Ultra-rapid and highly efficient enrichment of organic pollutants via magnetic mesoporous nanosponge for ultrasensitive nanosensors

Currently, owing to the single-molecule-level sensitivity and highly informative spectroscopic characteristics, surface-enhanced Raman scattering (SERS) is regarded as the most direct and effective detection technique. However, SERS still faces several challenges in its practical applications, such as the complex matrix interferences, and low sensitivity to the molecules of intrinsic small cross-sections or weak affinity to the surface of metals. Here, we show an enrichment-typed sensing strategy with both excellent selectivity and ultrahigh detection sensitivity based on a powerful porous composite material, called mesoporous nanosponge. The nanosponge consists of porous β-cyclodextrin polymers immobilized with magnetic NPs, demonstrating remarkable capability of effective and fast removal of organic micropollutants, e.g., ~90% removal efficiency within ~1 min, and an enrichment factor up to ~103. By means of this current enrichment strategy, the limit of detection for typical organic pollutants can be significantly improved by 2~3 orders of magnitude. Consequently, the current enrichment strategy is proved to be applicable in a variety of fields for portable and fast detection, such as Raman and fluorescent sensing.


Reviewer #1 (Remarks to the Author):
The manuscript titled 'Ultra-rapid and highly efficient enrichment of organic pollutants via magnetic nanoparticles/mesoporous nanosponge compounds for ultrasensitive nanosensors' written by Zhang et al presented a strategy for enrichment sensing of organic pollutants based on a powerful porous composite material, consisting of magnetic nanoparticles immobilized porous β-cyclodextrin polymers. The results showed that the ~90% removal efficiency can reach within ~1 min and the polymer adsorbent can be easily recycled from water and re-dispersed in ethanol so that the target molecules in the cavity of adsorbent is concentrated, with an enrichment factor up to ~103. The manuscript is carefully written and the logic is clear, so I think it belongs to Nature Communication with the following issues being addressed:

In the introduction section, the author introduced the background of surface-enhanced Raman scattering (SERS) and its limitations in detecting organic pollutants. I think this work has mainly reported a sample preparation method and doesn't face to resolve the scientific problems about SERS, so it is better to add some current sample preparation methods for organic pollutants detection based on magnetic nanoparticle and reduce some description about SERS limitations.
Response: Thanks for the suggestion from the reviewer. With the enlightenment from such questions, we have rewrote the manuscript. In fact, it seems surprising that after fifty decades, SERS has not yet been widely used in practical applications. This is owing to the fact that, besides the stability of SERS substrates and reproducibility of spot-to-spot, SERS still faces two major bottlenecks in commercial market. The first is the low detection sensitivity to the molecules of intrinsic small cross-sections or weak affinity to metal surface. The second is the interference from the complex matrices in real-sample detection. With the revised version, we reconsidered the work and highlighted the capability of current detection strategy in selective adsorption to interested molecules, particularly from the complex matrix in the real-sample environment. This point is the key scientific problems about SERS. Certainly, the current enrichment strategy can be used in a wide detection protocols such as SERS, fluorescence, UV-vis, even Mass Spectrometry, Chromatography, and others.
Thus, in the revised version, we added some experiments and discussions about the SERS detection in real-sample environments, inspired by the comments from reviewers. The correlated 2 revised parts in the Introduction section 2 and 3 has been updated. Figure 2d, the FT-IR spectrum of the MN-PCDP displayed a new peak at 1265 cm - Response: We appreciate the suggestion from the reviewer. First, we are regretful for a mistake about the description of FT-IR peak at 1265 cm -1 . This peak in relation to the C-F group was existing in TFT, but the signal intensity of this absorption peak at 1265 cm -1 is weaker than that in TFT owing to the partial replacement of F, implying that the β-CD has been crosslinked with TFT. Second, as a comparison, in the revised version, we chose anther cross-linking agent (epichlorohydrin, EPI), which is the most extensively studied β-CD polymer for water purification, to combine the β-CD. As shown in Supplementary Fig. 6, the removal efficiency of BPA by MNEPI-CDP was much lower than MNP-CDP. The correlated revised parts of the manuscript are shown as following:

In
Page 5, Paragraph 1: The signal intensity of absorption peak at 1265 cm -1 in relation to C-F stretching vibration is weaker than that in TFT owing to the partial replacement of F, 23, 24 implying that the β-CD has been crosslinked with TFT ( Supplementary Fig.2).
Page 6, Paragraph 1: In this work, different cross-linking agent, e.g. epichlorohydrin (EPI) is compared ( Supplementary Fig. 7). As shown in Supplementary Fig. 8, the removal efficiency of BPA by MNEPI-CDP in 1 min is 19.5%, which is much lower than MN-PCDP. (1 mg mL -1 ). Figure 3a, the time-dependent adsorptions of various organic micropollutants adsorbed by MN-PCDP usually contain aromatic model compounds, does the absorption for organic molecules which are rich in alkane chain also work? And can the author also explain the mechanism about the high removal efficiency?

In
Response: Thanks for the advice from the reviewer. We added some absorption experiments of organic molecules with alkane chain to explain the mechanism about the high removal efficiency.
The correlated revised parts of the manuscript are shown as following: Page 6, Paragraph 2: As is known, the hydroxyl groups of β-CD are located at the outer surface of the molecule, that is, primary hydroxyls at the narrow side and secondary hydroxyls at the wider side, which makes β-CD water-soluble but simultaneously generates an inner cavity that is relatively hydrophobic. 27 Because of their hydrophobic interior cavity, β-CD can either partially or entirely accommodate suitably sized lipophilic low molecular weight molecules or even polymers. 28 For example, MN-PCDP exhibits a remarkable adsorption capability and selectivity a b 4 for most aromatics and some chain compounds, as shown in Fig. 2a and Supplementary Fig. 10-11. Furthermore, by means of particular treatments such as changing pH value of solution, 29 the adsorption feature of molecules can be tuned. Thus, the MN-PCDP mesoporous nanosponge will display a wide applicability and selectivity in a variety of molecules.  Figure 13, the fluorescence spectra of BPA before and after the enrichment of MN-PCDP adsorbent seems different and there are many spikes in Figure S13a and it is very smooth in Figure S13b, does it completely caused by the counts difference or exist other reasons?

In Supplementary
The author may unify it in both spectrums.
Response: Thanks for your reminder. In the Supplementary Figure 13, the concentration of BPA in the fluorescence spectra after the enrichment of MN-PCDP adsorbent was the original concentration of BPA. In this circumstance, after the enrichment of MN-PCDP adsorbent, the "concentration" of BPA was much high than the original. Thus, the Figure S13b

Reviewer #2 (Remarks to the Author):
The manuscript is about the fabrication of a polymer based on the combination of beta-CD and

