TargetSpecific Capture of Environmentally Relevant ...

DOI: 10.1002/chem.201502021

Full Paper

& Nanoparticles

Target-Specific Capture of Environmentally Relevant Gaseous Aldehydes and Carboxylic Acids with Functional Nanoparticles McKenzie L. Campbell,[a] Fernanda D. Guerra,[b, c] Jhilmil Dhulekar,[b] Frank Alexis,*[b, d] and Daniel C. Whitehead*[a] Abstract: Aldehyde and carboxylic acid volatile organic compounds (VOCs) present significant environmental concern due to their prevalence in the atmosphere. We developed biodegradable functional nanoparticles comprised of poly(d,l-lactic acid)-poly(ethylene glycol)-poly(ethyleneimine) (PDLLA-PEG-PEI) block co-polymers that capture these VOCs by chemical reaction. Polymeric nanoparticles (NPs) preparation involved nanoprecipitation and surface functionalization with branched PEI. The PDLLA-PEG-PEI NPs were characterized by using TGA, IR, 1H NMR, elemental

Introduction The environment has been greatly affected by the rapid pace of industrialization and the increasing concentration of volatile organic compounds that are released. Volatile organic compounds (VOCs) are examples of compounds with low vapor pressures that are emitted into the atmosphere from sources divided into two categories: biogenic (i.e., mainly vegetative processes) and anthropogenic.[1, 2] Although biogenic sources emit approximately ten times more VOCs than anthropogenic sources, in urban areas, anthropogenic VOCs often dominate and therefore are of concern to the human population.[3] VOCs

[a] M. L. Campbell,+ Prof. Dr. D. C. Whitehead Department of Chemistry, Clemson University 467 Hunter Laboratories, Clemson, SC 29634 (USA) E-mail: [email protected]

[b] F. D. Guerra,+ J. Dhulekar, Prof. Dr. F. Alexis Department of Bioengineering, Clemson University 203 Rhodes Annex, Clemson, SC 29634 (USA) Fax: (+ 1) 864-656-4466 E-mail: [email protected] [c] F. D. Guerra+ CAPES Foundation, Ministry of Education of Brazil Braslia - DF 70040-020 (Brazil) [d] Prof. Dr. F. Alexis Institute of Biological Interfaces of Engineering Department of Bioengineering, Clemson University 401-2 Rhodes Engineering Research Center, Clemson, SC 29634 (USA) [+] These authors contributed equally to this work. Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201502021. It contains elemental analysis of PDLLA-PEG-COOH and PDLLA-PEG-PEI NPs, TGA data for biodegradation study, and 1H NMR spectra for PDLLA-PEG-COOH and PDLLA-PEG-OMe NPs. Chem. Eur. J. 2015, 21, 14834 – 14842

analysis, and TEM. The materials feature 18, 28, and 38 amines on their surface, capable of capturing aldehydes and carboxylic acids from gaseous mixtures. Aldehydes were captured by a condensation reaction forming imines, whereas carboxylic acids were captured by acid/base reaction. These materials reacted selectively with target contaminants obviating off-target binding when challenged by other VOCs with orthogonal reactivity. The NPs outperformed conventional activated carbon sorbents.

include a variety of reactive functional groups, such as aldehydes, carboxylic acids, alcohols, amines, amides, and aromatic compounds.[3] VOC emissions comprising short-chain carboxylic acids and aldehydes arise from both vehicular exhaust and the atmospheric photochemical oxidation of olefin and hydrocarbon emissions.[4–7] Further, the global daily use of cook stoves and fireplaces contribute to the emission of carbonyl compounds along with other compounds resulting from incomplete combustion of biomass and fossil fuels, which in turn undergo atmospheric oxidation to aldehydes and ketones.[8–10] Additionally, some aldehydes and carboxylic acid contaminants are also observed in enclosed environments, such as homes and apartments, due to various sources, including paints, aerosols, and wood products.[11] High concentrations of these VOCs are known irritants with the U. S. Environmental Protection Agency (EPA) listing thirteen carboxylic acids and aldehydes/ketones as hazardous air pollutants under the 1990 Clean Air Act Amendments.[12–14] Additionally, the EPA lists three of those aldehydes/ ketones as priority pollutants.[15] Aldehydes are potent mucosal membrane, eye, skin, and respiratory irritants, even causing bronchial asthma symptoms including several reports of full asthma attacks.[8, 11, 12, 16] Additionally, VOCs are known for their low, often unpleasant, odor thresholds: below one parts per billion (ppb) in some cases.[13] As was previously discussed, atmospheric reactions of primary emissions can form newly hazardous compounds. For example, the reaction of formaldehyde with atmospheric HCl[9] generates bis(chloromethyl)ether, a suspected carcinogen.[17, 18] Both volatile organic aldehydes and carboxylic acids are also implicated in the atmospheric generation of light-scattering aerosols, which contribute to increasing smog problems in urbanized areas.[5, 7, 8]

