Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Linked Reactivity at Mineral-Water Interfaces Through Bulk Crystal Conduction

Linked Reactivity at Mineral-Water Interfaces Through Bulk Crystal Conduction Linked Reactivity at Mineral-Water Interfaces Through Bulk Crystal Conduction Svetlana V. Yanina and Kevin M. Rosso Science 320, 218 (2008); DOI: 10.1126/science.1154833 This copy is for your personal, non-commercial use only. If you wish to distribute this article to others, you can order high-quality copies for your colleagues, clients, or customers by clicking here. Permission to republish or repurpose articles or portions of articles can be obtained by following the guidelines here. The following resources related to this article are available online at www.sciencemag.org (this information is current as of December 9, 2013 ): Updated information and services, including high-resolution figures, can be found in the online version of this article at: http://www.sciencemag.org/content/320/5873/218.full.html Supporting Online Material can be found at: http://www.sciencemag.org/content/suppl/2008/03/06/1154833.DC1.html A list of selected additional articles on the Science Web sites related to this article can be found at: http://www.sciencemag.org/content/320/5873/218.full.html#related This article cites 31 articles, 2 of which can be accessed free: http://www.sciencemag.org/content/320/5873/218.full.html#ref-list-1 This article has been cited by 27 article(s) on the ISI Web of Science This article has been cited by 5 articles hosted by HighWire Press; see: http://www.sciencemag.org/content/320/5873/218.full.html#related-urls This article appears in the following subject collections: Geochemistry, Geophysics http://www.sciencemag.org/cgi/collection/geochem_phys Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright 2008 by the American Association for the Advancement of Science; all rights reserved. The title Science is a registered trademark of AAAS. Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from www.sciencemag.org www.sciencemag.org www.sciencemag.org www.sciencemag.org www.sciencemag.org www.sciencemag.org on December 9, 2013 on December 9, 2013 on December 9, 2013 on December 9, 2013 on December 9, 2013 on December 9, 2013 REPORTS boundary shows maximum Ir concentrations ranging from 21. J. T. Wasson, G. W. Kallemeyn, Philos. Trans. R. Soc. concentration was similar to the modern value, 0.1 to 87 ng/g in 85 sections from deep marine to London Ser. A 325, 535 (1988). because Eocene estimates of Os burial flux (35) and continental depositional environments (4). Os residence time made in this work are very similar 22. A doubling of seawater Os concentration to 20 pg/kg 6. J. M. Luck, K. K. Turekian, Science 222, 613 (1983). would erase the Os deficit. to recent values (36–38). 7. C. Koeberl, S. B. Shirey, Palaeogeogr. Palaeoclimatol. 35. T. K. Dalai, G. E. Ravizza, B. Peucker-Ehrenbrink, Earth 23. The diameter D is estimated assuming a spherical Palaeoecol. 132, 25 (1997). projectile D (km) = {2 × [mass Os × 3/[Os] Planet. Sci. Lett. 241, 477 (2006). impactor impactor /3 8. T. Meisel, U. Krähenbähl, M. A. Nazarov, Geology 23, 36. T. K. Dalai, G. E. Ravizza, Geochim. Cosmochim. Acta 70, (ng/g) × r (kg/m )× 4P] }/1E4 with mass impactor 313 (1995). Os derived from (18). [Os] is the Os 3928 (2006). impactor impactor 9. R. Frei, K. M. Frei, Earth Planet. Sci. Lett. 203, 691 37. S. Levasseur, J.-L. Birck, C. J. Allègre, Earth Planet. Sci. concentration in the impactor (21), and r is the impactor (2002). density of the impactor (25). Lett. 174, 7 (1999). 10. D. G. Pearson et al., Geol. Soc. Am. 31, 123 (1999). 24. R. Tagle, P. Claeys, Geochim. Cosmochim. Acta 69, 2877 38. R. Oxburgh, Earth Planet. Sci. Lett. 159, 183 (1998). 11. G. Quitté et al., Meteorit. Planet. Sci. 42, 1567 (2007). (2005). 39. K. A. Farley, A. Montanari, E. M. Shoemaker, 12. B. Peucker-Ehrenbrink, G. E. Ravizza, A. W. Hofmann, 25. R. L. Korotev, Washington Univ. (St. Louis); available at C. S. Shoemaker, Science 280, 1250 (1998). 187 188 Earth Planet. Sci. Lett. 130, 155 (1995). http://meteorites.wustl.edu/id/density.htm. 40. For example, the lowest K-T boundary Os/ Os ever 13. G. Ravizza, B. Peucker-Ehrenbrink, Science 302, 1392 26. In the calculations (23), an Os concentration ([Os] ) reported (0.137) (8) corresponds to a projectile size of impactor (2003). of 490 ng/g (21) and an average density (r )of greater than 6 km in diameter. On the basis of the impactor 14. The marine Os isotope record, like the Ir fluence 3.35 g/cm (25) were representative of an L chondrite unusually large Os concentration at this section, we approach, is not well suited to detecting the impact of believe that this section does not accurately reflect the meteorite (Popigai LEI event), whereas an Os concentration 3 187 188 differentiated projectiles, such as achondrites with of 807.5 ng/g (21) and an averaged density of 3 g/cm true Os/ Os concentration of seawater immediately siderophile-element concentrations that are two to three after the K-T event. Instead, we suspect a substantial (25) were used as representative of a carbonaceous orders of magnitude lower than chondrites (7). chondrite (28) for the K-T event. inventory of particulate impact-derived Os. 15. M. F. Horan, R. J. Walker, J. W. Morgan, J. N. Grossman, 41. J. W. Morgan, M. F. Horan, R. J. Walker, J. N. Grossman, 27. The Ir-based estimate is calculated by knowing the Ir A. E. Rubin, Chem. Geol. 196, 5 (2003). fluence (3, 20), chondritic Ir concentration (Os/Ir = 1.08) Geochim. Cosmochim. Acta 59, 2331 (1995). 16. J. Whitehead, D. A. Papanastassiou, J. G. Spray, 42. G. S. Collins, H. J. Melosh, R. A. Marcus, Meteorit. Planet. (15), average density of a chondrite (25), and Earth R. A. F. Grieve, G. J. Wasserburg, Earth Planet. Sci. Lett. global Ir inventory. Sci. 40, 817 (2005). 181, 473 (2000). 43. F. T. Kyte, Deep-Sea Res. II 49, 1049 (2002). 28. F. T. Kyte, Nature 396, 237 (1998). 17. Methods and data table are available as supporting 29. G. S. Collins, K. Wunnemann, Geology 33, 925 (2005). 44. J. E. T. Channell et al., Geol. Soc. Am. Bull. 115, 607 (2003). material on Science Online. 30. The Ir and Os isotope-based projectile-size estimates 45. H. Pälike et al., Sci. Res., Proc. Ocean Drill. Prog. (Ocean 18. The increase in the seawater Os reservoir after an probably represent the integrated signal from both the Drilling Program, College Station, TX, 2005), vol. 199. impact event is determined by knowing (i) the Os Popigai and the Chesapeake Bay impacts, because spherules 46. We thank the ODP for providing the samples, D. Vonderhaar concentration in modern seawater (~10 pg/kg) (19) and believed to derive from both events are found within the for technical assistance, E. Scott for his advice and expertise on (ii) the estimated mass of the seawater (~1.4 × 10 kg). single Ir peak at ODP 1090 (20). The smaller estimated meteorites, and R. Smith for support. Comments by three The product of these gives the mass of Os in the ocean size for the Chesapeake Bay projectile (29) suggests anonymous reviewers greatly improved the paper. This work before an impact [mass Os (ng)]. The mass of Os that >90% of the Os and Ir released to the environment was supported by NSF grants OCE and EAR to G.E.R. and B.P.-E. sw derived from the impactor [mass Os (ng)] is impactor was derived from the Popigai event. 187 188 (f/1 – f)×mass Os ;where f =( Os/ Os – 31. A. D. Anbar, G. J. Wasserburg, D. A. Papanastassiou, sw postimpact Supporting Online Material 187 188 187 188 187 188 Os/ Os )/( Os/ Os – Os/ Os ). preimpact impactor preimpact P. S. Andersson, Science 273, 1524 (1996). www.sciencemag.org/cgi/content/full/320/5873/214/DC1 The fractional increase in the seawater Os reservoir is (mass 32. C.-T. A. Lee, G. J. Wasserburg, F. T. Kyte, Geochim. Materials and Methods Os +massOs )/(mass Os ) sw impactor sw . Cosmochim. Acta 67, 655 (2003). SOM Text 19. S. Levasseur, J.-L. Birck, C. J. Allègre, Science 282, 272 33. D. S. Ebel, L. Grossman, Geology 33, 293 (2005). Table S1 (1998). 34. In the steady-state, the relation between reservoir size 20. F. T. Kyte, S. Liu, Lunar Planet. Sci. XXXIII, 1981 N, removal flux F, and residence time t is t = N/F. 12 November 2007; accepted 29 February 2008 (abstr.) (2002). This relation implies that the Eocene seawater Os 10.1126/science.1152860 face planes, exchange of mass or electron equiv- Linked Reactivity at Mineral-Water alents between sites of differing potential energy at different locations on any given crystal is typ- ically assumed to be negligible. This assumption Interfaces Through Bulk is nonetheless questionable for the widespread group of minerals that are electrical semiconduc- Crystal Conduction tors. For example, iron oxides often have moder- ate to low electrical resistivity (1) and have been Svetlana V. Yanina and Kevin M. Rosso* studied as electrode materials for decades (2–4). Iron oxide crystal surfaces are chemically reactive The semiconducting properties of a wide range of minerals are often ignored in the study of their with water and ions, leading to solution-dependent interfacial geochemical behavior. We show that surface-specific charge density accumulation charging behavior that differs from one surface reactions combined with bulk charge carrier diffusivity create conditions under which interfacial type to the next; differing points of zero charge electron transfer reactions at one surface couple with those at another via current flow through the for proton adsorption is but one example (5, 6). crystal bulk. Specifically, we observed that a chemically induced surface potential gradient This difference should give rise to a surface elec- across hematite (a-Fe O ) crystals is sufficiently high and the bulk electrical resistivity sufficiently 2 3 tric potential gradient (Dy ) across any crystal low that dissolution of edge surfaces is linked to simultaneous growth of the crystallographically that has two or more structurally distinct faces distinct (001) basal plane. The apparent importance of bulk crystal conduction is likely to be exposed to solution. In principle, this gradient generalizable to a host of naturally abundant semiconducting minerals playing varied key roles in can bias the diffusion of charge carriers (7, 8). soils, sediments, and the atmosphere. Hence, conditions could exist when the gradient across a single crystal is sufficiently large and the he chemical behavior of mineral-water in- at these interfaces have probed the interaction of electrical resistivity of the material sufficiently terfaces is central to aqueous reactivity in water and relevant dissolved ions with crystallo- Tnatural waters, soil evolution, and atmo- graphically well-defined mineral surfaces. The Chemical and Materials Sciences Division, Pacific Northwest spheric chemistry and is of direct relevance for pursuit so far has been dominated by the assump- National Laboratory, Post Office Box 999, MSIN K8-96, Richland, maintaining the integrity of waste repositories tion that distinct surfaces of any given crystal WA 99352, USA. and remediating environmental pollutants. Tra- behave independently of each other. Except by *To whom correspondence should be addressed. E-mail: ditionally, explorations of fundamental reactions diffusion through the solution phase or across sur- kevin.rosso@pnl.gov 218 11 APRIL 2008 VOL 320 SCIENCE www.sciencemag.org REPORTS 2+ low that interfacial electron transfer reactions at (10, 11); reported bulk resistivities range from Fe in solution, a range of potential-determining 2 6 one surface couple with those at another by a 10 to 10 ohm·m(1). When hematite is sub- ions, electron transfer across hematite-solution current flowing spontaneously through the crys- jected to oxygen-limited aquatic environments, interfaces, and the possibility of moving electron tal bulk. The situation is analogous to galvanic particularly in acidic conditions, it can be reduc- equivalents through the crystal bulk. metal corrosion, but instead of spatially dis- tively dissolved according to Eq. 1: Development of a potential gradient Dy of ordered anodic and cathodic electron transfer significant magnitude across the crystal requires + − 2+ sites, the anode and cathode are spatially con- Fe O +6H +2e → 2Fe +3HO(1) selective interaction between potential-determining 2 3 (aq) 2 fined to crystallographically distinct surface planes ions and specific hematite surfaces. We focus here and are therefore physically separable for mea- which has a standard reduction potential E°~0.7V on roles of protons (low pH) and oxalate as a surement. We demonstrate operability of these (8, 12). The fundamental reaction central to the representative dicarboxylic acid. The hematite conditions for iron oxide, uncover their effects on overall process is (001) basal surface is structurally distinct from the surface chemical behavior, and make the case any edge surfaces. In water, the (001) surface is 3+ − 2+ that, in nature, different surfaces of certain abun- Fe + e → Fe (2) terminated predominantly by doubly coordinated (s) (aq) dant crystals are inextricably linked. hydroxyls (17–19) that are relatively inert to the 2+ We examined hematite (a-Fe O ) because it is Sorbed Fe from the aqueous phase is capable of protonation and deprotonation reactions needed 2 3 3+ a wide band gap semiconductor (band gap 1.9 to reducing hematite Fe in this system (13–15), for charge accumulation. Smaller populations of 2.3 eV) (1, 3) and the most stable form of iron yielding an iron redox cycle in which no net re- more reactive singly coordinated and triply co- oxide under dry oxidizing conditions; it is ordinated hydroxyls, capable of positive charge duction occurs. Introduction of dicarboxylic acids extremely common in nature (9). It has the such as oxalate causes net dissolution by chelating accumulation, are associated with terminal Fe 3+ corundum structure type based on hexagonal close surface Fe (ligand assisted dissolution); it also groups (19). Terminal Fe groups with low coor- packed oxygen planes in which 2/3 of the available enhances reduction, possibly through the forma- dination to the underlying surface can be easily 3+ 2+ octahedral cavities are occupied by Fe .This tion of ternary surface complexes with Fe ,for chelated by oxalate anions to form negatively (aq) structure gives rise to anisotropic electrical resis- example (16). This collective chemistry is a good charged mononuclear bidentate inner-sphere sur- tivity that is higher in the basal plane than along the test case for our main hypothesis because it face complexes (20, 21). In contrast, many remain- trigonal axis by up to four orders of magnitude involves a source of electron equivalents from ing low-index surfaces of hematite crystals such as (012) are dominated by higher-coordinated Fe (17, 22). This higher-coordinated Fe is more dif- Fig. 1. AFM images of representative A E ficult to chelate yet bears singly and/or triply co- hematite surfaces of the natural samples ordinated hydroxyls for charge accumulation. In before (A to D) and after (E to H) general therefore, we expect that, relative to other reaction at 75°C for 12 hours in the pH 2+ hematite surfaces, the (001) surface should show range of 2 to 3 in 1 mM Fe and 10 mM weaker pH-dependent charge accumulation, an ob- oxalate solution free of oxygen. Initial servation increasingly confirmed by recent data surface morphologies [(A) to (D) 10 mm and theory (23, 24), and stronger interaction with by 10 mm deflection images] for (A) a oxalate anions that increases with decreasing pH. (001) surface, (B) a (hk0) surface, (C) a (012) surface, and (D) a (113) surface We performed several experiments to poten- are extremely flat and in all cases except tiometrically measure Dy for specific hematite B F (hk0) crystallographically well ordered surfaces and to determine its effect on their 2+ into terrace-and-step structures at the behavior in Fe and oxalate solutions. We chose micrometer scale. At higher resolution a large natural specular hematite crystal with [(B) inset, 2 mmby2 mm deflection im- well-defined surfaces that could be isolated for age], the (hk0) surface is microfaceted study. The crystal was low in impurity content with edge terminations. Reacted surfaces (25), a natural n-type semiconductor (1), and had [(E) to (H), 20 mmby20 mmdeflection a room temperature electrical resistivity of 10 images with corresponding topographic ohm·m as measured by the four-point probe height images in the insets] show a sharp method. Generating replicate samples required distinction between (E) the (001) surface cutting specific crystallographic surfaces from C G behavior and (F) the (hk0) surface, (G) the the crystal as rectangular prism-shaped specimens, (012) surface, and (H) the (113) surface. which also yielded a vicinal surface type along Large hematite pyramids of uniform cut edges. For example, we prepared millimeter- orientation nucleate as islands on the sized oriented prisms exposing two (001) sur- initially flat (001) surface (E), achieving faces on the top and bottom of the prism and average heights of 200 nm and average four orthogonal (hk0) vicinal sides (25). Anneal- lateral diameters on the order of a mi- ing in air under conditions where hematite is crometer. In contrast, all other surfaces the only stable iron oxide effectively cleans and dissolve. Dissolution of the (hk0) surface organizes the surfaces without modifying the bulk yields large irregular pits and coarsens the D H electrical conductivity. This procedure yields faceted appearance of the surface (F). Dissolution of (012) and (113) surfaces highly organized (001) surfaces, with accompany- yields large crystallographically controlled ing (hk0) vicinal surfaces that are microfaceted etch pits combined with smaller scale with stable edge terminations (Fig. 1, A and B). roughening [(G) and (H), respectively]. Similarly, prism specimens bearing (012) and (113) surfaces (Fig. 1, C and D) with accompany- ing vicinal surfaces were prepared. To determine the magnitude of Dy and the 2+ roles of Fe and oxalate solution components, www.sciencemag.org SCIENCE VOL 320 11 APRIL 2008 219 REPORTS we measured the open-circuit potential (E )in pseudo-pyramidal morphology of uniform orien- lected area diffraction measurements along [001] OCP four solution types. The E is the electrode rest tation. Images at early stages show the island transects of this sample type (fig. S1), along with OCP potential relative to a standard reference elec- growth of these features on the initially flat (001) x-ray photoelectron spectroscopy, x-ray diffrac- trode. Changes in the E are directly related to surface (Fig. 1E). After 12 hours, the reaction tion, and energy dispersive x-ray spectroscopy, OCP changes in y (26, 27), which inturnis sensitive yielded merged pyramid-covered (001) surfaces confirmed that the grown material is structurally to surface complexation reactions with our with peak-to-valley heights averaging 200 nm and compositionally a-Fe O of identical orien- 2 3 + − 2+ potential-determining ions H ,Cl ,Fe ,and and pyramid bases approaching a micrometer in tation as the underlying material without detect- oxalate species (28). The measurements were per- width, imparting a distinct matte appearance to able impurities. The line of intersection of formed at room temperature at effectively con- the reacted (001) surface visible to the naked eye. apparent pyramid “facets” with the (001) plane stant ionic strength under anaerobic conditions Transmission electron microscopy (TEM) and se- is consistent with lines of {012}/{001} intersec- (25). We focused these measurements on the (hk0) H C O 2 2 4 (001) and the accompanying (hk0) vicinal surface (001) H C O 2 2 4 A B type. The observed approximately linear pH de- (hk0) H C O + Fe (II) 2 2 4 pendence, with predominantly negative slopes, is 0.6 0.6 (001) H C O + Fe (II) 2 2 4 consistent with the accumulation of positive sur- face charge with decreasing pH (Fig. 2). As ex- 0.4 0.4 pected, in