Slashdot Mirror


UK University Creates First Inkjet-Printed Graphene Circuit

MrSeb writes with this ExtremeTech excerpt aimed at the graphene enthusiasts out there: "You can add another crazy characteristic to graphene's ever-expanding list of 'wonder material' properties: It can now be used to create flexible, transparent thin-film transistors. ... using an inkjet printer. The discovery comes from researchers at the University of Cambridge, UK, who were trying to ameliorate the lackluster performance of existing inkjet-printed electronics. It's already possible to print standard CMOS transistors using different ferroelectric polymer inks, but the resultant circuit is so slow that it can't actually function as a computer. If graphene could replace or augment the interconnects or transistors, these circuits would be a lot faster — and that's what these Cambridge engineers have done. Furthermore, if you didn't think that was cool enough, the graphene-based ink that they've developed is transparent, too."

60 comments

  1. Pros and cons by hamsolo474 · · Score: 2

    well i think this is a great idea, printable circuits would make it much easier for the electronics hobbiest to build electronics. perhaps with printable circuits people might begin printing their circuits to look like their schematics. On the cons side its this sort of confusing technology that delayed me and i think stopped other people all together from attempting electronics, i was always interested in electronics but whenever i would crack open a commerical device to look at the circuits and attempt to gain some form of understanding as to how it works, the circuit would be laid out in some space saving non reabable/understandable design. this makes the circuits look overly complex and thus scares away people who otherwise might try their hand at it.

    1. Re:Pros and cons by Anonymous Coward · · Score: 0

      Where is the reference to IBM?

    2. Re:Pros and cons by wierd_w · · Score: 4, Interesting

      I can see a rather clever toy being made using this technology.

      Imagine for a moment, that you have a simple robot chasis like hardware platform, with a flat ribbin cable type connector that is really really wide (like the kind that attaches the keyboard inside a laptop, or the ribbon going to the optical head assy of a dvd drive, only much wider.). It is meant to accept all the electronic innards of the robot other than actuators and external ports and battery pack on a printed sheet of slightly textured plastic, that has a commercially inlaid edge connector on one side that fits the really wide flat ribbon connector, and on the other a specially notched corner, or some other orientation feature.

      The sheet is meant to be cleaned, like a dry erase board, but with a special solvent. This makes the sheets reusable.

      The printer loads the sheets, which could be printed on both sides if desired, and prints circuits and traces that lead to the inlaid metal edge connector pads.

      The idea is as a hobby electronics toy for enthusiasts and advanced school children. A kind of "build your own robot" kit with modular servos and sensors that attach to the simplified chasis a little like legos, and with an extremely low cost roll your own logic board that you print on the special plastic sheet. Since the circuit components are all inkjetted, component cost would be minimal: just the cost of the ink.

      Anything from novelty antennas to special logic could be printed on the sheets.

      I could really see such a toy being a fantastic learning aid, especially in a school setting. The limiting factors would be feature size, circuit impedence, and physical realestate on the printed sheet. While probably slower than an arduino, it would offer quite a bit more customization options, especially if it came with some design software that had sample patterns for apus, memory cell circuits, and various logic gates and simple IC units.

    3. Re:Pros and cons by mister_playboy · · Score: 1

      Since the circuit components are all inkjetted, component cost would be minimal: just the cost of the ink.

      Heh, have you priced any inkjet cartridges recently? :)

      --
      Do what thou wilt shall be the whole of the Law ::: Love is the law, love under will
    4. Re:Pros and cons by wierd_w · · Score: 1

      Considering that graphene is realy inexpensive (you produce quite a bit od it when writing with a standard pencil)...

      You are quite right about the issue though. The idea is to keep that cost low, or the toy wouold be substantially less popular. The idea is to make money on the accessories, not the consumables. (New sensors, actuators, battery p0acks, etc.)

    5. Re:Pros and cons by Anonymous Coward · · Score: 0

      is it really graphene you're producing or just multiple layers (at best) of graphite?

  2. UK University??? by Anonymous Coward · · Score: 1, Interesting

    I don't need someone to tell me where Cambridge is.

