 184; Dow Corning)

Polystyrene (Acros Organics 178890250)

Toluene

Trypsin-EDTA or Versen EDTA

Equipment

Adhesive tape or sonicator (see Step 12)

Argon (optional; see Step 25)

Beaker, plastic

Beaker, small

Centrifuge, tabletop

Coverslips, glass (24 x 24 mm) or tissue culture polystyrene dishes

See the section on Substrate Preparation.

Flasks, cell culture (75-cm²)

Forceps

Hot plate or oven preset to 100°C

Incubator, humidified, 5% CO₂

Laminar flow hood

Mask aligner (e.g., Süss MicroTec MA750) with a vacuum mask holder

A simpler and much cheaper alternative can be made from a UV lamp (Delolux 03S, 400 W) mounted with a 310- to 460-nm band pass filter and a custom-made vacuum mask holder.

Microscope, inverted

Nitrogen gas (optional; see Steps 1, 7)

Optical mask (Delta Mask)

The optical mask is a quartz plate coated with a chromium layer that has been etched to obtain the desired patterns of opaque and transparent regions. A file with the desired micropatterns must be provided to the manufacturer. If the micropattern features are greater than several tens of micrometers in their smallest dimension, a high-resolution plastic transparency can be used instead. The plastic transparency is less expensive but has limited spatial resolution.

Ovens preset to 60°C, 70°C

Parafilm

Petri dish sealed with Parafilm (for storing the resist master and stamps)

Photoresist master

If a master is not produced in-house, then purchase one from a commercial manufacturer (e.g., Biotray or Advance Reproductions Corporation). If the experiment does not require pattern resolution better than 10 µm, the mask may be ordered as an inexpensive transparency film from a number of companies (e.g., CAD/Art Services).

Pipettes

Plasma cleaner or UV/ozone cleaner (<200 nm) (see Step 17)

Scalpel

Silicon wafer, 100 mm (e.g., MEMC Electronic Materials) or glass (borosilicate) wafer, 100 x 100 x 0.175 mm (Schott D263T)

The glass wafers are thinner but much cheaper.

Spin coater (e.g., Laurell Technologies Corporation WS-400-6NPP-LITE)

Spoon, fork, or pipette, plastic (see Step 8)

Tissue (optional; see Step 15)

Tubes, 50-mL culture

Tweezers

Ultrasonic bath

Vacuum desiccator

Video recorder (optional; see Step 33)

METHOD

Fabrication of the Master

This part requires specialized equipment and training; thus, it is best to fabricate the master in collaboration with a nanofabrication facility. However, a relatively simple lithography facility can be set up under a chemical hood. Fabrication of a master need only be done once, because it can be reused indefinitely to produce the elastomeric stamps. The desired photoresist master can also be purchased from companies that fabricate custom-made microstructured masters on demand.

This protocol provides guidelines for the various parameters needed to fabricate a photoresist master. Details of this photolithography protocol, however, must be adapted to the specific needs of each laboratory (e.g., size of the microfeatures, distance between features, incubation times and temperatures). Note that the production of a master with microfeatures ranging from 1 µm to several hundred micrometers is a routine procedure in any microfabrication facility. The example in this protocol is best suited for features with a minimum dimension of 3-10 µm (but up to several tens of micrometers is okay) and a distance between features of <50 µm. Figure 1 provides an illustration of the steps required to fabricate a photoresist master.

View larger version (13K): [in this window] [in a new window]

Figure 1. The steps involved in producing molds by photolithography. (Circled numbers correspond to numbered steps in the Method.)

Photoresist Layer Coating

1. Clean the silicon or the glass wafer in an ultrasonic bath for 10 min in acetone. Dry it in a flow of filtered air or nitrogen.