14834

Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper Environmental pollution has become a global concern, and providing clean air and water remains a challenge. Conventional technologies that have been used to treat organic and toxic waste include adsorption, biological oxidation, chemical oxidation, and incineration. With the growth of nanotechnology, there is excellent potential for the fabrication of nanomaterials with large surface-to-volume ratio, high chemical reactivity, and unique functionalities to treat pollutants.[19] Nanomaterials play a large role in environmental remediation and have been used for various applications, such as the treatment of natural waters, soils, sediments, industrial and domestic wastewater, mine tailings, and polluted air.[3] Nanomaterials are extremely versatile; they have been employed previously as adsorbents,[20, 21] catalysts,[22] and sensors[23] owing to Figure 1. A) Schematic representation of PDLLA-PEG-COOH non-functionalized and PDDLA-PEG-PEI functionalized NP structures. Inner core formed by PDLLA chain; outer shell formed by PEG chain and carboxylic acid reactive their unique properties. Nano- sites in non-functionalized NP. B) Imine and ammonium carboxylate formation from reaction of amines present in materials can be functionalized the PDLLA-PEG-PEI NPs with aldehydes and carboxylic acid target molecules, respectively. by coating techniques or chemical modification to improve surface and optical properties, avoid aggregation, and eliminate CO2 adsorption in direct capture from ambient air and flue gas the interaction between the nanomaterials and biological subwith reversible CO2 desorption capabilities.[34, 35] By installing [19] stances. a branched amine on the surface of our self-assembled polyA variety of studies have exploited the use of nanomaterials meric nanoparticles (NPs), we surmised that aldehydes might for the remediation of VOCs in an effort to decrease air pollube captured by means of a condensation reaction to form an tion.[19, 24, 25] Examples include the use of metal and metal oxide imine, whereas the carboxylic acids might form ammonium nanomaterials,[24] dendrimers,[19] carbon nanomaterials,[25] and carboxylates through acid/base reaction (Figure 1 B). Detailed polymer nanocomposites.[26] The target-specific capturing of herein is a full account of our efforts toward the development compounds from gaseous mixtures is a significant and difficult and evaluation of these materials in the context of remediating problem, because off-target fouling of sorbents might limit gaseous contaminants bearing aldehyde and carboxylic acid their utility. Therefore, a broad impact might be achieved with functionality. the development of a method that can selectively capture compounds of different functionalities from complex gaseous Results and Discussion mixtures of various concentrations. Herein, we report a versatile and modular platform that can selectively target the capture Polymeric nanoparticles (NPs) were synthesized and functionalof aldehyde and carboxylic acid functional group classes in the ized for capturing target gaseous molecules of the aldehyde gas phase. We designed functional nanoparticles comprised of and carboxylic acid functional group classes. Poly(d,l-lactic poly(d,l-lactic acid)-poly(ethylene glycol)-poly(ethyleneimine) acid)-poly(ethylene glycol)-carboxylic acid (PDLLA-PEG-COOH) (PDLLA-PEG-PEI) block co-polymers that present a branched block co-polymer was synthesized, and PDLLA-PEG-COOH NPs polyamine functionality on their surface (Figure 1 A). Recent were obtained through the solvent-evaporation technique.[36] work involving the selective capture of aldehydes and CO2 by PDLLA-PEG-COOH NPs were reacted with branched poly(ethyusing amino-functionalized mesoporous silicates highlights the leneimine) (PEI) to obtain PDLLA-PEG-PEI NPs (Figure 1 A), impact that amine-containing nanomaterials afford on targetthrough an amide conjugation reaction with 1-ethyl-3-(3-dimeing gases.[27–32] The Jones group has disclosed elegant studies thylaminopropyl)carbodiimide (EDC). The PEI polymer was utilizing poly(ethylenimine)-capped mesoporous silicates for chosen to functionalize PDLLA-PEG-COOH NPs based on the Chem. Eur. J. 2015, 21, 14834 – 14842

www.chemeurj.org

14835


Full Paper presence of a suite of primary, secondary, and tertiary amines in its structure. The two features that distinguish our NPs from the materials developed during the course of the pioneering work of Jones and co-workers[27–35] are both results of our design strategy: 1) our materials are based upon a biodegradable and environmentally friendly PDLLA polymer platform, which is critical to eliminate environment contamination with metal silicate materials; and 2) our EDC-mediated NP capping strategy is modular and tunable.

Nanoparticle characterization NMR spectroscopy, TGA, and FTIR techniques were performed for non-functionalized PDLLA-PEG-COOH, PDLLA-PEG-OCH3 NPs and functionalized PDLLA-PEG-PEI particles. We selected PDLLA-PEG-OCH3 NPs as a non-reactive control to test the specific reactivity of the PDLLA-PEG-PEI NPs. Specifically, NPs displaying terminal methoxy functional groups were not expected to exhibit reactivity with the targeted aldehyde and carboxylic acid contaminants. NMR chemical shifts are shown in Table 1. Although PDLLA-PEG-OCH3 and PDLLA-PEG-COOH NPs did not have a resonance appearing at approximately 2.6 ppm, this resonance was observed for PDLLA-PEG-PEI NPs, indicating the presence of PEI in the functionalized NPs. An overlay of the 1H NMR spectra of PDLLA-PEG-COOH and PDLLA-PEG-PEI NPs, as well as of PEI, is presented in Figure 2. By comparing the spectra observed for PDLLA-PEG-COOH and PDLLA-PEG-PEI NPs, a new resonance was observed at approximately 2.6– 2.7 ppm for PEI NPs compared to non-functionalized NPs. This resonance was attributed to the chemical shift of the methylene hydrogens of the PEI backbone, suggesting that the PDLLA-PEG-COOH NPs were successfully functionalized with PEI through the EDC coupling. It was also possible to verify the presence of PEI by the observation of a strong absorbance appearing at approximately 3000 cm¢1 in the FTIR spectrum (Figure 3). This peak is attributed to N¢H stretching frequencies due to the presence of 18 and 28 amines in the functionalized NPs. Furthermore, TGA analysis of PDLLA-PEG-COOH, PDLLA-PEG-OCH3, PDLLA-PEG-PEI NPs, and PEI polymer demonstrated that PDLLA-PEG-COOH NPs were successfully functionalized with PEI, as shown in Figure 4. It was observed that PDLLA-PEG-PEI NPs exhibit a different thermal degradation profile compared to the two non-functionalized NP controls. A thermal degradation was evident at approximately 200 8C for the PDLLA-PEG-PEI NPs, whereas for the two non-functionalized materials, the thermal degradation was not observed until 300 8C. This different profile arises from the presence of PEI on the functional NPs, which caused a decrease in the initial deg-