    1. Re:UK University??? by dbIII · · Score: 4, Informative

      The US also has a Cambridge and built a University there.

    2. Re:UK University??? by Anonymous Coward · · Score: 1

      Interesting. I wonder where they go the idea?

    3. Re:UK University??? by Anonymous Coward · · Score: 4, Funny

      It's a franchise. Australia is due to open one soon.

    4. Re:UK University??? by ioshhdflwuegfh · · Score: 1

      The US also has a Cambridge and built a University there.

      Thanks dbill, that was like really insightful and stuff.

    5. Re:UK University??? by MasterPatricko · · Score: 4, Informative

      The first president (Henry Dunster), the first benefactor (John Harvard), and the first schoolmaster (Nathaniel Eaton) of Harvard were all Cambridge University alumni, as was the then ruling (and first) governor of the Massachusetts Bay Colony, John Winthrop.

      --
      I'd tell a UDP joke, but you may not get it. I'd tell a TCP joke, but I'd have to keep repeating it until you got it.
    6. Re:UK University??? by dbIII · · Score: 1

      Thanks for the sarcasm, but it's an international forum here and it's not obvious to everyone.

    7. Re:UK University??? by Anonymous Coward · · Score: 0

      I think the 'UK' part of the title made it made painfully obvious to which Cambridge the submitter is referring.

      Given the international reputation of Cambridge University, I'm surprised by this comment, even from American readers. It's comparable to your Harvard over here.

    8. Re:UK University??? by Anonymous Coward · · Score: 1

      Cambridge University is almost 1000 years old and is one of the worlds best Universities. If it was MIT the title wouldn't be "A US University" would it?

    9. Re:UK University??? by PReDiToR · · Score: 1

      It's not "international," it's Americo-centric.

      This is not a criticism.

      And even the high-up fruit that we get on this site might not be aware of the knowledge centres of Airstrip One.

      --

      Do not meddle in the affairs of geeks for they are subtle and quick to anger
    10. Re:UK University??? by Anonymous Coward · · Score: 0

      The US also has a Cambridge and built two Universities there.

      FTFY.

  3. Available at Office Depot by The+Clockwork+Troll · · Score: 5, Funny

    I hope they are using Genuine EPSON Graphene Inks.

    --

    There are no karma whores, only moderation johns
  4. Doesn't surprise me by jd · · Score: 3, Interesting

    This was a matter of time, given that they could do regular transistors and graphene is much simpler. What will be interesting, however, is what people do with this. There have been all kinds of interesting attempts at novel engineering (liquid crystal memories, for example) that either never got funded to completion or ended up going nowhere.

    The latter is as it should be. There is no point in research if you know the results beforehand. It is merely fraud to only do what you know will work. The former is typical bigotry against those who actually do the real work in society.

    --
    It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
    1. Re:Doesn't surprise me by ioshhdflwuegfh · · Score: 1

      Maybe not very well said, but you got good points there...

    2. Re:Doesn't surprise me by ultranova · · Score: 3, Insightful

      What will be interesting, however, is what people do with this.

      Probably nothing. Inkjets tend to have their jets clog, which in turn leads to imperfections which render the circuit inoperative. While this can in part be helped by printing larger features, the resolution is already pretty low, and that makes the effective resolution - and thus the amount of components you can fit on a paper - even lower.

      In short, in all but the most specialized applications it's better to just get a regular chip.

      --

      Forget magic. Any technology distinguishable from divine power is insufficiently advanced.

    3. Re:Doesn't surprise me by Vellmont · · Score: 1


      In short, in all but the most specialized applications it's better to just get a regular chip.

      If by "specialised" you mean new, I'd agree with you. I'm sure people said the exact same thing when the CPU first was created, but nobody would consider a CPU specialised any more. What's specialised now could easily become the largest number of chips produced very quickly.

      If this technology pans out as described, these little chips aren't going to replace silicon chips, they'll be used for applications where a silicon chip is either impractical because of expense, or too fragile. Silicon chips have been around far too long and developed far too much for anything to directly replace them.

      --
      AccountKiller
  5. Just an OS issue by Anonymous Coward · · Score: 0

    "but the resultant circuit is so slow that it can't actually function as a computer"

    They just need to upgrade to Windows 7.

