Research ArticleChemistry

Regulating the reactivity of black phosphorus via protective chemistry

See allHide authors and affiliations

Science Advances  11 Nov 2020:
Vol. 6, no. 46, eabb4359
DOI: 10.1126/sciadv.abb4359

Abstract

Rationally regulating the reactivity of molecules or functional groups is common in organic chemistry, both in laboratory and industry synthesis. This concept can be applied to inorganic nanomaterials, particularly two-dimensional black phosphorus (BP) nanosheets. The high reactivity of few-layer (even monolayer) BP is expected to be “shut down” when not required and to be resumed upon application. Here, we demonstrate a protective chemistry–based methodology for regulating BP reactivity. The protective step initiates from binding Al3+ with lone pair electrons from P to decrease the electron density on the BP surface, and ends with an oxygen/water-resistant layer through the self-assembly of hydrophobic 1,2-benzenedithiol (BDT) on BP/Al3+. This protective step yields a stabilized BP with low reactivity. Deprotection of the obtained BP/Al3+/BDT is achieved by chelator treatment, which removes Al3+ and BDT from the BP surface. The deprotective process recovers the electron density of BP and thus restores the reactivity of BP.

INTRODUCTION

Precisely tuning the properties of nanomaterials to obtain desired characteristics is one of the most important goals of nanoscience. Within this scope, rationally regulating the reactivity of nanomaterials is critical for future multistep programmable processing and applications. Some nanomaterials (or some certain parts) need to be protected to reduce their reactivity under certain specific conditions and to restore their activity after successful deprotection (13).

Scientists have, in the past few decades, proposed plenty of efficient and selective protection-deprotection strategies toward regulating the reactivity of various functional groups in organic chemistry (4, 5). Usually, the functional group (or organic molecule) is linked with the protective group via various chemical/physical interactions (typically covalent bond) to reduce its reactivity, so that the protected functional group (or molecule) can survive in the subsequent steps. Subsequently (after the steps involving the attack of other functional groups), the protective group is removed, restoring the original functional group or molecule. This protective strategy is pervasive in multistep organic syntheses, such as natural product synthesis, solid-phase peptide synthesis, and polymer synthesis (4, 6, 7). Unfortunately, these well-established organic protective-deprotective processes are hardly applicable in inorganic nanomaterials owing to the lack of functional groups in inorganic nanomaterial surfaces, their irreversible agglomeration, surface reconstruction, and particle etching during the complex protective-deprotective process (8, 9). An efficient and facile approach to regulate the reactivity of inorganic nanomaterials remains elusive.

Black phosphorus (BP), a rising star in postgraphene two-dimensional (2D) nanomaterials, is known for its tunable bandgap (from ~0.3 to ~2.0 eV), its good compromise between charge carrier mobility and current on/off ratios, its broadband absorption from the visible to the mid-infrared range (1012), and its excellent biocompatibility (13). These attractive properties position BP to be suitable for application in optoelectronics and biomedical areas (10, 13). However, the high chemical reactivity of BP and oxygen/water leads to BP degradation under ambient conditions, causing functional failure of BP (14, 15). Recent research reveals that the reactivity of BP is directly related to the lone pair electrons of the P atom (internal factor) and the surrounding oxygen/water (external factor) (14). The lone pair electrons of each P atom contribute to the high electron density on the BP surface (16, 17), which gives BP strong reducibility. The surrounding oxygen/water can easily attach to the highly reactive surface of BP and react to form PxOy. To protect BP, a conceivable strategy could be to decrease its surface electron density and prevent oxygen/water from accessing the BP surface. On the basis of this idea, considerable efforts have been made to protect BP from ambient degradation. However, despite the rapid progress in the effective protection of BP, a practical method to deprotect passivated BP has not been developed. It is even more difficult to develop a method that combines protection and deprotection processes to switch BP from a passivated state to a reactive state in response to environmental changes (i.e., passivation by storing under ambient condition, and reactivity resuming to facilitate further functionalization or degradation when necessary) (13, 1720).