Figure 2. Overlaid 1H NMR spectrum of PDLLA-PEG-COOH, PDLLA-PEG-PEI NPs and PEI. [D6]DMSO was used as solvent, and its residual hydrogen peak appeared at 2.5 ppm, as was observed in the spectrum referent to PDLLAPEG-COOH NPs. PEI was the molecule used to functionalize PDLLA-PEGCOOH NPs and its resonance also appeared at 2.5 ppm, but it is broader than the residual solvent peak. Successful functionalization of NPs was observed with the broadening of the DMSO solvent peak, caused by the presence of PEI compared with the non-functionalized material.

Figure 3. FTIR spectrum of synthesized nanoparticles. Conjugation reaction of non-functionalized PDLLA-PEG-COOH nanoparticles with PEI is indicated by the appearance of N¢H stretching bands at approximately 3000 cm¢1.

Table 1. 1H NMR shifts of PDLLA-PEG-COOH, PDLLA-PEG-PEI, and PDLLA-PEG-OCH3 NPs.

-CH(CH3)- -C(=O)¢CH(CH3)- -CH2CH2¢O- PEI PDLLA-PEG-COOH 1.462 PDLLA-PEG-PEI 1.434 PDLLA-PEG-OCH2 1.468

5.181 5.173 5.183

Chem. Eur. J. 2015, 21, 14834 – 14842

3.507 3.507 3.508

www.chemeurj.org

DMSO

– 2.504 2.729 and 2.618 2.505 – 2.502

14836

radation temperature of the functionalized material, resulting in a combination of the profiles observed for both PDLLA-PEG-COOH NPs and PEI. The PEI units present in the PDLLA-PEG-PEI NPs cause the material to exhibit thermal degradation at a lower temperature, because PEI itself initially thermally degrades at 100 8C. Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper Degradation profile

Figure 4. Thermogravimetric analysis of thermal decomposition of non-functionalized PDLLA-PEG-COOH, non-functionalized PDLLA-PEG-OCH3 nanoparticles, functionalized PDLLA-PEG-PEI NPs, and PEI. The temperature at which the first thermal degradation of PEI functionalized NPs was observed is lower when compared to both non-functionalized NPs due to the presence of PEI in the functionalized material.

The particle size of the PDLLA-PEG-COOH and PDLLA-PEGPEI NPs were assessed by using a transmission electron microscope (TEM) operating at 100.0 kV beam (Figure 5).[36] The PDLLA-PEG-COOH NPs have spherical shape and are under

A degradation study was done by using a sample of the PDLLA-PEG-PEI NPs. The functionalized NPs were obtained as described in the method section and the resulting freeze-dried solid was kept under two different temperature environments, 25 and 35 8C, during a period of 28 days in the absence of water. Samples were collected for six time points: day 0, 1, 3, 6, 14, and 28. The thermogravimetric analysis results of the time points collected for this study indicated no appreciable degradation over the period of 28 days, indicating that the resulting powder material was stable under the two temperatures analyzed (Figure S3 in the Supporting Information). We elected to initially evaluate the biodegradation profile of the PDLLA-PEG-PEI NPs in solid form, because the materials were evaluated as freeze-dried powders in the vapor capture assays described below. Further, practically relevant applications including packed-bed filter cartridges would certainly employ freeze-dried solids in lieu of NP solutions. From the degradation analysis, we concluded that adventitious (i.e., atmospheric) concentrations of water at both room temperature or at 35 8C does not affect the structural integrity of our materials. Nevertheless, previous degradation studies by others have demonstrated that PLA-PEG based NPs are indeed biodegradable in aqueous solution.[37] Vapor assays After thorough characterization of the synthesized PDLLA-PEGPEI NPs, we set out to evaluate their ability to capture gaseous vapors comprised of aldehyde and carboxylic acid functional groups. In our standard assay, a 10 mg sample of freshly prepared PDLLA-PEG-PEI NPs was suspended on a tissue paper barrier above a 1 mL aliquot of target analyte in a GC vial (Figure 6 A), and the NPs were allowed to interact with the vapor portion of the analyte sample for 30 minutes. Headspace analysis was conducted by gas chromatography (FID detection). The GC

Figure 5. TEM images obtained for A) PDLLA-PEG-COOH NPs and B) PDLLAPEG-PEI NPs.