  6. Cool, but effective? by pryoplasm · · Score: 2

    I like the idea of a transparent circuit, however when it comes to troubleshooting it or trying to find a break in it, transparent might not be the best thing...

    still, interesting. I am wondering if they can add something to the graphene to make it more visible without taking away some of graphenes properties...

    --
    Those who live by the sword, get shot by those who live by the gun...
    1. Re:Cool, but effective? by reiisi · · Score: 3, Insightful

      Have you ever tried troubleshooting a microprocessor off the assembly line?

      I think you're thinking in terms of replacing the circuit board etching kits for hobbiests. Sure, transparency would put a minor kink in that use, but not a show stopper.

      (And I'm sure we can both think of cool things to do with the transparency.)

      It would be interesting to know how much various pigments interferred with the electrical properties, of course. Home printing of circuit boards (circuit papers?) would definitely make a lot of projects easier, particularly when the active components can also be printed on, perhaps in the same pass.

      Lo-res LSI.

      --
      Computer memory is just fancy paper, CPUs just fancy pens with fancy erasers; the 'net is just a fancy backyard fence.
    2. Re:Cool, but effective? by Anonymous Coward · · Score: 0

      I like the idea of a transparent circuit, however when it comes to troubleshooting it or trying to find a break in it, transparent might not be the best thing...

      If you crack open the case of a modern CPU or microcontroller you will see that it's not too easy to find the different parts either.
      You can usually find the internal memory (Cache in the case of CPU's and Flash/SRAM in the case of microcontrollers.) pretyy easy since they usually cover several square millimeters with the same pattern repeated over and over again but to be able to identify the other parts you pretty much need to have seen/constructed that particular design before.
      With transparant electronics it would be possible to first print color-coded background for the different CPU parts before printing the electronics on top of this.
      While it doesn't have any application in consumer electronics it would be pretty cool for educational purposes. That way you could easily find and measure on a microcontroller during code execution to see how signals propagate through the ALU each clock cycle.

    3. Re:Cool, but effective? by GaryOlson · · Score: 1

      If transparent in visible light, is the circuit visible in UV light or other wavelengths?

      --
      Every mans' island needs an ocean; choose your ocean carefully.
  7. target apps by reiisi · · Score: 2

    6809 processor on a sheet of paper.

    And could we squeeze in a 6821, and a 6847? (Still lots of NTSC TVs not in the land fills. Intel deserves to die for HDCP.)

    Capacitive keyboard? (Not a whole lot worse than the chiclet keyboards.)

    --
    Computer memory is just fancy paper, CPUs just fancy pens with fancy erasers; the 'net is just a fancy backyard fence.
    1. Re:target apps by Lazarian · · Score: 1

      Even better: print out an OLED screen along with a fully functional Radio Shack Color Computer. You could play Dungeons of Daggorath right on a sheet of paper.

      I love that machine. I still have one :)

    2. Re:target apps by Almost-Retired · · Score: 4, Interesting

      "63C09 processor on a sheet of paper."

      There, I fixed it for you. You see the 6309 was the result of Hitachi obtaining a permission to build a 6809 that was a functional clone of the 6809, but had to call it something else other than its being a cmos 6809 chip, and were enjoined from ever saying there was a difference other than the resultant design was in cmos.

      But something odd happened on the way to the foundry. When the 6809's op-codes were put into a map whose x/y was the value of the nibbles, there was quite a number of 'holes', so Hitachi filled them in with enhanced feature op-codes. Op-codes it remained for the users to discover because even when disclosed, Hitachi was prevented from even confirming their existence. Such things as several more accumulator registers (E,F, and Q which meant A,B,E,F combined for 32 bits), and where the 6809 had some mul functions that gave 16 bit answers, the 6309 had a new 16x16 multiply with a 32 bit answer, and a divide of 16 bits into 32 bits with a 16 bit answer and a 16 bit remainder in 39 clocks worst case. Loads and stores could be 32 bits wide, a slight speedup because it skiped an instruction fetch in the middle, and a switchable ability to pipe line the next instruction fetch during an otherwise idle bus cycle which made it about 10% faster because you actually had to turn it off by a write to the MD register.

