The authors have declared that no competing interests exist.
Time-of-flight secondary ion mass spectrometry (ToF-SIMS) analysis was used to qualitatively and quantitatively assess the distribution of permethrin insecticide on the surfaces and interiors of Olyset long-lasting insecticidal net (LLIN) fibers. Total insecticide content in LLINs has been established using many analytical methods. However, it is important to quantify the bioavailable portion residing on the fiber surfaces for incorporated LLINs. ToF-SIMS is a very surface sensitive technique and can directly image the spatial distribution of permethrin insecticide on the surface of Olyset fibers. Surface permethrin appeared as patchy deposits which were easily removed by acetone and reappeared after several days as interior permethrin migrated (bloomed) from the fiber interior. After a wash/incubation cycle, permethrin deposits were more diffuse and less concentrated than those on the as-received fibers. ToF-SIMS is particularly sensitive to detect the Cl- ion, which is the characteristic ion of permethrin. Ion implantation and quantification of dopants using SIMS is well established in the semiconductor industry. In this study, quantitative depth profiling was carried out using 35Cl- ion implantation to correlate secondary ion yield with permethrin concentration, yielding a limit of detection of 0.051 wt% for permethrin. In some cases, surface concentration differed greatly from the fiber interior (>1 μm below the surface). Two- and three-dimensional mapping of Cl at sub-micrometer resolution showed permethrin to be dissolved throughout the fiber, with about 2 vol% residing in disperse, high-concentration domains. This suggests that these fibers fall into the class of monolithic sustained-release devices. It is expected that ToF-SIMS can be a valuable tool to provide insight into the insecticide release behavior of other LLIN products, both current and future.
Insecticide-treated mosquito netting has been a mainstay of malaria control programs worldwide for more than 15 years. From 2004 to 2014, more than one billion long-lasting insecticidal nets (LLINs) in the form of bednets have been delivered into malaria-endemic regions [
Two approaches are used to manufacture LLINs. One is to topically apply a durable insecticide formulation onto netting made with multifilament polyester yarn. The alternative is to mix insecticide into a polymer, either high-density polyethylene (HDPE) or polypropylene (PP), melt-spin the polymer into fibers and then convert these fibers into netting. This latter approach, insecticide “incorporation,” is intended to produce nets having insecticide present throughout the bulk of the fiber. In principle, the fiber acts as a sustained-release device, presenting a portion of the insecticide on its surface where it can interact with an alighting mosquito. If the insecticide is stripped from the fiber surface by e.g., washing or abrasion, it is replenished by spontaneous migration of additional insecticide from the interior [
Analytical methods for measuring the total insecticide content in LLINs have been established [
A suitable chemical assay for insecticide on fiber surfaces has been elusive. Tami
Azondekon
Lacking in these investigations has been a method to directly observe and quantify insecticide on the surfaces of LLIN fibers. Such capability would serve to confirm that a given assay procedure for removing surface insecticide (e.g., solvent washing or controlled rubbing) achieves complete removal without disturbing the subsurface concentration. It would also help characterize the rate and extent of insecticide migration from the subsurface to surface during regeneration. Furthermore, it could lead to a more thorough understanding of the dynamics of incorporated-insecticide nets, assisting in the development of improved products in the future.
The present study examines the use of time-of-flight secondary ion mass spectrometry (ToF-SIMS) to probe insecticide on the surfaces and interiors of LLIN fibers. During ToF-SIMS analysis, a sharply focused ion beam is used to sputter molecules from the top 1–2 monolayers of a sample, creating (secondary) ions that are then directed into a mass spectrometer for chemical species identification and quantification. The incident ion beam is typically focused to a diameter of 0.3 μm, and the total area investigated is correspondingly small, typically 50 μm × 50 μm. If the beam is moved laterally stepwise, the specimen can be scanned to create a two-dimensional map of chemical species at sub-micrometer resolution. This analysis can be carried out on a fiber surface or on the cross-section of a cut fiber. Furthermore, a second ion beam can be used to sputter off successive layers (1–2 nm each) of the sample, allowing direct analysis of the underlying material, leading to a three-dimensional depth profile.
