I hope this page will eventually contain all the known data about the SU-8 photoresist. Thus, I need your help! Send
me all your data number 16525, and I will include it in this page, with credits, of course!
The SU-8 is a negative, epoxy-type, near-UV photoresist based on EPON SU-8 epoxy resin (from Shell Chemical) that has been originally developed, and patented (US Patent No. 4882245 (1989) and others) by IBM. This photoresist can be as thick as 2 mm and aspect ratio >20 have been demonstrated with a standard contact lithography equipment. These astounding results are due to the low optical absorption in the UV range which only limits the thickness to 2 mm for the 365nm-wavelength where the photo-resist is the most sensitive (i.e., for this thickness 100% absorption occurs).
Of course LIGA still yield better results but low-cost application will undoubtedly benefit from this resist that is well suited for acting as a mold for electroplating because of its relatively high thermal stability (Tg>200C for the cross-linked (i.e., exposed) resist).
Two companies have now bought a license from IBM to sell the photoresist :
MicroChem Corp. (previously named Microlithography Chemical Corp.), 1254 Chestnut Street, Newton, MA 02164-1418, USA, Tel: +1 617 965-5511, Fax: +1 617 965-5818 under the name NANO
SU-8 ### with different viscosities (SU-8 5; SU-8 10; SU-8 25; SU-8 50; SU-8 100),
SOTEC MICROSYSTEM 11 avenue des Baumettes, 1020 RENENS, Switzerland, Tel: +41 21 693 67 51, +41 21 635 95 34, Fax: +41 21 635 94 41 under the name SM10#0 with different viscosities
(SM1040; SM 1060; SM1070) and provide also a version with low coefficient of thermal expansion (SM20#0).
In France, a representative from MCC is Chimie Tech Services
(Claude Denis, Annabel Nerozzi), 7, rue Marcelin Berthelot - Zone Industrielle, 92762 Antony Cedex, Tel: +33 01 55 59 55 75 FAX: +33 01 55 59 55 90
Using the litterature reference, the following tables, based on an original compilation from Loren St Clair,
gives some properties of the resist. We begin with the known mechanical properties:
Table 1 : Mechanical properties of the SU-8 photoresist.
Characteristics
Value
Conditions
Reference
Modulus of elasticity : E
4.02 GPa
in tension, postbaked at 95C, screw tensile testing machine
for 0 - 400 mic. thick SU-8 film coated on a 3" 375 mic. thick Si wafer,
prebaked, illuminated, postbaked at 95C and brought back to room temperature (20C)
Here are some general properties that can not fit in a table, and are deducted mostly from experiment. They should not be taken for more than that :-)
An aspect ratio larger than 1:50 has been obtained... but repeatability may be a problem :-) Typical aspect ratio would be 1:20, and 1:25 for good process.
SU-8 seems to have a good biocompatibility. In collaboration with neuroscientists working in the domain of electrophysiology, they use to culture biological cells on SU-8 and it seems that there is a 'good' biocompatibility of this material. The SU-8 photoepoxy is used as passivation layer for measurement electrodes in the bottom part of a cell culture chamber. [Marc Heuschkel, EPFL]. Thank you Marc!
The SU-8 is very difficult (impossible?) to reflow when it is cross-linked. At 210C it does not reflow after 21h (!) but at 220C it already changes colour and turns black... This seems to contradict the table above where Tg ~ 200C (LaBianca95a) but is consistent with result for other negative resists where degradation occurs before reflow because the resist is heavily cross-linked. For positive resist this is not a problem because it is always possible to insulate it to decrease the cross-linkage... [Ren Yang] Thank you Yang!
Here are all the data gathered over the years pertaining to the SU-8 processing. These datas complete the general information you may obtain from the literature, and are mostly tricks. But they do help a lot :-) Contact in bracket gives the source of the info. The textis almost verbatim from the source, with only very slight edition when it was needed. When some details are 'not available', it is mostly because of some proprietary information that could not be divulged openly. Sometimes, SU-8 processing looks more like an art than like an exact science :-)
MicroChem has set-up a SU-8 bulletin board that allow to ask question and get response, and including many useful informations...
