Research Article www.acsami.org
Pyrolyzable Nanoparticle Tracers for Environmental Interrogation and Monitoring Jason R. Cox,*,† Mohammed Alsenani,† Scott E. Miller,‡ James A. Roush,‡ Rena Shi,† Hooisweng Ow,† Sehoon Chang,† Anthony A. Kmetz,† Shannon L. Eichmann,† and Martin E. Poitzsch† †
Aramco Research Center-Boston, Aramco Services Company, 400 Technology Square, Cambridge, Massachusetts 02139, United States ‡ 908 Devices Inc., 27 Drydock Avenue, Boston, Massachusetts 02210, United States S Supporting Information *
ABSTRACT: Environmental tracing applications require materials that can be detected in complex fluids composed of multiple phases and contaminants. Moreover, large libraries of tracers are necessary in order to mitigate memory effects and to deploy multiple tracers simultaneously in complex oil fields. Herein, we disclose a novel approach based on the thermal decomposition of polymeric nanoparticles comprised of styrenic and methacrylic monomers. Polymeric nanoparticles derived from these monomers cleanly decompose into their constituent monomers at elevated temperatures, thereby maximizing atom economy wherein the entire nanoparticle mass contributes to the generation of detectable units. A total of ten unique single monomer particles and three dual-monomer particles were synthesized using semicontinuous monomer starved addition polymerization. The pyrolysis gas chromatography-flame ionization detection/mass spectrometry (GC-FID/MS) behavior of these particles was studied using high-pressure mass spectrometry. The programmable nature of our methodology permits simultaneous removal of contaminants and subsequent identification and quantification in a single analytical step. KEYWORDS: pyrolysis GC-MS, thermally degradable polymers, nanoparticles, enhanced oil recovery, colloidal stability, high-pressure mass spectrometry
■
INTRODUCTION The importance of hydrocarbon feedstocks to both the energy and materials landscapes necessitates the development of efficient and practical approaches to crude oil recovery.1,2 The complexity of subterranean fluid pathways coupled with the placement of numerous injection and production wells require extensive optimization of injection rates in order to maximize oil recovery.3,4 This is accomplished, in part, through use of tracer tests wherein connections between wells are identified via the detection of a known compound at one or multiple wells after incorporation into the injection fluid.5−7 Tracers find utility in applications ranging from hydrocarbon production to groundwater management and monitoring.8−10 With field tests of underground tracers dating back to 1905, their continued use is validation of the information they can provide in the highly under-sampled subsurface.10 Additionally, partitioning tracer tests can be used to determine the quantity of residual oil (or NAPL in environmental parlance) in a well’s zone of interrogation for enhanced recovery. These types of tests are often used in environmental remediation to remove harmful chemicals in the vadose zone of the subsurface.11,12 They also are common in the development of chemically enhanced oil recovery projects.12,13 The aforementioned examples demonstrate how © XXXX American Chemical Society
tracers help serve the purpose of understanding the distribution of geological features and fluids through the injection of well characterized tracers. This process, however, is not without challenges. Specifically, the heterogeneous nature of the reservoir fluid matrix presents a major issue, produced fluids contain multiple phases abundant with ions, dissolved organic matter (DOM), and hydrocarbons.14 The hydrocarbon component is particularly troublesome as crude oil is known to exhibit strong fluorescence across the visible spectrum, thereby requiring implementation of a separation step prior to analysis in order to rid the spectrum of background signal that might interfere with tracer detection.15 Elegant analytical approaches exploiting high-performance liquid chromatography (HPLC) separation before either fluorescence or mass spectroscopic analysis have become the method of choice for tracer identification and quantification.16−18 However, these analyses generally are not field-portable and require sophisticated instrumentation. To resolve these issues, intricate solutions have been devised through the utilization of Received: December 14, 2016 Accepted: March 14, 2017 Published: March 14, 2017 A
DOI: 10.1021/acsami.6b16050 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 1. (a) Schematic diagram depicting the β-scission mechanism governing the thermal degradation of polymer nanoparticle tracers. (b) Cartoon illustrating the various steps of the tracer analysis including removal of interfering fluids, degradation of the tracer, and subsequent analysis via GC-MS.
volatile components such as hydrocarbons or water from the sample matrix prior to pyrolysis; (2) once pyrolysis of the nanoparticle is achieved, the monomers volatize out of the resulting salt matrix eliminating interference or scaling from the saline fluids; (3) the commercial availability of a variety of functionalized styrenic and methacrylic monomers facilitates the development of a large library of unambiguously detectable barcodes; and (4) the decomposition temperature of polymeric nanoparticles is a function of the polymer backbone linkages thus offering an additional dimension of barcoding capability. Moreover, the comprehensive process of removing hydrocarbons and water, degrading nanoparticles into monomers, separation of components via GC and analysis via FID/MS is completed in less than 7 min, a time scale significantly shorter than that of traditional tracer analyses. Herein, we detail our efforts toward developing a library of polymer nanoparticle tracers and a detection methodology that leverages three dimensions of barcoding capability, monomer mass, retention time, and polymer degradation temperature. The use of nanoparticles for tracing applications is advantageous compared to small molecule tracers since the transport and retention characteristics of the nanoparticle tracers are dictated by the coating material. With nanoparticles, modulation of the core should yield similar behavior among the library of barcodes. For small molecules, each individual compound must be vetted through a battery of tests prior to deployment.