100 nm in diameter, confirming that the synthesized materials were indeed within the nanoscale range (Figure 5 A). After functionalization, the PDLLA-PEG-PEI NPs present similar morphology and size as shown in Figure 5 B, exhibiting an average diameter of approximately 60 nm. Elemental analysis was also obtained for PDLLA-PEG-COOH and PDLLA-PEG-PEI nanoparticles, confirming the presence of nitrogen on the surface of the functionalized nanoparticles (Figures S1 and S2 in the Supporting Information).

Chem. Eur. J. 2015, 21, 14834 – 14842

www.chemeurj.org

Figure 6. A) Cartoon diagram of the vapor capture studies. B) Cartoon diagram for the competition vapor assay showing competitive capture of the hexanal analyte over 1-nonene.

14837


Full Paper headspace concentration of the analyte was compared between samples treated with PDLLA-PEG-PEI NPs for 30 minutes and untreated control headspace samples. Data were collected in sextuplicate and evaluated for statistical significance by using a one-tailed Student’s T test.

Single-vapor assays We first investigated the capture of hexanal and hexanoic acid (Figure 7; 1 and 2). The PDLLA-PEG-PEI NPs effected a 98 % reduction (P < 0.0005) of the headspace vapors of hexanal samples compared to untreated hexanal controls (Figure 8 A). PDLLA-PEG-COOH (i.e., carboxylic acid capped) and PDLLAPEG-OCH3 (i.e., methoxy capped) NPs were evaluated as controls. We expected that these materials, presenting non-compatible surface functional groups, would fail to significantly reduce the headspace vapor of the target analytes. In the event, these control NPs exhibited only slight reduction (6 %; P < 0.05 and 9 % (statistically insignificant), respectively) in hexanal headspace vapors, possibly due to weak electrostatic adsorption phenomena. However, hexanoic acid vapors were reduced by 90 % (P < 0.0005) when exposed to the PDLLAPEG-PEI NPs (Figure 8 B). Next branched molecules, 2-methylbutyraldehyde (Figure 7, 3) and 3-methylbutanoic acid (Figure 7, 4) were investigated to assess whether steric factors within the substrate would hinder capture by the PDLLA-PEG-PEI NPs. Exposure to the PDLLAPEG-PEI NPs gave an 81 % reduction (P < 0.0005) of the 2-

Figure 8. Single vapor assays: A) Average GC peak area reduction for hexanal after treatment with PDLLA-PEG-PEI NPs, PDLLA-PEG-COOH NPs, and PDLLA-PEG-OMe NPs. B) Average GC peak area reduction for hexanoic acid after treatment with PDLLA-PEG-PEI NPs. C) Average GC peak area reduction for 2-methylbutyraldehyde after treatment with PDLLA-PEG-PEI NPs. D) Average GC peak area reduction for 3-methylbutanoic acid after treatment with PDLLA-PEG-PEI NPs. E) Average GC peak area reduction for butyraldehyde after treatment with PDLLA-PEG-PEI NPs. F) Average GC peak area reduction for butyric acid after treatment with PDLLA-PEG-PEI NPs. G) Average GC peak area reduction for formaldehyde after treatment with PDLLA-PEG-PEI NPs. H) Average GC peak area reduction for octanal after treatment with PDLLA-PEG-PEI NPs. I) Average GC peak area reduction for 1-nonene after treatment with PDLLA-PEG-PEI NPs. J) Average GC peak area reduction for 1butanol after treatment with PDLLA-PEG-PEI NPs.

Figure 7. List of compounds and corresponding vapor pressures (20 8C) used in vapor capture assays. Chem. Eur. J. 2015, 21, 14834 – 14842

www.chemeurj.org

methylbutyraldehyde (Figure 8 C) and a 76 % reduction (P < 0.005) of the 3-methylbutanoic acid (Figure 8 D). Smaller molecular weight aldehyde and carboxylic acid congeners with higher vapor pressures were used to illustrate the ability of the PDLLA-PEG-PEI NPs to capture more volatile com-

