letters to nature agnostics. P. Gottig and R Ramachandran hel corrections hank G. Bourenkov and H. bartunik. and g. leonard with synchrotron data collection at DESY BW6(Hamburg) and ESRF ID14-4(Grenoble), respectively. Competing interests statement Self-assembled monolayer organic he authors declare that they have no competing financial interests. field-effect transistors ence should be addressed to H B. (e-mail: hbsepbiochem. mpg. de).The ates of the tricorn protease have been deposited in Protein Data Bank under Jan Hendrik Schon, Hong Meng &Zhenan Bao on code 1K32 Nare4l3,713-716(2001) The values of the transconductance in table l and in the tey 715, second paragraph) are incorrect. The values should be addendum by ten. The data plotted in Figs 2 and 3 are correct conclusions are not affected An efficient room-temperature silicon-based light-emitting diode Wai Lek Ng, M. A Lourenco, R. M. Gwilliam, S Edain, G shao Ordered nanoporous arrays of carbon K P Homewood supporting high dispersions of Silicon light-emitting diodes (led) show light emission at the platinum nanoparticles bandgap energy of silicon with efficiencies approaching those of standard Ill-Vemitters: 0. 1% for planar devices(our Letter) and sang Hoon Joo, Seong Jae Choi, lwhan Oh, Juhyoun Kwak, about 1% when total internal reflection is minimized by surface Zheng Liu, Osamu Terasaki& Ryong Ryoo turing. We point out here an additional example of a silico device also show t emission at the bandgap the authors described devices made by the SACMos-3 a plas. However, and focus the We inadvertently omitted to cite an earlier reference bulk of the paper on visible emission under reverse (G Che, B Lakshmi, E R. Fisher and C. R. Martin No they also report briefly on a device operated under forward bias 349; 1998), which was published in 1995(and not 20 inted) ving efficiencies of around 0.01%, although no explanation of the Also, our suggestion that using the pores in a microporous material mechanism is given. It is now becoming clear that crystalline silicon, as templates could be a way in which to produce nanoscale materials when appropriately engineered, is capable of supporting efficient has been discussed before(see, for example, C. R. Martin Science light emission, opening up many significant applications. O 266, 1961-1966(1994)and J. C Hulteen C R. Martin J Mater. I. Green, M. A Shao, l, Wang, A, Reece, P. L.& Gal, M. Efficient silicon light-emitting diodes. Natur Chem.7,1075-1087(1997) 412,805-808(2001) Kramer, I. et aL Light-emitting devices in ind 527-533(1993) erratum The timing of the last deglaciation in North atiantic climate records Warm tropical sea surface Claire Waelbroeck, Jean-Claude Duplessy, Elisabeth Michel, Laurent Labeyrie, Didier Paillard Josette Duprat temperatures in the Late cretaceous Nature412,724-727(2001) and Eocene epochs We directly used the observed leads of sea surface temperature wi Paul N Pearson, Peter W. Ditchfield, Joyce Singan Katherine g. harcourt-Brown christ aspect to air temperature( dated in calendar years), whereas the air Richard K. olsson Nicholas j Shac Mike A ha with reservoir ages computed as the difference between marine and atmospheric ages. Taking this into consideration, apparent Nature413,481-487(2001) surface-water ages are 1, 180+ 630 to 1, 880+ 750 years at the end of the Heinrich I surge event(14, 500 years BP)and 930+ 250 to In this Article, the temperature scale in Figure 3i should have been 1,050+ 230 years at the end of the Younger Dryas cold episode. This the same as in Figure 3g s not change the discussion and conclusions 470 A@2001 Macmillan Magazines Ltd ATURE VOL 414/22 NOVEMBER 20011letters to nature 470 NATURE | VOL 414 | 22 NOVEMBER 2001 | www.nature.com Acknowledgements We acknowledge ®nancial support of the Deutsche Forschungsgemeinschaft and of Roche Diagnostics. P. GoÈttig and R. Ramachandran helped with biochemical analyses. We thank G. Bourenkov and H. Bartunik, and G. Leonard for help with synchrotron data collection at DESY BW6 (Hamburg) and ESRF ID14-4 (Grenoble), respectively. Competing interests statement The authors declare that they have no competing ®nancial interests. Correspondence should be addressed to H.B. (e-mail: hbs@biochem.mpg.de). The coordinates of the tricorn protease have been deposited in Protein Data Bank under accession code 1K32. ................................................................. addendum An ef®cient