Comm.
Response: Thanks for your comments. After receiving the comments from referees, we noticed that probablly, referees provided such commens is because that we cannot demonstrate the novelty and significance of this work clearly in the introduction part. In fact, after carfully read the suggested two papers, i.e. Science of the Total Environment 728 (2020) 138789 and Journal of Chromatography A, 1503 (2017) 1-11), we would like to highlight the concept of extraction in mentioned two papers and our "enrichment" in our manuscript. In addition, with the enlightenment from referees, we have noticed that the selectivity and the performance in detection of complex matrix in real-sample environment indeed are very important. Thus, in the revised version, we have updated these data, and rewrote the manuscript. In the following, we shall explain in details.
1) Concerning the concept of extraction in mentioned two papers and our "enrichment" in our 8 manuscript. I have read the extraction related references (including the suggested papers, Science of the Total Environment 728 (2020) 138789 and Journal of Chromatography A, 1503 (2017) 1-11). Yes, they indeed used the similar porous materials. However, reviewer is interesting of "the selectivity or the performance of the materials" since extraction is to select target molecule from complex system. E.g., extraction (adsorption) of organic molecules (microcystins or PAHs) from environmental water or soil sample in suggested two papers. Selectivity is the first importance in this kind of work, i.e. selective adsorption interested molecules from multiple-molecules system.
They did not notice to improve the removal speed, and enrichment capablity, thus you may see that in above two papers, one used nearly one hour to extract and 7 min to desorb and the other paper used high concentration adsorption materials and desorbed in a large volume of desorption solvent. The extraction process if operation into a traditional solid state extraction cartridges, the separating rate is relatively slow. Normally, the speed of extraction is 0.5-5 mL/min, which can not meet our current study motivation with very fast enrichment operation.
Extraction seems to be more the work for environmental scientists, e.g. water or soil treatment.
However, in our current work, we are aiming our work to develop a new detection strategy for trace pollutants, and we paid our attention to the enrichment capability and fast operation, (in the following, we shall list the novelty and advance of this work), which is very important for the practical application in protable, fast, on site detection. Thus, our enrichment strategy is different from classical extraction process.
With the enlightenment of referees, we reconsidered and updated the significance and novelty of this work, and we want to emphasize here: we explored a very effective enrichment and detection strategy with remarkable performance including 1) ultra-fast removal speed within ten second to ~1 min, 2) high removal efficiency up to ~90%, 3) ultra-fast desorption speed, within ten seconds to ~1 min, 4) high desorption efficiency up to ~100%,

5) selectivity to interested target molecules thus benefit to the detection of real-samples free from interference of impurities or fluorescence background (this point is new enlightened by referees),
6) totally high enrichment factor thus high increased sensitivity up to 10(3) level within 2~3 9 min operation period which is very critical for the practical application in portable, fast, on site detection. 7) wide application potential for SERS, fluorescence, UV-vis and other sensing strategies.
In addition, to improve our enrichment capability and fast operation, we referred a recent reported best processes to prepare the mesoporous -cyclodextrin polymer published in Nature, 2016, 529, 190, cited in our manuscript in ref 20 in the revised version. We added magnetic particles into this best mesoporous CD polymer. If you compare our results, you may found we demonstrated the best and carefully designed protocols so that we realized (repeated here again) above seven performance.
In the revised version, with the enlightenment from referees, we have updated the data and added the new experiments on the selectivity and the performance in detection of complex matrix in real-sample environment, which are indeed very important.
In our revised version of manuscript, we have added Figure 5, and related text description, and also some data in supporting information. These include,