pure electrolyte solution the (001) surface showed a less-negative slope relative to that of the (hk0) surface, consistent with a lower 0.2 0.2 (hk0) KCl density of charge accumulation sites on the (001) (001) KCl surface (Fig. 2A). In contrast, oxalate anions bind (hk0) KCl + Fe(II) 0.0 0.0 preferentially to the (001) surface with decreasing (001) KCl + Fe(II) pH, even to the point of sign reversal in the slope 2+ (Fig. 2B). Addition of Fe to either solution 0 1 2 3 4 5 0 1 2 3 4 5 shows that its primary effect is to lower the over- pH pH all potential for both the (001) and (hk0) surfaces Fig. 2. The pH dependence of open-circuit poten- KCl without substantially modifying the pH depen- tials with respect to the normal hydrogen elec- dence (Fig. 2, A and B). Taking E – E as KCL + Fe (II) (hk0) (001) 0.4 trode (NHE) for (001) and (hk0) surfaces in (A) H C O an estimate of Dy , in the presence of oxalate and 0 2 2 4 pure electrolyte solution (10 mM KCl) with and 2+ H C O + Fe (II) 2+ irrespective of the presence of Fe we found (aq) 2 2 4 without Fe (1 mM FeCl )and in (B) electrolyte (aq) 2 0.2 that the potential gradient is large and positive, on solution (10 mM KCl) with oxalate (10 mM) with 2+ the order of tenths of volts below pH = 3 (Fig. and without Fe (1 mM FeCl ). The pH depen- (aq) 2 0.0 2C). Under these conditions, we expect that mo- dence of the potential difference between (001) bile electrons acting as majority carriers in hem- and (hk0) surfaces in the four solution types is -0.2 atite would be directed by Dy from the (001) shown in (C). Error bars indicate ± two standard surface to the (hk0) surface. E measurements deviations from a linear trend. OCP -0.4 directly between identical surface types [e.g., E – E ](25) showed no significant voltage. (001) (001) 0 1 2 3 4 5 To examine the effects of Dy of this sign and pH magnitude on the surface chemical behavior, we examined surfaces of the oriented prisms from the Fig. 3. Schematic diagram summariz- same sample by using atomic force microscopy ing the observed reaction behavior for (AFM) before and after anaerobic reaction with hematite crystals showing (A) (001) py- 2+ Fe -oxalate solutions. Thermostated batch ves- ramidal growth coupled to (hk0) disso- sels were used with temperatures ranging from lution, (B) (001) and (hk0) dissolution room temperature to 75°C and pH ranging from 2 for selectively sealed two-crystal cases 2+ to3(25). Fe -oxalate concentrations consistent where the same surface area and type with previously published experiments that estab- as in (A) are exposed to solution, and 3+ lish net dissolution in terms of Fe release on (C) (001) pyramidal growth coupled to (aq) fine-grained powders were used (16, 29, 30). Col- (hk0) dissolution facilitated by a con- ducting paste connection between two lectively, these conditions were selected in keeping crystals. with Eq. 1 while also accelerating surface trans- formations into a more easily observable time frame. Light was excluded in all cases to avoid oxalate acting as a reductant. Equilibrium thermo- dynamic calculations along with Eh measure- ments at run conditions confirm that all our 2+ reaction conditions lie within the Fe stabil- (aq) ity field (25). Hematite was the limiting reactant; total dissolution would retain undersaturation with respect to any possible iron oxide phases. AFM examination of (001) surfaces after re- sealed surface action runs showed remarkable features. In every [001] exposed surface case, for both natural and synthetic samples, sealed conducting paste (001) surfaces were overgrown with a hexagonal 220 11 APRIL 2008 VOL 320 SCIENCE www.sciencemag.org E (V), [E -E ] E (V vs. NHE) cell (hk0) (001) OCP E (V vs. NHE) OCP REPORTS tion, but the interplanar angle varies with run growths on the (001) surface (fig. S3B). Therefore, with coupling mediated by charge transport from duration and does not correspond to low-index the (001) pyramidal overgrowths do not form by (001) to (hk0) surfaces through the crystal bulk. planes in hematite. The large size, morphologic precipitation of ferric iron. Furthermore, we deduce The process involves preferential net oxidative 2+ symmetry, and mutual orientation of these pyra- that chemical processes at the (001)-solution in- adsorption of Fe at the (001)-solution inter- (aq) 3+ mids require homoepitaxy, that is, growth of terface causing pyramidal growth during reaction face and valence interchange with structural Fe additional hematite on hematite. are facilitated by solid contact between the (001) at that surface (Fig. 4). At temperatures of inter- In contrast, all other surfaces examined show and (hk0) surfaces; that is, these surfaces must be est (room temperature and higher), bulk charge features characteristic of dissolution. For exam- on thesamecrystal. transport is sufficiently facile to support a small ple, the four (hk0) vicinal sides of prism samples The behavior strongly suggests that bulk charge current through the bulk. Net electron equiv- bearing (001) surfaces on top and bottom ex- transport provides the link between the two types alents injected into the (001) surface follow an hibited fine-scale pitting and roughening (Fig. of surfaces. As a further test, we again prepared electrically biased random walk through the 1F). The (012) and (113) surfaces of prism sam- two crystals with partial exposure of (001) on one crystal to (hk0) surfaces. At (hk0) exit points, 3+ 2+ ples show development of etch pits at various and (hk0) surfaces on the other, except this time internal reduction of Fe to Fe solubilizes length scales and with symmetry corresponding with an electrical connection between them (Fig. and releases iron into solution. This circuit is to crystallographic orientation (Fig. 1, G and H). 3C). A crystal exposing only (001) surfaces was driven by the Dy gradient generated across the We observed identical behavior under the same connected to a crystal beneath exposing only crystal from divergent charge accumulation at conditions with use of synthetic tabular hematite (hk0) surfaces by electrically conductive colloidal structurally distinct surface types. The sign and crystals bearing primarily (001) and (012) sur- Ag paste, which was subsequently cured, sealed magnitude of Dy , the conductivity of the natu- faces, in which case no surface preparation by off from contact with solution using additional ral crystal, and the growth rates of the pyramidal annealing was required (fig. S2). The conclusion epoxy, and tested for ohmic behavior by resis- islands are all mutually consistent. For example, is that the hematite (001) surface grows under our tivity measurements. In this design, the crystals taking Dy = 0.2 V at pH = 2, a temperature- conditions whereas all other surfaces sampled are effectively wired together by the (001)-Ag- adjusted electrical resistivity = 10 ohm·m for dissolve, as represented schematically in Fig. 3A. (001) junction between them. Reaction in this 75°C (31, 32), and an electron transport path We designed experiments using the prism wired two-crystal configuration proceeds as if the length of 1 mm, the maximum amount of addi- samples to test whether or not pyramid islands crystals were one; pyramidal hematite grows on tional hematite expected on the (001) surface in are deposited on the (001) surface by precipita- the exposed (001) surface of the upper crystal 12 hours is a layer ~100 nm thick, the same order 3+ tion of trace Fe from solution (16). In these (fig. S3C), whereas the four (hk0) sides of the of magnitude as that observed. Surface potential– experiments, two prism samples were used in the lower crystal dissolve (Fig. 3C). Therefore, the driven charge carrier diffusivity has been invoked reaction vessel instead of one. Four (hk0) vicinal nature of the interaction between the (001) and qualitatively to explain microscopic oxide trans- sides of one crystal were sealed with an inert vicinal surfaces that gives rise to the pyramidal formation processes before (13, 33, 34) but not epoxy (25), leaving two (001) surfaces exposed, growth of hematite (001) during reaction derives on the length scale examined here nor with sur- whereas on the other crystal the two (001) sur- from bulk charge transport. Surface diffusion face specificity. Given the observation that the faces were sealed, leaving four vicinal surfaces along the hematite-solution interface was ruled (001) surface continues to grow beyond the co- exposed. Collectively, the two crystals expose the out by painting a ring of sealant on a (001) alescence of the pyramidal islands, at the atomic same six kinds of surfaces to solution as in the surface so that only bulk transport could access scale the pyramidal (001) morphology must runs above with one crystal, in the same relative the circumscribed region, and within that region retain the essential structural and therefore proportion and surface area, but they involve (001) hematite island growth also occurred (fig. S4). chemical characteristics that give rise to the surfaces that are physically separated from the The collective behavior of the system is potential of the initial (001) surface. Further- (hk0) surfaces (Fig. 3B). In this case, the results therefore suggestive of two distinct but coupled more, the observed process does not preclude of reaction runs show only dissolution features interfacial processes: growth at (001) by traditionally held spatially localized dissolution on all exposed surfaces, including (001) (fig. S3A). in the hematite system. Rather, the evidence 2+ 2+ 3+ − The (001) pyramidal morphology does not de- Fe → Fe → Fe + e (3) suggests that the processes operate in parallel (aq) (001) (001) velop in this separated two-crystal configuration. and that the behavior based on the electrical The same experiment performed on samples in and dissolution of edge surfaces, for example, circuit through the crystal dominates when which the pyramidal morphology had already (hk0) surfaces, by chemical requirements that establish a large been grown on (001) before sealing its (hk0) enough surface electric potential gradient are 3+ − 2+ 2+ Fe + e → Fe → Fe (4) (hk0) (hk0) (aq) sides showed dissolution of the pyramidal over- met. The finding provides insight into the reduc- Fig. 4. Schematic diagram depict- tive transformation of iron oxides, which is im- ing the inferred coupled interfacial portant in the biogeochemical cycling of iron in electron transfer process operative nature and the removal of iron oxide films in (Aqueous solution) under our conditions for the hema- industry. Because this finding can be easily gen- tite single crystals. The chemically eralized to a host of naturally abundant semi- self-induced surface potential gra- (001) (hk0) conducting transition metal oxide and sulfide dient across the crystal directs minerals capable of dominating the interfacial current flow through the bulk. The surface area in soils, sediments, and among current is facilitated by sufficiently atmospheric particles, its implications are fairly low electrical resistivity in a process widespread. Of immediate impact is the concept 2+ 2+ 3+ − that is fed by net injection of electron Fe Fe Fe + + aq (001) (001) that the reactivity of any given surface on such equivalents at (001) surfaces and net materials can be coupled to that of another sur- release of electron equivalents at (hk0) (Hematite bulk) + face, with a dependence on crystal morphology surfaces. 