Here, we develop a protective chemistry–based strategy (Fig. 1) for rationally regulating the reactivity of BP. We begin by binding the BP with Al3+ ions to decrease its surface electron density, effectively decreasing its reducibility. Then, the hydrophobic 1,2-benzenedithiol (BDT) molecule assembles into a dense array on the surface of BP/Al3+ via the Al─S bond, which effectively isolates the nanocomposite from oxygen/water. This protective process offers an ultrastable BP complex (BP/Al3+/BDT), which can be stable under ambient conditions even for 2 months without its key physical/chemical characteristics being altered. Contrary to previous reports, this ultrastable BP/Al3+/BDT can be deprotected by chelator treatment [typically EDTA-tetrasodium (EDTA-4Na)]. This is possible because of the stronger binding affinity between Al3+ and EDTA-4Na in BP that enables the removal of Al3+ and BDT layers from the BP surface. The removal of the Al3+ and BDT layers restores the high surface electron density of BP, resuming its reactivity. To prove this concept, we used the deprotected BP in a degradation study. Expectedly, it exhibited the same behavior as original BP.

Fig. 1 Schematic illustration of regulating the reactivity of BP via protective chemistry.

Protective step 1: Binding Al3+ ions with lone pair electrons on the surface of P atoms decreases surface electron density of BP, leading to a reduced chemical reactivity of BP. Protective step 2: Self-assembly of the hydrophobic dense array on the BP surface isolates BP from surrounding oxygen/water. Deprotective step: Removal of Al3+ ions and hydrophobic dense array on the BP surface by a chelating agent. The treatment recovers the electron density of BP, restoring the original reactivity of the deprotected BP. BDT, 1,2-benzenedithiol; EDTA-4Na, EDTA-tetrasodium.

RESULTS

Synthesis and characterization of BP/Al3+/BDT

Bulk BP was prepared and characterized following a methodology highlighted in a previous report (fig. S1, A to C) (21). BP nanosheets were obtained via the sonication of powdered bulk BP in N,N-dimethylformamide. Scanning electron microscopy and transmission electron microscopy (TEM) images show that the size of the BP is 858.6 ± 89.1 nm (fig. S1, D and E). High-resolution TEM images show a single-crystal BP nanosheet with a lattice spacing of 2.56 Å, which assigns to the (111) plane of BP (fig. S1F). Atomic force microscopy (AFM) analysis reveals that the thickness of BP is 2.65 ± 0.27 nm, which implies that there are four to six individual phosphorene layers (fig. S1G) (21). X-ray diffraction (XRD) and Raman spectra demonstrate the same crystal characteristic of BP as its bulk form (fig. S1, H and I).

The protective step starts by binding Al3+ to the BP surface. BP/Al3+ is obtained by the simple mixing of BP and AlCl3 in an ethanol solution at room temperature. In our case, besides Al3+, a wide range of metal ions were systematically screened (see note S1). Noble metal ions (such as Au3+, Ag+, and Pd2+) can form a redox pair with BP and quickly react with BP to form noble metal nanoparticles on the BP surface (fig. S2, A to I). Heavy metal ions (such as Cu2+, Zn2+, Ni+, Co2+, Mn2+, Fe2+, and Sn4+) and light metal ions (such as Na+, K+, Mg2+, Ca2+, Al3+, and Ti4+) can form similar BP/metal ion complexes. However, compared to Al3+, most of them show the weaker ability for passivating BP (fig. S2K). A more detailed discussion about the effect of charge and radius on interaction strength can be found in fig. S3 and note S1. Some metal ions (typically Ti4+ and Sn4+) undergo fast hydrolysis, which is not conducive for regulating the reactivity of BP (fig. S3, A to G). Therefore, Al3+ ions are selected to form BP/Al3+ coordination complexes because they have a strong electron-withdrawing ability, relatively high stability, and low reactivity over other metal ions (2224). The changes in zeta potential suggest the successful binding of Al3+ ions on BP (fig. S3D). X-ray photoelectron spectroscopy (XPS) characterization provides further evidence for the successful attachment of Al3+ ions to the BP surface (fig. S3I).