14838


Full Paper pounds with similar efficiency to the less volatile hexanal and hexanoic acid. Butyraldehyde (Figure 7, 5) has a vapor pressure of 83.1 mmHg at 20 8C, which is approximately eight times that of hexanal (10 mmHg at 20 8C). When exposed to the PDLLA-PEG-PEI NPs, an 86 % reduction (P < 0.0005) in butyraldehyde vapor was observed (Figure 8 E). Butyric acid (Figure 7, 6) was used as the smaller acid analogue with a vapor pressure of 0.43 mmHg in comparison to 0.18 mmHg at 20 8C for hexanoic acid. Treatment with the functionalized nanoparticles afforded an 88 % reduction (P < 0.0005) of the butyric acid vapor (Figure 8 F). Formaldehyde (Figure 7, 7) vapors were also consumed at 69 % (P < 0.0005) as shown in Figure 8 G highlighting our material’s ability to sequester extremely volatile small molecules. In contrast to the more volatile aldehydes tested, octanal (2 mmHg at 20 8C) (Figure 7, 8) was investigated to probe the capability for our materials to capture a less volatile aldehyde contaminant. When exposed to the PDLLA-PEG-PEI NPs, the octanal vapor concentration was reduced by 84 % (P < 0.0005; Figure 8 H). The final two single vapor assays sought a proof of concept for the chemoselectivity of our PDLLA-PEG-PEI NPs. We challenged our PDLLA-PEG-PEI NPs with 1-nonene (Figure 7, 8), a linear 9-carbon molecule bearing an alkene functional group. We surmised that our amine-functionalized PDLLA-PEG-PEI NPs would fail to capture 1-nonene to an appreciable extent, owing to the lack of compatible reactivity between the amine functionality on the NPs and the alkene functional group on the target analyte. Figure 8 I shows an overall retention of the nonene vapor after exposure to the PDLLA-PEG-PEI NPs with only a non-statistically significant reduction of 14 %, presumably due to non-reactive adsorption mediated by electrostatic interactions of the target analyte with the surface of the PDLLA-PEG-PEI NPs. Further, the concept was extended to a more polar substrate, 1-butanol, which returned a statistically insignificant 5 % reduction after exposure to our material. These results are important for two reasons. First, it lends credence to our proposed mechanisms for the capture of the targeted aldehyde and carboxylic acid analytes. Secondly, it demonstrates that our PDLLA-PEG-PEI NPs are able to avoid offtarget binding. Competition assays To further illustrate the chemoselectivity of our PDLLA-PEG-PEI NPs, a competition assay was conducted in which both hexanal and 1-nonene were introduced to the reaction chamber (Figure 6 B), and given 30 minutes to vaporize and react with the PDLLA-PEG-PEI NPs. In this experiment, we expected to see preferential binding of the aldehyde, hexanal, through our predicted reactivity. Further, we expected that 1-nonene, containing an incompatible alkene functional group, would fail to react with the PDLLA-PEG-PEI NPs and thus would not be captured. Analysis revealed a 75 % reduction (P < 0.0005) of the hexanal vapor concentration along with an approximately 1.4X increase (P < 0.05) in the gas-phase portion of 1-nonene present after treatment (Figure 9 A). These phenomena arise from Chem. Eur. J. 2015, 21, 14834 – 14842

www.chemeurj.org

Figure 9. Competition assays: A) Average GC peak area reduction for hexanal and 1-nonene after treatment with PDLLA-PEG-PEI NPs. B) Average GC peak area reduction for hexanoic acid and 1-nonene after treatment with PDLLA-PEG-PEI NPs. C) Average GC peak area reduction for hexanal and hexanoic acid after treatment with PDLLA-PEG-PEI NPs. D) Average GC peak area reduction for hexanal and octanal after treatment with PDLLA-PEG-PEI NPs.

the selective reduction of the hexanal vapor in the sample chamber by selective adsorption onto the PDLLA-PEG-PEI NPs. Re-equilibration of the closed system results in a larger vapor concentration of 1-nonene after hexanal capture, accounting for the enhanced 1-nonene signal after NP treatment. Next, we probed the reactivity of hexanoic acid in a competitive system

14839


Full Paper with 1-nonene. The PDLLA-PEG-PEI NPs afforded a 71 % reduction (P < 0.0005) of the hexanoic acid with a statistically insignificant 10 % reduction of the 1-nonene (Figure 9 B). This result is particularly compelling given that 1-nonene is approximately 33 times more volatile than hexanoic acid (cf. vapor pressures in Figure 7). Next, hexanal and hexanoic acid were treated simultaneously to demonstrate the concurrent capture of aldehyde and carboxylic acid analytes. Treatment with PDLLA-PEG-PEI NPs effected a simultaneous 90 % (P < 0.0005) and 69 % (P < 0.0005) reduction of headspace vapors for hexanal and hexanoic acid, respectively (Figure 9 C). The comparatively inferior capture of the hexanoic acid is likely due to the lower vapor pressure of hexanoic acid (0.18 mmHg at 20 8C) compared to hexanal (10 mmHg at 20 8C). Additionally, hexanal and octanal vapors were exposed concurrently, and the observed reductions (hexanal 87 %, P < 0.0005; octanal 52 %, P < 0.0005) for the two vapors followed similar trends to previous competition results: the less volatile octanal had a lower percent reduction compared to the more volatile hexanal (Figure 9 D).

Aldehyde capture mediated by imine formation Finally, 1H NMR spectroscopy studies were conducted to probe the mechanism of aldehyde capture. Specifically, we wanted to confirm the formation of the putative imine bond to further rule out any non-specific adsorption of the target analyte by electrostatic interactions. In this experiment, we treated PDLLA-PEG-PEI NPs with a spectroscopically simple aldehyde analyte, pivaldehyde. A partial 1H NMR spectrum resulting from the interaction between the PDLLA-PEG-PEI NPs and pivaldehyde is shown in Figure 11. The diagnostic appearance of

Charcoal control study Next, we sought to demonstrate an advantage of our PDLLAPEG-PEI NPs over conventional sorbents, such as activated charcoal. To this end, we conducted a control assay by using activated charcoal (20–40 mesh particle size) to assess the capability of this material toward the capture of hexanal and hexanoic acid gas-phase analytes. Activated carbon is a routinely used sorbent for various applications including flue gas treatment processes.[38] When hexanal was treated with the activated carbon, a 24 % reduction (P > 0.0005) of the vapor was observed (Figure 10 A; presumably mediated by non-specific electrostatic interactions), whereas the PDLLA-PEG-PEI NPs afforded a superior 98 % (P < 0.0005) reduction (Figure 8 A). As can be seen from Figure 10 B, our studies show a statistically insignificant 5 % reduction of hexanoic acid when exposed for the same period of time to the activated charcoal in comparison to the 90 % reduction (P < 0.0005; Figure 8 B) after exposure to our PDLLA-PEG-PEI NPs.

