Eric Hellebrand - Research Interests



Melting in the mantle

Abyssal peridotites are mantle rocks that are exposed on the ocean floor along mid-ocean ridges. They are widely believed to be the residues of the partial melting process that leads to the eruption of mid-ocean ridge basalt (MORB). Traditionally, the chemical compositions of these residual peridotites are used to estimate how much basaltic components were extracted during partial melting, and even to obtain information about the physics of melting and melt migration. However, recent osmium and hafnium isotopic studies on these abyssal peridotites have shown that many abyssal peridotites appear to have seen partial melting long (>100 Myrs) before recent upwelling under the spreading ridge.

Os isotopic composition of abyssal peridotites: old melt extraction

Whole rock Al2O3 is a measure of the extent of melting (F): low at high-Al, high-F at low Al contents. Os isotope ratios can provide some age information, because 187Os is the radiogenic decay product of 187Re. The depleted peridotites in this Gakkel Ridge peridotite study have low Os isotope ratios, which requires rhenium removal (by melting) more than 2 billion years ago. From Liu et al. (Nature, 2008)


These finding raise all sorts of fundamental questions, some of which I am currently trying to address:

  1. What proportion of the abyssal peridotites has not undergone recent partial melting but inherited its depletion from an old melting event?
  2. How can we reconcile the abundant old peridotite depletion with nearby volcanic eruptions? In other words, what are the MORB sources and residues?
  3. Do abyssal peridotites provide us with a representative picture of the top of the melting column?
  4. How can we use refractory peridotites to learn more about the composition and origin of migrating melts?

While the importance and implications of these observations are fiercely debated at the moment, it is evident that abyssal peridotite studies offer a view into the upper mantle that is not seen in erupted MORB.


The lower oceanic crust



The lower oceanic crust is formed by crystallization of the melts that leave the mantle. Since the lower crust is normally overlain by the upper crust (who would have thought that...) it is difficult to get a good picture of what it's made of, and how exactly it's formed. So we have to drill into the oceanic crust, and we go to ophiolites.

In its most simplified form, the lower crust is composed of cumulate minerals, crystallized from melts that end up erupting and forming the upper crust. Chemically, that means that a cumulate is not the same as a melt, as illustrated by this plot. The composition of a cumulate depends on the relative proportion of the crystallizing minerals and which elements they incorporate into their structure. Because of their sample- and grain-scale heterogeneities, studying plutonic rocks properly is far more time consuming that peridotites.

cumulate vs melt


One fundamentally new result of recent lower crustal studies, is that some of the most primitive cumulates are quite probably mantle peridotites transformed into olivine-rich troctolites by the reaction with intruding gabbroic melts. This was proposed after studying fresh lower crustal rocks recovered by ocean drilling (IODP Leg 304/305 at Atlantis Massif, Mid-Atlantic Ridge 30N) and later supported by field mapping and petrology in Italian ophiolites.

Olivine-rich troctolite

Clearly, if these former mantle peridotites are a significant constituent of the lower oceanic crust, formed by "adding" 1 part of melt to 3 parts of mantle, the mantle-crust mass balance models need to be reevaluated.




On of the reasons I mainly use micro-analytical techniques is the strong seawater overprint that the ocean floor rocks have seen during their slow cooling history from mantle melting conditions to ocean bottom temperatures. That's why abyssal peridotites are often called "abysmal peridotites". Frequently, more than 90 percent of all primary mantle minerals are replaced by low-temperature minerals, such as serpentine and, less frequently, carbonate. In order to reconstruct the mantle melting history of these rocks, we have avoid the distorting contribution of these secondary minerals.

Transmitted and reflected light image of a serpentinized peridotite

In order to obtain major element compositions of the different minerals in abyssal peridotites, I use an electron microprobe (aka EPMA). Typically, this technique has a spatial resolution of one micron. For trace element abundances, I use an ion microprobe (aka SIMS). That produces little craters in the sample surface (black spots on the image above), with a diameter of about 15 microns. Both techniques are available at the University of Hawaii.

Analyzing depleted peridotites requires a different approach than fresh MORB glasses. A modern laser-ablation ICP-MS lab can acquire trace element abundances of nearly the entire periodic table in less than 10 minutes per spot. Next to nothing is known about the behavior of the most highly incompatible (and fluid-mobile) elements in residual mantle peridotites due to these analytical limitations. The ultra-low abundances, alteration along cracks and inclusion, and potential contamination by sample preparation can only be tackled by SIMS, and only in small focused increments.

Trace elements in some Gakkel Ridge MORB vs peridotite cpx

Currently, I am refining the setup to analyze Be and B by SIMS, getting detection limits better than 50 and 300 ppt, respectively. In the ppb-level concentration range, the accuracy is limited by lack of proper standards, so we are working hard on that too. Other elements, notably Ba, U and Nb will be targeted next using a different analytical setup. Considering that all plagioclase-free and vein-free abyssal peridotites are homogeneous on a grain scale for other lithophile trace elements, a sequence of different trace element groups measured in spots next to each other, will provide invaluable insights in the behavior of the highly incompatible lithophile trace elements in the oceanic mantle.

If you are a student and you're interested in any of these aspects, please contact me for further information. The established field of mid-ocean ridge research is undergoing some drastic paradigm shifts at the moment, and it's a fantastic time to be involved in this.