Subsequently, a layer of BDT attaches to the surface of BP/Al3+ via self-assembly to further strengthen the protection of BP. BDT was chosen as the protective layer for the following reasons. (i) It can form an orderly molecular array on the selected substrate owing to its rich π electron density and hydrophobic nature (25, 26); (ii) the thiol group, in this case, is more suitable than other functional groups such as carboxyl, oxhydryl, and amino groups for the assembly of a hydrophobic layer on the surface of BP/Al3+ (note S2 and fig. S4); (iii) BDT exhibits less conformational freedom over the linear n-alkane thiol ligands and thus can assemble into a denser monolayer on the substrate (fig. S4E) (27, 28); and (iv) theoretically, the BDT monolayer is sufficiently thin (~0.50 nm, based on AFM images in fig. S1G and Fig. 2), and as such, it will have little influence on the physical/chemical properties of the coated BP.

Fig. 2 Characterization of BP/Al3+/BDT.

(A) TEM image. (B) AFM (height profile along the white line) image. (C) STEM–energy-dispersive x-ray spectroscopy (EDX) elemental mapping images. (D) High-angle annular dark-field (HAADF) image. (E) Magnified HAADF image taken from the selected area in (D). a.u.: arbitrary units. (F) Selected-area electron diffraction (SAED) pattern of BP and BP/Al3+/BDT. (G) FTIR spectra of BP, BP/Al3+, BP/Al3+/BDT, and BDT. (H) 1H NMR spectra of BP, BP/Al3+/BDT, and BDT. (I) Thermogravimetric curves of BP and BP/Al3+/BDT. ppm, parts per million.

The morphology of the obtained BP/Al3+/BDT was investigated. TEM (Fig. 2A) images show that the morphology of BP/Al3+/BDT has a 2D nanostructure without observable defects on its surface. AFM images show that the thickness of BP/Al3+/BDT is 3.75 ± 0.22 nm (Fig. 2B). These suggest that the obtained BP/Al3+/BDT does not show a notable morphological difference from the original BP nanosheets (fig. S1, E and G). The XRD pattern of BP/Al3+/BDT gives the same feature peaks like that of the original BP, indicating that the crystal structure was unaltered (fig. S4H). The conductivity of BP is also preserved after the protective treatment (fig. S4I).

The surface configuration of BP/Al3+/BDT was studied by scanning TEM (STEM). High-angle annular dark-field (HAADF)–STEM images and energy-dispersive x-ray spectroscopy (EDX) analysis of BP/Al3+/BDT reveal the uniformity of the distribution of P, Al, and S over the whole nanosheet (Fig. 2C). The HAADF image of BP/Al3+/BDT in Fig. 2D indicates a lattice constant of 0.256 nm, which is consistent with the original BP nanosheets. Figure 2E shows the HAADF image of the enlarged area in Fig. 2D (dashed yellow rectangle). The spots with relatively high contrast (labeled with dashed white circle) located at the central area of the P atom (with low contrast) hexagons can be assigned to the Al3+ ions. The Z-contrast intensity distribution (Fig. 2E, inset, corresponding to the selected area labeled with dashed green rectangle) discloses a P-Al periodic pattern, which further suggests that the Al3+ ions favor a central location in P hexagons. Figure 2F shows the selected area electron diffraction (SAED) pattern of BP and BP/Al3+/BDT. In comparison to BP, the diffraction spots associated with the (001) and (021) lattice planes of BP/Al3+/BDT are almost extinct (labeled with dashed red circle), while the (111) lattice plane of BP/Al3+/BDT gets enhanced (labeled with dashed green circle). The difference in diffraction spots between BP and BP/Al3+/BDT is attributed to the difference in electron beam scattering and interference, further confirming the binding of Al3+ to BP.