Figure 11. A) Scheme of pivaldehyde reacting with the PDLLA-PEG-PEI NPs resulting in an imine bond with an imine methine proton resonance of 7.5 ppm. B) 1H NMR evidence for imine bond formation indicating capture of aldehyde functionality with the PDLLA-PEG-PEI NPs.

a new singlet at d = 7.5 ppm suggests the presence of an imine proton within the aldehyde-treated PDLLA-PEG-PEI NP sample.[39] Any contribution of the possible hemi-aminal tetrahedral intermediate was ruled out by D2O treatment of the NMR sample, which failed to induce loss of the new singlet at 7.5 ppm by means of proton/deuteron exchange. Further, this NMR experiment offered an opportunity for us to probe the adsorption capacity of our PDLLA-PEG-PEI NPs. We prepared a standard curve relating the integration of the aldehyde methine proton of pivaldehyde to concentration of the analyte (see the Supporting Information). Using this relationship, the capacity of the material for the adsorption of pivaldehyde under saturated vapor conditions was calculated to be 0.21 mmol g¢1.

Conclusion

Figure 10. Control study: A) Average GC peak area reduction for hexanal treated with activated charcoal. B) Average GC peak area reduction for hexanoic acid treated with activated charcoal. Chem. Eur. J. 2015, 21, 14834 – 14842

www.chemeurj.org

We have presented the preparation, characterization, and evaluation of PDLLA-PEG-PEI NPs capable of selectively capturing environmental contaminants of broad concern bearing aldehyde and carboxylic acid functional groups in the gas phase. Our material showed reduction of aldehyde and carboxylic

14840


Full Paper acid vapors greater than 69 and 76 %, respectively, with reductions of up to 98 % in some cases. Further, we demonstrated that our NPs were capable of effecting the simultaneous capture of mixtures of aldehydes and carboxylic acids, as well as mixtures of two different aldehydes. Additionally, our NPs were capable of selectively capturing target aldehyde and carboxylic acid contaminants even when challenged by comparably or more volatile non-targeted vapors. Finally, the NPs disclosed herein significantly outperform conventional sorbents, such as activated carbon in the absorption target analytes. The significant advantage of our strategy over current methods arises from the ability to tailor the surface functionality of the nanomaterials for a specific target analyte from vapor mixtures. Future efforts will focus on the evaluation of further generations of these promising NPs for the remediation of other environmental contaminants of broad concern by taking advantage of the uniquely modular nature of our functional nanomaterials.

remove residual catalyst and PEG. The organic phase was collected and dried over Na2SO4, concentrated by rotary evaporation to remove solvent, and precipitated in cold methanol. The same ringopening procedure was used to synthesize PDLLA-PEG-OCH3, using d,l-lactide as the monomer, toluene as the solvent, tin(II) 2-ethylhexanoate as the catalyst, but using HO-PEG-OCH3 as the initiator. NMR (Figures S4 and S5 in the Supporting Information), FTIR, and TGA analyses were performed for synthesized materials.

Nanoparticles synthesis and characterization The conjugation reaction was carried out after nanoprecipitation of PDLLA-PEG-COOH polymeric NPs using poly(ethylenimine) solution and N-(3-dimethylaminopropyl)-N-ethylcarbodiimide hydrochloride (EDC) to obtain PDLLA-PEG-PEI NPs. Two different types of non-functionalized NPs were obtained from a PDLLA-PEG block copolymer; PDLLA-PEG-COOH NPs were synthesized by using PDLLAPEG-COOH co-polymer and PDLLA-PEG-OCH3 NPs were synthesized by using PDLLA-PEG-OCH3 co-polymer. NMR, FTIR, and TGA analyses were performed for synthesized materials. Note: the PDLLAPEG-PEI nanoparticles were not amenable to GPC analysis.

Experimental Section Synthesis materials

Gas-chromatography materials

d,l-lactide (PURASORB DL) was supplied by Purac Biomaterials. Tin(II) 2-ethylhexanoate, sodium sulfate, anhydrous toluene, methanol, chloroform, poly(ethylenimine) solution (PEI, average MW 1300 Da by LS, 50 wt. % in H2O) and N-(3-dimethylaminopropyl)N-ethylcarbodiimide hydrochloride were supplied by Sigma–Aldrich. Acetonitrile was supplied by Fisher Scientific. Hydroxyl polyethylene glycol carboxyl (HO-PEG-COOH, MW = 3500 Da) and methoxypolyethylene glycol (HO-PEG-OCH3, MW = 5000 Da) were supplied by JenKem Technology. NMR spectra were recorded on a Bruker Avance 300 MHz NMR. Chemical shifts are reported in parts per million (ppm), and spectra are referenced to residual solvent peaks. TGA was performed on a TA Instruments Hi-Res TGA 2950 thermogravimetric analyzer under nitrogen from 25 to 600 8C at 20 8C min¢1. FTIR analysis was performed with a Nicolet Magna 550 with NicPlan FT-IR Microscope and Mapping Stage. Mw, Mn, and polydispersity (PDI) of PDLLA-PEG-COOH and PDLLA-PEG-OCH3 co-polymers were obtained by using gel permeation chromatography technique on a Waters 1525 Binary HPLC/GPC system equipped with autosampler, column heater, and RI Detector. The experiments were conducted by using Waters Styragel HR 4E THF and Waters Styragel HR 1 THF columns, using a 100 % THF mobile phase at flow rate of 1 mL min¢1. Spectra were collected with RI detector and analyzed by using Breeze operating software. Mw, Mn, and PDI data for PDLLA-PEG-COOH and PDLLA-PEG-OCH3 are given in the Supporting Information.