      The net result for the color computers OS9 operation system was, when combined with some bug fixing, nearly twice as fast at the same clock speed. FWIW, I have one in my 'coco3', running right now. We call it Nitros9 now. Community supported.

      If this 'sheet of paper' can get to a 2 mhz clock speed, I can see newer developments made even easier already.
      A notebook coco3 for instance. Just turn the 'page' to run a different test configuration. :)

      Cheers, Gene

    3. Re:target apps by reiisi · · Score: 1

      Wish I could afford to ship mine across the ocean. :-/

      (Wish I had a place to keep it here. Not sure if my kids would appreciate it, though.)

      --
      Computer memory is just fancy paper, CPUs just fancy pens with fancy erasers; the 'net is just a fancy backyard fence.
  8. Just wondering by Anonymous Coward · · Score: 2, Interesting

    So, will this lead to downloadable (based on open source hardware and OS) devices? And downloadable designs for your devices as well?

  9. Transparent by Arancaytar · · Score: 2

    You could now print an invisible computer onto a piece of paper?

    More awesome than can be described.

    1. Re:Transparent by bobaferret · · Score: 1

      I can see the teachers now requesting that students subject their paper to powerful EM pules before using it to take a test. Or printing a wifi enabled cameras on the TP in the girls bathroom.

  10. Hate to be a downer, but... by Anonymous Coward · · Score: 2, Funny

    I'll bet it causes cancer.

    1. Re:Hate to be a downer, but... by Anonymous Coward · · Score: 0

      No, you've got it all wrong. You need to smear graphite all over your hands while you handle your food cooked in a teflon coated skillet while sitting on your freshly scotchgarded sofa. THEN you'll have cancer.

    2. Re:Hate to be a downer, but... by GaryOlson · · Score: 1

      I'll bet it causes cancer.

      Only if printed on white mice at a circuit density 1000 times higher than normal human circuits.

      --
      Every mans' island needs an ocean; choose your ocean carefully.
  11. Really? by king+neckbeard · · Score: 1

    Inkjet printers are pretty much the bane of all technology. I would strongly prefer that they be used as little as possible.

    --
    This is my signature. There are many like it, but this one is mine.
    1. Re:Really? by wild_berry · · Score: 1

      You like your OLED display? That's inkjet-printed. I figure that Cambridge University has tamed the Inkjet -- the OLED was discovered there, too.

    2. Re:Really? by Anonymous Coward · · Score: 0

      Wave your magic wand, click your ruby slippers three times and it will be so.

  12. University of Kentucky by ZombieBraintrust · · Score: 1

    It is especially confusing because the University of Kentucky is referred to as UK in the states. Though I guess UK University would be redundant in that instance.

  13. Not alumni by Kupfernigk · · Score: 2
    Although Cambridge has given up and now uses the American term in fundraising, they were not in fact U of C alumni (and nor am I, to declare an interest). That's because if you go to Cambridge and graduate, you remain a member of the University for life or until the Senate votes you off, though obviously with no official title and unpaid. Same with Oxford.

    Technically, therefore, Harvard was founded by members of the University of Cambridge.

    The old joke is that graduates of Oxford intend to run the country (which they almost invariably do) while graduates of Cambridge don't care who runs the country so long as they get to do something interesting. Given how immensely successful Harvard has been, the rule obviously applied in this case.

    --
    From scarped cliff or quarried stone she cries "A thousand types are gone, I care for nothing, no not one."
  14. This + 3d printing by nurb432 · · Score: 1

    Opens up a lot of possibilities ( once the bugs are worked out ). Not only can you make simple mechanical devices, but now you could create control systems for them too.

    One of these days PirateBay will be more than music and video .. but torrents for actual 'stuff'..

    --
    ---- Booth was a patriot ----
  15. Cheap ones perhaps by Kupfernigk · · Score: 1
    Have you ever used a high end inkjet printer like a Designjet? Or seen an Indigo working? Or even a high end office inkjet printer like a Ricoh Gelsprinter? (Ours is well over 20000 pages so far and has never blocked or thrown a bad page).