2. Spin coat the wafer as follows:

i. Place the wafer on a spin coater.

ii. Cover the central half of the wafer’s surface with photoresist.

iii. Spin coat the photoresist onto the wafer at 500 rpm for 30 sec.

iv. Spin coat the photoresist onto the wafer at 2000 rpm for 30 sec.

v. Dry the wafer under a hood for 10 min.
The photoresist will be ~5 µm in height, which is suitable for 4- to 10-µm-wide features. When using a different photoresist for smaller features, this spin-coating step will have to be modified.

3. Bake the photoresist layer onto the wafer using either a hot plate or an oven for 1 min at 100°C. Cool the wafer on the bench for ~15 min until it returns to room temperature.

Photoresist Lithography

4. Place the photoresist layer and the optical mask into contact with one another on the mask aligner or on the custom-made vacuum mask holder. Illuminate with the UV lamp (UV power 45 mW/cm² at 365 nm) for 10 sec.

5. Dilute the stock developer using one part developer and four parts H₂O. Develop the photoresist in diluted developer for 90-120 sec. Rinse the wafer in a distilled H₂O bath, which stops the development process. Dry the resist master with a flow of filtered air.
Because the photoresist used in this protocol is "positive," the illuminated areas are dissolved in the developer, creating the microfeatures (a negative photoresist would yield the opposite result).
Do not post-bake the master. During the development step, microfeatures can be observed through a microscope using reflected light. If the patterns have not been completely revealed, extend the development time. The UV illumination time, the developer concentration, and the development time in Steps 4 and 5 depend upon the resist used and on its thickness. It may be necessary to adjust these.

Photoresist Master Surface Coating

6. Silanize the resist master in order to prevent it from adhering to the PDMS during stamp fabrication.

i. Place the resist master into a vacuum desiccator along with a small beaker containing a few drops of chlorotrimethylsilane.

ii. Turn on the vacuum for 30 min to induce the formation of chlorotrimethylsilane vapors.

iii. Turn off the vacuum, keep the desiccator sealed, and let the resist master silanize overnight.

7. Place the resist master for 30 min in a 70°C oven to complete the silanization.
Store the resist master away from dust (a Petri dish sealed with Parafilm works well). Handle the master with care. Be careful not to touch the patterned area with anything that could scratch and destroy it (e.g., use forceps). Clean it only with pressurized air or nitrogen.

Micropatterned Substrate Fabrication

Fabrication of the elastomeric stamp and microcontact printing of the micropattern can be performed in any biology laboratory without the need for specialized equipment. Two substrates for micropatterning are presented: the glass coverslip, which ensures the best optical quality, and the tissue culture treated polystyrene dish. Because in most cases protein adhesion is better on oxidized polystyrene than on glass, glass coverslips must be coated with a thin layer of polystyrene before protein and cells are applied. On bare glass, cells have been observed to rip adherent proteins, like fibronectin, off the glass.

Stamp Fabrication
See Figure 2 for an overview of PDMS stamp production.

View larger version (9K): [in this window] [in a new window]

Figure 2. PDMS stamp production.

8. Combine PDMS and the curing agent (both are included in the Sylgard 184 kit) in a 10:1 ratio in a plastic beaker and mix thoroughly using a plastic spoon, fork, or pipette. This step will generate bubbles in the elastomer.

9. Degas the PDMS mixture under vacuum to remove all of the air bubbles. Alternatively, centrifuge the mixture for 5 min at 3000 rpm.
Avoid an excessively strong vacuum, which can cause the mixture to spill out of the beaker.

10. Cast a 5-mm thick layer of degassed PDMS mixture onto the resist master. Cure it for 2 h in a 60°C oven.

11. Gently peel off the PDMS layer.
The microstructures can be seen at the surface of the PDMS layer when viewed through a microscope.
See Troubleshooting.