The formation of BP/Al3+/BDT was further verified by Fourier transform infrared (FTIR) spectroscopy and proton nuclear magnetic resonance (1H NMR) (Fig. 2, G and H). In comparison with BP, the FTIR spectra of BP/Al3+/BDT show six substantial characteristic bands at 3400, 1637, 1563, 1430, 1024, and 770 cm−1, respectively (Fig. 2G). The characteristic bands at 3400 and 1637 cm−1 are assigned to the O─H stretching vibration (29). A similar characteristic band is also observed in the FTIR spectrum of BP/Al3+, suggesting that a few ─OH groups were attached to Al3+ in the synthesis of BP/Al3+ (30). The other characteristic bands at 1563, 1430, 1024, and 770 cm−1 are attributed to the BDT molecule, suggesting the presence of BDT molecule in the complex. The S─H stretching vibration in the spectrum of BDT is found at 2653 cm−1, where the FTIR spectrum of BP/Al3+/BDT shows a flat curve (Fig. 2G, red line) (31). The disappearance of the S─H stretching vibration in the FTIR spectrum of BP/Al3+/BDT is evidence of bond formation between Al and S. Figure 2H shows 1H NMR spectra of BP, BP/Al3+/BDT, and BDT. A single peak assigned to hydrogen in the S─H group is observed in the BDT at 3.6 parts per million (ppm). Contrastingly, no such peaks are observed at 3.6 ppm for BP/Al3+/BDT (Fig. 2H, red line) (32). This further confirms the formation of the Al─S bond. In addition, two chemical shifts of the H in the benzene ring of BDT are observed after the self-assembly slightly shifts to 7.16 ppm (7.07 for original BDT) and 7.48 ppm (7.36 for original BDT) (Fig. 2H, inset). This can be attributed to the covalent interaction of BDT and Al3+ ions (33). FTIR and 1H NMR characterization provide robust evidence for the formation of the BP/Al3+/BDT complex. The Raman spectrum of BP/Al3+/BDT further supports the existence of BDT on the BP surface (fig. S4J). Thermogravimetric analysis reveals that the mass ratio of BP:Al:BDT is approximately 11:1:3 (Fig. 2I). BP is fully covered by Al3+/BDT according to the theoretical calculation (theoretical ratio of BP:Al:BDT is 10:1:3, and the mass ratio of BP:Al:BDT is directly affected by the layer number of BP; see fig. S5, A to C, and note S3 for calculation details).

Reactivity of BP/Al3+/BDT

The reactivity (to oxygen/water) of BP/Al3+/BDT was investigated via a polarizing microscope, TEM, XPS, and ultraviolet-visible (UV-vis) spectroscopy. At the initial stage, the polarizing optical microscope images of both as-prepared BP and as-prepared BP/Al3+/BDT showed a perfectly clean and flat surface (Fig. 3, A1 and B1). TEM images of these two samples showed the same 2D nanosheet structures without observable defects (Fig. 3, A1 and B1, insets). The surface of BP exhibited rough and small topographic protrusions (hereafter termed “bubbles”) after 1 day of ambient exposure (Fig. 3A2). The surface became rougher, and the bubble size increased when the exposure time was extended to 7 days (Fig. 3A3). The corresponding TEM images show the evolution process of structural destruction and surface bubble growth (Fig. 3, A1 to A3, insets). These results suggest that BP is oxidized after ambient exposure for 1 day and heavily oxidized after 7 days. Contrary to BP, the surface of BP/Al3+/BDT remains almost unaltered after 60 days of ambient exposure (Fig. 3B). Furthermore, a crystal structure is observed for BP/Al3+/BDT with a lattice spacing of 2.56 Å, which is indexed to the (111) plane of BP even after 1 year of ambient exposure. This result is further supported by XRD, Raman spectra, water contact angle, and zeta potential characterizations, demonstrating the long-term ambient stability of BP/Al3+/BDT (see fig. S5D and note S3 for details).

Fig. 3 Characterization of degraded BP and BP/Al3+/BDT under ambient conditions.

Polarizing microscope images of (A) bulk BP (0, 1, and 7 days) and (B) bulk BP/Al3+/BDT (0, 30, and 60 days). Insets: Corresponding TEM images. Scale bars, 200 nm. (C and D) HR-XPS spectra of P 2p peaks for BP and BP/Al3+/BDT with ambient exposure for various durations. (E and F) UV-vis spectra of BP and BP/Al3+/BDT dispersed in water for various durations. Insets: variation of the UV-vis absorption ratios at 470 nm (A/A0) of BP (A0: original value).