    You really cannot compare a disposable bit of plastic with what inkjet technology is capable of. It's like saying that a BMW 5 series is useless for getting anywhere because Chinese kids' tricycles suck.

    --
    From scarped cliff or quarried stone she cries "A thousand types are gone, I care for nothing, no not one."
  16. What does speed have to do with computation? by mark-t · · Score: 1

    The purist in me has a problem with the wording in the summary and first article.... I mean, it's one thing to say that the resultant circuits are too slow for any practical application as computers, but speed would have absolutely no bearing on whether or not computations can actually be performed with it, making the notion of the circuits being too slow to function as a computer a wholly worthless statement that even at best is very misleading, and at worst outright false.

  17. The first? by Anonymous Coward · · Score: 1

    In Thunderbird the partial headline read 'inkjet-printed Grap' and so to me I thought it was...
    'UK University Creates First Injet-Printed Crap', but people print crap all the time, what is different about their printed crap?

  18. Cambridge vs Cambridge... by feranick · · Score: 1

    The real first Cambridge is the UK one. Cambridge, MA, USA was founded by students of the University of Cambridge, UK in the new world. Harvard was established first. Much later, MIT decided to make Cambridge. MA its own home. Despite these two giants, the University of Cambridge is a world leading university. Discounting it as a "UK University???" only shows utter ignorance.

  19. Re: hao are y by Anonymous Coward · · Score: 0

    ni hao sha gua!

  20. Before you get too excited... by Anonymous Coward · · Score: 0

    Graphene isn't a semiconductor in the conventional sense, so the transistor-like structures that can be created with it don't have a full-off state. You can do some very interesting things with them, but nothing that looks like VLSI digital circuitry, unless you include conventional semiconductors as well. You'll more likely see it in analog applications, or as a component of VLSI circuits that include silicon as well.

  21. Re: hao are y by Anonymous Coward · · Score: 0

    prease to show papers, gweilo

  22. 6809 by Anonymous Coward · · Score: 0

    I remember those 6809 can be found in the Web Camera, so don't be suprised by this news.

  23. rogue engineers at hitachi by reiisi · · Score: 1

    Yeah, I thought the 6309 was cool, too.

    But it was not the direction Motorola intended to take the 6809, and may have played a part in their decision to leave the 6809 and return to single-stack architecture in their embedded products.

    OS-9/6809 was cool, too.

    I spent (wasted?) a lot of time trying to build an OS/run-time that combined the position independence of OS-9 with a split stack parameter passing paradigm. The 6809 just didn't quite have enough resources for that. (And the 6309 took the design the wrong direction for that.)

    Still fascinated by the design of the 6809, still not sure why.

    --
    Computer memory is just fancy paper, CPUs just fancy pens with fancy erasers; the 'net is just a fancy backyard fence.
    1. Re:rogue engineers at hitachi by Almost-Retired · · Score: 1

      (And the 6309 took the design the wrong direction for that.)

      Maybe, but having those additional commands and register resources sure made optimizing the os9 code a bunch easier. Enough that it is about twice as fast as os9 was now because we fixed some bugs in the process, then ported the bugfixes back to the 6809 versions. We now build nitros9 for either cpu from the same src's.

        I did the initial and next couple of rbf.mn conversions myself. Plus I re-wrote the serial mouse driver just 3 years ago. The project is now called nitros9, and disk images to install it and make new boot disks are downloadable from the sourceforge site, and all the tools to rebuild it for winders/mac/linux are available from OpenCoco too. That is called "Toolshed". I keep a repo clone of both here on this machine.

      Which direction would you have taken the 6309?

      Thanks & Cheers, Gene

    2. Re:rogue engineers at hitachi by reiisi · · Score: 1

      Well, now I don't remember. Did the 6309 expand DP?

      Anyway, I'd have started by expanded the index registers, to make it possible to directly address a reasonably wide pixel display buffer and such. (Per the 68HC16 or whatever that was/is.) Widened DP to make it possible to reasonably use it to base process local variables in a large process space. Added another index register (and kept the U) to ease direct execution of p-machine or Forth kinds of intermediate code. Stuff in that vein.