Insecticide-incorporated LLIN fibers were expected to be compatible with ToF-SIMS analysis for several reasons. First, the area of spot analysis can be as small as 300 nm, which is up to 700 times smaller than the 120–200 μm diameter of an individual fiber, thus making two-dimensional surface chemical mapping and depth profiling conceivable. Second, the insecticides of interest are halogenated, and would be expected to generate high yields of either chlorine or bromine ions for detection. Furthermore, the insecticide concentration in fresh LLINs of 0.2–2 wt%, depending on the insecticide, is well within the detectable range for halogenated compounds by ToF-SIMS. On the other hand, ToF-SIMS analysis of fiber surfaces presents some challenges. The rounded fiber surfaces cause spectral peak broadening, reducing mass resolution, due to differential travel times of secondary ions from the sample to the detector. In addition, the extremely small region of analysis requires that each specimen must be analyzed at multiple locations in order to ensure representative sampling.
This report focuses on only one commercial LLIN product, Olyset Net (Sumitomo Chemical Co., Ltd., Tokyo, Japan). This product is a permethrin-incorporated HDPE monofilament net and was the first LLIN to be fully recommended by WHOPES [
Image obtained at 5 kV after sputter coating the specimen with gold palladium.
The strong peaks at m/z 35 and 37 correspond to the two naturally occurring stable isotopes of chlorine. The peak at m/z 37 may also include a contribution from C3H- ion. Note that the ratio of 35Cl- and 37Cl- appears to be higher than their naturally abundant 3:1 ratio in the spectrum. For pristine permethrin, this is due to saturated 35Cl- and 37Cl- ion counts to the detector. On pure Permethrin and as received Olyset fiber, 35Cl saturation is observed. However, this signal saturation does not occur for washed and incubated samples because the permethrin concentration is significantly lower. For the Olyset fiber surface, the contribution of C3H- at m/z 37 from HDPE is significant, but the instrument has sufficient mass resolution to separate 37Cl- and C3H-. Separation of 37Cl- and C3H- cannot be shown on this plot, they are too close in mass. This plot is intended to show a wider mass range. However, if we zoom in at the m/z 37 region, we clearly see two individual peaks. The C4H3- ion at
ToF-SIMS has been used previously to observe the distribution of low molecular weight additives in polymer matrices [
For this study, a quantification method was developed [
Olyset net samples were obtained from Sumitomo Chemical Co. Ltd. (Tokyo, Japan).
ToF-SIMS analysis was performed using a TOF.SIMS 5 (IONTOF, Münster, Germany) instrument. A 25 keV, 0.4 pA Bi3+ analytical beam was used for imaging and quantitative analysis, while a 3 keV, 20 nA Cs+ beam was used to sputter the sample for depth profile analysis. A 45° angle of incidence was used for both beams. For depth profiling, a 120 μm x 120 μm crater was sputtered and analysis was carried out on a 50 μm x 50 μm area at the center of the crater. Typical mass resolution for the analyses was 4000 at 29
Calibration for quantitative analysis was carried out using a 35Cl- ion implantation method described previously [
Quantification of these results was challenging because the rounded and somewhat rough surface topography of the fibers caused secondary ions of a given
To test the effects of sample pressing and sputtering on the secondary ion emissions, measurements were made on 20 random locations on an Olyset net specimen as received (i.e., unpressed and unsputtered), and on a specimen obtained from pressed adjacent area. Afterwards, both specimens were sputtered and measurements were repeated. The results (
Each plot represents readings from 20 random locations on a single sample specimen. Analysis area of each location is 500 μm × 500 μm.
Unpressed | Unpressed, sputtered | Pressed | Pressed, sputtered | |
---|---|---|---|---|
Number of locations analyzed (N) | 20 | 20 | 20 | 20 |
Median | 4.53 | 4.34 | 4.07 | 3.13 |
Mean | 4.66 | 4.63 | 3.89 | 3.55 |
Standard Deviation | 2.22 | 1.28 | 1.66 | 1.94 |
Relative Standard Deviation (%) | 47.6 | 27.7 | 42.8 | 54.7 |
95% Confidence Interval for Mean | 3.69–5.63 | 4.07–5.20 | 3.16–4.62 | 2.70–4.40 |
Chlorine regions are shown in green. Analysis area of each location is 400 μm × 400 μm.
