nai

Project Progress

1. Background

Theme 4 work at JHU/APL using laser time-of-flight mass spectrometry (TOF-MS) techniques continues in collaboration with the Goddard Center for Astrobiology (GCA) team and external partners. We seek to contribute to research on the following major questions within the GCA:

1. What are the fine scale morphology and composition of comet nuclei?


2. Do comets comprise an enabling inventory of complex pre-biotic organics?


3. How should we approach the characterization of this inventory?

Laser desorption (LD) methods are complementary to ongoing non-laser GCA work at GSFC; they sample distinct yet related sets of compounds from complex samples. Our work is also a link between analyses performed with similar facility instrumentation at other labs in that we can measure quantitatively the scientific consequences of miniaturization or other resource limitations. The specific objectives of the laboratory-based effort at APL are to:

1. Examine organics in standards, cometary analogs, and meteorites with LDMS;


2. Develop an LDMS sensitivity-selectivity database for high mass, refractory organics in various matrices and optimize the method;

Compare LD-based “prompt ionization” with LD + laser post-ionization (REMPI) for use as organic chemical mapping technique; and

3. Determine optimal combined LDMS and pyrolysis GCMS (and other) analyses of common samples that advise mission design.

2. Progress Summary

During 2005-2006 we have made substantial progress in both (i) technique and instrument development; and (ii) sample and data analysis.

2.1. Technique and Instrument Development

NAI participation at APL supports basic and applied scientific research that both motivates and benefits from ongoing instrument development efforts, research projects, and mission conceptualization supported under various other programs. Mass spectrometers already in place or under development for multiple purposes have had primary support from (completed) PIDDP, ASTID, and MIDP grants. The NAI work seeks to investigate and optimize these techniques for application to analysis of organics in primitive solar system samples (in meteorites, on missions, or in returned samples). Starting from existing technique infrastructure, we are utilizing and exploring the analytical flexibility available in certain platforms, such as with the choice of laser desorption/ionization methods, voltage modes, sample configurations, and sample manipulation approaches. Only modest hardware modifications are required by any particular choice, however the choices themselves must be made with great care so we can understand how the science results are affected by them.

Several experimental approaches are available with the modifications being tested in this effort:

1. LDI-TOF-MS (LDMS) — A single laser pulse is used for desorption and ionization. Generally much fragmentation occurs but high-mass parent peaks are still present at low levels in many cases. LDMS is readily miniaturized and space qualifiable.


2. LD-EPI-TOF-MS — Laser-desorbed neutrals are post-ionized with electron beam. It is sensitive to high-IP and electronegative species, and slightly more complex than LDMS.


3. MALDI-TOF-MS — The analyte is intimately mixed (dissolved) in a laser-absorbing matrix (e.g., organic acid, Co powder), and then LDI is applied. High mass biomolecules (10-100 kDa) readily survive. MALDI is very sample prep dependent but potentially flyable.


4. REMPI-MS (μL²MS) — In resonance enhanced multi-photon ionization (REMPI), one laser desorbs neutrals; a second laser post-ionizes, with 2 selected to preferentially ionize chosen chemicals. Laboratory versions have proven sensitivity to organics (e.g., PAHs) on fine spatial scale. A highly miniaturized version is in development at APL.


5. RIMS — Resonance ionization mass spectrometry (RIMS) is like REMPI, but multiple laser wavelengths are selected to match transitions to particular metastable electronic excited states of individual species (tunable or multiple fixed-frequency laser method).

{{ 1 }}

Figure 1 shows the most recent test system for all the above methods that is under development and analysis with partial support of NAI funds. The mass spectrometer sensor itself (Fig. 1b) was originally developed in collaboration with GSFC for application on Mars. The total system, now made more flexible and optimized for the broader GCA objectives, is almost complete and will be used productively to this end in the coming year and beyond. The following instrument development tasks were supported in part by NAI in 2005-2006:

{{ 2 }}

1. Completion of LDMS into the “Tower TOF” vacuum system (Figure 2a). The orientation is vertical to permit analysis of loose samples (e.g., powders) when coupled to various sample manipulation systems. Metallic seals are used to achieve lower base pressures for higher signal-to-noise for trace high-mass organics. The sample holding chamber has viewports to permit lateral incidence post-ionization.