Degradation of BP yields a product of PxOy and, lastly, produces phosphate anions (BP→PxOy→PO43−) (15). With degradation, the content of PxOy on the BP surface or the content of PO43− in the BP aqueous dispersion are conceivably increased. High-resolution XPS (HR-XPS) spectra of P 2p were used to determine the evolution of the content on the PxOy surface during the degradation process of both BP and BP/Al3+/BDT under ambient conditions. As shown in Fig. 3 (C and D), both BP and BP/Al3+/BDT show two peaks. One peak is visible at 128.5 to 131.5 eV and is assigned to P, while the other is visible at 132.5 to 135.2 eV and is assigned to PxOy. At the initial stage, both as-prepared BP and BP/Al3+/BDT display a high-intensity P peak and low-intensity PxOy peak, respectively. After ambient exposure for 7 days, the peak intensity of BP decreases (P: from 87.3 to 10.7%), while the peak intensity of PxOy increases (PxOy: from 12.7 to 89.3%) simultaneously (Fig. 3C). In contrast, for BP/Al3+/BDT, the peak intensity of P exhibited no significant changes (P: from 89.9 to 76.3%) even after ambient exposure for 60 days (Fig. 3D), while the peak intensity of PxOy slightly increased (PxOy: from 10.1 to 23.7%). In addition, the intensity of P─P/P─O for BP/Al3+/BDT is very close to that of the original BP, indicating that the BP/Al3+/BDT offers reliable protection to improve the stability. XPS analysis results were consistent with those of polarizing microscopy. This indicates that the stability of BP/Al3+/BDT is superior to that of BP.

To further address the degradation of both BP and BP/Al3+/BDT, we detected the amount of PO43− in BP dispersion and BP/Al3+/BDT dispersion by UV-vis [see experimental procedures in the Supplementary Materials and fig. S5 (E and F) for details] (21). For the initial dispersion, the absorbance intensity of both BP and BP/Al3+/BDT at 470 nm is roughly the same (Fig. 3, E and F), indicating the same concentration of BP in these two solutions. With increasing dispersion time, this absorbance intensity of BP gradually dwindles, and the absorbance intensity of PO43− at 710 nm increases simultaneously (Fig. 3E and fig. S5G). After incubating in aqueous solution for 7 days, the absorbance intensity of BP at 470 nm (A) decreased by 95.5% compared to the original value (A0) (Fig. 3E, inset), while the absorbance intensity of PO43− at 710 nm increased by 93.7% compared to the original value (fig. S5G). These results reveal the fast degradation of BP in aqueous solution. The final concentration of PO43− (6.6 μg/ml) in the degraded solution is close to the initial concentration of BP (6.8 μg/ml), which is consistent with the UV-vis observation. Contrarily, for BP/Al3+/BDT aqueous dispersion, the UV-vis absorbance intensity of BP/Al3+/BDT and PO43− shows no significant changes after incubating for 60 days (Fig. 3F and fig. S5H). UV-vis spectra prove that the stability of BP/Al3+/BDT is superior to BP.

The above results indicate that our protective strategy through BP/Al3+/BDT successfully embeds BP with an ultrastability and reduces its reactivity. Our strategy relies on the coordinated interaction between Al3+ and BP, which is expected to be stronger than that induced by noncovalent functionalization (34, 35). Furthermore, the BDT hydrophobic layer provides a dense barrier to oxygen/water. Therefore, both internal and external influencers for BP degradation are minimized, rendering an ultrastable BP in comparison to the BP passivated by other methods (table S1). The BP/Al3+/BDT can even survive some harsh oxidation conditions. As shown in fig. S5 (I and J), BP/Al3+/BDT can remain stable in solutions containing strong oxidants (such as noble metal salt water solution HAuCl4, H2PdCl4, and AgNO3) for 8 days, while the as-prepared BP reacts with noble metal salts immediately.

