Loading ...
Sorry, an error occurred while loading the content.

Solar advances from Australia

Expand Messages
  • penakaskin
    Up to 92% savings of the silicon wafers are achieved with Sliver cell technology. Extremely light cell with power to weight ratio greater than 1500 W/kg have
    Message 1 of 8 , Jan 13, 2004
    • 0 Attachment
      Up to 92% savings of the silicon wafers are achieved with Sliver cell
      technology.

      Extremely light cell with power to weight ratio greater than 1500 W/kg
      have also demonstrated.

      There have pdf on that linked page, all juice is in it.

      http://solar.anu.edu.au/pages/epilift.html
    • terence_ms
      This seem like the way to go. Silicon is much more environmentaly friendly than some of the other toxic crap used nowdays for solar cells. And the other big
      Message 2 of 8 , Jan 13, 2004
      • 0 Attachment
        This seem like the way to go. Silicon is much more environmentaly
        friendly than some of the other toxic crap used nowdays for solar
        cells.

        And the other big plus where it claims it uses 90% less silicon means
        that the energy used to produce the cells is going to be much less
        thereby making the energy payback period much shorter.

        If I had money, I back this lot all the way.

        -Terence
        --- In energyresources@yahoogroups.com, "penakaskin" <penakaskin@y...>
        wrote:
        >
        > Up to 92% savings of the silicon wafers are achieved with Sliver
        cell
        > technology.
        >
        > Extremely light cell with power to weight ratio greater than 1500
        W/kg
        > have also demonstrated.
        >
        > There have pdf on that linked page, all juice is in it.
        >
        > http://solar.anu.edu.au/pages/epilift.html
      • Roger Arnold
        I downloaded and read the pdf. Wow! This appears to be a very significant development. Perhaps not just for solar PV either. From what I know of IGBTs and
        Message 3 of 8 , Jan 13, 2004
        • 0 Attachment
          I downloaded and read the pdf. Wow! This appears to be a very significant
          development. Perhaps not just for solar PV either. From what I know of
          IGBTs and power MOSFETs, it should be adaptable to improve cost /
          performance of electrical power switches. That would reduce the cost and
          improve efficiency of DC-DC converters and motor controllers. Hello, R2D2?

          Roger Arnold
          Sunnyvale, CA

          ----- Original Message -----
          From: "penakaskin" <penakaskin@...>
          >
          > Up to 92% savings of the silicon wafers are achieved with Sliver cell
          > technology.
          >
          > Extremely light cell with power to weight ratio greater than 1500 W/kg
          > have also demonstrated.
          >
          > There have pdf on that linked page, all juice is in it.
          >
          > http://solar.anu.edu.au/pages/epilift.html
        • mduffin3
          ... means ... slivers are going to be a bitch to handle and assemble into arrays on a high volume production basis, and even worse for a bifacial array.
          Message 4 of 8 , Jan 14, 2004
          • 0 Attachment
            --- In energyresources@yahoogroups.com, "terence_ms"
            <terence_ms@y...> wrote:
            > This seem like the way to go. Silicon is much more environmentaly
            > friendly than some of the other toxic crap used nowdays for solar
            > cells.
            >
            > And the other big plus where it claims it uses 90% less silicon
            means
            > that the energy used to produce the cells is going to be much less
            > thereby making the energy payback period much shorter.
            >
            > If I had money, I back this lot all the way.
            >
            >It looks like an interesting step, but there are problems. Those
            slivers are going to be a bitch to handle and assemble into arrays on
            a high volume production basis, and even worse for a bifacial array.
            Assembly will also eat up a good bit of the energy saving. The energy
            saving at the level of a single cell might be as high as 4:1, it
            won't be near 10:1. At the system level it may be near 2;1, which is
            great progress but not near what the casual reader would infer.
            The big gain might be in cost, especially for non-concentrating
            arrays. I would guess that this development could get us near or
            below $2.00/watt, or maybe near 10 cents/kWh. Murray
          • lawrence_01749
            Murray, I looked at the site and pictures but I could not make out what s the secret sauce to this breakthrough . Maybe I missed something. What do they do
            Message 5 of 8 , Jan 14, 2004
            • 0 Attachment
              Murray, I looked at the site and pictures but I could not make out
              what's the secret sauce to this "breakthrough". Maybe I missed
              something. What do they do that enables them to use 1/10 the
              silicon of typical wafers sawn from boule?