Results for washed and regenerated samples are shown for four samples in
Initial readings were carried out after samples were pressed and sputtered. Analysis area of each location is 500 μm × 500 μm.
Sample | N | Median | Mean | Standard Deviation | Relative Standard Deviation (%) | 95% Confidence Interval for the Mean | % Change in Mean vs Initial | |
---|---|---|---|---|---|---|---|---|
A | Initial | 20 | 2.95 | 3.25 | 1.06 | 32.7 | 2.79–3.72 | - |
Washed | 20 | 1.80 | 1.80 | 0.53 | 29.4 | 1.57–2.04 | -44.6 | |
4d incubated | 20 | 2.47 | 2.97 | 1.26 | 42.5 | 2.42–3.52 | -8.8 | |
10d incubated | 20 | 2.88 | 2.91 | 0.92 | 31.7 | 2.50–3.31 | -10.7 | |
B | Initial | 20 | 5.29 | 5.36 | 1.59 | 29.6 | 4.66–6.06 | - |
Washed | 20 | 0.97 | 0.89 | 0.30 | 33.5 | 0.76–1.03 | -88.3 | |
4d incubated | 20 | 3.96 | 3.75 | 1.70 | 45.3 | 3.00–4.49 | -30.1 | |
10d incubated | 20 | 3.14 | 3.56 | 1.53 | 43.0 | 2.89–4.23 | -33.6 | |
C | Initial | 20 | 4.29 | 5.00 | 1.76 | 35.3 | 4.22–5.77 | - |
Washed | 10 | 1.50 | 1.61 | 0.38 | 23.8 | 1.37–1.85 | -67.7 | |
4d incubated | 20 | 4.09 | 4.17 | 1.70 | 40.9 | 3.42–4.91 | -16.6 | |
10d incubated | 20 | 3.86 | 3.92 | 0.71 | 18.1 | 3.61–4.23 | -21.5 | |
D | Initial | 20 | 4.40 | 5.25 | 2.23 | 42.4 | 4.27–6.22 | - |
Washed | 20 | 1.27 | 1.32 | 0.34 | 25.9 | 1.17–1.47 | -74.8 | |
4d incubated | 20 | 2.97 | 3.19 | 1.01 | 31.8 | 2.75–3.64 | -39.2 | |
10d incubated | 20 | 3.72 | 3.90 | 0.81 | 20.7 | 3.54–4.25 | -25.8 |
Incubating the washed samples at 30°C resulted in a substantial recovery of detectable 35Cl on the fibers, consistent with migration of permethrin from the subsurface to the surface. In all cases, recovery was partial with the mean 35Cl levels being -8.8% to -39.2% of the initial levels after 4 days incubation and -10.7% and -33.6% after 10 days. It is apparent from these data that the migration of permethrin to the fiber surfaces was essentially completed within 4 days at this temperature.
The results above show the relative levels of permethrin on the surfaces of Olyset fibers and the micrometer-scale lateral variability in these levels.
Depth profiles of unimplanted Olyset fibers are shown in
The fiber was sputtered with 20 nA 3keV Cs+ over 120 μm × 120 μm and the secondary ions were analyzed with 0.3 pA 25 keV Bi3+ after each cycle of sputtering over 50 μm × 50 μm from the center of crater area. Permethrin (PM) concentrations shown for the depleted nets were obtained using conventional gas chromatography.
If there was contamination due to handling, an increase at the surface would be expected where a depletion is actually observed. It is also known that the typical signature for a surface contaminant would be a very sharp drop in concentration as the ion beam sputters through the contaminant.