2. A compact, high-vacuum compatible XYZ motion stage dedicated to this instrument has been assembled and integrated (Figure 2b). In-vacuum lateral motion precisions of less than 10 μm are achieved with a cross-roller bearing translation stage assembly coupled to a welded bellows. Maximum XY travel of ±12.5 mm permits larger samples (such as 1” electron microprobe mounts) to be examined without having to break vacuum. This is valuable because related inclusions in meteorites where organics may be associated can be several mm across and/or separated by several mm.


3. The process of breaking vacuum takes time and can change sensitivity/instrument bias. Sample holding components and sample application techniques were further developed for analyzing a wide variety of minimally “prepared” materials (not dissolved, extracted, concentrated, etc.). The purpose is to extract as much information as possible from such samples, and to recognize intrinsic limitations, balanced against the advantages of simplicity and spatial context (of in situ analysis). The quality of spectra obtained from neat samples is very dependent on the local electric field (shape and material of holder), sample surface roughness, absorptivity, and any associated mounting materials or chemicals (H₂O, adhesives, matrix). In addition to the etched Si grooves described previously, we have additionally compared the performance of LDMS analysis between “loose” and “deposited” ground samples. Loose samples, requiring minimal preparation, are typically loaded in cylindrical wells (Figure 3a) in a probe tip. Though they are normally lightly packed to eliminate air pockets, etc., these samples are usually rough on the scale of the laser spot, so that point-by-point analysis traverses a range of local surface conditions, which affect absorptivity, intensity, etc. Deposited samples, loaded onto flat surfaces, insulating (Figure 3c) or metallic (Figure 3d), in a water “slurry” suspension by pipetter, followed by air drying, are typically much smoother. The primary difference in spectral quality is that the droplet sample spectra are more reproducible than the loose sample spectra, which exhibit “drop outs” and signal loss over a run. However, individual (single shot) spectra do not appear to be very different; mass resolution is apparently reduced in raw averages of loose spectra due to peak shifts and not necessarily from an intrinsic (single shot) ion formation volume effect. This observation suggests that superior data processing capabilities such as storage and subsequent post-processing of all laser spectra may realize improvements. As opposed to these “bulk” analyses, we have separately noted and studied the importance of microanalysis for detection of organics in complex samples. Even in “ground” samples (Figure 3b), regions of size comparable to the laser spot can remain intact and may contain a majority of organic compounds of interest due to association with mineral phases.

{{ 3 }}

2.2. Sample and Data Analyses

A number of analyses have been conducted to understand further the sensitivities of various laser TOF-MS approaches to detection of high mass organics in neat samples, and to compare those with parallel studies at GSFC with GCMS, LCMS, and other techniques. In support of Objective 2, LDMS organics database, we have continued to (i) analyze systematically a sequence of both procedural blanks and standard reference materials; (ii) record new spectra from geostandard and meteorite samples under identical conditions as the blanks and references; and (iii) evaluate the accuracy of high sensitivity elements such as Li that may be tracers of aqueous alteration on parent bodies of carbonaceous meteorites and of past liquid transport on Mars.

2.2.1. Simple Background Samples

{{ 4 }}

Chemically simple samples such as finely ground silica are being used as backgrounds for organic concentration series as well as proxies for silicate matrix host materials in which only trace levels of organics are expected (on small bodies or on Mars). Figure 4 shows example spectra from the DS-TOF instrument at 355 nm laser desorption from such a sample. At low laser intensity, only Na and K are found to ionize efficiently, as expected. These have low ionization potentials (IPs) and are abundant in the sample. The metallic (Cr, Fe) peaks are likely from the trace metallic component of this sample, though we cannot rule out desorption of the stainless steel sample holder by scattered light from these data. As the laser intensity is increased (bottom of Figure 4), these peaks grow in size and broaden, and new low-IP elements, present at low trace levels, appear in the spectrum. Li, Rb, and Cs are expected only at parts-per-million by weight (ppmw) or below, however they are readily desorbed and ionized. As each of these elements has significant geochemical value, this very high sensitivity will be more thoroughly evaluated for relevance to astrobiology objectives. Other peaks are either atomic ions, cluster ions, or trace organic impurities. This sample thus serves as a reasonable procedural blank for most concentration series.