2+ Fe as a whole. This phenomenon should apply to 3+ Fe + natural crystals in the environment as well as those selectively cut, broken, or otherwise pre- 3+ − − 2+ 2+ Fe + + Fe Fe (hk0) (hk0) aq + pared for laboratory study. www.sciencemag.org SCIENCE VOL 320 11 APRIL 2008 221 REPORTS 15. P. Larese-Casanova, M. M. Scherer, Environ. Sci. Tech. 33. N. M. Dimitrijevic, D. Savic, O. I. Micic, A. J. Nozik, References and Notes 1. R. T. Shuey, Semiconducting Ore Minerals, vol. 4 of 41, 471 (2007). J. Phys. Chem. 88, 4278 (1984). 16. D. Suter, C. Siffert, B. Sulzberger, W. Stumm, 34. J. P. Jolivet, E. Tronc, J. Colloid Interface Sci. 125, 688 Developments in Economic Geology (Elsevier, Amsterdam, 1975). Naturwissenschaften 75, 571 (1988). (1988). 35. This research was supported by the U.S. Department of 2. J. H. Kennedy, K. W. Frese, J. Electrochem. Soc. 125, 723 17. V. Barron, J. Torrent, J. Colloid Interface Sci. 177, 407 (1996). (1978). 18. C. M. Eggleston et al., Geochim. Cosmochim. Acta 67, Energy (DOE), Office of Basic Energy Sciences, Geosciences Program. It was performed at the William 3. C. Gleitzer, J. Nowotny, M. Rekas, Appl. Phys. A Mat. Sci. 985 (2003). Proc. 53, 310 (1991). 19. T. P. Trainor et al., Surf. Sci. 573, 204 (2004). R. Wiley Environmental Molecular Sciences Laboratory 20. O. W. Duckworth, S. T. Martin, Geochim. Cosmochim. (EMSL) at the Pacific Northwest National Laboratory 4. B. A. Balko, K. M. Clarkson, J. Electrochem. Soc. 148, E85 (2001). Acta 65, 4289 (2001). (PNNL). The EMSL is funded by the DOE Office of 21. T. H. Yoon, S. B. Johnson, C. B. Musgrave, G. E. Brown, Biological and Environmental Research. PNNL is 5. P. Venema, T. Hiemstra, P. G. Weidler, W. H. van operated by Battelle for the DOE under contract Riemsdijk, J. Colloid Interface Sci. 198, 282 (1998). Geochim. Cosmochim. Acta 68, 4505 (2004). 22. J. R. Rustad, E. Wasserman, A. R. Felmy, Surf. Sci. 424, DE-AC06-76RLO 1830. We gratefully acknowledge the 6. F. Gaboriaud, J. Ehrhardt, Geochim. Cosmochim. Acta 67, assistance of C. Wang for TEM; B. Arey for scanning 967 (2003). 28 (1999). 23. T. Hiemstra,W. H.Van Riemsdijk, Langmuir 15, 8045 (1999). electron microscopy; D. McCready for pole reflection 7. C. G. B. Garrett, W. H. Brattain, Phys. Rev. 99, 376 x-ray diffraction; Y. Lin for access to electrochemistry (1955). 24. P. Zarzycki, Appl. Surf. Sci. 253, 7604 (2007). 8. P. Mulvaney, V. Swayambunathan, F. Grieser, D. Meisel, 25. Materials and methods are available on Science Online. apparatus; and A. Felmy, E. Ilton, and J. Amonette for comments on an early version of this J. Phys. Chem. 92, 6732 (1988). 26. M. J. Avena, O. R. Camara, C. P. Depauli, Colloid Surf. 9. R. M. Cornell, U. Schwertmann, The Iron Oxides: 69, 217 (1993). manuscript. Structure, Properties, Reactions, Occurrence and Uses 27. N. Kallay, T. Preocanin, J. Colloid Interface Sci. 318, 290 (VCH, Weinheim, Germany, 2003). (2008). Supporting Online Material 10. T. Nakau, J. Phys. Soc. Jpn. 15, 727 (1960). 28. J. A. Davis, R. O. James, J. O. Leckie, J. Colloid Interface www.sciencemag.org/cgi/content/full/1154833/DC1 11. N. Iordanova, M. Dupuis, K. M. Rosso, J. Chem. Phys. Sci. 63, 480 (1978). Materials and Methods 29. B. Zinder, G. Furrer, W. Stumm, Geochim. Cosmochim. 122, 144305 (2005). Figs. S1 to S4 12. J. S. LaKind, A. T. Stone, Geochim. Cosmochim. Acta 53, Acta 50, 1861 (1986). References 30. S. Banwart, S. Davies, W. Stumm, Colloid Surf. 39, 303 (1989). 961 (1989). 13. P. Mulvaney, R. Cooper, F. Grieser, D. Meisel, Langmuir 31. S. Kerisit, K. M. Rosso, Geochim. Cosmochim. Acta 70, 4 January 2008; accepted 25 February 2008 1888 (2006). Published online 6 March 2008; 4, 1206 (1988). 14. A. G. B. Williams, M. M. Scherer, Environ. Sci. Tech. 38, 32. S. Kerisit, K. M. Rosso, J. Chem. Phys. 127, 124706 10.1126/science.1154833 (2007). Include this information when citing this paper. 4782 (2004). other regions (3–6), Madagascar has complex, Aligning Conservation Priorities often nonconcordant patterns of microendemism among taxa (12–17), rendering the design of ef- ficient protected-area networks particularly diffi- Across Taxa in Madagascar with cult (4, 6). We collated data for endemic species in six major taxonomic groups [ants, butterflies, High-Resolution Planning Tools frogs, geckos, lemurs, and plants (table S1)], using recent robust techniques in species distribution 1,2 1,2 3 4 5 6 C. Kremen, *† A. Cameron, † A. Moilanen, S. J. Phillips, C. D. Thomas, H. Beentje, modeling (18, 19) and conservation planning 6 7 8 9 10 11 12 J. Dransfield, B. L. Fisher, F. Glaw, T. C. Good, G. J. Harper, R. J. Hijmans, D. C. Lees, (20, 21) to produce the first quantitative conserva- 13 14 15 2 16 E. Louis Jr., R. A. Nussbaum, C. J. Raxworthy, A. Razafimpahanana, G. E. Schatz, tion prioritization for a biodiversity hot spot with 17 18 19 9 M. Vences, D. R. Vieites, P. C. Wright, M. L. Zjhra this combination of taxonomic breadth (2315 species), geographic extent (587,040 km ), and Globally, priority areas for biodiversity are relatively well known, yet few detailed plans exist to spatial resolution (30–arc sec grid = ~0.86 km ). direct conservation action within them, despite urgent need. Madagascar, like other globally Currently, an important opportunity exists to recognized biodiversity hot spots, has complex spatial patterns of endemism that differ among influence reserve network design in Madagascar, taxonomic groups, creating challenges for the selection of within-country priorities. We show, in an given the government’s commitment, announced analysis of wide taxonomic and geographic breadth and high spatial resolution, that at the World Parks Congress in 2003, to triple its multitaxonomic rather than single-taxon approaches are critical for identifying areas likely to existing protected-area network to 10% coverage promote the persistence of most species. Our conservation prioritization, facilitated by newly (22). Toward this goal, our high-resolution anal- available techniques, identifies optimal expansion sites for the Madagascar government’s current ysis prioritizes areas by their estimated contribu- goal of tripling the land area under protection. Our findings further suggest that high-resolution tion to the persistence of these 2315 species and multitaxonomic approaches to prioritization may be necessary to ensure protection for biodiversity identifies regions that optimally complement the in other global hot spots. existing reserve network in Madagascar. We input expert-validated distribution models pproximately 50% of plant and 71 to analysis of these patterns is required to allocate for 829 species and point occurrence data for the 82% of vertebrate species are concen- conservation resources most effectively (7, 8). remaining species [those with too few occur- Atrated in biodiversity hot spots covering To date, only a few quantitative, high- rences to model, called rare target species (RTS)] only 2.3% of Earth’s land surface (1). These resolution, systematic assessments of conserva- into a prioritization algorithm, Zonation (20, 21), irreplaceable regions are thus among the highest tion priorities have been developed within these which generates a nested ranking of conservation global priorities for terrestrial conservation; rea- highly threatened and biodiverse regions (9, 10). priorities (23). Species that experienced a large sonable consensus exists on their importance This deficiency results from multiple obstacles, proportional loss of suitable habitat (range reduc- among various global prioritization schemes that including limited data or access to data on species tion) between the years 1950 and 2000 were given identify areas of both high threat and unique distributions and computational constraints on higher weightings [equation 2 of (23), (24)]. We biodiversity (2). The spatial patterns of species rich- achieving high-resolution analyses over large evaluated all solutions [defined here as the ness, endemism, and rarity of different taxonomic geographic areas. We have been able to over- highest-ranked 10% of the landscape to match groups within priority areas, however, rarely align come each of these obstacles for Madagascar, a the target that Madagascar has set for conservation and are less well understood (3–6). Detailed global conservation priority (1, 2, 11). Like many (22)] in two ways: (i) percent of species entirely 222 11 APRIL 2008 VOL 320 SCIENCE www.sciencemag.org http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Science Unpaywall

Linked Reactivity at Mineral-Water Interfaces Through Bulk Crystal Conduction

ScienceMar 6, 2008

Loading next page...