12. Using a microscope and transmitted light, locate the region of interest containing the desired micropattern geometries. With a scalpel, excise 1-cm² stamps. Larger regions can be excised, but stamps larger than 1 cm² are not easy to handle.
Store stamps in a closed container away from dust (a Petri dish sealed with Parafilm works well). Stamps can be kept for months before being used. Adhesive tape can be used to clean dusty stamps, although, for small features, it is better to use stamps that have been cleaned by sonication.

Substrate Preparation
See Figure 3 for an overview of preparation of a polystyrene-coated glass coverslip.

View larger version (10K): [in this window] [in a new window]

Figure 3. Preparation of a polystyrene-coated glass coverslip.

The substrate described here is a glass coverslip coated with polystyrene. A tissue culture polystyrene dish can also be used. A short plasma or UV/ozone activation step (if the equipment is available) will enhance protein adhesion and cell attachment (Van Kooten et al. 2004). Proteins can be covalently bound to one another on polystyrene using 1-ethyl-3-(3-dimethyl aminopropyl) carbodiimide (EDC) and N-hydroxysuccinimide (NHS) (Grabarek and Gergely 1990). For a useful handbook on conjugation techniques, see Hermanson (2008).

13. Wash the glass coverslip with ethanol. Dry the coverslip with filtered air or let it dry in the hood.
Sonicate the coverslip in ethanol to optimize dust removal.

14. Spin coat the coverslip as follows:

i. Place the coverslip on the spin coater.

ii. Deposit a few drops of the adhesion promoter AP6000 onto the glass.

iii. Spin coat the coverslip for 10 sec at 100 rpm (to cover the coverslip), then for 20 sec at 1000 rpm (to form a thin layer).

iv. Dry the AP6000 layer.
The adhesion promoter ensures that the polystyrene will attach to the glass.

15. Coat with polystyrene as follows:

i. Prepare a solution of 2% polystyrene solution in toluene (2%) and place a few drops of it on the AP6000-treated coverslip.

ii. Spin coat for 10 sec at 100 rpm (to coat the coverslip), then for 20 sec at 1000 rpm (to form a thin layer of polystyrene).

iii. After the spin coating, dry the polystyrene layer.
If a spin coater is not available for use in Steps 14 and 15, spread the drops manually by handling the coverslip with tweezers and displacing the drop with an orbital rotation of the coverslip. After coating the coverslip, remove excess liquid with a tissue. If this alternative method is used, reduce the polystyrene to 0.5% so that the polystyrene layer will not be too thick.
Glass coverslips covered with polystyrene can be stored at room temperature for days. Oxidation of the substrate (Step 17) should be carried out immediately before the substrate is stamped (Step 20).

Microcontact Printing
See Figure 4 for an overview of stamping fibronectin and backfilling with PLL-g-PEG.

View larger version (16K): [in this window] [in a new window]

Figure 4. Stamping fibronectin and backfilling with PLL-g-PEG.

16. Place a 20-µL drop of 50 µg/mL fibronectin in PBS onto the microstructured surface of a 1-cm² stamp. Spread the fibronectin solution across the stamp surface with the tip of the pipette by moving it toward each corner of the stamp. Leave the inked stamp for ~30 min to allow the fibronectin to adsorb to the PDMS.
Spreading the fibronectin is necessary because the PDMS is hydrophobic. The surface of the stamp can be gently touched without damaging it.

17. Transfer the polystyrene-coated coverslip (from Step 15) into a plasma cleaner. Turn on the pump to create a vacuum in the chamber. Allow a low flow of oxygen into the chamber and apply an oscillating electric field at 30 W for 10 sec.
Tissue culture polystyrene dish substrates can be printed on without plasma oxidation, but plasma treatment (as described here) significantly improves printing and passivation efficiencies. Polystyrene Petri dishes (not activated for tissue culture) can also be used, provided they are activated by plasma treatment.
Stamp oxidized substrates (Step 20) within 15 min of activating them.
Substrate activation can also be performed in a UV/ozone cleaner ( < 200 nm).