              While you're at it, look at Evergreen Solar, in my neighborhood.
              They claim to use half the silicon of sawn wafers, pulling up a
              silicon ribbon from melt.

              Dick Lawrence

              --- In energyresources@yahoogroups.com, "mduffin3" <murrayv@m...>
              wrote:
              > --- In energyresources@yahoogroups.com, "terence_ms"
              > <terence_ms@y...> wrote:
              > > This seem like the way to go. Silicon is much more environmentaly
              > > friendly than some of the other toxic crap used nowdays for solar
              > > cells.
              > >
              > > And the other big plus where it claims it uses 90% less silicon
              > means
              > > that the energy used to produce the cells is going to be much
              less
              > > thereby making the energy payback period much shorter.
              > >
              > > If I had money, I back this lot all the way.
              > >
              > >It looks like an interesting step, but there are problems. Those
              > slivers are going to be a bitch to handle and assemble into arrays
              on
              > a high volume production basis, and even worse for a bifacial
              array.
              > Assembly will also eat up a good bit of the energy saving. The
              energy
              > saving at the level of a single cell might be as high as 4:1, it
              > won't be near 10:1. At the system level it may be near 2;1, which
              is
              > great progress but not near what the casual reader would infer.
              > The big gain might be in cost, especially for non-concentrating
              > arrays. I would guess that this development could get us near or
              > below $2.00/watt, or maybe near 10 cents/kWh. Murray
            • mduffin3
              ... Conventionally the horizontal surface of the wafer is all the surface you have for the PV. The thickness of the wafer is far greater than needed (like
              Message 6 of 8 , Jan 15, 2004
              • 0 Attachment
                --- In energyresources@yahoogroups.com, "lawrence_01749"
                <lawrence_01749@y...> wrote:
                > Murray, I looked at the site and pictures but I could not make out
                > what's the secret sauce to this "breakthrough". Maybe I missed
                > something. What do they do that enables them to use 1/10 the
                > silicon of typical wafers sawn from boule?
                >
                > While you're at it, look at Evergreen Solar, in my neighborhood.
                > They claim to use half the silicon of sawn wafers, pulling up a
                > silicon ribbon from melt.
                >
                Conventionally the horizontal surface of the wafer is all the surface
                you have for the PV. The thickness of the wafer is far greater than
                needed (like maybe a factor of 10), but "thinness" is limited by saw
                control and mechanical strength. They are selectively etching clean
                through the wafer on the 111 crystal orientation with the etch trench
                orders of magnitude thinner than a saw kerf, and then turning the
                resulting slivers on their side so that the wafer thickness is
                converted to surface area. The slivers are also much thinner than a
                wafer because there is no saw control issue and much smaller
                mechanical strength requirement. You want to use relatively thick
                wafers to maximize the benefit, but the thickness will be limited by
                etch rate and control.
                Think of a one square ft. 2" thick board. You have 144 sq. in. of
                surface (on one side). Now saw that board into strips 1/10th in.
                thick, with a saw kerf of say 1/50th inch.Now lay the strips on their
                side.You will have about 118 strips, each 12" by 2" or 2800 sq. in.,
                a 20x increase in surface area. If you can start with a 4" thick
                board you get 40x.
                However their slivers are extremely thin, not very wide and
                potentially orders of magnitude longer than they are wide, although
                they must control the form factor. ImaGINE mounting these slivers
                only at there ends and separated by their own thickness for bifacial
                irradiation, considering how small and fragile they are. and you will
                have thousands of them from a single wafer. Neat idea, but as I said,
                mounting will be a bitch.
                Also the portion of the boule that can be used is limited by
                resistivity that is not uniform along it's length. You dope the
                wafers , but want to use a single recipe for mass production, so the
                starting resisitivity is limited in range. These guys are doping the
                vertical wall of the etch trench, which is very uniform for a run of
                wafers, and because of the huge increase in surface area they can
                adjust the recipe for a different wafer batch, thus using more of the
                boule. Murray
              • lawrence_01749
                thanks Murray, good explanation. As you say, handling, mounting, and wiring these slivers gotta be a bitch, would take some fairly complex automated handling
                Message 7 of 8 , Jan 15, 2004
                • 0 Attachment
                  thanks Murray, good explanation.