 
/lp/unpaywall/linked-reactivity-at-mineral-water-interfaces-through-bulk-crystal-Ep6jYqCwjp

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0036-8075
DOI
10.1126/science.1154833
Publisher site
See Article on Publisher Site

Abstract

Linked Reactivity at Mineral-Water Interfaces Through Bulk Crystal Conduction Svetlana V. Yanina and Kevin M. Rosso Science 320, 218 (2008); DOI: 10.1126/science.1154833 This copy is for your personal, non-commercial use only. If you wish to distribute this article to others, you can order high-quality copies for your colleagues, clients, or customers by clicking here. Permission to republish or repurpose articles or portions of articles can be obtained by following the guidelines here. The following resources related to this article are available online at www.sciencemag.org (this information is current as of December 9, 2013 ): Updated information and services, including high-resolution figures, can be found in the online version of this article at: http://www.sciencemag.org/content/320/5873/218.full.html Supporting Online Material can be found at: http://www.sciencemag.org/content/suppl/2008/03/06/1154833.DC1.html A list of selected additional articles on the Science Web sites related to this article can be found at: http://www.sciencemag.org/content/320/5873/218.full.html#related This article cites 31 articles, 2 of which can be accessed free: http://www.sciencemag.org/content/320/5873/218.full.html#ref-list-1 This article has been cited by 27 article(s) on the ISI Web of Science This article has been cited by 5 articles hosted by HighWire Press; see: http://www.sciencemag.org/content/320/5873/218.full.html#related-urls This article appears in the following subject collections: Geochemistry, Geophysics http://www.sciencemag.org/cgi/collection/geochem_phys Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright 2008 by the American Association for the Advancement of Science; all rights reserved. The title Science is a registered trademark of AAAS. Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from www.sciencemag.org www.sciencemag.org www.sciencemag.org www.sciencemag.org www.sciencemag.org www.sciencemag.org on December 9, 2013 on December 9, 2013 on December 9, 2013 on December 9, 2013 on December 9, 2013 on December 9, 2013 REPORTS boundary shows maximum Ir concentrations ranging from 21. J. T. Wasson, G. W. Kallemeyn, Philos. Trans. R. Soc. concentration was similar to the modern value, 0.1 to 87 ng/g in 85 sections from deep marine to London Ser. A 325, 535 (1988). because Eocene estimates of Os burial flux (35) and continental depositional environments (4). Os residence time made in this work are very similar 22. A doubling of seawater Os concentration to 20 pg/kg 6. J. M. Luck, K. K. Turekian, Science 222, 613 (1983). would erase the Os deficit. to recent values (36–38). 7. C. Koeberl, S. B. Shirey, Palaeogeogr. Palaeoclimatol. 35. T. K. Dalai, G. E. Ravizza, B. Peucker-Ehrenbrink, Earth 23. The diameter D is estimated assuming a spherical Palaeoecol. 132, 25 (1997). projectile D (km) = {2 × [mass Os × 3/[Os] Planet. Sci. Lett. 241, 477 (2006). impactor impactor /3 8. T. Meisel, U. Krähenbähl, M. A. Nazarov, Geology 23, 36. T. K. Dalai, G. E. Ravizza, Geochim. Cosmochim. Acta 70, (ng/g) × r (kg/m )× 4P] }/1E4 with mass impactor 313 (1995). Os derived from (18). [Os] is the Os 3928 (2006). impactor impactor 9. R. Frei, K. M. Frei, Earth Planet. Sci. Lett. 203, 691 37. S. Levasseur, J.-L. Birck, C. J. Allègre, Earth Planet. Sci. concentration in the impactor (21), and r is the impactor (2002). density of the impactor (25). Lett. 174, 7 (1999). 10. D. G. Pearson et al., Geol. Soc. Am. 31, 123 (1999). 24. R. Tagle, P. Claeys, Geochim. Cosmochim. Acta 69, 2877 38. R. Oxburgh, Earth Planet. Sci. Lett. 159, 183 (1998). 11. G. Quitté et al., Meteorit. Planet. Sci. 42, 1567 (2007). (2005). 39. K. A. Farley, A. Montanari, E. M. Shoemaker, 12. B. Peucker-Ehrenbrink, G. E. Ravizza, A. W. Hofmann, 25. R. L. Korotev, Washington Univ. (St. Louis); available at C. S. Shoemaker, Science 280, 1250 (1998). 187 188 Earth Planet. Sci. Lett. 130, 155 (1995). http://meteorites.wustl.edu/id/density.htm. 40. For example, the lowest K-T boundary Os/ Os ever 13. G. Ravizza, B. Peucker-Ehrenbrink, Science 302, 1392 26. In the calculations (23), an Os concentration ([Os] ) reported (0.137) (8) corresponds to a projectile size of impactor (2003). of 490 ng/g (21) and an average density (r )of greater than 6 km in diameter. On the basis of the impactor 14. The marine Os isotope record, like the Ir fluence 3.35 g/cm (25) were representative of an L chondrite unusually large Os concentration at this section, we approach, is not well suited to detecting the impact of believe that this section does not accurately reflect the meteorite (Popigai LEI event), whereas an Os concentration 3 187 188 differentiated projectiles, such as achondrites with of 807.5 ng/g (21) and an averaged density of 3 g/cm true Os/ Os concentration of seawater immediately siderophile-element concentrations that are two to three after the K-T event. Instead, we suspect a substantial (25) were used as representative of a carbonaceous orders of magnitude lower than chondrites (7). chondrite (28) for the K-T event. inventory of particulate impact-derived Os. 15. M. F. Horan, R. J. Walker, J. W. Morgan, J. N. Grossman, 41. J. W. Morgan, M. F. Horan, R. J. Walker, J. N. Grossman, 27. The Ir-based estimate is calculated by knowing the Ir A. E. Rubin, Chem. Geol. 196, 5 (2003). fluence (3, 20), chondritic Ir concentration (Os/Ir = 1.08) Geochim. Cosmochim. Acta 59, 2331 (1995). 16. J. Whitehead, D. A. Papanastassiou, J. G. Spray, 42. G. S. Collins, H. J. Melosh, R. A. Marcus, Meteorit. Planet. (15), average density of a chondrite (25), and Earth R. A. F. Grieve, G. J. Wasserburg, Earth Planet. Sci. Lett. global Ir inventory. Sci. 40, 817 (2005). 181, 473 (2000). 43. F. T. Kyte, Deep-Sea Res. II 49, 1049 (2002). 28. F. T. Kyte, Nature 396, 237 (1998). 17. Methods and data table are available as supporting 29. G. S. Collins, K. Wunnemann, Geology 33, 925 (2005). 44. J. E. T. Channell et al., Geol. Soc. Am. Bull. 115, 607 (2003). material on Science Online. 30. The Ir and Os isotope-based projectile-size estimates 45. H. Pälike et al., Sci. Res., Proc. Ocean Drill. Prog. (Ocean 18. The increase in the seawater Os reservoir after an probably represent the integrated signal from both the Drilling Program, College Station, TX, 2005), vol. 199. impact event is determined by knowing (i) the Os Popigai and the Chesapeake Bay impacts, because spherules 46. We thank the ODP for providing the samples, D. Vonderhaar concentration in modern seawater (~10 pg/kg) (19) and believed to derive from both events are found within the for technical assistance, E. Scott for his advice and expertise on (ii) the estimated mass of the seawater (~1.4 × 10 kg). single Ir peak at ODP 1090 (20). The smaller estimated meteorites, and R. Smith for support. Comments by three The product of these gives the mass of Os in the ocean size for the Chesapeake Bay projectile (29) suggests anonymous reviewers greatly improved the paper. This work before an impact [mass Os (ng)]. The mass of Os that >90% of the Os and Ir released to the environment was supported by NSF grants OCE and EAR to G.E.R. and B.P.-E. sw derived from the impactor [mass Os (ng)] is impactor was derived from the Popigai event. 187 188 (f/1 – f)×mass Os ;where f =( Os/ Os – 31. A. D. Anbar, G. J. Wasserburg, D. A. Papanastassiou, sw postimpact Supporting Online Material 187 188 187 188 187 188 Os/ Os )/( Os/ Os – Os/ Os ). preimpact impactor preimpact P. S. Andersson, Science 273, 1524 (1996). www.sciencemag.org/cgi/content/full/320/5873/214/DC1 The fractional increase in the seawater Os reservoir is (mass 32. C.-T. A. Lee, G. J. Wasserburg, F. T. Kyte, Geochim. Materials and Methods Os +massOs )/(mass Os ) sw impactor sw . Cosmochim. Acta 67, 655 (2003). SOM Text 19. S. Levasseur, J.-L. Birck, C. J. Allègre, Science 282, 272 33. D. S. Ebel, L. Grossman, Geology 33, 293 (2005). Table S1 (1998). 34. In the steady-state, the relation between reservoir size 20. F. T. Kyte, S. Liu, Lunar Planet. Sci. XXXIII, 1981 N, removal flux F, and residence time t is t = N/F. 12 November 2007; accepted 29 February 2008 (abstr.) (2002). This relation implies that the Eocene seawater Os 10.1126/science.1152860 face planes, exchange of mass or electron equiv- Linked Reactivity at Mineral-Water alents between sites of differing potential energy at different locations on any given crystal is typ- ically assumed to be negligible. This assumption Interfaces Through Bulk is nonetheless questionable for the widespread group of minerals that are electrical semiconduc- Crystal Conduction tors. For example, iron oxides often have moder- ate to low electrical resistivity (1) and have been Svetlana V. Yanina and Kevin M. Rosso* studied as electrode materials for decades (2–4). Iron oxide crystal surfaces are chemically reactive The semiconducting properties of a wide range of minerals are often ignored in the study of their with water and ions, leading to solution-dependent interfacial geochemical behavior. We show that surface-specific charge density accumulation charging behavior that differs from one surface reactions combined with bulk charge carrier diffusivity create conditions under which interfacial type to the next; differing points of zero charge electron transfer reactions at one surface couple with those at another via current flow through the for proton adsorption is but one example (5, 6). crystal bulk. Specifically, we observed that a chemically induced surface potential gradient This difference should give rise to a surface elec- across hematite (a-Fe O ) crystals is sufficiently high and the bulk electrical resistivity sufficiently 2 3 tric potential gradient (Dy ) across any crystal low that dissolution of edge surfaces is linked to simultaneous growth of the crystallographically that has