18. Aspirate off the fibronectin drop and immediately add a large drop of PBS to the surface before it dries. Repeat the PBS wash two more times to remove all of the unadsorbed fibronectin.

19. Remove the PBS drop and let the stamp dry in the airflow of a hood (it will take from a few seconds up to 1 min). The stamp is ready for printing when the surface appears dry while looking at light reflecting from it.
See Troubleshooting.

20. Grasp the dry stamp with tweezers, invert the stamp, and place the microstructured surface in contact with the activated substrate (from Step 17). Briefly apply gentle pressure with the tweezers to ensure good contact between the stamp and the substrate. Leave the stamp in contact with the substrate for 1 min.
This step is absolutely critical for success and requires practice to achieve satisfactory results.

21. Remove the stamp and immerse it into H₂O in a 50-mL culture tube.

22. Prepare 0.1 mg/mL PLL-g-PEG solution in 10 mM HEPES (pH 7.4). Immerse the printed substrate in this solution for 30 min. To minimize the amount of PLL-g-PEG used, place a 100-µL drop on the substrate and cover it with a piece of Parafilm.
The PLL-g-PEG renders the nonprinted areas resistant to cell adhesion.
See Troubleshooting.

23. During PLL-g-PEG adsorption, clean the stamp. Heat the culture tube containing the stamp at 60°C and sonicate it in an ultrasonic bath for 15 min. Then sonicate the stamp in ethanol for 15 min.
This cleaning procedure might not remove all the protein from the stamp. If another type of protein is to be stamped, it is better to use a new stamp.

24. Dry the clean stamp under a hood for 1 h. Return the stamp to its storage container.

25. Wash the PLL-g-PEG-grafted substrate with PBS for 2 min and then for 10 min.
The substrate is now ready for cell deposition. If not used immediately, it can be stored immersed in PBS or dried and stored under argon at 4°C for up to 1 wk.

Cell Deposition

Micropatterned substrates prepared using this protocol have been used successfully with the following cells: HeLa-B, RPE1, MCF10A, MCF7, NIH3T3, HepaRG, MDCK, and human mesenchymal stem cells, as well as dendritic cells derived from murine bone marrow.

26. Wash adherent cells in PBS. Detach the cells with trypsin-EDTA (for 5-10 min, depending on the dilution used) or Versen EDTA.

27. Add complete culture medium to the flask and collect the cells by centrifugation for 3 min at 1500 rpm.

28. Remove the supernatant and resuspend the cells in complete culture medium at 50,000 cells/mL.

29. Add the cell solution to the micropatterned substrate (from Step 25). The final cell density should be ~10,000 cells/cm². Place the cell-covered substrate into a tissue culture incubator.
The incubation time depends upon the cell line (e.g., 15-30 min for RPE1 and 30-60 min for HeLa-B).

30. Check the cells under a microscope to determine if a sufficiently large proportion of cells have attached to the micropatterns.

31. Remove unattached cells with a gentle flow of medium added to one side of the dish and aspirated from the other side.
Keep the cells moist; do not allow them to become exposed to the air. The physicochemical properties of PEG can cause dewetting, leading the attached cells to quickly dry out.

32. Return the attached cells to the incubator. Let them spread fully (1-5 h, depending on the cell type).

33. Fix or video-record the cells.
See Figure 5 for examples of cells attached to micropatterns.

View larger version (47K): [in this window] [in a new window]

Figure 5. Images of cells on micropatterns. (Top) Fluorescently labeled fibronectin, (middle) HeLa cells under phase contrast, (bottom) overlay. (Right) Enlargements (2.5X) of cells shown in the images on the left. Scale bar, 100 µm.

See Troubleshooting.

TROUBLESHOOTING

Problem: Separating the PDMS layer from the photoresist master destroys the master.

[Step 11]

Solution: Consider the following:

1. Curing the PDMS mixture (Step 10) for more than 2 h at 60°C can induce PDMS binding to the photoresist, which weakens the resist layer and results in destruction of the photoresist master when the two layers are separated. Avoid curing for more than 2 h.