SOTEC MICROSYSTEM has compiled a lot of interesting trick about the processing of the SU-8, that you should read carefully! For example they explain that the thin solid crust that is often observed when dealing with thick layer of photoresist, causing loss of resolution, is due to the heating of the surface induced by high dose (200 mJ/cm2) and long exposure time. A way to alleviate the problem is to divide the exposure in 10 to 15 seconds intervals with a 60 seconds waiting time in between. Of course, as with other photoresist, the exact exposure time is shown to depend on the substrate reflectivity, but, in most case, a bit of over-exposure is not a problem here (negative resist).
Electronic Visions has compiled another interesting document on the SU-8 processing describing a complete process in details, with thickness (10 µm - 450 µm) versus spinner speed (300 rpm - 1500rpm) charts for all commercial blend of SU-8 (SU-8 5 to SU-8 100), pre-exposure time versus thickness (from 30 min at 10 µm to... 7 hours at 450 µm!), exposure (broad band UV source 350 nm - 450 nm) dose versus thickness (from 150 mJ/cm2 (measured at 465 nm) at 10 µm to 1400 mJ/cm2 at 450 µm) and some useful tricks. The document is in PDF format... but too big to include on this page. It is not on their website, but has been placed on the Bulletin Board about the SU-8 created by MicroChem. Alternatively you may ask them for the application notes on SU-8 [ Eva Wimmesberger, Electronics Vision Co.]. Meanwhile, here is a short extract, the table of layer thickness at a spin
speed of 900 rpm for the different blend of SU-8 (static dispense, about 1ml of SU-8 per inch diameter of the wafer to coat, ie 4'' wafer = 4 ml):
SU-8
Thickness (µm)
5
12
10
30
25
58
50
150
100
320
A complete process example with 70% solid resist (?) [S. Basrour, J.-C.
Jeannot & V. Soumann, LPMO/CNRS,
IMFC, Universite de Franche-Comte, Besancon, France, Translated from French: F. Chollet]. Thank you Skandar!
165 mJ/cm2 (with a standard contact UV insulator for microelectronics)
Post bake
15 min at 95°C
Relaxation time
wait 10 min
Development
place the sample in pure PGMEA (propylene glycol methyl ether acetate) during 5 min
rinse with isopropanol (IPA)
if some white stains remain, replace sample in PGMEA for 5 min
repeat until no more white stain appear
Remarks
The photo-resist is VERY SENSITIVE to the planarity of the devices used during the whole process (hot plate, UV insulator(?)...) (e.g., reflow occur on the hot plate) and for good results, the equipment has to be carefully leveled. The photo-resist is very hard to strip after cross-linking. The insulated photo-resist soften during hard bake while the
non-insulated photo-resist remains hard (?). The side walls are 'very' smooth and 'vertical'.
A series of tricks from a long time user: MCC itself :-)
Soft bake: For high aspect ratio imaging, try longer soft bake times in order to reduce the retained solvent level. This may help since you will probably need to have longer develop times. By reducing the solvent level you reduce the risk of exposed resist loss, swelling, adhesion failure, etc. (However, for low aspect ratio structures it is often desirable to work with a bit more solvent, since it speeds up the development time and makes the film easier to strip.) Exposure:Since you have reduced the solvent level you may need to increase your exposure dose. PEB: Some SU-8 users report that extended PEB times at lower temperatures produce better results than shorter times at higher temperatures. Some even go as low as 60 C in a convection oven. Other suggestions include ramping the temperature up, or starting at a lower initial temperature (around 50C). The Tg of SU-8 before PEB is about 55C, the idea is to produce some cross-linking prior to your primary PEB temperature which is frequently between 90-95C, in order to prevent flowing. You might also consider ramping the temperature down after PEB to reduce stress and cracking. Develop: Many SU-8 users have commented that this was their most challenging process step. For high aspect ratio features it is necessary to use medium to strong agitation. One SU-8 user reported that he actually suspends his wafer on a stir rod and immerses it in a beaker of developer. This gives him the best results (clears out deep vias and trenches). Obviously this isn't a high volume production worthy process, but may be helpful in a lab. Also consider using GBL, the SU-8 solvent as a predip prior to immersion in PGMEA. GBL is more aggressive than PGMEA and may clean up your sidewalls since it will remove partially cross-linked SU-8. I would recommend no more than 2-3 minutes, then go immediately into PGMEA. One word of caution; if you are under cross-linked GBL will chew up your structures. Another idea is to consider megasonics. This technique is commonly used in x-ray LIGA processes and has been reported to work well with SU-8. Rinsing: I'd be interested in hearing more on this subject. We have had no success with IPA or DI water. IPA seems to create white spots, which can not be subsequently removed in PGMEA. IPA may also cause surface cracking due to the high evaporation/cooling rate. DI water seems to be death. Any ideas?
