Progress 06/01/23 to 05/31/24
Outputs
Target Audience:The target audience of this research will be the soil, ecological, agricultural, and agronomic research communities. These communities iwll be reached by publication in high-profile, peer-reviewed journals, and by presentations at research conferences such at the American Geophysical Union Fall Meeting and the Soil Science Society of America meeting. Further, this work will target graduate, undergraduate, and high school researchers, to expose them to new concepts and capabilties, expand their skillset, and engage them in new research. This will occur through experiential learning opportunities and laboratory instruction. During this reporting period, we attended both of the above meetings and discussed the technology, although we have not yet presented results. One of our target audiences has been young scientists, and we worked toward that goal this period. We hired an intern (high school student attending college in the Fall)work with us during Summer 2024, and some of her efforts have been in service of this project. We intend to re-hire her for Summer work in 2025. We are also looking to potentially bring a U. Arizona graduate or undergraduate to work at Aerodyne for a long period of time. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?There have been several opportunities for training and development during this reporting period. First, we brought on an intern (high school senio/college freshman) that worked with us over the summer 2024, trying out different injection methods. This was her first experience in a lab, so she learned a wide variety of skills, from solution-making to glassware cleaning, to software writing and data analysis. We will also be inviting her to join us next summer. In addition, one scientists on the team, Dr. Guo has recieved extensive trainign on mass spectrometric analysis, allowing us to quickly work up data. Further, we have had several scientists attend scientific conferences. We have not present ed any results from this projectm yet, but we have used the opportunity to build relationships with potential collaborators and informally discuss our work. We anticipate that in the next reporting period we will be presenting work and potentially writing papers. How have the results been disseminated to communities of interest?As discussed above, although we have not yet formally disseminated results (talks, poster, papers), we have had discussions with potential future collaborators and other scientists that have given us insights into potentially interesting science that could happen with the technology we're developing. What do you plan to do during the next reporting period to accomplish the goals?In line with the proposed timeline, in the next year we plan to pursue several interesting research avenues. First, we will use MD probe arrays to explore how soil microbial commuities respond to the exudation of nutrients from a simulated root. This will require building small scale probe arrays and placing them in soil (and sterile sand), mapping out gradients. We will engage our co-PI Li Li (Penn. State) to model the transport of these solutes. We will also take small soil sampled for biological analysis (DNA/RNA) by co-PI Meredith. We will continue development of the rhizobox, which will allow us to steer a root next to an MD probe to allow us to see real root exudation. We expect this accomplishemnt to start near the end of next reporting period.
Impacts
What was accomplished under these goals?
The focus of this period of the project was on developing and refining the detection and sampling method. Central to the success of the project is to ensure that we can reliably and quantitatively measure a wide range of solutes using the field-able Vocus chemical ionization mass spectrometer. Detection approach The detection approach builds upon mass spectrometric methods aimed at measuring gaseous organic compounds using chemical ionization (CI). For this project we utilized the Aerodyne Vocus proton transfer reaction mass spectrometer. This time-of-flight instrument is a powerful tool that is capable of measure parts per trillion levels of gas concentration. This type of instrument has very rarely been used for liquid sampling, so a novel liquid sampling unit had to be developed and tested. During this period we designed and built the new inlet, and tested it on a variety of solutes that might be found in soil, including sugars, amino acids, and organic acids. The general design of the inlet relies upon vaporization of a discrete liquid sample, followed by rapid transport to the Vocus inlet. Thermal stability is important in this setup, so the design consisted of a stream of pre-heated nitrogen (avoiding oxygen that could lead to thermal oxidation) passing over a hot sample injection head, then carrying the vaporized solute to the Vocus inlet within 100 milliseconds. During this testing we optimized the vaporization by varying nitrogen and injection head temperature, nitrogen flow rate, injection head and needle material, injection flow rate and volume, transfer tubing material and coatings, Vocus inlet temperature, Vocus inlet flow rate. These parameters were optimized for maximum signal to noise, time response, and ease of cleaning. After optimizing these parameters, we observe nearly linear responses to a wide range of chemical classes. However, we observed a non-linearity in the response of micromolar concentrations of some organic acids such as glutaric and maleic acid. We trace this to a need for alower pH of the solution, allowing for the organic acids to re-neutralize with excess protons in order to vaporize as neutral molecules. This required