• Published on

  • View

  • Download

Embed Size (px)


<ul><li><p>ANALYTICAL </p><p>ANOMALOUSLY NIGO RARE - EARTH ELEMENT ABUNDANCES IN HAWAIIAN LAVAS </p><p>R. V. Fodor Department of Marine, Earth, and </p><p>Atmospheric Sciences North Carolina State University Raleigh, NC 27695-8208 </p><p>Gbor Dobosi Geochemical Research Hungarian Academy of Sciences Budapest, Hungary 1112 </p><p>G. R. Bauer Department of Land and Natural Resources Division of Water Resource Management Honolulu, HI 96813 </p><p>The compositions of the lavas that comprise the Hawaiian Islands are often used as models for understand-ing volcanism and the chemical char-acteristics of oceanic volcanoes. From a chemical viewpoint, Hawaii is an ideal paradigm because of its location far from continental crust. Hawaii's oceanic setting tells us that the mag-mas that erupted as lavas to con-s t ruc t the i s l ands did not pa s s through potentially contaminating material. Chemical compositions of Hawai i an lavas , therefore , com-pletely represent the source from which they were derived: peridotite rock of the upper mantle. </p><p>Chemical and isotopic analyses of Hawaiian volcanic rocks have become routine laboratory procedures. In the interpretation and modeling of data for a suite of rocks, we often focus on </p><p>the significance of the abundance of certain trace elements, such as Zr, Nb, Th, and Ce, and on the impor-tance of isotope ratios, such as 87Sr/ 86Sr, which may vary by &lt; 0.00005 from one sample to another. </p><p>The objectives of these analytical studies include the identification of some geologically important charac-teristics of the Earth , such as the chemical nature of the upper mantle beneath each volcano, compositional changes in the mantle during volcano construction, and the chemically and isotopically varied mant le compo-nents tha t may have mixed before and during magma generation. An-other objective is to define the pro-cesses by which magmas change composition during storage in sub-volcano reservoirs before eruption. </p><p>D u r i n g t h e 1980s n u m e r o u s geochemists and igneous petrologists produced highly refined models for </p><p>Hawaiian magmatism. However, in-formation from new chemical studies of Hawaiian lavas is not always com-patible with existing models. In this article we will discuss one such study of a suite of lavas from the island of Kahoolawe that contain rare-ear th element abundances unlike those ever expected for Hawaiian volcanic rocks. </p><p>Chemical characteristics of lavas The lavas of the 15 volcanoes of the eight major islands (Niihau, Kauai, Oahu, Molokai, Lanai, Kahoolawe, Maui, and Hawaii) originated as ba-salt composition magmas produced by partial melting of the mantle. The melting occurred at depths of - 6 0 -90 km. Over the past 5 million years, the magmas erupted to construct broad, shield-like volcanoes reaching far above the sea floor to form the Hawaiian Islands. </p><p>In their sampling of the island la-vas, geologists and geochemists note the s t ra t igraphie positions of the samples on the volcanoeswhether they represent the main body of the volcano, the shield edifice, lavas that filled a caldera at the summit of the volcano, or other features such as small cones made of cinders or spat-ter that may mark the final erup-tions of a volcano. </p><p>Hundreds of analyses of Hawaiian volcanic rocks have established the detailed chemical makeup of the in-</p><p>0003-2700/92/0364-639A/$02.50/0 1992 American Chemical Society </p><p>ANALYTICAL CHEMISTRY, VOL. 64, NO. 11, JUNE 1, 1992 639 A </p><p>Niihau </p><p>Kauai </p><p>Oahu </p><p>Lanai </p><p>Molokai Maui </p><p>Hawaii mmrm </p></li><li><p>ANALYTICAL APPROACH </p><p>dividual volcanoes and helped to characterize their mantle source ma-terial. The rare-ear th elements are an important part of these evalua-tions. Part icularly valuable is the geochemical behavior of ra re-ear th elements during the partial melting of peridotite rock to form basaltic magmas and during crystallization of those magmas to form basaltic rocks. For example, because heavy ra re -earth elements (Yb and Lu) preferen-tially partition into certain minerals, the role of a mantle phase such as garnet in the production of basalt can be evaluated from the rare-earth el-ement content in lavas. Also, because light ra re -ear th elements (La and Ce) have different compatibility with crystallizing basaltic minerals than do middle (Eu) and heavy rare-earth elements, we can determine the type and amount of minerals that crystal-lized in basaltic magma during vol-cano development. </p><p>Rare-earth elements in lavas </p><p>To decipher geochemical signatures and to improve petrogenetic (rock or-igin) models for ocean island volca-n ism, r a r e - e a r t h e lement a b u n -dances are determined in all newly collected suites of Hawaiian lavas. Data are acquired primarily by using neutron activation analysis (NAA). </p><p>The eight rare-earth elements La, Ce, Nd, Sm, Eu, Tb, Yb, and Lu are usually determined as a set for ba-saltic rock. Typically, the concentra-tions (in parts per million) are nor-m a l i z e d to a v e r a g e v a l u e s for chondritic meteorites and plotted as rare-earth element patterns (Figure 1). Normalizing the data to chon-drites provides reference to material that approximates the earth's mantle composition. It also enables one to graphically compare rare-earth ele-ment concentrations among different rocks by eliminating the Oddo-Har-kins effect of higher concentrations for elements with even atomic num-bers. </p><p>The patterns for various composi-tional types of Hawaiian lavas have been characterized (1) and can be reasonably well predicted for a par-ticular rock after its major element composition is determined. In light of this, we believed that our geochemi-cal investigation of Kahoolawe Island would be fairly routine and the trace element compositions somewhat pre-dictable. </p><p>Kahoolawe Island </p><p>Prior to our work, Kahoolawe Island was the only Hawaiian volcano that </p><p>had not been studied geochemically. It is managed by the U.S. military, and access is restricted. Our collec-tion of ~ 200 samples was obtained with permission from U.S. Navy per-sonnel at Pearl Harbor. Providing helicopter transportation and demo-lition experts who escorted us to en-sure our safety, the Navy allowed us to comb the island. </p><p>Our first assessment of Kahoolawe basaltic rocks (2) yielded some lava rare-ear th element abundances that were dramat ica l ly different from those expected. These rock samples were collected from lava vents that represent the last episodes of Ka-hoolawe volcanism. Many rocks were enriched with a greater rare-ear th element content than that found in other Hawaiian lavas, including the main shield lavas of Kahoolawe (Fig-ure la). For example, the concentra-tion of La, which would typically be 13-15 ppm in the basaltic rocks ex-amined , was 1 0 0 - 2 0 0 ppm. The </p><p>Figure 1. Patterns obtained using NAA. (a) Rare-earth element patterns for a typical shield lava and for two rare-earth element- and Y-enriched lavas of Kahoolawe Island. Star indicates a negative Ce anomaly, (b) Rare-earth element patterns for four samples from different parts of the same lava flow on Kahoolawe. The pattern labeled "normal" is for a portion of lava believed to be free of rare-earth element and Y enrichment. The great differences in rare-earth element abundances (from "normal" to "highly enriched") indicate that the rare-earth element abundances are not evenly distributed throughout the lava. Patterns are constructed by dividing part-per-million values for the lavas by the average rare-earth element abundances of chondritic meteorites. </p><p>rocks with high rare-ear th element abundances also had unusually high Y concentrationsin one case, up to five times greater than expected. </p><p>Additionally, some Kahoolawe la-vas had low Ce abundances relative to the high amounts of the remaining ra re -ea r th elements studied. Such occurrences of comparatively low Ce are referred to as negative Ce anom-alies (Figure la). </p><p>The anomalously high rare-ear th elements and Y values were particu-larly intriguing because Kahoolawe had been a bombing target for the U.S. Navy since World War II. The lengthy history of bombing made us wonder whether the unusual rare-earth element and Y concentrations were related to trace elements in ex-plosives. However, a recent publica-tion (3) citing a lava on Oahu with similar r a r e -ea r th element and Y concentrat ions dissuaded us from pursuing this hypothesis. </p><p>Possible explanations for rare-earth element and Y enrichments We were left with three alternative geochemical hypotheses. The first was that the upper mantle contains a r a re phase t h a t is enriched wi th rare-ear th elements and Y and occa-sionally is included in the part ia l me l t ing t h a t produces Hawa i i an magmas. The second theory involved assimilation of some unusual rare-ear th e lement /Y-bear ing mater ia l (perhaps from the marine environ-ment) by some of the magmas as they ascended from the mantle to the sur-face. Finally, we considered second-ary processes such as surface weath-ering and exposure to hydrothermal solutions. We gave this idea lowest priority because the rocks in ques-tion did not appear to be weathered or altered, or to have been exposed to hydrothermal activity. Furthermore, rare-ear th elements and Y are not necessarily mobilized in rocks and concentrated elsewhere during the commonly observed geologic pro-cesses. </p><p>We also found problems with the first two hypotheses. The rare-earth element and Y abundances of the anomalously enriched rocks did not vary systematically with other trace elements in geochemically similar rocks, nor did the r a re -ea r th ele-ments vary systematically among themselves. This suggested tha t a single, special mant le phase t ha t melted to yield rare-ear th element/ Y-enriched magmas had not existed. Because the rocks were not enriched </p><p>640 A ANALYTICAL CHEMISTRY, VOL. 64, NO. 11, JUNE 1, 1992 </p><p>(a) Enriched lavas </p><p>Typical shield lava </p><p>Atomic number </p><p>(b) Highly enriched </p><p>Normal </p><p>Slightly enriched </p><p>Atomic number </p></li><li><p>in elements likely to be abundant in a seawater environment, such as Sr and Mn, we discarded the idea of ass imi la ted r a r e - e a r t h e l emen t /Y-bearing marine material before