Solvents, reagents, starting materials, and product GC standards were purchased from commercial sources and used without purification. Gas chromatography (GC) analyses were conducted by using a Shimadzu GC-2014 gas chromatograph equipped with a Shimadzu AOC-20i auto injector and a flame ionization detector (FID). The GC was equipped with a 30 m Õ 0.25 mm Õ 0.25 mm Zebron ZB-WAX Plus capillary GC column. Agilent Technologies gas chromatography vials with septum screw caps (1.5 mL total volume) were used in the analysis assays.

Splitless method temperature profile for vapor assays GC analyses were carried out within the following parameters: inlet temperature 250.0 8C; splitless injection at 30.9 mL min¢1; injector sampling depth 10 mm; column flow 1.33 mL min¢1, constant pressure; carrier gas helium; FID temperature 225 8C; temperature program 40 8C for 5 min, 50 8C min¢1 ramp to 200 8C, hold for 5 min.

Methodology for vapor-assay analysis by gas chromatography General GC procedure for vapor assays including 1) standard vapor areas for each substrate followed by/and 2) functionalized nanoparticle formulation reactivity with each individual vapor substrate.

Polymer synthesis and characterization Oven vacuum was used to dry reagents at 28 mmHg pressure. PDLLA-PEG-COOH co-polymer was synthesized by ring opening polymerization using hydroxyl polyethylene glycol carboxyl as the initiator and tin(II) 2-ethylhexanoate as the catalyst. Briefly, HOPEG-COOH, d,l-lactide, and Na2SO4 were vacuum dried in a roundbottom 3-neck flask overnight before use. HO-PEG-COOH and d,llactide were dissolved by stirring in 120 8C anhydrous toluene under N2 gas. After 15 min, tin(II) 2-ethylhexanoate was added to the reaction vessel, purged with N2, and heated at reflux for 12 h. Next, the polymer product was extracted in chloroform/water to Chem. Eur. J. 2015, 21, 14834 – 14842

www.chemeurj.org

General procedure for standard vapor-area assay by GC analysis The opening of a 1.5 mL GC vial was covered with a 5 Õ 5 cm piece of Kimwipe tissue paper. Using a glass stir rod, a small sample well was made with the Kimwipe by gently applying pressure with the tip of the glass stir rod. A vial cap was secured on the vial and a 1 mL injection of the volatile liquid substrate was introduced into the vial. After a 30 minute vaporization equilibrium time, the vial was subjected to GC analysis as described above.

14841


Full Paper General procedure for functionalized nanoparticle assays by GC analysis By using the previously described process for formation of a well within the GC vial, 10 mg of the functionalized nanoparticle was added into the Kimwipe sample well and then secured with a vial cap. A 1 mL injection of the designated volatile substrate was introduced into the vial and allowed to vaporize and subsequently react with the solid nanoparticles for 30 min. Upon completion of the 30 minute reaction time, the vial was subjected to GC analysis.

Acknowledgements The authors gratefully acknowledge the Fats & Proteins Research Foundation, Inc., the Poultry Protein and Fat Council, and the Clemson University Animal Co-Products Research and Education Centerfor generous support. The authors also thank the CAPES Foundationand the Ministry of Education of Brazilfor a graduate fellowship for F.G. The authors acknowledge that the GPC was performed on facilities supported by a grant from NIGMS of the National Institutes of Health (award number 5P20M103444-07). Finally, the authors thank the Clemson Electron Microscope Facility and Ms. Kim Ivey for technical assistance pertaining to nanoparticle characterization. Keywords: gaseous pollutants · nanoparticles poly(ethyleneimine) · poly(lactic acid) · polymers

·

[1] J. Kesselmeier, M. Staudt, J. Atmos. Chem. 1999, 33, 23 – 88. [2] R. Atkinson, J. Arey, Chem. Rev. 2003, 103, 4605 – 4638. [3] A. Vaseashta, M. Vaclavikova, S. Vaseashta, G. Gallios, P. Roy, O. Pummakarnchana, Sci. Technol. Adv. Mater. 2007, 8, 47 – 59. [4] M. Possanzini, V. Di Palo, A. Cecinato, Atmos. Environ. 2002, 36, 3195 – 3201. [5] K. Kawamura, I. R. Kaplan, Environ. Sci. Technol. 1987, 21, 105 – 110. [6] K. Kawamura, L. L. Ng, I. R. Kaplan, Environ. Sci. Technol. 1985, 19, 1082 – 1086. [7] Y. Yokouchi, Y. Ambe, Atmos. Environ. 1986, 20, 1727 – 1734. [8] J. Zhang, K. R. Smith, Environ. Sci. Technol. 1999, 33, 2311 – 2320. [9] F. Lipari, J. M. Dasch, W. F. Scruggs, Environ. Sci. Technol. 1984, 18, 326 – 330. [10] J. J. Schauer, M. J. Kleeman, G. R. Cass, B. R. Simoneit, Environ. Sci. Technol. 2001, 35, 1716 – 1728. [11] F. RanciÀre, C. Dassonville, C. Roda, A. Laurent, Y. Le Moullec, Momas, Sci. Total Environ. 2011, 409, 4480 – 4483. [12] L. M. Marnett, Health Effects of Aldehydes and Alcohols in Mobile Source Emissions; Health Effects Institute; National Academy Press: Washington, DC, 1988, 579 – 603.