2. Silanization of the resist master (Steps 6 and 7) might have failed. This will result in strong binding of the cured PDMS to the master. Before making a new master, check on a blank one that the silane used for the coating is still functional. Silanes tend to degrade, and the silane solution should be renewed regularly.

Problem: Salt crystals form on the surface of the elastomeric stamp as the PBS dries.

[Step 19]

Solution: This may happen if an excess of PBS is left to dry on the stamp. Avoid this by washing the stamp once with H₂O before drying.

Problem: The polystyrene layer detaches and floats at the surface when the coverslip is placed in aqueous solution.

[Step 22]

Solution: Consider the following:

1. Treat the coverslip with adhesion promoter (Step 14) prior to applying the polystyrene.

2. Affix the polystyrene to the coverslip by placing the polystyrene-coated coverslip under a UV lamp for 5 min at = 185 nm and 4 x 60 W (Yan and Harnish 2003).

Problem: Cells do not grow uniformly and/or they grow significantly beyond the boundaries of the microfeatures.

[Step 33]

Solution: Consider the following when stamping the substrate in Step 20:

1. Insufficient pressure might result in poor printing. Apply pressure with the tweezers both uniformly and for a longer period of time.

2. Excessive pressure will induce sagging of the stamp, causing proteins to print outside of the features. Apply pressure more gently.

3. Do not move the stamp once it has touched the surface. Both application and removal of the stamp should be done with precision and great care.

DISCUSSION

Microfabrication techniques as applied to cell biology now have a long and successful history (see Folch and Toner [2000] and Whitesides et al. [2001] for reviews of many micropatterning techniques developed from the 1970s to the 1990s). Accompanying the rapid development of biological applications (in cell biology, tissue engineering, cell co-cultures, bioassays, and biosensors, among other fields), there has been a huge burst of technical papers in the last 10 years that expand the utility of micropatterning techniques to new substrates (glass, plastics, hydrogels, and elastomers), additional molecules, cell types, and into three dimensions.

There are many alternative techniques, but five main processes dominate the field of micropatterning:

1. Photolithography and lift-off (as well as other stencil-type methods)

2. Laser/electron beam etching

3. Micro- and nanoprinting

4. UV-based chemistry

5. Microcontact printing

Each method has drawbacks and advantages, so that the method chosen will depend upon the application. In a biology laboratory, independence from the expense and complexity of a specialized microfabrication facility may be a priority. Thus, photolithography and other stencil-type methods should be avoided. Although these methods provide patterns with very good spatial resolution and the methods are quite versatile, each substrate has to be made in a clean room using highly toxic chemicals. Laser beam and electron beam etching are rather easy to implement, provided that a microscopy facility is available with a dedicated microfabrication microscope. Although the method is versatile (the size and shape of features can be readily altered, and multiple patterns of different proteins can be generated), it is a slow method, producing low numbers of substrates having only very small areas covered with micropatterns. The same is true for nanoprinting, which is versatile but slow. Microprinting using regular printers requires very slow printing rates to achieve satisfactory spatial resolution.

Two methods are left for biologists wanting to do simple micropatterning: microcontact printing and UV-based chemistry. UV-based techniques often require special surface chemistry and photomasks, which can be expensive when very high resolution is required (at low resolution, only a transparency is needed). Both UV-based techniques and microcontact printing require specialized software to design the masks. For a recent contribution to UV-based techniques for two-dimensional surface micropatterning, see Azioune et al. (2009), which includes an introduction that reviews the field.