eruption. </p><p>Identifying the rare-earth element/ Y-bearing phase The key to finding the origin of the ra re -ea r th element and Y enrichment in the basalts was to determine how these elements were contained. That is, which basalt mineral phases housed high amounts of rare-ear th elements and Y, and how did they occur in the rocks? Was it a phenocryst phase (mineral grain visible without magnification), or was it microscopic and hidden in the groundmass of the basalts? Our detailed microscopic examination of the rare-ear th element/ Y-rich samples had revealed nothing special about their mineral assemblages. All observed phases had been expected to be present in these rocks. </p><p>We decided tha t electron micro-probe analysis would be ideal for locating the phase because we would be able to scan the sample using an electron beam while the detectors were optimized for ra re -ear th elements and Y. By viewing the sample during investigation, the slightest hint of La or Y X-rays created by the beam falling on a r a r e -ea r th element/Y-rich phase would reveal the location of the phase in the rock. </p><p>This method of examination could not, however, be done using the standard procedure for mineral analysis (using a l -3- electron beam), because the chances were small for locating a phase t ha t may be only micrometers in size. The investigation required optimizing the spectrometers for La and Y, enlarging the electron beam to a - 200- diameter, and scanning a polished specimen of r a r e - e a r t h e lement /Y- enriched lava until a La or Y signal was detected. The search was completed in seconds. The groundmass of the lava sample examined was rich with 10-30^m-s ized grains hidden amid the normal assemblage of pyroxene, plagioclase, and Fe-Ti oxides. </p><p>From t h a t po in t on, we drew sketches of all the located rare-earth element/Y-rich grains and the surrounding fields of view of the specimen as we saw them through the microprobe op t ica l s y s t e m . T h e s e sketches would help us later locate and study the rare-earth element/Y-rich phases with the aid of a polarizing microscope. </p><p>Optical microscopy revealed irreg-</p><p>SWAGELOK Tube Fittings ...your widest selection of* </p><p>sizes, shapes, and materials leak-tight performance, low in-service cost fractional sizes 1/16" to 2" metric sizes 2mm to 25mm (male and female connectors </p><p>available in tube to NPT or BSP/ISO threads) all machinable metals and plastics locally available from Authorized Sales &amp; Service Representatives </p><p>SWAGELOK Co., Solon, Ohio 44139 / SWAGELOK Canada Ltd., Ontario </p><p>Male Connector </p><p>Bulkhead Male Connector </p><p>Male Elbow </p><p>45 Male Elbow </p><p>Male Run Tee </p><p>Male Branch Tee </p><p>Female Connector </p><p>Bulkhead Female Connector </p><p>Female Elbow </p><p>Female Run Tee </p><p>Female Branch Tee </p><p>Union </p><p>Reducing Union </p><p>Bulkhead Union </p><p>Union Elbow </p><p>Union Tee </p><p>Union Cross </p><p>SWAGELOKIoMUnlon </p><p>SWAGELOK lo AN Bulkhead Union </p><p>Reducer </p><p>Bulkhead Reducer </p><p>SWAGtLOK 10 AN Adapter </p><p>Port Connector </p><p>Reducing Port Connector </p><p>SWAGELOKtoTube Socket Welb Union </p><p>SWAGELOK ID Tube Sockel Weld Elbow </p><p>SWAGELOK to Male Pipe Weld Connector </p><p>SWAGELOK to Male Pipe Weld Elbow </p><p>O-Seal Male Connector Pipe Thread </p><p>O-Seal Straight Thread Connector </p><p>g) 1989 SWAGELOK Co., all rights reserved PF 2 087 </p><p>SAE/MS Positionable Male Elbow </p><p>SAE/MS Male Connector </p><p>SAE/MS Positionable Male Run Tee </p><p>45 SAE/MS Positionable Male Elbow </p><p>SAE/MS Positional) le Male Branch Tee </p><p>Cap Plug </p><p>Heat Exchanger Tee </p><p>Bore-Through Male Connector Used As Thermocouple Connector </p><p>KN Tube Fittings For Use on Polyethylene Tubing </p><p>CHROMATOGRAPH FITTINGS </p><p>Union </p><p>Union Tee </p><p>Zero Volume Column End Fitting </p><p>SWAGELOK 10 Female SWAGELOK Union T U B E F I T T I N G S </p><p>CIRCLE 125 ON READER SERVICE CARD </p><p>ANALYTICAL CHEMISTRY, VOL. 64, NO. 11, JUNE 1, 1992 641 A </p></li><li><p>ANALYTICAL APPROACH </p><p>ularly shaped, high-relief, somewhat fibrous-looking grains. We were unfamiliar with the properties of this phase as viewed under polarized light, but the examination nevertheless provided an unders tanding of how certain Kahoolawe basaltic lavas served as hosts to high ra re -earth element and Y levels. We could then place the lava sample back into the electron microprobe for quantitative determination of the material previously unknown in Hawai ian volcanic rocks. </p><p>Enrichment by secondary processes Concurrent with our research, F. A. Frey of the Massachusetts Institute of Technology performed neutron activation rare-earth element analysis of additional samples from the same lava that we were examining with an electron microprobe. The NAA results showed t ha t the anomalous r a r e - e a r t h e lement and Y abundances were not evenly distributed throughout the lava. </p><p>Examined from various places on </p><p>the island, the same lava flow conta ined r a r e - e a r t h element and Y contents rang ing from normal to highly enriched levels for basaltic ro...</p></li></ul>