delayed luminescent species such as lanthanide-ligand pairs embedded within nanoparticle matrixes or inorganic phosphors with persistent emission which can be detected using fluorescence spectroscopy.19−22 In these cases, the issue of background hydrocarbon signal is mitigated by the extended time scale of luminescence originating from lanthanide ions or phosphors in comparison to the short-lived fluorescence of hydrocarbon components in the crude oil. Nonetheless, these materials are expensive, difficult to synthesize in large quantities, and limited by the number of potential barcodes that could be deployed. The barcoding aspect is critical given the number of unique and differentiable tracers required for simultaneous injection into a large, multiwell oil field. Inspired by self-immolative systems wherein the complete decomposition of a macromolecule into its constituent monomers is induced by external stimuli,23−26 we embarked on a program to develop inexpensive and scalable tracer nanoparticles that emulate this behavior and exhibit the requisite stability to survive the harsh subterranean conditions of oil reservoirs. In comparison to traditional nanoparticle tracer systems in which the detectable element is often encapsulated in silica or other carrier matrixes, platforms that undergo programmable decomposition or self-immolation maximize detectability by fragmenting the entire nanoparticle mass into potentially detectable units which can then be analyzed via mass spectrometry (MS) and flame ionization detection (FID). Mass spectrometry offers an exquisite balance between sensitivity and resolution, thus affording the opportunity to detect a variety of barcodes at low levels. Our approach leverages the efficient depolymerization of certain types of polymers, specifically polystyrenes and polymethacrylates, into their constituent monomers at elevated temperatures as shown schematically in Figure 1.27−29 The resulting monomers are separated and detected using GC-FID/MS. The key features of this methodology are as follows: (1) polymeric nanoparticles exhibit limited volatility compared to their monomers permitting selective removal of
■
EXPERIMENTAL SECTION
Materials. All chemicals were obtained from Fisher Scientific (Fair Lawn, NJ) and used as received unless noted otherwise. Polyethylenimine (25 kDa) was obtained from Sigma-Aldrich (St. Louis, MO). Styrenic and methacrylic monomers were passed through a short alumina column to remove inhibitors before use. Crude oil was a gift from Saudi Aramco (Dhahran, KSA). Water was double-deionized using a Millipore Milli-Q system to produce 18 MΩ deionized water. Sealable 5 mL microwave vials (CG-4920-01) were obtained from Chemglass, Inc. (Vineland, NJ) and used as received. B
DOI: 10.1021/acsami.6b16050 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces Stock electrolyte solutions were prepared by dissolving the appropriate amount of electrolyte into DI H2O followed by filtration through a 0.2 μm nylon filter to remove any impurities or dust particles. Tangential flow filtration was performed using a KrosFlo Research Iii TFF system from Spectrum Laboratories, Inc. (Rancho Dominguez, CA). General Procedure for the Synthesis of Polymeric Nanoparticles. The procedure consists of the following steps: (1) polymer nanoparticle synthesis, (2) coating with PEI followed by functionalization with glycidol and cross-linking, and (3) characterization via dynamic light scattering, GC-FID/MS, and thermogravimetric analysis. Polymer nanoparticles were prepared using a procedure reported by Sajjadi with slight modification.30 In this approach, 60 mL of sodium dodecyl sulfate (SDS) (2.55 wt % in DI water) was added to a round-bottom flask and left to degas with N2 for 15 min. Next, 100 mg of ammonium persulfate was added to the flask and left to dissolve under an N2 purge. After dissolution of the initiator, the temperature was raised to 90 °C and held for 15 min. Next, a syringe containing the styrenic monomer was connected through a tube to the flask, and 1 mL of monomer solution was injected at 0.02 mL/min using a programmable syringe pump (see Figure S1 for an image of the reaction apparatus). After addition of monomer, the reaction was allowed to proceed for 30 min and the reaction was then cooled to room temperature. PEI-Glycidol Coating Procedure. The 10 wt % 25 kDa branched polyethylenimine (PEI) stock solution in pH 5 acetate buffer (0.7 mL of glacial acetic acid and 0.6 g of potassium hydroxide per liter) was prepared. A volume of 10 mL of 1.7 wt % as synthesized PS nanoparticle with sodium dodecyl sulfate surfactant on the surface was diluted into 10 mL of deionized water and placed in an addition funnel. In a round-bottom flask, 10 mL of 100 mg/mL stock PEI solution was diluted into 10 mL of pH 5 acetate buffer and magnetically stirred. Into this PEI solution, nanoparticles from the previous step were added dropwise using the addition funnel. PEI-coated nanoparticles were collected and pH adjusted to ∼8 using 1.0 M hydrochloric acid and placed in an