Reviewer #3 (Remarks to the Author):
My major concern is this study's novelty. Above of all, the concept of target enrichment is not novel, and magnetic nanoparticles functionalized with β-cyclodextrin have been extensively that probablly, referees provided such commens is because that we cannot demonstrate the novelty and significance of this work clearly in the introduction part. In fact, after carfully investigate the comments, two ponits are including, one is that referees said the similar materials have been reported, and the other is that this kind materials has been used for sensing.
Concerning above two points, in the revised version, we have reworte the introduction, and supplemented new experiments and added them.
For the materials, we want to explain two concept of "extraction" in some publicaitons and our "enrichment" in our manuscript. I have read many extraction related references. Yes, they indeed used the similar porous materials. However, these materials in these publications were mainly used for extraction, which is to select target molecule from complex system. E.g., extraction (adsorption) of organic molecules (microcystins or PAHs) from environmental water or soil sample in suggested two papers. Selectivity is the first importance in this kind of work, i.e.
selective adsorption interested molecules from multiple-molecules system. They did not notice to improve the removal speed, and enrichment capablity, thus you may see that in above two papers, one used nearly one hour to extract and 7 min to desorb and the other paper used high concentration adsorption materials and desorbed in a large volume of desorption solvent. The extraction process if operation into a traditional solid state extraction cartridges, the separating rate is relatively slow. Normally, the speed of extraction is 0.5-5 mL/min, which can not meet our current study motivation with very fast enrichment operation. Extraction seems to be more the work for environmental scientists, e.g. water or soil treatment.
However, in our current work, we are aiming our work to develop a new detection strategy for trace pollutants, and we paid our attention to the enrichment capability and fast operation, (in the following, we shall list the novelty and advance of this work), which is very important for the practical application in protable, fast, on site detection. Thus, our enrichment strategy is different from classical extraction process.
With the enlightenment of referees, we reconsidered and updated the significance and novelty of this work, and we want to emphasize here: we explored a very effective enrichment and detection strategy with remarkable performance including 1) ultra-fast removal speed within ten second to ~1 min, 15 2) high removal efficiency up to ~90%, 3) ultra-fast desorption speed, within ten seconds to ~1 min, 4) high desorption efficiency up to ~100%, In addition, to improve our enrichment capability and fast operation, we referred a recent reported best processes to prepare the mesoporous -cyclodextrin polymer published in Nature, 2016, 529, 190, cited in our manuscript in ref 20 in the revised version. We added magnetic particles into this best mesoporous CD polymer. If you compare our results, you may found we demonstrated the best and carefully designed protocols so that we realized (repeated here again) above seven performance.

5) selectivity to interested target molecules thus benefit to
Concering the comments of referee "this kind materials has been used for sensing", we have carefully studied the licteratures and according to our knowledge, the organic pollutant sensing from literatures based on the magnetic nanoparticles functionalized with β-cyclodextrin, can be divided into two categories. One is that some magnetic materials decorated with βcyclodextrin monomer to build up such as electrochemical sensing. In this case, they did not use crosslinked β-cyclodextrin polymers and the absorption capacity of β-cyclodextrin monomer was much less than the crosslinked β-cyclodextrin polymers. Moreover, they did not carry out enrichment related study. The other kind of sensing using this kind of materials is to use crosslinked β-cyclodextrin polymers, but they use for liquid chromatography (HPLC) or photoluminescence spectroscopy sensing, e.g. solid-phase extraction adsorbent. Again, they are studying the extraction and selectivity of target molecule from complex pollutants.
Above two kinds of sensing are quite different as our enrichment and concentration-strategy for detection. More importantly, up to now, we have not found the reports of SERS or fluorescence related sensing based on the current reported highly effective enrichment strategy of magnetic 16 nanoparticles immobilized with crosslinked porous β-CD polymer.
Actually, like the nature paper we cited in Ref. 20 in the revised version, the crosslinked porous β-CD polymer have also reported before this work, but this work again reported an improved and excellent performance. Thus we think it is valuable to be published again in Nature.
For our work, besides of above senven remarkable performance. We cited this excellent method in Nature, and we added new magnetic design and obtain excellent enrichment properties, and seslectivity capablity to eliminate the interference of complex matrix in the real-sample detection environment.
Therefore, the novelty and significance of our manuscript are still there and we are believing it is valuable to be considered for publish in Nature Communications.
------------As for "the proposed method failed to offer a significantly improvement with respect to sensitivity and reproducibility compared to other related methods." We listed seven remarkable performance of the current strategy. Actually, item-2), 4), and 6) are related to this questions. Using the current protocol, we realized high efficient adsorption up to ~90%, selective seperation by means of the effects of host-guest inclusion and magnet, desorption with neearly ~100% efficiency, thus the total enrichment capability up to ~1000 times, hence an improved and increased sensitivity up to 10(3) level can be obtained. Particularly, within 2~3 min operation period which is very critical for the practical application in portable, fast, on site detection. These data have been shown in Figure 4 and Figure 5 according to the detection method of SERS or fluorescence specstroscopies.
As for the reproducibility, in the Fig. S22, we provided the adsorption efficiency of different experiment times and consecutive regeneration cycles, indicating the great reproducibility.
In addition, with the enlightenment from referees, we have added many new data in the revised version and we rewrote the manuscript, we added the new results on the selectivity to interested target molecules, thus benefit to the detection of real-samples free from interference of impurities or fluorescence background (Presented in new Figure 5 and supporting information 19-