Chem. Eur. J. 2015, 21, 14834 – 14842

www.chemeurj.org

[13] Y. Hoshika, Anal. Chem. 1982, 54, 2433 – 2437. [14] The Clean Air Acts Amendments of 1990 List of Hazardous Air Pollutants, US Environmental Protection Agency, http://www.epa.gov/ ttnatw01/orig189.html (accessed Apr 27, 2015). [15] Priority Pollutants, US Environmental Protection Agency, http://water.epa.gov/scitech/methods/cwa/pollutants.cfm (accessed Apr 27, 2015). [16] A. P. Altshuller, Atmos. Environ. Part A 1993, 27, 21 – 32. [17] Council. National Research Formaldehyde and Other Aldehydes. Board on Toxicology and Environmental Health Hazards. National Academy Press: Washington, DC, 1981. [18] C. C. Yao, Miller, G. C. U. S. Department of Health, Education, and Welfare(NIOSH), Cincinnati, OH, 1979, Publication No. 79 – 118. [19] M. M. Khin, A. S. Nair, V. J. Babu, R. Murugan, S. Ramakrishna, Energy Environ. Sci. 2012, 5, 8075 – 8109. [20] G. P. Lithoxoos, A. Labropoulos, L. D. Peristeras, N. Kanellopoulos, J. Samios, I. G. Economou, J. Supercrit. Fluids 2010, 55, 510 – 523. [21] W. Chen, L. Duan, D. Zhu, Environ. Sci. Technol. 2007, 41, 8295 – 8300. [22] A. Di Paola, E. Garca-Lûpez, G. Marc, L. Palmisano, J. Hazard. Mater. 2012, 211, 3 – 29. [23] P. V. Kamat, D. Meisel, C. R. Chim. 2003, 6, 999 – 1007. [24] G. Shan, S. Yan, R. D. Tyagi, R. Y. Surampalli, T. C. Zhang, Pract. Period. Hazard. Toxic Radioact. Waste Manage. 2009, 13, 110 – 119. [25] X. Ren, C. Chen, M. Nagatsu, X. Wang, Chem. Eng. J. 2011, 170, 395 – 410. [26] X. Zhao, L. Lv, B. Pan, W. Zhang, S. Zhang, Q. Zhang, Chem. Eng. J. 2011, 170, 381 – 394. [27] A. Nomura, C. W. Jones, ACS Appl. Mater. Interfaces 2013, 5, 5569 – 5577. [28] A. Nomura, C. W. Jones, Chem. Eur. J. 2014, 20, 6381 – 6390. [29] G. Qi, Y. Wang, L. Estevez, X. Duan, N. Anako, A. Li, W. A. Park, C. W. Jones, E. P. Giannelis, Energy Environ. Sci. 2011, 4, 444 – 452. [30] M. A. Alkhabbaz, P. Bollini, G. S. Foo, C. Sievers, C. W. Jones, J. Am. Chem. Soc. 2014, 136, 13170 – 13173. [31] Y. Kuwahara, D. Kang, J. R. Copeland, N. A. Brunelli, S. A. Didas, P. Bollini, C. Sievers, T. Kamegawa, H. Yamashita, C. W. Jones, J. Am. Chem. Soc. 2012, 134, 10757 – 10760. [32] S. Choi, J. H. Drese, C. W. Jones, ChemSusChem 2009, 2, 796 – 854. [33] J. C. Hicks, J. H. Drese, D. J. Fauth, M. L. Gray, G. Qi, C. W. Jones, J. Am. Chem. Soc. 2008, 130, 2902 – 2903. [34] S. Choi, J. H. Drese, P. M. Eisenberger, C. W. Jones, Environ. Sci. Technol. 2011, 45, 2420 – 2427. [35] P. Bollini, S. Choi, J. H. Drese, C. W. Jones, Energy Fuels 2011, 25, 2416 – 2425. [36] Note: The dark sections of TEM image B are an artifact of the uranyl staining process. [37] E. Pridgen, F. Alexis, O. C. Farokhzad, R. Langer, Methods in Bioengineering Book Series Nanoscale Bioengineering and Nanomedicine, 1st edition, 2009, Artech House, Boston, MA. [38] P. S. Yi, W. J. Hong, M. L. Ping, Adv. Mater. Res. 2014, 1049 – 1050. [39] G. J. Karabatsos, S. S. Lande, Tetrahedron 1968, 24, 3907 – 3922.

Received: May 22, 2015 Published online on September 1, 2015

14842