addition funnel. Prior to treatment with glycidol, 1 mL of the cross-linker, 1,4-butanediol diglycidyl ether, was added to 20 mL deionized water in a round-bottom flask and homogenized. The PEI-coated nanoparticle solution was subsequently added dropwise into the crosslinker solution and stirred overnight. To quench the reaction, 10 mL of 2.0 M tris buffer was added to the solution and stirred for 1 h prior to subjection to tangential flow filtration purification to remove excess PEI. To further glycidylate the PEI-coated PS nanoparticles, 2 mL of glycidol was added into 20 mL of PEI-coated nanoparticles (approximately 0.85 wt % concentration) and stirred overnight. To quench this reaction, 5 mL of 2.0 M tris buffer was added to the solution and let stir for at least 1 h before the purification via tangential flow filtration. Colloidal Stability Testing Protocol. Solution Preparation. The 100 ppm solutions of the coated nanoparticles were prepared in either seawater or low-salinity brine (see Table S1 for compositions). The solutions were placed inside 5 mL microwave vials which were subsequently crimp-sealed with PTFE lined aluminum septa and degassed with N2 for 5 min. Three replicates of each nanoparticle solution were placed in a thermostat regulated oven operating at 103 °C. An additional set of duplicates of each nanoparticle solution were kept at room temperature to serve as control samples. Determination of Hydrodynamic Diameter using Dynamic Light Scattering (DLS). After the specified time interval, solutions were removed from the oven and allowed to cool to room temperature prior to analysis. Vials were decrimped and the solutions were transferred to polystyrene cuvettes (BI-SCP, Brookhaven Instruments Co.) for DLS analysis. Dynamic light scattering experiments (intensity average) were performed using a Brookhaven NanoBrook system (Brookhaven Instruments Co.) operating at a measurement angle of 90°. The CONTIN algorithm was used for fitting of the autocorrelation functions. Hydrodynamic diameters were calculated using the Stokes− Einstein equation. Each sample was measured in triplicate with an acquisition time of 2 min/sample and a count rate of at least 500 kcps. Differences in salt solution viscosity were not adjusted or accounted
for in the measurements, thus the relative change in hydrodynamic diameter is most informative. After measurement, the samples were transferred back to the 5 mL microwave vials, sealed, and placed in the oven or benchtop (room temperature samples). Pyrolysis. Pyrolysis was conducted using an AS 5250 pyrolysis autosampler from CDS Analytical. Samples are housed in a thin closed-end tube filled with quartz wool. Liquid samples (usually 1 μL) are injected into the wool and held in place by capillary forces. Tubes are manually loaded into the pyrolyzer’s autosampler, which drops samples into the heating compartment, which is then purged with inert gas (here helium). The pyrolyzer then operates in two successive heating stages: drying and pyrolysis. After purging, the exhaust valve shifts to a transfer line to expel waste, and the sample is dried at a temperature and duration set by the operator, allowing for the removal of unwanted materials that volatilize at temperatures below the pyrolysis temperature. Two separate pyrolysis methods were used. For styrenic nanoparticles, the drying temperature was set to 400 °C for 20 s followed by pyrolysis at 800 °C for 15 s. For methacrylic nanoparticles, the drying temperature was set to 350 °C for 15 s followed by pyrolysis at 400 °C for 15 s. In both cases, the temperature ramp rate was ballistic. Temperature ramp studies for SDS were conducted similar to the procedure reported for nanoparticles with the exception that only a single pyrolysis temperature was used and this temperature was incremented from 300 to 800 °C. Particle Detection using GC-MS/FID. A prototype G908 portable GC-MS instrument from 908 Devices was used for analyses. The G908 consists of a ballistic gas chromatograph capable of rapid temperature profiles combined with multiple, parallel detectors. The G908 achieves column heating rates of up to 5 °C/s via direct column heating. The detectors used were a flame ionization detector (FID) and 908 Devices miniature mass spectrometer. Chromatographic separations were performed on an MXT-1 column (4 m, 0.25 mm i.d., 0.25 μm df) using a temperature ramp from 40 to 300 °C at 1 °C/s and then holding 300 °C for 90 s for a total run time of 6 min. The output flow of material from the column is then split between two detectors: the MS and the FID. The input flow rate was approximately 3.5 mL/min for each detector. The FID response provided reproducible and adequately time-separated signals suitable for peak identification and detection for the investigated markers using the GC method detailed above. The MS detector uses a microscale 3D ion trap operating at pressures between 1 and 10 Torr depending on the application. These operating pressures enable extremely fast system start-up times (5−10 s) and very small low-power (
