Seite - 122 - in Photovoltaic Materials and Electronic Devices

tobe less than0.02Ω. As this ismuchlower thanthe typical series resistance in the cell, the impact isnegligible. 7 Ohm  cm2  [28].  However,  the  specific  contact  resistance  between  TCO  and  Mo  was  found  to  be  three  orders  of  magnitude  lower,  in  the  range  of  10−5  Ω  cm2  [29].  From  the  specific  contact  resistance  (RSCR),  the  contribution  of  the  contact  resistance  to  the  overall  resistance  in  the  cell  can  be  estimated.  We  calculated  the  contact  resistance  for  a  1  cm2  cell.  This  was  done  for  different  widths  of  overlap  between  the  TCO  and  the  Mo,  as  shown  in  the  TCO/Mo  contact  width  in  Figure  8a.  For  a  1  cm2  solar  cell,  typical  total  series  resistances  are  between  1  and  2  ohm.  For  two  specific  contact  resistances  (Rscr),  the  contact  resistance  was  calculated  to  be  less  than  0.02  Ω.  As  this  is  much  lower  than  the  typical  series  resistanc   in  the  cell,  the  impact  is  negligible.  Figure  8.  Contact  resistance  for  a  1  cm2  cell  as  a  function  of  the  width  of  the  contact  area  between  the  TCO  and  the  Mo  (a)  and  the  metal  busbar  and  the  TCO  (b).  The  legend  shows  the  specific  contact  resistance  (Rscr,  in  Ω  cm2).  The  cell  length  is  5  mm.  For  printed  lines,  the  specific  contact  resistance  between  the  metal  and  the  TCO  is  between  0.01  and  0.05  Ω  cm2.  At  present,  the  factors  underlying  the  contact  resistance  is  under  research  in  our  group  and  include  the  ink  material,  the  ink  curing  conditions  and  the  TCO  sheet  resistance.  The  contribution  of  the  contact  resistance  was  calculated  for  a  variety  of  various  busbar/TCO  overlap  (contact)  widths  and  specific  contact  resistances.  Clearly,  Figure  8b  shows  that  the  contact  resistance  is  much  higher  than  for  the  TCO/Mo  case  and  can  give  a  significant  contribution  to  the  overall    series  resistance.  For  the  case  of  the  metallic  interconnect  combined  with  the  metallic  finger  grid,  the  contact  resistance  was  also  calculated.  In  this  case,  the  busbar  width  was  taken  to  be  50  μm  and  the  finger  widths  of  20,  60  and  100  μm  were  used  with  a  finger  spacing  is  0.7,  1.6  and  2.1  mm,  respectively.  As  a  result  of  the  higher  contact  area  between  the  metal  and  the  TCO,  the  contact  resistance  drops  to  values  below  1.5  Ohm,  as  shown  in  Figure  9.  This  makes  the  system  more  robust  against  the  occurrence  of  contact  resistance  between  the  printed  metal  and  the  TCO,  although  it  is  not  negligible.  Figure8. Contact resistance fora1cm2 cellasa functionof thewidthof thecontact area between the TCO and the Mo (a) and the metal busbar and the TCO (b). The legend shows the specific contact resistance (Rscr, inΩ cm2). The cell length is5mm. For printed li es, the specific ont ct resistance between the metal and the TCOisbetween0.01and0.05Ωcm2. Atpresent, the factorsunderlyingthecontact resistance is under research in our group and include the ink material, the ink curing conditionsandtheTCOsheet resistance. Thecontributionof thecontact resistance wa c l ulated for a v riety of v rious sbar/TCO overlap (contact) widths and specific contact resistances. Clearly, Figure 8b shows that the contact resistance is much higher than for the TCO/Mo case and can give a significant contribution to theoverall series resistance. For the case of the metallic interconnect combined with the metallic finger grid, thecontact resistancewasalsocalculated. In thiscase, thebusbarwidthwastakento be50µmandthefingerwidthsof20,60and100µmwereusedwithafingerspacing is 0.7, 1.6 and 2.1 mm, respectively. As a result of the higher contact area between the metal and the TCO, the contact resistance drops to values below 1.5 Ohm, as shown in Figure 9. This makes the system more robust against the occurrence of contact resistancebetweentheprintedmetalandtheTCO,althoughit isnotnegligible. Theeffectof thecontact resistanceonthecellperformancewascalculatedfor the casewithout andwith metallicgrid. For the case withoutmetallic grid, the data are presented in Figure 10 for a scribe with of 150µm and 350µm. We have used therangeofspecificcontact resistancebetween0.01and0.1Ωcm2. Theblack lines indicate thecasewithoutcontact resistance (TCOinterconnect). Aspecificcontact resistance of 0.01Ω cm2 has only minimal impact on the cell efficiency. However, for higher specific contact resistances, the impact is larger and the efficiency drops several absolute percent for the highest specific contact resistances calculated. For a 122