[Scott Heidemann, MicroChem Corp.]. Thank you Scott!
Sticking trick quite often, the SU-8 layer sticks to the mask during exposure (collected by Paddy on the MEMS-list).
Even we had this problem of mask sticking. It mainly relates to prebake. We do prebake in a oven (not hot plate) at 90C for 25 to 30 mins. We leave it for at least 10-15 mins in air before exposing it in the mask aligner. That appears to work alright. We never had the problem of cracking. It appears that the resist is under stress for some reason in your case. You may need to optimise the exposure and PEB times.
I had trouble with a ridge all around the wafer, that was thicker and sticky. I had to remove it 'manually' in some cases... but I believe you don't have this problem. With thick photoresist (not SU-8) I used to simply cover the wafer with a half petri dish on the hot plate, the increased baking (ie, IR get reflected) on the resist surface (remember a hotplate bake sample from the bottom) was sufficient to suppress sticking problem I experienced previously. Alternatively, you may wish to try a convection oven for the softbake, it bakes the sample more uniformly.
Cracking tricks
cracking often appears after PEB (collected by Paddy on the MEMS-list).
The problem of the cracks is due to two things :
1) unsufficent exposure dose. Try to expose the layers with increasing dose: 10 mJ - 100mJ - 1000mJ. You will see that the quality of the surface change.
2) the cracks are due to the shrinkage of the SU-8. This effect is increased by exposing large areas of the wafer but if you adapt the design of your mask it disappears.
SOTEC MICRO has introduced a blend of SU-8 with a filler that effectively reduce the coefficient of thermal expansion of the photoresist from 50ppm/K to 21 ppm/K. This brings it closer to the CTE of glass (about 10 ppm/K) and should decrease the stress related to CTE mismatch, thus allow for fewer cracks.
Rinsing trick After the prescribed (by a registered Voodoo doctor) development time, I rinse the front and back side of the wafer w/ fresh PGMEA (MicroChem Developer), blow dry the back side, put it back on the spin coater, pool more PGMEA on the frontside and spin dry at 2-3 kRPM for about 15 seconds. You can actually watch as it dries. This seems to
do a very good job at removing any excess solvent and thus any suspended polymer [Thomas Long, Georgetown University's Physics Dept]. Thank you Thomas!
Removing trick
SOTEC microsystem has introduced a new serie of liquid remover the SMST-S/M/H that effectively remove cross-linked SU-8 without affecting metals (the different version S, M or H of the SMST remover depends on the existing metals on the structure, the S version for Cu structure is very slow)
ashing of the photoresist at a temperature larger than about 600C leaves very few residue... if the structure can sustain it!
SU-8 is virtually insoluble in most chemicals if well crosslinked. You would have to use RIE to etch off the SU-8. However only O2 Plasma does not work well. Try including 'some' CF4 (details not available :-). One thing that was observed is that if you have metal with SU-8 (cured at 200 Deg C) SU-8 is removed without any CF4. Generally you are able to remove SU-8 with O2 plasma and CF4. Then the wafers go through an ultrasonic scrub clean to remove residues.[Paddy]. Thank
you Paddy!
Here is a list of further links to more information on research done with the resist. Enjoy.