two or more structurally distinct faces distinct (001) basal plane. The apparent importance of bulk crystal conduction is likely to be exposed to solution. In principle, this gradient generalizable to a host of naturally abundant semiconducting minerals playing varied key roles in can bias the diffusion of charge carriers (7, 8). soils, sediments, and the atmosphere. Hence, conditions could exist when the gradient across a single crystal is sufficiently large and the he chemical behavior of mineral-water in- at these interfaces have probed the interaction of electrical resistivity of the material sufficiently terfaces is central to aqueous reactivity in water and relevant dissolved ions with crystallo- Tnatural waters, soil evolution, and atmo- graphically well-defined mineral surfaces. The Chemical and Materials Sciences Division, Pacific Northwest spheric chemistry and is of direct relevance for pursuit so far has been dominated by the assump- National Laboratory, Post Office Box 999, MSIN K8-96, Richland, maintaining the integrity of waste repositories tion that distinct surfaces of any given crystal WA 99352, USA. and remediating environmental pollutants. Tra- behave independently of each other. Except by *To whom correspondence should be addressed. E-mail: ditionally, explorations of fundamental reactions diffusion through the solution phase or across sur- kevin.rosso@pnl.gov 218 11 APRIL 2008 VOL 320 SCIENCE www.sciencemag.org REPORTS 2+ low that interfacial electron transfer reactions at (10, 11); reported bulk resistivities range from Fe in solution, a range of potential-determining 2 6 one surface couple with those at another by a 10 to 10 ohm·m(1). When hematite is sub- ions, electron transfer across hematite-solution current flowing spontaneously through the crys- jected to oxygen-limited aquatic environments, interfaces, and the possibility of moving electron tal bulk. The situation is analogous to galvanic particularly in acidic conditions, it can be reduc- equivalents through the crystal bulk. metal corrosion, but instead of spatially dis- tively dissolved according to Eq. 1: Development of a potential gradient Dy of ordered anodic and cathodic electron transfer significant magnitude across the crystal requires + − 2+ sites, the anode and cathode are spatially con- Fe O +6H +2e → 2Fe +3HO(1) selective interaction between potential-determining 2 3 (aq) 2 fined to crystallographically distinct surface planes ions and specific hematite surfaces. We focus here and are therefore physically separable for mea- which has a standard reduction potential E°~0.7V on roles of protons (low pH) and oxalate as a surement. We demonstrate operability of these (8, 12). The fundamental reaction central to the representative dicarboxylic acid. The hematite conditions for iron oxide, uncover their effects on overall process is (001) basal surface is structurally distinct from the surface chemical behavior, and make the case any edge surfaces. In water, the (001) surface is 3+ − 2+ that, in nature, different surfaces of certain abun- Fe + e → Fe (2) terminated predominantly by doubly coordinated (s) (aq) dant crystals are inextricably linked. hydroxyls (17–19) that are relatively inert to the 2+ We examined hematite (a-Fe O ) because it is Sorbed Fe from the aqueous phase is capable of protonation and deprotonation reactions needed 2 3 3+ a wide band gap semiconductor (band gap 1.9 to reducing hematite Fe in this system (13–15), for charge accumulation. Smaller populations of 2.3 eV) (1, 3) and the most stable form of iron yielding an iron redox cycle in which no net re- more reactive singly coordinated and triply co- oxide under dry oxidizing conditions; it is ordinated hydroxyls, capable of positive charge duction occurs. Introduction of dicarboxylic acids extremely common in nature (9). It has the such as oxalate causes net dissolution by chelating accumulation, are associated with terminal Fe 3+ corundum structure type based on hexagonal close surface Fe (ligand assisted dissolution); it also groups (19). Terminal Fe groups with low coor- packed oxygen planes in which 2/3 of the available enhances reduction, possibly through the forma- dination to the underlying surface can be easily 3+ 2+ octahedral cavities are occupied by Fe .This tion of ternary surface complexes with Fe ,for chelated by oxalate anions to form negatively (aq) structure gives rise to anisotropic electrical resis- example (16). This collective chemistry is a good charged mononuclear bidentate inner-sphere sur- tivity that is higher in the basal plane than along the test case for our main hypothesis because it face complexes (20, 21). In contrast, many remain- trigonal axis by up to four orders of magnitude involves a source of electron equivalents from ing low-index surfaces of hematite crystals such as (012) are dominated by higher-coordinated Fe (17, 22). This higher-coordinated Fe is more dif- Fig. 1. AFM images of representative A E ficult to chelate yet bears singly and/or triply co- hematite surfaces of the natural samples ordinated hydroxyls for charge accumulation. In before (A to D) and after (E to H) general therefore, we expect that, relative to other reaction at 75°C for 12 hours in the pH 2+ hematite surfaces, the (001) surface should show range of 2 to 3 in 1 mM Fe and 10 mM weaker pH-dependent charge accumulation, an ob- oxalate solution free of oxygen. Initial servation increasingly confirmed by recent data surface morphologies [(A) to (D) 10 mm and theory (23, 24), and stronger interaction with by 10 mm deflection images] for (A) a oxalate anions that increases with decreasing pH. (001) surface, (B) a (hk0) surface, (C) a (012) surface, and (D) a (113) surface We performed several experiments to poten- are extremely flat and in all cases except tiometrically measure Dy for specific hematite B F (hk0) crystallographically well ordered surfaces and to determine its effect on their 2+ into terrace-and-step structures at the behavior in Fe and oxalate solutions. We chose micrometer scale. At higher resolution a large natural specular hematite crystal with [(B) inset, 2 mmby2 mm deflection im- well-defined surfaces that could be isolated for age], the (hk0) surface is microfaceted study. The crystal was low in impurity content with edge terminations. Reacted surfaces (25), a natural n-type semiconductor (1), and had [(E) to (H), 20 mmby20 mmdeflection a room temperature electrical resistivity of 10 images with corresponding topographic ohm·m as measured by the four-point probe height images in the insets] show a sharp method. Generating replicate samples required distinction between (E) the (001) surface cutting specific crystallographic surfaces from C G behavior and (F) the (hk0) surface, (G) the the crystal as rectangular prism-shaped specimens, (012) surface, and (H) the (113) surface. which also yielded a vicinal surface type along Large hematite pyramids of uniform cut edges. For example, we prepared millimeter- orientation nucleate as islands on the sized oriented prisms exposing two (001) sur- initially flat (001) surface (E), achieving faces on the top and bottom of the prism and average heights of 200 nm and average four orthogonal (hk0) vicinal sides (25). Anneal- lateral diameters on the order of a mi- ing in air under conditions where hematite is crometer. In contrast, all other surfaces the only stable iron oxide effectively cleans and dissolve. Dissolution of the (hk0) surface organizes the surfaces without modifying the bulk yields large irregular pits and coarsens the D H electrical conductivity. This procedure yields faceted appearance of the surface (F). Dissolution of (012) and (113) surfaces highly organized (001) surfaces, with accompany- yields large crystallographically controlled ing (hk0) vicinal surfaces that are microfaceted etch pits combined with smaller scale with stable edge terminations (Fig. 1, A and B). roughening [(G) and (H), respectively]. Similarly, prism specimens bearing (012) and (113) surfaces (Fig. 1, C and D) with accompany- ing vicinal surfaces were prepared. To determine the magnitude of Dy and the 2+ roles of Fe and oxalate solution components, www.sciencemag.org SCIENCE VOL 320 11 APRIL 2008 219 REPORTS we measured the open-circuit potential (E )in pseudo-pyramidal morphology of uniform orien- lected area diffraction measurements along [001] OCP four solution types. The E is the electrode rest tation. Images at early stages show the island transects of this sample type (fig. S1), along with OCP potential relative to a standard reference elec- growth of these features on the initially flat (001) x-ray photoelectron spectroscopy, x-ray diffrac- trode. Changes in the E are directly related to surface (Fig. 1E). After 12 hours, the reaction tion, and energy dispersive x-ray spectroscopy, OCP changes in y (26, 27), which inturnis sensitive yielded merged pyramid-covered (001) surfaces confirmed that the grown material is structurally to surface complexation reactions with our with peak-to-valley heights averaging 200 nm and compositionally a-Fe O of identical orien- 2 3 + − 2+ potential-determining ions H ,Cl ,Fe ,and and pyramid bases approaching a micrometer in tation as the underlying material without detect- oxalate species (28). The measurements were per- width, imparting a distinct matte appearance to able impurities. The line of intersection of formed at room temperature at effectively con- the reacted (001) surface visible to the naked eye. apparent pyramid “facets” with the (001) plane stant ionic strength under anaerobic conditions Transmission electron microscopy (TEM) and se- is consistent with lines of {012}/{001} intersec- (25). We focused these measurements on the (hk0) H C O 2 2 4 (001) and the accompanying (hk0) vicinal surface (001) H C O 2 2 4 A B type. The observed approximately linear pH de- (hk0) H C O + Fe (II) 2 2 4 pendence, with predominantly negative slopes, is 0.6 0.6 (001) H C O + Fe (II) 2 2 4 consistent with the accumulation of positive