As shown in this protocol, with microcontact printing, once a photoresist master (a mold) is available to produce the stamps, no special equipment and no special chemistry is needed to produce the patterns. In addition, microcontact printing is well suited for generating features to control cell adhesion geometry and cell shape (i.e., patterns having minimal dimensions of several micrometers). It is also easy to implement in a cell biology laboratory. Microcontact printing can be applied to many types of substrates, although the quality of the patterns produced depends critically upon the quality of the contact between the stamp and the substrate. Microcontact printing is usually used to transfer molecules to a substrate, but it can also be used for etching (Kandere-Grzybowska et al. 2005). The transferred molecule can be used to bind other molecules, such as cell adhesion molecules, or can itself be a cell adhesion molecule. Alternatively, the transferred molecule can be a protein- or cell-repelling molecule, like polyethylene oxide. Microcontact printing, therefore, offers a large spectrum of potential variations.

To optimize the transfer of molecules from the stamp to the surface, it was proposed in the 1990s to use thiol chemistry on gold (a recent version of this protocol was recently released by Ostuni et al. [2009]). This very efficient method became a standard in the field; however, it requires gold-covered substrates, which are rare among substrates of interest to biologists. In addition, it is not optimal for low-light live-cell fluorescent microscopy. Direct microcontact printing on regular cell culture substrates is a relatively simple and versatile alternative method. A variety of surface treatments can be applied to the regions of the substrate that should repel cells or protein regions. A variety of other treatments can be used to optimally bind cell adhesion molecules. Options include simple adsorption, covalent binding, electrostatic interactions and hydrophobic/hydrophilic interactions. These choices will be guided by the types of cells used (and consequently the type of adhesion molecule used), the size of the patterns, and the time during which cells should be kept on the patterns. Some very simple techniques (e.g., direct patterning on bare glass without any backfilling with repellent molecule) will work well with cells that do not bind on bare glass and do not pull too strongly on their adhesion molecules. For other cells, very good repellent molecules should be used, and very strong binding of the adhesion molecules to the substrate is required. For a discussion of some of these issues, see Fink et al. (2007). Direct microcontact printing works with many cell types, keeping them confined for several days (although not weeks). PLL-g-PEG is a cell/protein-repellent molecule that binds readily and with strong affinity to glass and activated polystyrene, by electrostatic interaction of poly-L-lysine chains.

REFERENCES

Azioune A, Storch M, Bornens M, Théry M, Piel M. 2009. Simple and rapid process for single cell micro-patterning. Lab Chip 9: 1640–1642.[Medline]

Fink J, Théry M, Azioune A, Dupont R, Chatelain F, Bornens M, Piel M. 2007. Comparative study and improvement of current cell micro-patterning techniques. Lab Chip 7: 672–680.[Medline]

Folch A, Toner M. 2000. Microengineering of cellular interactions. Annu Rev Biomed Eng 2: 227–256.[Medline]

Grabarek Z, Gergely J. 1990. Zero-length crosslinking procedure with the use of active esters. Anal Biochem 185: 131–135.[Medline]

Hermanson GT. 2008. Bioconjugate Techniques, 2nd ed. Academic Press, Burlington, MA.

Kandere-Grzybowska K, Campbell C, Komarova Y, Grzybowski BA, Borisy GG. 2005. Molecular dynamics imaging in micropatterned living cells. Nat Methods 2: 739–741.[Medline]

Ostuni E, Whitesides GM, Ingber DE, Chen CS. 2009. Using self-assembled monolayers to pattern ECM proteins and cells on substrates. Methods Mol Biol 522: 183–194.[Medline]

Van Kooten TG, Spijker HT, Busscher HJ. 2004. Plasma-treated polystyrene surfaces: Model surfaces for studying cell-biomaterial interactions. Biomaterials 25: 1735–1747.[Medline]

Whitesides GM, Ostuni E, Takayama S, Jiang X, Ingber DE. 2001. Soft lithography in biology and biochemistry. Annu Rev Biomed Eng 3: 335–373.[Medline]

Yan M, Harnish B. 2003. A simple method for the attachment of polymer films on solid substrates. Advanced Materials 15: 244–248.