22).
Therefore, I sincerely hope this revised version with many supplemented data can be considered and accept for publication as soon as possible.
The authors addresses my comments very well. I learned a lot from it. Thanks you, I recommend the publication in Nature Communications.
Reviewer #2 (Remarks to the Author): I think that the authors have made a lot of effort in improving the manuscript, they have included further data regarding the selectivity and multiplexing capabilities of the sensing and extraction platform. But from my point of view, the work is not relevant enough to publish in this journal. I think that it is more suitable for a sister journal.
Reviewer #3 (Remarks to the Author): By introducing magnetic nanoparticles to an established system (Nature, 2016, 529, 190-194), this study achieved ultra-rapid and highly efficient enrichment of organic pollutants. Magnetic SERS has been extensively studied and applied in organic pollutant sensing (Journal of Environmental Sciences, 2019, 80, 14-34). Compared with those related researches, the proposed study mainly focused on removal speed, and enrichment capability. However, sensitivity and spectral reproducibility are two crucial factors for SERS-based nanosensors, which are more important and should not be neglected. The enrichment ability of such materials makes sense for me, but as an ultrasensitive nanosensor, the following issues haven't been addressed in the revised manuscript. 1. No significant improvement in sensitivity has been achieved in this study, comparing with other magnetic SERS-based sensors for organic pollutants. 2. The authors did nothing to improve the SERS spectral reproducibility, so the accuracy of the proposed method is not convincing. 3. In my opinion, a SERS nanosensor that combines ultrahigh sensitivity (achieved by high enrichment ability and signal enhancement ability), high spectral reliability, and fast removal speed, is publishable in the high-level journal of Nature Communication.

Reviewer #1 (Remarks to the Author):
The authors addresses my comments very well. I learned a lot from it. Thanks you, I recommend the publication in Nature Communications.
Response: Thank you very much for the reviewer's comments and suggestions.

Reviewer #2 (Remarks to the Author):
I think that the authors have made a lot of effort in improving the manuscript, they have included further data regarding the selectivity and multiplexing capabilities of the sensing and extraction platform. But from my point of view, the work is not relevant enough to publish in this journal. I think that it is more suitable for a sister journal.
Response: Thank you very much for the reviewer's comments, just as the recommendation from reviewer-1 and -3, we believe a highly efficient and ultrafast enrichment protocol and a high sensitivity, fast operation, and good repeatability of SERS detection is important to the field and valuable to be published in Nature Communications.