MicroChem Corp., one of the SU-8 manufacturer, has created a bulletin board dedicated to SU-8. A good source of information... and a good place to
ask question! There is not that much activity on the Board, actually... maybe is it related to the slowness of the MCC server :-) or to the secrecy of MEMS researcher :-( That's sad because it is a tool that could be helpfull...
The MEMS-list at the ISI, a discussion group where you can ask for everything you want about MEMS and specially about SU-8.
Here is a link to a paper from the original IBM
team working on the SU-8 (albeit the original use of the SU-8 is for high resolution patterns and hence for very thin layers)
The EPFL/IBM team SU-8 page with a lot picture to convince you to use the SU-8... or to simply use their technology for your molding or microfluidic application [Hubert Lorenz, MIMOTEC]
An EPFL/Faculty of Medecine team has interesting results for biological (and example of different cell cultutre grows on the SU-8) and buried microchannels applications [Marc Heuschkel, EPFL]
Another page from the EPFL, from the team who has the longest experience on the photoresist, with very interesting processing tips, details about the chemistry and on the SU-8 blend sold by SOTEC MICROSYSTEMS. A must read! [Louis J. Guerin, EPFL, SOTEC]
Here is a list of commercial products or commercial application related to the SU-8. If you don't have time to develop your own process... it is a very good alternative :-) Still, you may want to contact the companies cited to get some tricks for free... or sell yours :-) Oh, BTW, they are of course listed at no charge, so if you want to be there just drop me a line :-)
MicroChem Corp. (previously named Microlithography Chemical Corp.) is the oldest manufacturer of the SU-8 photo-resist, licensed by IBM. The people in charge of the SU-8 are Scott Heidemann and Mark Shaw. They are usually very helpful, do not hesitate to ask them your questions :-)
SOTEC MICROSYSTEMS is a spin-off company from the EPFL. It has been founded by Dr Louis J. Guerin and Marc Heuschkel. They are specialized in the fabrication of SU8 structures, where the SU8 is used a permanent material and is not stripped away at the end. Their main products includes microfluidic structures, packaging and microcoils. They also bought a license from IBM and they sell the SU-8 photoresist with their own blend. The address of SOTEC MICROSYSTEMS is: SOTEC MICROSYSTEMS, 11 avenue des Baumettes, 1020 Renens, Switzerland [Louis J. Guerin, SOTEC]
MIMOTEC is a spin-off company from the EPFL. It has been founded by Hubert Lorenz. It deals with the commercialisation of micromolds fabricated by a UV-LIGA process using SU-8. For this process they have built a small clean room unit fully dedicated to SU-8 microfabrication [Hubert Lorenz, MIMOTEC]
Technics, Inc., an RIE equipment manufacturer in Pleasanton, CA has been looking at RIE processes to strip SU-8 for the past 6-8 months (1998). They are reporting 4-5um/minute removal rate for hard baked SU-8 patterns that contain plated up structures using
an isotropic process. They are also working on anisotropic processes. The primary contacts are Chris Ropoulos, V.P. sales, and Marco Traversa, member of technical staff [Scott Heidemann, MicroChem Corp.]
Electronic Visions is an well-known Austrian manufacturer of mask aligner and resist spinner that has developped a complete and very useful set of process condition to be used with their equipment for all SU-8 resist blends, that you shouldask for :-) [Loren StClair, Electronics Visions].
Karl Suss is a manufacturer of mask aligners and resist spinner working with SU-8 [Scott Heidemann, MicroChem Corp.].
Further very useful information (and the previous one :-) from the IBM and the EPFL teams (and a few others...) may be found in the following papers (please contact me if you know other references :-) This list was established with the help of Loren St CLAIR, Hubert LORENZ, Arnaud BERTSCH, Marc HEUSCHKEL, Patrick MOUNAIX, and LING Zhong-Geng. It is used by Microchem in their SU-8 bulletin board, thus it must be a good list :-)
D. Stumbo, and J. Wolfe, "Ion exposure characterization of a chemically amplified epoxy resist", J. Vac. Scien. Technol. B 11 (1993) : 2432-2435 : an early paper for those interested in proton exposure with SU-8 (thin layer).