                  As you say, handling, mounting, and wiring these slivers gotta be a
                  bitch, would take some fairly complex automated handling equipment,
                  maybe some modified wirebonding equips from the semi packaging
                  industry. After all that I'm not convinced they would be better off
                  than Evergreen pulling a ribbon out of the melt. Do they have to
                  use diffusion doping, or can they do that with ion beams?

                  Dick L

                  --- In energyresources@yahoogroups.com, "mduffin3" <murrayv@m...>
                  wrote:
                  > --- In energyresources@yahoogroups.com, "lawrence_01749"
                  > <lawrence_01749@y...> wrote:
                  > > Murray, I looked at the site and pictures but I could not make
                  out what's the secret sauce to this "breakthrough". Maybe I missed
                  something. What do they do that enables them to use 1/10 the
                  silicon of typical wafers sawn from boule?
                  > >
                  > > While you're at it, look at Evergreen Solar, in my
                  neighborhood. They claim to use half the silicon of sawn wafers,
                  pulling up a silicon ribbon from melt.
                  > >
                  > Conventionally the horizontal surface of the wafer is all the
                  surface you have for the PV. The thickness of the wafer is far
                  greater than needed (like maybe a factor of 10), but "thinness" is
                  limited by saw control and mechanical strength.
                • mduffin3
                  ... off ... Evergreen has a nice looking process but for some reason I could not download their data sheets. My computer locks up. They claim to have reached
                  Message 8 of 8 , Jan 16, 2004
                  • 0 Attachment
                    --- In energyresources@yahoogroups.com, "lawrence_01749"
                    <lawrence_01749@y...> wrote:
                    > thanks Murray, good explanation.
                    >
                    > As you say, handling, mounting, and wiring these slivers gotta be a
                    > bitch, would take some fairly complex automated handling equipment,
                    > maybe some modified wirebonding equips from the semi packaging
                    > industry. After all that I'm not convinced they would be better
                    off
                    > than Evergreen pulling a ribbon out of the melt. Do they have to
                    > use diffusion doping, or can they do that with ion beams?
                    >
                    > Dick L
                    >
                    Evergreen has a nice looking process but for some reason I could not
                    download their data sheets. My computer locks up. They claim to have
                    reached 15% conversion efficiency, but I would guess they don't spec.
                    more than 12. In principle their pull from the melt should be single
                    crystal silicon, but that thin strip is going to cool very rapidly,
                    leaving a lot of crystal imperfections. The guys in Australia have
                    experimented with surface treatments like selective etching of
                    pyramidal faces to increase surface and capture of reflected light,
                    thus raising effective efficiency. I wonder if Evergreen has tried
                    such ideas. It seems like the Evergreen thin film is about the
                    thickness of a conventional wafer, so they get the doubled Si
                    utilization by eliminating the saw kerf.
                    The Australian approach gets easily a 10x rather than 2x Si
                    utilization, etching is a very low energy process, and the slivers
                    will have superior thermal conductivity due to their thinness, so
                    they have some big pluses. Also by using thick wafers they reduce saw
                    wastage.
                    I was thinking about the mounting problem, and have decided it is not
                    so bad. My first thought was pick and place, but that is probably not
                    the way to go. There are a couple of metal alloys that heve a
                    coefficient of thermal expansion very close to that of silicon, and
                    have excellent thermal conductivity. Those slivers are symmetric, so
                    it doesnt matter which face is up. After etching, the slivers could
                    be laid out as they fall and have one surface passivated for
                    protection. They could then be loaded in a shaker and distributed
                    with uniform orientation onto a continuous strip of metal that was
                    selectively pretreated with soft solder. the strip would then go
                    through a low temperature furnace to complete the soldering. The
                    result woud be a nice strip to use in, say, a 20x concentrator. If
                    the passivation was photoresist or equivalent it could then be
                    removed in a low energy plasma. Surface treatment to increase
                    efficiency would probably still be practical.
                    To answer your question, from the paper I think they must use
                    diffusion, because they must dope the junction before separating the
                    slivers, or they wouldn't be bifacial, and there is no way to implant
                    in those trenches.
                    I'm starting to think that this idea is a real barnburner. The above
                    described assembly method, if practical, is both simple and
                    relatively low energy, and the resulting strips would be rugged and
                    easy to handle. Murray
                  Your message has been successfully submitted and would be delivered to recipients shortly.