sur- face charge with decreasing pH (Fig. 2). As ex- 0.4 0.4 pected, in pure electrolyte solution the (001) surface showed a less-negative slope relative to that of the (hk0) surface, consistent with a lower 0.2 0.2 (hk0) KCl density of charge accumulation sites on the (001) (001) KCl surface (Fig. 2A). In contrast, oxalate anions bind (hk0) KCl + Fe(II) 0.0 0.0 preferentially to the (001) surface with decreasing (001) KCl + Fe(II) pH, even to the point of sign reversal in the slope 2+ (Fig. 2B). Addition of Fe to either solution 0 1 2 3 4 5 0 1 2 3 4 5 shows that its primary effect is to lower the over- pH pH all potential for both the (001) and (hk0) surfaces Fig. 2. The pH dependence of open-circuit poten- KCl without substantially modifying the pH depen- tials with respect to the normal hydrogen elec- dence (Fig. 2, A and B). Taking E – E as KCL + Fe (II) (hk0) (001) 0.4 trode (NHE) for (001) and (hk0) surfaces in (A) H C O an estimate of Dy , in the presence of oxalate and 0 2 2 4 pure electrolyte solution (10 mM KCl) with and 2+ H C O + Fe (II) 2+ irrespective of the presence of Fe we found (aq) 2 2 4 without Fe (1 mM FeCl )and in (B) electrolyte (aq) 2 0.2 that the potential gradient is large and positive, on solution (10 mM KCl) with oxalate (10 mM) with 2+ the order of tenths of volts below pH = 3 (Fig. and without Fe (1 mM FeCl ). The pH depen- (aq) 2 0.0 2C). Under these conditions, we expect that mo- dence of the potential difference between (001) bile electrons acting as majority carriers in hem- and (hk0) surfaces in the four solution types is -0.2 atite would be directed by Dy from the (001) shown in (C). Error bars indicate ± two standard surface to the (hk0) surface. E measurements deviations from a linear trend. OCP -0.4 directly between identical surface types [e.g., E – E ](25) showed no significant voltage. (001) (001) 0 1 2 3 4 5 To examine the effects of Dy of this sign and pH magnitude on the surface chemical behavior, we examined surfaces of the oriented prisms from the Fig. 3. Schematic diagram summariz- same sample by using atomic force microscopy ing the observed reaction behavior for (AFM) before and after anaerobic reaction with hematite crystals showing (A) (001) py- 2+ Fe -oxalate solutions. Thermostated batch ves- ramidal growth coupled to (hk0) disso- sels were used with temperatures ranging from lution, (B) (001) and (hk0) dissolution room temperature to 75°C and pH ranging from 2 for selectively sealed two-crystal cases 2+ to3(25). Fe -oxalate concentrations consistent where the same surface area and type with previously published experiments that estab- as in (A) are exposed to solution, and 3+ lish net dissolution in terms of Fe release on (C) (001) pyramidal growth coupled to (aq) fine-grained powders were used (16, 29, 30). Col- (hk0) dissolution facilitated by a con- ducting paste connection between two lectively, these conditions were selected in keeping crystals. with Eq. 1 while also accelerating surface trans- formations into a more easily observable time frame. Light was excluded in all cases to avoid oxalate acting as a reductant. Equilibrium thermo- dynamic calculations along with Eh measure- ments at run conditions confirm that all our 2+ reaction conditions lie within the Fe stabil- (aq) ity field (25). Hematite was the limiting reactant; total dissolution would retain undersaturation with respect to any possible iron oxide phases. AFM examination of (001) surfaces after re- sealed surface action runs showed remarkable features. In every [001] exposed surface case, for both natural and synthetic samples, sealed conducting paste (001) surfaces were overgrown with a hexagonal 220 11 APRIL 2008 VOL 320 SCIENCE www.sciencemag.org E (V), [E -E ] E (V vs. NHE) cell (hk0) (001) OCP E (V vs. NHE) OCP REPORTS tion, but the interplanar angle varies with run growths on the (001) surface (fig. S3B). Therefore, with coupling mediated by charge transport from duration and does not correspond to low-index the (001) pyramidal overgrowths do not form by (001) to (hk0) surfaces through the crystal bulk. planes in hematite. The large size, morphologic precipitation of ferric iron. Furthermore, we deduce The process involves preferential net oxidative 2+ symmetry, and mutual orientation of these pyra- that chemical processes at the (001)-solution in- adsorption of Fe at the (001)-solution inter- (aq) 3+ mids require homoepitaxy, that is, growth of terface causing pyramidal growth during reaction face and valence interchange with structural Fe additional hematite on hematite. are facilitated by solid contact between the (001) at that surface (Fig. 4). At temperatures of inter- In contrast, all other surfaces examined show and (hk0) surfaces; that is, these surfaces must be est (room temperature and higher), bulk charge features characteristic of dissolution. For exam- on thesamecrystal. transport is sufficiently facile to support a small ple, the four (hk0) vicinal sides of prism samples The behavior strongly suggests that bulk charge current through the bulk. Net electron equiv- bearing (001) surfaces on top and bottom ex- transport provides the link between the two types alents injected into the (001) surface follow an hibited fine-scale pitting and roughening (Fig. of surfaces. As a further test, we again prepared electrically biased random walk through the 1F). The (012) and (113) surfaces of prism sam- two crystals with partial exposure of (001) on one crystal to (hk0) surfaces. At (hk0) exit points, 3+ 2+ ples show development of etch pits at various and (hk0) surfaces on the other, except this time internal reduction of Fe to Fe solubilizes length scales and with symmetry corresponding with an electrical connection between them (Fig. and releases iron into solution. This circuit is to crystallographic orientation (Fig. 1, G and H). 3C). A crystal exposing only (001) surfaces was driven by the Dy gradient generated across the We observed identical behavior under the same connected to a crystal beneath exposing only crystal from divergent charge accumulation at conditions with use of synthetic tabular hematite (hk0) surfaces by electrically conductive colloidal structurally distinct surface types. The sign and crystals bearing primarily (001) and (012) sur- Ag paste, which was subsequently cured, sealed magnitude of Dy , the conductivity of the natu- faces, in which case no surface preparation by off from contact with solution using additional ral crystal, and the growth rates of the pyramidal annealing was required (fig. S2). The conclusion epoxy, and tested for ohmic behavior by resis- islands are all mutually consistent. For example, is that the hematite (001) surface grows under our tivity measurements. In this design, the crystals taking Dy = 0.2 V at pH = 2, a temperature- conditions whereas all other surfaces sampled are effectively wired together by the (001)-Ag- adjusted electrical resistivity = 10 ohm·m for dissolve, as represented schematically in Fig. 3A. (001) junction between them. Reaction in this 75°C (31, 32), and an electron transport path We designed experiments using the prism wired two-crystal configuration proceeds as if the length of 1 mm, the maximum amount of addi- samples to test whether or not pyramid islands crystals were one; pyramidal hematite grows on tional hematite expected on the (001) surface in are deposited on the (001) surface by precipita- the exposed (001) surface of the upper crystal 12 hours is a layer ~100 nm thick, the same order 3+ tion of trace Fe from solution (16). In these (fig. S3C), whereas the four (hk0) sides of the of magnitude as that observed. Surface potential– experiments, two prism samples were used in the lower crystal dissolve (Fig. 3C). Therefore, the driven charge carrier diffusivity has been invoked reaction vessel instead of one. Four (hk0) vicinal nature of the interaction between the (001) and qualitatively to explain microscopic oxide trans- sides of one crystal were sealed with an inert vicinal surfaces that gives rise to the pyramidal formation processes before (13, 33, 34) but not epoxy (25), leaving two (001) surfaces exposed, growth of hematite (001) during reaction derives on the length scale examined here nor with sur- whereas on the other crystal the two (001) sur- from bulk charge transport. Surface diffusion face specificity. Given the observation that the faces were sealed, leaving four vicinal surfaces along the hematite-solution interface was ruled (001) surface continues to grow beyond the co- exposed. Collectively, the two crystals expose the out by painting a ring of sealant on a (001) alescence of the pyramidal islands, at the atomic same six kinds of surfaces to solution as in the surface so that only bulk transport could access scale the pyramidal (001) morphology must runs above with one crystal, in the same relative the circumscribed region, and within that region retain the essential structural and therefore proportion and surface area, but they involve (001) hematite island growth also occurred (fig. S4). chemical characteristics that give rise to the surfaces that are physically separated from the The collective behavior of the system is potential of the initial (001) surface. Further- (hk0) surfaces (Fig. 3B). In this case, the results therefore suggestive of two distinct but coupled more, the observed process does not preclude of reaction runs show only dissolution features interfacial processes: growth at (001) by traditionally held spatially localized dissolution on all exposed surfaces, including (001) (fig. S3A). in the hematite system. Rather, the evidence 2+ 2+ 3+ − The (001) pyramidal morphology does not de- Fe → Fe → Fe + e (3) suggests that the processes operate in parallel (aq) (001) (001) velop in this separated two-crystal configuration. and that the behavior based on the electrical The same experiment performed on samples in and dissolution of edge surfaces, for example, circuit through the crystal dominates when which the pyramidal morphology had already (hk0) surfaces, by chemical requirements that establish a large been grown on (001) before sealing its (hk0) enough surface electric potential gradient are 3+ − 2+ 2+ Fe + e → Fe → Fe (4) (hk0) (hk0) (aq) sides showed dissolution of the pyramidal over- met. The finding provides insight into the reduc- Fig. 4. Schematic diagram depict- tive transformation of iron oxides, which is im- ing the inferred coupled interfacial portant in the biogeochemical cycling of iron in electron transfer process operative nature and the removal of iron oxide films in (Aqueous solution) under our conditions for the hema- industry. Because this finding can be easily gen- tite single crystals. The chemically eralized to a host of naturally abundant semi- self-induced surface potential gra- (001) (hk0) conducting transition metal oxide and sulfide dient across the crystal directs minerals capable of dominating the interfacial current flow through the bulk. The surface area in soils, sediments, and among current is facilitated by sufficiently atmospheric particles, its implications are fairly low electrical resistivity in a process widespread. Of immediate impact is the concept 2+ 2+ 3+ − that is fed by net injection of electron Fe Fe Fe + + aq (001) (001) that the reactivity of any given surface on such equivalents at (001) surfaces and net materials can be coupled to that of another sur- release of electron equivalents at (hk0) (Hematite bulk) + face, with a dependence on crystal morphology surfaces. 