N. LaBianca, and J. Delorme, "High aspect ratio resist for thick film applications", in Proc. SPIE vol. 2438, SPIE, (1995) : 846-852 : the first (?) attempt to use thick layer of SU-8. By the IBM team.
N. LaBianca, J. Gelorme, K. Lee, E. Sullivan,and J. Shaw, "High aspect ratio optical resist chemistry for MEMS applications", in Proc. 4th Int. Symp. on Magnetic Materials, Processes, and Devices, The Electrochem. Soc., 95-18 (1995), pp.386-396
K. Lee, N. LaBianca, S. Rishton, and S. Zohlgharnain, "Micromachining applications for a high resolution ultra-thick photoresist", J.
Vac. Scien. Technol. B 13 (1995) : 3012-3016 : the first (?) attempt to use thick layer of SU-8 for micromachining, describing the use as a mask for deep silicon etching and as a mold for nickel electroplating. By the IBM team.
H. Lorenz, M. Despont, N. Fahrni, N. Labianca, P. Vettiger, and P. Renaud, "EPON SU-8 : A low-cost negative resist for MEMS", in Proc. of Micro Mechanics Europe'96, Barcelona, (1996) : 32-35 : first paper of the IBM/EPFL team.
J. M. Shaw, J. D. Gelorme, N. C. LaBianca, W. E. Conley, and S. J. Holmes, "Negative photoresists for optical lithography", IBM Journal of Research and Development 41(1997) : 81-94 : entertaining paper about the genesis of the chemistry for the SU-8 and some application written by the inventors of the resist (online paper, great!)
M. Despont, H. Lorenz, N.Fahrni, J. Brugger. P. Renaud, and P. Vettiger, "High aspect ratio ultrathick, negative-tone near-UV photoresist for MEMS applications", in Proc. MEMS'97, IEEE, Nagoya, (1997) : 518-522 : ... and the SU-8 is reveled to an incredulous MEMS world :-)
H. Lorenz, M. Despont, M. Fahrni, N. LaBianca, P. Vettiger, and P. Renaud , "SU-8: a low-cost negative resist for MEMS", J. Micromech. Microeng 7(1997) : 121-124 : First journal paper of the EPFL team, from the work presented at the Micro Mechanics Europe'96 conference. Gives some deposition parameters and a simple mechanical characterization.
L. Dellmann, S. Roth, C. Beuret, G. Racine, H. Lorenz, M. Despont, P. Renaud, P. Vettiger, and N. de Rooij, "Fabrication process of high aspect ratio elastic structures for piezoelectric motor applications", in Proc. Transducers 1997, Chicago, (1997) : 641-644
L. Guerin, M. Bossel, M. Demierre, S. Calmes, and P. Renaud, "Simple and low cost fabrication of embedded microchannels by using a new thick-film photoplastic", in Proc. Transducers 1997, Chicago, (1997) : 1419-1422
A. Bertsch, H. Lorenz, and P. Renaud, "Combining microstereolithography and thick resist UV lithography for 3D microfabrication", in Proc. MEMS'98, IEEE, Heidelberg, (1998) : 18-23 : The title
explains it all...
B. Eyre, J. Blosiu and D. Wiberg, "Taguchi Optimization for the processing Epon SU-8 resist", in Proc. MEMS'98, IEEE, Heidelberg, (1998) : 218-222 : gives optimized process variables such as softbake time, exposure time, post-exposure time, develop time and substrate type for SU-8 with 73% solvent that were obtained with the Taguchi method for three typical thickness 50mic., 100 mic., and 220 mic.
H. Lorenz, M. Despont, N.Fahrni, J. Brugger. P. Renaud, and P. Vettiger, "High aspect ratio ultrathick, negative-tone near-UV photoresist and its applications for MEMS", Sens. & Act. A A64(1998) : 33-39 : Expanded version of the MEMS'97 paper. It describes process, results (aspect ratio up to 18, thickness
up to 1.2 mm) and application of the SU-8 (including the use of the resists as a structural material (photoplastic) and multi-level patterning).