Combining the lateral and depth resolution capabilities of ToF-SIMS makes it possible to generate three-dimensional concentration information of the species of interest. This is done by producing a two-dimensional map of the secondary ion emissions at successive sputtering levels. A 256 x 256 point scan carried out over a 50 μm x 50 μm area of a fiber at 13 sputtering levels (surface plus 12 subsurface levels to a maximum depth of 1.89 μm) produced 851968 spatially discrete values of 35Cl- secondary ion intensity (
With this information, it is possible to quantitatively describe the Cl concentration variance within the 50 μm x 50 μm x 1.89 μm volume of the fiber being analyzed. Recognizing that the total Cl content within this volume is
Using this expression, it is possible to calculate that 50% of the Cl in the sample resides in regions having Ix ≤ 18.1. Likewise, only 3.3% of the Cl is residing in regions having Ix ≥ 50. Therefore, despite the presence of high concentration domains in the fiber, the large volume region of low concentration carries most of the total Cl load.
Although they occupy a small volume fraction of the fiber, discrete domains having high 35Cl- secondary ion intensities were found throughout.
Data is plotted to simulate the view from the fiber surface into the direction of the core. Data points lying closest to the surface (depth < 0.5μm) are circled for clarity.
Olyset nets are designed to have sufficient permethrin insecticide on the fiber surface to kill susceptible mosquitoes on contact. If the permethrin is removed by washing, handling, etc., additional permethrin is expected to spontaneously migrate from the fiber interior to the surface, restoring insecticidal performance. ToF-SIMS images from this study confirm that permethrin is present on as-received Olyset fiber surfaces, and following its removal with acetone, permethrin reappears at the surface within 4 days at 30°C. Although the return of insecticidal performance after washing has been reported in numerous publications, [
Permethrin observed on the surfaces of as-received Olyset fibers was qualitatively and quantitatively different from that on fibers that were washed and incubated. At the dimensional scale of analysis, permethrin did not appear as a uniform coating, but instead as discrete, high concentration patches in both the as-received and the washed/incubated cases. This made it necessary to take measurements on many points (usually 20) of each specimen in order to ensure that a sampling was representative. This non-uniformity may be caused by the emergent permethrin moving laterally to form pools. For unwashed fibers, the RSDs of normalized 35Cl- secondary ion intensities ranged from 32.7% to 54.7%. Although the permethrin on washed/incubated fibers appeared more diffuse than the as-received fibers, the RSDs were still quite high (18.1–45.3%). The mean level of permethrin on the surfaces of washed/incubated fibers was essentially the same for all four net samples tested, even though the initial concentrations varied significantly. In each case, the permethrin level on the surfaces of the washed/incubated fibers was lower than the initial levels, but the percent recovery varied widely because of the variation in the initial concentrations. No significant change was seen in the mean surface concentrations between 4 and 10 days of incubation, indicating that the blooming process had reached equilibrium within 4 days.
The differences seen in permethrin on the surfaces of as-received vs. washed/incubated fibers can be explained by considering the differences in thermal and mechanical stress conditions during the blooming process for each case. During fiber manufacture, permethrin is presumably fully dissolved in the polymer melt at an elevated temperature. As this melt is extruded, cooled and stretched, the polymer partially crystallizes, forcing the permethrin to concentrate in the amorphous regions of the fiber. It is during this process that a portion of the permethrin segregates to the fiber surface. The rate and extent of segregation will depend on the temperature, cooling rate and stretching forces applied, and variations in these are probably the reason for the variations in mean surface concentrations between nets as received. Mechanical stresses during knitting and product storage temperatures may also contribute to these variations. In contrast, permethrin blooming in the washed/incubated samples took place under identical, quiescent conditions at a well-controlled temperature. In this case, the extent of blooming was essentially the same from net to net, and lower than that observed before washing. An immediate consequence of this analysis is the understanding that the initial concentration of permethrin on the surface of the net is likely to bear little relevance to concentration after the first wash/recovery cycle. For this reason, quality assessment of this and similar products should probably include insecticidal potency after the first wash/blooming cycle, instead of relying on data from pristine samples.