us to modify the injection approach such that a small concentration of HCl solution was added into the sample (~1 mM). This provided high linearity of across a wide range the compounds. Sampling Approach Once the vaporization and detection and approach were established, we developed a sampling system that utilized microdialysis (MD) probe arrays. To achieve this we utilized cutting edge technology from VICI Valco in the form of very low-volume multiselector valves, allowing for up to 16 probes to be sampled, and low-flow liquid pumps. The latter have been very valuable as a replacement for syringe pumps, which have a quite limited volume for the flow precision we need. In this setup, clean water is flowed through the MD probe, through a multiselector valve, and into a small (5 uL) sample loop. Once the sample loop is filled it is opened to the injection head and its contents are pushed into the injection head at a fast flow rate to rapidly vaporize and measure the sample. The entire sequence is automated and scheduled, allowing for hands-off, unattended sampling. Multiple aspects of this sampling approach have been tested to maximize collection efficiency of the MD probes and minimize complexity. This work is still ongoing. Central to this is the need to identify the optimal sampling timing and flows. Thus far these parameters have included perfusate flow rate, sample transfer time, sample volume, and tubing material. One extremely important aspect was tubing material. Polyether ether ketone (PEEK) is an industry standard for low-volume sample transfer, as it is relatively inert and tubing with extremely low inner diameter (ID) can be manufactured (<0.25 mm). We found during testing that while the PEEK tubing was inert during rapid sample transfer, if the sample resided in the tubing for more than 2 minutes, it began to degrade the sample. The amino acids (here alanine and glycine) were particularly susceptible to this, but the organic acids were as well. This was especially true in the sample loop itself. Alanine in a PEEK sample loop would degrade quickly. On the contrary, when the PEEK loop was replaced by PTFE, the sample could stay in the loop for over 3 hours without any signs of degradation. As such we are currently exploring suppliers of PTFE tubing with a small ID. As part of this work, we also explored two more aspects of the sample extraction. First, most MD applications push perfusate through the probe and into a collection volume (in our case a sample loop). However, with the fluid pumps it is possible to actively draw flow, rather than push. We have therefore tested the approach where the sample is drawn (not pushed) out of the MD probe and into the sample loop. This is a simpler approach from a flow and control perspective the a perfusate-push approach. Preliminary results indicate that this "pull" mode showed similar recovery rates as in "push" mode. We also tested two approaches to drawing sampling into the sample loop. First, the sample can slowly fill the transfer lines until it reaches the sample loop. While simple, the amount of sample required is dependent not only on the sample loop volume but also on the transfer line volume. For example, over 1 meter of 0.125 mm (0.005") ID transfer tubing is 12 uL, which when including the sample loop then requires 17 uL of sample. Two thirds of the sample is therefore never loaded into the sample loop and flushed out after the measurement. Further, if the perfusate flow rate is 0.5 uL/min, it requires 34 minutes to perform the sample extraction, during which there may be sample losses on the tubing. To address this we developed a plug-flow approach, where the required sample is drawn out of the probe into the transfer tube. When slightly more has been extracted than is required by the sample loop, the flow is increased to quickly transport that volume into the sample loop. We have tested this, drawing 0.5 uL/min from the MD probe, then rapidly transferring it to the loop at 10 uL/min. For a 5 uL sample loop, the entire sequence requires ~10-12 minutes. This has the added advantage that it allows for much longer transfer tubes between the probe and the instrument. Current experimental The measurement system is nearing readiness for true soil water measurements with one month. The detection approach has been demonstrated to be highly reproducible and sensitive. There are several tests yet to run. Most importantly, we will test whether we can use a double-common selector valve to continuously flow a small amount of perfusate through the probe, in order to allow a very slow flow (<<1 uL/min) through the probe while it is not being sampled, to encourage full equilibration through the membrane. In this configuration, in the course of ~2 hours between a given probe measurement, new sample is extremely slowly being pulled into the transfer line, such that when that probe is finally addressed for measurement, the equilibrated sample can be quickly drawn into the sample loop. We are currently testing the viability of this, as it will allow the most equilibrated measurement possible. Our first experiments in soil will make use of a MD probe array that allow for artificial dosing of the soil with realistic quantities of solutes. The solute gradient emanating from the dosing point will be measured with MD probes spaced away from the probe with 1 mm-resolution. We will first be testing this in sand, where we expect no uptake of solutes, then extend into real soil samples. We will sieve the real-world soil in order to ensure a fine texture that is amenable to measurement of mm-scale gradients. This array has entered final design stages and will be ready within several weeks.
Publications
|