M. Heuschkel, L. Guérin, B. Buisson, D. Bertrand, and P. Renaud, "Buried microchannels in polymer for delivering of solutions to neurons in a network", Sensors & Actuators: B. Chemical 48/1-3 (1998) : 356-361
W. Flack, W. Fan, and S. White, "Optimization and characterization of ultrathick photoresist films", in Proc. Advances in Resist Technology and Processing XV, Proc. SPIE Vol. 3333, SPIE (1998) : 1288-1303 : a paper from Ultratech Stepper describing experiment with the SU-8 and their 1Xstepper. They concluded that SU-8(25) showed no photo-sensitivity in the gh-line spectrum (390-450 nm) and they used i-line stepper (355-375 nm) to pattern line with an aspect ratio of 1.2 (sic) in a 25mic. thick layer giving a fully detailed process ... (online paper, great!)
J. Thorpe, D. Steenson, and R. Miles, "High frequency transmission line using micromachined polymer dielectric", Electron. Lett. 34 (1998) : 1237-1238 : using SU-8 as insulation dielectric for mm-wave transmission line in copper between 20 GHz and 40 GHz.
H. Lorenz, M. Despont, P. Vettiger, and P.Renaud, "Fabrication of photoplastic high-aspect ratio microparts and micromolds using SU-8 UV resist", Microsyst. Technol. 4 (1998) : 143-146 : more example of photoplastic use and complete description of the patented MIMOTEC process where the SU-8 is used to create metallic multilevel mold by repeated coating and patterning followed by electroplating and removal of the resist.
L. Dellmann, S. Roth, C. Beuret, L. Paratte, G.-A. Racine, H. Lorenz, M. Despont, P. Renaud, P. Vettiger, and N. de Rooij, "Two steps micromoulding and photopolymer high-aspect ratio structuring for applications in piezoelectric motor components", Microsyst. Technol. 4 (1998) : 147-150 : complex use of the SU-8 with a detailed process.
H. Lorenz, M. Laudon, and P. Renaud, "Mechanical characterization of a new high-aspect-ratio near UV-photoresist", Microelec. Engin. 41/42 (1998) : 371-374 : first attempt at a comprehensive mechanical characterization of the SU-8.
C. Malek, "Mask prototyping for ultra-deep X-ray lithography: preliminary studies for mask blanks and high-aspect-ratio absorber patterns", in Proc. SPIE vol. 3512, SPIE (1998) : 277-285 : investigates the SU-8 for electroplating thick X-LIGA mask absorber. They also patterned SU-8 with X-ray and found it 'over one order of magnitude' more sensible than PMMA (X 25 with 50 mic. thick film) and, to present thickness-independant incident dose. An interesting point to note, with X-ray you don't have reflection from the substrate and hence no standing wave in the resist, yielding really smooth sidewalls...
S. Arscott, F. Garet, P. Mounaix, L. Duvillaret, J.-L. Coutaz, and D. Lippens, "Terahertz time-domain spectroscopy of films fabricated from SU-8", Electron. Lett.35 (1999) : 243-244 : measurement of the refractive index (n) and the absorption coefficient (a) of the SU-8 between 100 GHz and 1.6 THz. Moreover they show results of an interesting double exposure experiment, allowing to insulate different thickness of photoresist from a single layer!
M. Campbell, D. N. Sharp, M. T. Harrison, R. G. Denning, and A. J. Turberfield, "Fabrication of photonic crystals for the visible spectrum by holographic lithography", Nature 404 (2000) : 53 - 56 : the title says it : fabrication of 3-D photonic bandgap crystal working in the visible spectrum... but omits to tell it is using the SU-8! Will it be the killer application? :-)
Zhong-geng Ling, Kun Lian, Linke Jian, "Improved patterning quality of SU-8 microstructures by optimizing the exposure parameters", in Proc. SPIE vol. 3999, SPIE (2000) : 1019-1027 : at last, a thorough investigation of the SU-8 absorption in the UV and up to the NIR, leading to many process improvements (aspect ratio 25 with 360um thick layer). Very interesting!