It was noted that after washing for 1 min in acetone, a greatly diminished, but still measurable, 35Cl secondary ion emission was consistently observed. Although it may be due to incomplete removal of permethrin, it may also be due to the initial emergence of permethrin from the interior. The time between washing and data collection ranged from about 2 to 4 hours, and the rate of diffusion to the surface is expected to be highest at the beginning, so this is not an unreasonable possibility, especially given the sensitivity of this technique to Cl detection. Additional experiments to measure the rate of permethrin migration to the surface would serve to clarify this.
The phenomenon of low molecular weight additive migration from the bulk to the surface of polymers has been studied for many years, largely because of its role in the premature degradation of polymers from antioxidant and photostabilizer loss [
Blooming has been used to advantage in the design of certain types of monolithic controlled-release devices, in which an active ingredient (a.i.) is dispersed in a polymer well above its saturation concentration. The a.i. is present in the polymer as distinct dispersed phases embedded within a continuous phase consisting of polymer saturated with the a.i. [
Both two- and three-dimensional ToF-SIMS imaging confirm the existence of permethrin-rich domains of Olyset fibers, and that these domains are widely scattered and vary greatly in size. Surprisingly, these domains collectively contain only about 3 vol% of the permethrin within the fiber, based on the secondary 35Cl- emission intensities. The remainder is diffusely distributed (dissolved) in the regions between these domains. However, since only one site on a fiber was examined, this result may not be representative of the Olyset nets in general.
Quantitative depth profile analysis also confirmed that the concentration of permethrin in the bulk fiber is not a reliable indicator of the concentration at the surface. Concentration in the fiber subsurface (ca. 1 to 5 μm from the surface) was found to be mostly uniform with respect to depth, but in some cases deviated strongly between 1 μm and the surface. The quantitative depth profiling method described in this article was capable of determining permethrin concentration to a detection limit of 0.051wt%.
ToF-SIMS analysis has been demonstrated as a useful method for imaging and quantifying permethrin insecticide on the surface and in the interior of Olyset fibers. Using the ion-implantation for calibration, quantitative depth profiling was conducted with a limit of detection of 0.051 wt% permethrin. In some cases, the concentrations of permethrin within the first micrometer of the fiber surface were found to be significantly different than in the bulk.
The presentation of permethrin on the fiber surface was found to be qualitatively and quantitatively different for as-received fibers versus fibers that have been acetone-washed and incubated. In both cases, permethrin deposits were patchy, but somewhat more diffuse in appearance after washing/incubation. Permethrin levels on washed/incubated samples were consistently lower than as-received samples. These observations are likely due to the different set of conditions under which permethrin blooming occurred during melt spinning versus quiescent incubation. For this reason, it is recommended that the quality of insecticidal performance be assessed not only for freshly made nets, but also samples that have been washed and incubated, since surface permethrin levels may be significantly different in the two cases.
Two- and three-dimensional imaging suggest that permethrin is dissolved throughout the fiber, but also small domains of highly concentrated permethrin are dispersed throughout. In the fiber section studied, these domains collectively occupied only about 3% of the fiber volume. It is expected, based on models of monolithic controlled-release devices, that as permethrin is lost from the fiber, these domains will decrease in size. Once these domains disappear altogether, the driving force for blooming will cease, and the insecticidal power of the fiber will be lost, even if the fiber still contains significant insecticide. This underscores the need for quality assessment methods beyond the measurement of total insecticide in the net.
Although the present study focused on the Olyset permethrin/HDPE system, it is expected that other insecticide-incorporated LLINs can be evaluated similarly. Furthermore, as mosquitoes in malaria-endemic areas develop resistance to pyrethroid insecticides,[
The authors gratefully acknowledge the following persons for their assistance in the course of this work: Charles B. Mooney (North Carolina State University Analytical Instrumentation Facility) for the SEM micrograph used in
This work was performed in part at the Analytical Instrumentation Facility (AIF) at North Carolina State University, which is supported by the State of North Carolina and the National Science Foundation (award number ECCS-1542015). The AIF is a member of the North Carolina Research Triangle Nanotechnology Network (RTNN), a site in the National Nanotechnology Coordinated Infrastructure (NNCI).