Reviewer #3 (Remarks to the Author):
By introducing magnetic nanoparticles to an established system (Nature, 2016, 529, 190-194), this study achieved ultra-rapid and highly efficient enrichment of organic pollutants. Magnetic SERS has been extensively studied and applied in organic pollutant sensing (Journal of Environmental Sciences, 2019, 80, 14-34). Compared with those related researches, the proposed study mainly focused on removal speed, and enrichment capability. However, sensitivity and spectral reproducibility are two crucial factors for SERS-based nanosensors, which are more important and should not be neglected.
The enrichment ability of such materials makes sense for me, but as an ultrasensitive nanosensor, the following issues haven't been addressed in the revised manuscript.
1. No significant improvement in sensitivity has been achieved in this study, comparing with other magnetic SERS-based sensors for organic pollutants.
Answer: Thanks for the advice from the reviewer. In this manuscript, we realize highly efficient enrichment capability up to ~1000 times and the LODs of fluorescence detections for enriched molecules were enhanced by 2~3 orders. Depending on the reviewer comment, in the revised manuscript, the SERS sensitivities of our enrichment method were measured and compared with other reports. As for the SERS sensitivity for the organic pollutants, we chose a more common molecular, TMTD (also called thiram), as the probe molecule to explain the superiorities of our protocol. As shown in Fig. 4a-b, the LOD of TMTD after the enrichment of MN-PCDP is up to 5 fM ( Fig. 4c), that is superior to most of the magnetic SERS-based sensors (generally no better than 10 -12 M). Moreover, as for carbendazim and diquat, the detectabilities are obtained in 5 pM ( Supplementary Fig. 18) and 1 pM ( Supplementary Fig. 19), respectively, while the similar SERS sensor just achieve to ~nM level. Furthermore, the detectabilities of several analytes are summarized in Table 1 Fig. 17-20), which are much lower than most of the magnetic SERS-based sensors (Table 1).   Another SERS substrate, the wet-chemical synthesized Au NPs with addition of inorganic salt, was adopted to explain the consistency of SERS signal, which was the simplest and effective way in commercial detection platforms at present. For the aggregating approach with Au colloid, the reproducibility was consistent in aggregation states from different batches. As shown in Fig 4d and Supplementary Fig. 31-32, the detection probability is up to 100%. Third, the SERS detection about analyte with lower concentration in real-sample system was also tested. In Fig. 5e, carbendazim molecule with 1 nM in soil solution is distinctly detected by the selective enrichment of MN-PCDP.
The correlated revisions are shown as following: Page 9, Paragraph 1: Furthermore, multiple adsorption and desorption experiments by MN-PCDP for above five organic molecules are implemented to illustrate the reproducibility of this adsorbent. In Supplementary Fig. 21-23, Supplementary Table 5-6, the removal efficiencies and enrichment efficiencies of MN-PCDP adsorbent are excellent for target molecules with RSD less than 1% and 5%, respectively. The Raman detectable reproducibility of TMTD by hydrophobic slippery SERS platform is shown in Supplementary Fig. 24-30. In Fig. 4c, the Raman signals of TMTD characteristic peaks are acquired with 100% detection probability in 0.5 pM, ~65% in 50 fM and ~10% in 5 fM. In addition, the solution-based aggregation approach, a simplest and effective way in commercial detection platforms at present, is adopted to clarify the consistency of SERS signal. As shown in Fig   4d and Supplementary Fig. 31-32, the SERS signals display superior spectral reproducibility and uniformity with 100% detection probability and RSD value of ~5%, even at TMTD concentration of 10 -12 M.      3. In my opinion, a SERS nanosensor that combines ultrahigh sensitivity (achieved by high enrichment ability and signal enhancement ability), high spectral reliability, and fast removal speed, is publishable in the high-level journal of Nature Communication.
Answer: Thanks for the comment and very good suggestions from the reviewer. Following the suggestion, the revised manuscript has been obviously improved. After revision, we think the manuscript realize the review's criteria for publishing in the high-level journal. In this work, we show an enrichment-typed sensing strategy by using a powerful mesoporous nanosponge. Based on the excellent capturing selectivity of and fast removal capability for organic micropollutants from β-cyclodextrin polymers as well as the magnetic nanoparticles, great enrichment and detection performance on analyte are achieved, e.g. ~90% removal efficiency (RSD<1%) within ~1 min, concentrated and enriched from complex matrix with an enrichment factor up to ~10 3 (RSD<5%), ultrahigh detection sensitivity (0.5 pM with 100% detection possibility and 5 fM with ~10% detection possibility for TMTD molecule). By means of the current absorption strategy, the mesoporous nanosponge is used to separate and selectively enrich (2~3 orders of magnitude) target molecules from the real-sample system. Importantly, the current enrichment strategy is proved to be helpful tool in a variety of fields for portable and fast detection, such as UV-vis, Raman and fluorescent.
Therefore, we believe this revised version with many supplemented data can be considered and accepted for publication.