room-temperature silicon-based light-emitting diode Wai Lek Ng, M. A. LourencË o, R. M. Gwilliam, S. Ledain, G. Shao & K. P. Homewood Nature 410, 192±194 (2001). .................................................................................................................................. Silicon light-emitting diodes (LED) show light emission at the bandgap energy of silicon with ef®ciencies approaching those of standard III±V emitters: 0.1% for planar devices (our Letter) and about 1% when total internal re¯ection is minimized by surface texturing1 . We point out here an additional example of a silicon device also showing light emission at the bandgap2 . The authors described devices made by the SACMOS-3 process and focus the bulk of the paper on visible emission under reverse bias. However, they also report brie¯y on a device operated under forward bias giving ef®ciencies of around 0.01%, although no explanation of the mechanism is given. It is now becoming clear that crystalline silicon, when appropriately engineered, is capable of supporting ef®cient light emission, opening up many signi®cant applications. M 1. Green, M. A., Shao, J., Wang, A., Reece, P. J. & Gal, M. Ef®cient silicon light-emitting diodes. Nature 412, 805±808 (2001). 2. Kramer, J. et al. Light-emitting devices in industrial CMOS technology. Sensors Actuators A37±A38, 527±533 (1993). ................................................................. erratum Warm tropical sea surface temperatures in the Late Cretaceous and Eocene epochs Paul N. Pearson, Peter W. Ditch®eld, Joyce Singano, Katherine G. Harcourt-Brown, Christopher J. Nicholas, Richard K. Olsson, Nicholas J. Shackleton & Mike A. Hall Nature 413, 481±487 (2001). .................................................................................................................................. In this Article, the temperature scale in Figure 3i should have been the same as in Figure 3g. M ................................................................. corrections Self-assembled monolayer organic ®eld-effect transistors Jan Hendrik SchoÈn, Hong Meng & Zhenan Bao Nature 413, 713±716 (2001). .................................................................................................................................. The values of the transconductance in Table 1 and in the text (page 715, second paragraph) are incorrect. The values should be divided by ten. The data plotted in Figs 2 and 3 are correct and the conclusions are not affected. M ................................................................. correction Ordered nanoporous arrays of carbon supporting high dispersions of platinum nanoparticles Sang Hoon Joo, Seong Jae Choi, Ilwhan Oh, Juhyoun Kwak, Zheng Liu, Osamu Terasaki & Ryong Ryoo Nature 412, 169±172 (2001). .................................................................................................................................. We inadvertently omitted to cite an earlier reference alongside ref. 8 (G. Che, B. Lakshmi, E. R. Fisher and C. R. Martin Nature 393, 346± 349; 1998), which was published in 1995 (and not 2000 as printed). Also, our suggestion that using the pores in a microporous material as templates could be a way in which to produce nanoscale materials has been discussed before (see, for example, C. R. Martin Science 266, 1961±1966 (1994) and J. C. Hulteen & C. R. Martin J. Mater. Chem. 7, 1075±1087 (1997)). M ................................................................. correction The timing of the last deglaciation in North Atlantic climate records Claire Waelbroeck, Jean-Claude Duplessy, Elisabeth Michel, Laurent Labeyrie, Didier Paillard & Josette Duprat Nature 412, 724±727 (2001). .................................................................................................................................. We directly used the observed leads of sea surface temperature with respect to air temperature (dated in calendar years), whereas the air temperature calendar ages should have been converted into 14C ages, with reservoir ages computed as the difference between marine and atmospheric 14C ages. Taking this into consideration, apparent surface-water ages are 1,180 6 630 to 1,880 6 750 years at the end of the Heinrich 1 surge event (14,500 years BP) and 930 6 250 to 1,050 6 230 years at the end of the Younger Dryas cold episode. This does not change the discussion and conclusions. M © 2001 Macmillan Magazines Ltd