      OS-9 is cool, but I kept running into walls, subconsciously expecting the full Unix functionality in the tools. Didn't have the sense to just go out and buy a PC to use developing tools for OS-9, kept trying to bootstrap the tools on the native arch.

      Say, is there anywhere still making real (as opposed to gate array) 63/809s? I'd sure rather teach my kids microcontrollers on the 6809 than on the 8080.

      --
      Computer memory is just fancy paper, CPUs just fancy pens with fancy erasers; the 'net is just a fancy backyard fence.
    3. Re:rogue engineers at hitachi by Almost-Retired · · Score: 1

      >Well, now I don't remember. Did the 6309 expand DP?
      No, it was still an 8 bit register.
      >Anyway, I'd have started by expanded the index registers, to make it possible to directly address a reasonably wide pixel display buffer and such. (Per the 68HC16 or whatever that was/is.) Widened DP to make it possible to reasonably use it to base process local variables in a large process space. Added another index register (and kept the U) to ease direct execution of p-machine or Forth kinds of intermediate code. Stuff in that vein.

      And likely that would not have fit in the register decoding scheme. The extra registers/accumulators it got were, if my memory is correct:

      A mode register, called MD, setting or clearing a couple of bits of that could switch the speedup stuffs on/off, so that with everything off, which was not the default, it ran exactly like the 6809 to preserve timing loops and such. The default mode had the instruction prefetch enabled. There was another bit too, but I don't recall its function ATM.

      2 new 8 bit accumulator registers, E & F, with 16 bit load/store when addressed as W, and all 4 for 32 bit load/stores when addressed as Q which loaded or stored ABEF in that byte order. Data moves were a little faster since it didn't need a 2nd instruction fetch to load or store 32 bits.

      Level 2 os9 brought us the ability to address 2 megabytes of main memory, but tandy didn't build the hardware into the GIME chip for the last 2 bits, so the stock coco3 has only a 512k memory capability. Tony DiStefano made a kit that added the last 2 bits expanding that to 2024K (decimal) and I have one of those kits in my coco3. The paging scheme within the S-U-X-Y accessible range puts any 8k block of that main memory into any 8k block of the processors 64k address space. So when I wrote 'myram', a ramdisk, only an 8k block of that memory was mapped into cpu space at any one microsecond, but the whole thing was available transparently to a program using /r0 as its virtual disk drive. However it is not noticeably faster than /dd or /s1, which are a pair of 1Gb seagate hawk scsi drives accessed by a Cloud9 TC^3 controller in the mpi's slot 2 (base 0 numbering). To move a megabyte from either storage medium is about (6309) 11 or (6809) 13 seconds and is controlled by how fast the cpu can do a block move. The diff between myram and everybody else's is that myram self formats to whatever size in 8k blocks is set in the 'cyl' variable in the descriptor, doing this in about 50 milliseconds before allowing the first access, and the deiniz operation returns every byte to the free memory pool. Very transparent to the user.

      To get back to to 63x09, its being cmos meant little or no heating, so little that once I had pulled the psu stuffs out of mine and it is now powered by the same old AT psu that powers all the drives and the mpi, the heat rise there is, is right over the 2 dimms of memory, and is 2 degrees F!

      Looking at the waveforms around this C rated (4mhz) chip in mine, with a 100 mhz scope, tells me it should be capable of operating at a considerably higher speed than the max of 1.79 mhz it does run at. At one point several years ago I was running 2 monitors by way of a WP-RS card in slot 1, which would have isolated that video from the cpu clock speed, and subbing a vco for the crystal that I could crank up to 10 or 20 mhz. But I think that would have been doomed to fail the irreplaceable GIME from over heating, so that never got 'tested'. Putting the video into the GIME torpedoed its power budget, and if I were young enough (I'm 77 now) to consider making a coco4, that would have been excised and put into a dedicated video chip. Something with enough output transistors in the design to pull to either rail in say 1/4 of a clock cycle. The existing GIME has a video output rise or fall time nearly twice the NTSC time of 110 ns, or about 240ns. So it cannot even explore the full bandwidth of an NTSC monitor.