Mechanism for decreasing the reactivity of BP/Al3+/BDT

The reducing reactivity of BP/Al3+/BDT can be attributed to two factors: first, the binding of Al3+ to the BP surface, which results in an electron density shift from the BP surface to Al3+, rendering a lower chemical reactivity of BP/Al3+/BDT; second, the self-assembled hydrophobic dense array on the BP surface effectively isolates BP from oxygen and water, preventing further degradation. Decreasing electron density on the BP surface is revealed by XPS spectra and further supported by density functional theory (DFT) simulation. Full-scan XPS spectra (Fig. 4A) reveals the presence of the relevant elements (the signal of Si derives from the substrate). In the BP sample, the P 2p core-level XPS spectrum shows P 2p3/2 and P 2p1/2 doublet at 129.6 and 130.7 eV, respectively, corresponding to the characteristic of crystalline BP (Fig. 4B) (13, 21). In the BP/Al3+ sample, owing to Al-P interaction, P 2p3/2 and P 2p1/2 doublet appears at higher binding energy (shift from 129.6 to 130.2 eV and from 130.7 to 131.2 eV, respectively). The lone pair electrons from the P atom donate to Al3+, which reduces the electron density on the surface of BP (3s and 3p orbitals) (17, 20). The decreased electron density of the BP surface layer causes strong attractive interactions in the inner layer of the P atom (2p orbitals); therefore, the appearance of XPS signals goes to higher binding energy. After BDT functionalization, owing to the formation of Al─S bonds, electrons of S enter the empty orbitals of Al3+ (36). In comparison to BP/Al3+ (~74.6 eV), the XPS peak of Al 2p for BP/Al3+/BDT (~75.0 eV) appears at the higher binding energy (Fig. 4C). Meanwhile, partial electrons retrace from Al3+ to P, which leads to the P 2p3/2 and P 2p1/2 doublet of BP/Al3+/BDT shifting to the lower binding energy (Fig. 4B, red line).

Fig. 4 Mechanism of reactivity decrease in protected BP.

(A) Full XPS spectra of BP, BP/Al3+, and BP/Al3+/BDT. (B and C) HR-XPS spectra of P 2p and Al 2p. (D to F) Calculated NBO charge of P atom, Al3+ ion, and S atom. Structure model of (G1) BP/Al3+ and (G2) BP/Al3+/BDT. Computational mapping of electron density difference in (G3) BP/Al3+ and (G4) BP/Al3+/BDT. Green regions indicate increased electron density, and blue regions indicate decreased electron density. Contours are shown at the 0.0001 a.u. level. (H) Water contact angles of BP, BP/Al3+, and BP/Al3+/BDT.

DFT calculations were carried out to investigate the electron transfer during the binding of Al3+ to the BP surface. After geometry optimization, a BP/Al3+ and BP/Al3+/BDT complex combined by coordination interaction was generated without showing the H atom (Fig. 4, G1 and G2). To quantitatively analyze the charge transfer, we calculated natural bond orbital (NBO) charges of BP, BP/Al3+, and BP/Al3+/BDT (Fig. 4, D to F). After the binding of Al3+ to the BP surface, NBO charges for P atoms increased (Fig. 4D), while NBO charges for Al3+ ions decreased (Fig. 4E). These results verify that electron density shifting occurs from BP to Al3+. Theoretically, the electron density of the BP surface should experience a decrease owing to the electron transfer from P to Al3+. This hypothesis is confirmed by mapping the electron density of BP/Al3+ (Fig. 4G3). As expected, a decrease in electron density (blue area) is observed for BP, whereas an increase in electron density (green region) is observed for Al3+ (Fig. 4G3). After BDT functionalization, compared to BP/Al3+, NBO charges for the P atoms decrease slightly, while NBO charges for Al3+ ions remain almost unchanged. Meanwhile, in comparison to BDT, NBO charges for S atoms of BDT increase slightly in the presence of Al3+ ions (Fig. 4F). The variation of NBO charges strongly confirms the electron transfer from Al3+/BDT to BP. These results are also consistent with the electron density mapping of BP/Al3+/BDT. As shown in Fig. 4G4, the electron density of BP increases slightly (green region in Fig. 4G4), while the electron density of Al3+/BDT decreases slightly (blue area in Fig. 4G4).