2.2.2. Spiked Background Samples

{{ 5 }}

Simple experiments with powder mixtures of blanks and solid-phase small organic standards such as PAHs and carboxylic acids at various concentrations are being conducted systematically to understand sensitivity, matrix effects, instrument biases, and sample preparation factors in LDMS of organics in geological materials. For example, Figure 5 shows a multi-shot raw average of spectra from a physical admixture of the PAH phenanthrene at 0.1% by weight to silica sand. The laser intensity was kept low to minimize fragmentation. The parent peak at m/z 178 is immediately detected in the correct position. The mass resolution at this peak is modest, due to shifting of individual laser shot spectra, but sufficient to identify the mass accurately. The identity of the peak at m/z 83 labeled ‘X’ is unknown. Phenanthrene is a nonvolatile aromatic compound that is expected to be common in the “free PAH” soluble organic phase in comets and meteorites, but is challenging to detect directly by non-laser approaches. Only the parent mass is present as the molecular ion, implying that multiphoton bombardment, possibly enhanced by incipient inverse Bremsstrahlung electron bombardment in the desorption plume, is responsible for ion formation. The relative chemical stability of phenanthrene makes it much less likely to ionize by protonation or alkali attachment. In contrast, similar analysis of samples with admixtures of carboxylic acids shows different behavior. Figure 6 provides example spectra of benzoic acid in silica sand. Only minor if any signal is present at/near the parent mass m/z 122, while significant peaks are found for sodium benzoate and its —O neutral loss fragment. The substitution of H with Na or K is more typical in compounds with OH terminals than at H positions in pure hydrocarbons. Phenanthrene and benzoic acid are moderate-mass compounds, relatively non-volatile and semi-volatile, respectively, that could originate from either biotic or abiotic source organics. As such these studies are part of the larger GCA Theme 4 systematic evaluation of various mass spectrometric techniques for determining the sources of organics that might be found at comets and on Mars, such as by the Sample Analysis at Mars suite on MSL (PI P. Mahaffy, who leads GCA Theme 4 at NASA/GSFC).

{{ 6 }}

The lower portion of Figure 6 demonstrates the sensitivity of LDMS to higher mass “impurity” organics that are present in the as provided crystallite material. The 14 amu periodicity in this portion of the spectrum suggests these are primarily hydrocarbons, mixed between aliphatic and aromatic moieties, and possibly including higher order carboxylic acids. We are presently re-evaluating the background and will be re-running these standards following heat treatment to eliminate any potential source of organics in volatile phases.

2.2.3. Natural Analog Samples

Figure 7 shows data from an organic-bearing natural shale sample that is often used in established methods of organic analysis including derivatization and hydropyrolysis GCMS. Preliminary raw DS-TOF averages produce a great deal of ion signal, however the TOF jitter reduces the appearance of sharp peaks suggesting single-shot spectra may be more appropriate for such powdered samples. Qualitatively, the LDMS spectra do show less evidence of extensive high mass aromatics, compared to C-chondrites, consistent with n-alkane dominated IOM from algal source (Greenwood et al. 1998). Both jitter and chemical noise are currently limiting resolution of high-mass parent compounds in powder samples; extensive fragmentation at higher laser power suggests trying lower pressure and intensity. This also applies to analysis of silicate smokes (Fe and Mg based) obtained from N. Johnson and J. Nuth (GCA Theme 3), which are undergoing extensive analysis.

{{ 7 }}

We are additionally assessing the quantitative performance of LDMS for isotopic and elemental analysis of alkali metals, particularly Li and Rb, to which laser TOF-MS appears to be exceptionally sensitive. Preliminary UV LDMS data indicate “large” isotope variations (% level) might be recovered from unprepared samples (without matrix). We are currently characterizing the small instrument bias as a function of sample mounting scheme (probe and material details). As a preliminary result, we have looked at the raw ⁶Li/⁷Li ratio in a Li carbonate standard RM 8545 LSVEC. Li is an important element for tracing aqueous processing due to its very high mobility and strong relative isotope bias developed during traversal. An average over 3 spots and 3 laser pulse energies gave ⁶Li/⁷Li = 0.085, compared with LSVEC standard value of 0.0832 using a very crude peak analysis method. While the statistics are insufficient to determine precision as yet, we anticipate that the performance will be sufficient to capture typical variations of several % within even small regions in whole rock and meteorite samples. However, documenting this as an external calibration may require improvements in the laser desorption and ionization characteristics, such as using shorter pulse durations and wavelengths.