2+ Fe as a whole. This phenomenon should apply to 3+ Fe + natural crystals in the environment as well as those selectively cut, broken, or otherwise pre- 3+ − − 2+ 2+ Fe + + Fe Fe (hk0) (hk0) aq + pared for laboratory study. www.sciencemag.org SCIENCE VOL 320 11 APRIL 2008 221 REPORTS 15. P. Larese-Casanova, M. M. Scherer, Environ. Sci. Tech. 33. N. M. Dimitrijevic, D. Savic, O. I. Micic, A. J. Nozik, References and Notes 1. R. T. Shuey, Semiconducting Ore Minerals, vol. 4 of 41, 471 (2007). J. Phys. Chem. 88, 4278 (1984). 16. D. Suter, C. Siffert, B. Sulzberger, W. Stumm, 34. J. P. Jolivet, E. Tronc, J. Colloid Interface Sci. 125, 688 Developments in Economic Geology (Elsevier, Amsterdam, 1975). Naturwissenschaften 75, 571 (1988). (1988). 35. This research was supported by the U.S. Department of 2. J. H. Kennedy, K. W. Frese, J. Electrochem. Soc. 125, 723 17. V. Barron, J. Torrent, J. Colloid Interface Sci. 177, 407 (1996). (1978). 18. C. M. Eggleston et al., Geochim. Cosmochim. Acta 67, Energy (DOE), Office of Basic Energy Sciences, Geosciences Program. It was performed at the William 3. C. Gleitzer, J. Nowotny, M. Rekas, Appl. Phys. A Mat. Sci. 985 (2003). Proc. 53, 310 (1991). 19. T. P. Trainor et al., Surf. Sci. 573, 204 (2004). R. Wiley Environmental Molecular Sciences Laboratory 20. O. W. Duckworth, S. T. Martin, Geochim. Cosmochim. (EMSL) at the Pacific Northwest National Laboratory 4. B. A. Balko, K. M. Clarkson, J. Electrochem. Soc. 148, E85 (2001). Acta 65, 4289 (2001). (PNNL). The EMSL is funded by the DOE Office of 21. T. H. Yoon, S. B. Johnson, C. B. Musgrave, G. E. Brown, Biological and Environmental Research. PNNL is 5. P. Venema, T. Hiemstra, P. G. Weidler, W. H. van operated by Battelle for the DOE under contract Riemsdijk, J. Colloid Interface Sci. 198, 282 (1998). Geochim. Cosmochim. Acta 68, 4505 (2004). 22. J. R. Rustad, E. Wasserman, A. R. Felmy, Surf. Sci. 424, DE-AC06-76RLO 1830. We gratefully acknowledge the 6. F. Gaboriaud, J. Ehrhardt, Geochim. Cosmochim. Acta 67, assistance of C. Wang for TEM; B. Arey for scanning 967 (2003). 28 (1999). 23. T. Hiemstra,W. H.Van Riemsdijk, Langmuir 15, 8045 (1999). electron microscopy; D. McCready for pole reflection 7. C. G. B. Garrett, W. H. Brattain, Phys. Rev. 99, 376 x-ray diffraction; Y. Lin for access to electrochemistry (1955). 24. P. Zarzycki, Appl. Surf. Sci. 253, 7604 (2007). 8. P. Mulvaney, V. Swayambunathan, F. Grieser, D. Meisel, 25. Materials and methods are available on Science Online. apparatus; and A. Felmy, E. Ilton, and J. Amonette for comments on an early version of this J. Phys. Chem. 92, 6732 (1988). 26. M. J. Avena, O. R. Camara, C. P. Depauli, Colloid Surf. 9. R. M. Cornell, U. Schwertmann, The Iron Oxides: 69, 217 (1993). manuscript. Structure, Properties, Reactions, Occurrence and Uses 27. N. Kallay, T. Preocanin, J. Colloid Interface Sci. 318, 290 (VCH, Weinheim, Germany, 2003). (2008). Supporting Online Material 10. T. Nakau, J. Phys. Soc. Jpn. 15, 727 (1960). 28. J. A. Davis, R. O. James, J. O. Leckie, J. Colloid Interface www.sciencemag.org/cgi/content/full/1154833/DC1 11. N. Iordanova, M. Dupuis, K. M. Rosso, J. Chem. Phys. Sci. 63, 480 (1978). Materials and Methods 29. B. Zinder, G. Furrer, W. Stumm, Geochim. Cosmochim. 122, 144305 (2005). Figs. S1 to S4 12. J. S. LaKind, A. T. Stone, Geochim. Cosmochim. Acta 53, Acta 50, 1861 (1986). References 30. S. Banwart, S. Davies, W. Stumm, Colloid Surf. 39, 303 (1989). 961 (1989). 13. P. Mulvaney, R. Cooper, F. Grieser, D. Meisel, Langmuir 31. S. Kerisit, K. M. Rosso, Geochim. Cosmochim. Acta 70, 4 January 2008; accepted 25 February 2008 1888 (2006). Published online 6 March 2008; 4, 1206 (1988). 14. A. G. B. Williams, M. M. Scherer, Environ. Sci. Tech. 38, 32. S. Kerisit, K. M. Rosso, J. Chem. Phys. 127, 124706 10.1126/science.1154833 (2007). Include this information when citing this paper. 4782 (2004). other regions (3–6), Madagascar has complex, Aligning Conservation Priorities often nonconcordant patterns of microendemism among taxa (12–17), rendering the design of ef- ficient protected-area networks particularly diffi- Across Taxa in Madagascar with cult (4, 6). We collated data for endemic species in six major taxonomic groups [ants, butterflies, High-Resolution Planning Tools frogs, geckos, lemurs, and plants (table S1)], using recent robust techniques in species distribution 1,2 1,2 3 4 5 6 C. Kremen, *† A. Cameron, † A. Moilanen, S. J. Phillips, C. D. Thomas, H. Beentje, modeling (18, 19) and conservation planning 6 7 8 9 10 11 12 J. Dransfield, B. L. Fisher, F. Glaw, T. C. Good, G. J. Harper, R. J. Hijmans, D. C. Lees, (20, 21) to produce the first quantitative conserva- 13 14 15 2 16 E. Louis Jr., R. A. Nussbaum, C. J. Raxworthy, A. Razafimpahanana, G. E. Schatz, tion prioritization for a biodiversity hot spot with 17 18 19 9 M. Vences, D. R. Vieites, P. C. Wright, M. L. Zjhra this combination of taxonomic breadth (2315 species), geographic extent (587,040 km ), and Globally, priority areas for biodiversity are relatively well known, yet few detailed plans exist to spatial resolution (30–arc sec grid = ~0.86 km ). direct conservation action within them, despite urgent need. Madagascar, like other globally Currently, an important opportunity exists to recognized biodiversity hot spots, has complex spatial patterns of endemism that differ among influence reserve network design in Madagascar, taxonomic groups, creating challenges for the selection of within-country priorities. We show, in an given the government’s commitment, announced analysis of wide taxonomic and geographic breadth and high spatial resolution, that at the World Parks Congress in 2003, to triple its multitaxonomic rather than single-taxon approaches are critical for identifying areas likely to existing protected-area network to 10% coverage promote the persistence of most species. Our conservation prioritization, facilitated by newly (22). Toward this goal, our high-resolution anal- available techniques, identifies optimal expansion sites for the Madagascar government’s current ysis prioritizes areas by their estimated contribu- goal of tripling the land area under protection. Our findings further suggest that high-resolution tion to the persistence of these 2315 species and multitaxonomic approaches to prioritization may be necessary to ensure protection for biodiversity identifies regions that optimally complement the in other global hot spots. existing reserve network in Madagascar. We input expert-validated distribution models pproximately 50% of plant and 71 to analysis of these patterns is required to allocate for 829 species and point occurrence data for the 82% of vertebrate species are concen- conservation resources most effectively (7, 8). remaining species [those with too few occur- Atrated in biodiversity hot spots covering To date, only a few quantitative, high- rences to model, called rare target species (RTS)] only 2.3% of Earth’s land surface (1). These resolution, systematic assessments of conserva- into a prioritization algorithm, Zonation (20, 21), irreplaceable regions are thus among the highest tion priorities have been developed within these which generates a nested ranking of conservation global priorities for terrestrial conservation; rea- highly threatened and biodiverse regions (9, 10). priorities (23). Species that experienced a large sonable consensus exists on their importance This deficiency results from multiple obstacles, proportional loss of suitable habitat (range reduc- among various global prioritization schemes that including limited data or access to data on species tion) between the years 1950 and 2000 were given identify areas of both high threat and unique distributions and computational constraints on higher weightings [equation 2 of (23), (24)]. We biodiversity (2). The spatial patterns of species rich- achieving high-resolution analyses over large evaluated all solutions [defined here as the ness, endemism, and rarity of different taxonomic geographic areas. We have been able to over- highest-ranked 10% of the landscape to match groups within priority areas, however, rarely align come each of these obstacles for Madagascar, a the target that Madagascar has set for conservation and are less well understood (3–6). Detailed global conservation priority (1, 2, 11). Like many (22)] in two ways: (i) percent of species entirely 222 11 APRIL 2008 VOL 320 SCIENCE www.sciencemag.org

Journal

ScienceUnpaywall

Published: Mar 6, 2008

There are no references for this article.