Self-assembly of a hydrophobic dense array on the BP surface is another crucial factor that contributes to the enhancement of BP stability. Previous reports demonstrated that water and oxygen are key factors in the process of ambient degradation of BP (14). In our case, the BP surface was fully covered by Al3+. However, the Al3+ layer was not hydrophobic enough to prevent water diffusion (contact angle of 12.3° for BP and 24.8° for BP/Al3+, as shown in Fig. 4H), and the monolayer of Al3+ was too thin to block the penetration of oxygen/water. Assembly of the hydrophobic dense array increases the contact angle of the BP surface from 24.8° to 130.5°, which, in turn, strongly increases the hydrophobicity of the obtained BP/Al3+/BDT complex. As shown in Fig. 3 and fig. S3L, although the BP/Al3+ shows improved stability (see note S1), further functionalization with the BDT layer promotes the stability of the material over the BP/Al3+ complex even further. The hydrophobic surface of BP can effectively prevent contact between water and BP, decreasing water-induced BP degradation (27, 28).

Beyond the hydrophobic surface, a closed-packed array-like dense molecular film was formed, which effectively isolated BP from oxygen and water. Owing to the interactions among aromatic rings, the BDT could form a highly ordered closed-packed array on the surface of BP. The interspace between the BDT molecules was around 3.40 Å (37), which is slightly smaller than the size of O2 (~3.46 Å) and water (~3.50 Å) molecules (3840). Thus, oxygen and water were blocked from the molecular layer, preventing BP from being easily degraded by the environment. When BDT was replaced by 2-naphthalenethiol (NAT; a similar aromatic thiol with BDT) for self-assembly on the BP surface (fig. S6), the obtained BP/Al3+/NAT complex demonstrated a stability similar to that of BP/Al3+/BDT (fig. S6E). The enhanced stability of BP/Al3+/NAT can be attributed to the hydrophobic surface (the measured water contact angle was 122.8°) and the dense-packed NAT (fig. S6E, inset). However, when a mixture of hydrophobic molecules was used (BDT/NAT = 1/1; fig. S6C), the obtained BP complex was less stable than BP/Al3+/BDT or BP/Al3+/NAT (fig. S6, D to F). Mixed hydrophobic molecule coassembling on the BP surface can induce defects within the closed-packed array (fig. S6, G to I). Thus, although BP/Al3+/BDT-NAT achieved a similar hydrophobicity, water and oxygen invasion would take place at this defect site, inducing degradation of BP (fig. S6F, inset). Therefore, the dense-packed hydrophobic array on the BP surface is also an important factor in isolating oxygen/water for improving the stability of BP.

Deprotection of BP/Al3+/BDT

The ultrastable BP/Al3+/BDT can be deprotected by removal of Al3+ from the BP/Al3+/BDT surface, as shown in Fig. 5A. Here, the removal of Al3+ is realized when EDTA-4Na is added, which is a conventional metal ion chelator (41). The full methodology is described as follows: First, we assess the removal ability of Al3+ in EDTA-4Na aqueous solution. The BP/Al3+/BDT complex is immersed in EDTA-4Na aqueous solution with different concentrations. Then, the residue Al3+ ions on the BP/Al3+/BDT surface are detected via fluorescence photometry using 8-hydroxyquinoline (see the Supplementary Materials for details and fig. S7, A and B) (42). The emission peak at 510 nm, which is a characteristic emission of 8-hydroxyquinoline aluminum salt, disappeared gradually, indicating that Al3+ ions had been successfully removed from the BP/Al3+/BDT surface. The removed amount of Al3+ ions by EDTA-4Na is directly correlated to the concentration of EDTA-4Na. The concentration of EDTA-4Na was 5 mM (fig. S7C). Figure 5B shows that photoluminescence (PL) intensity at 510 nm (Al3+ residue in BP/Al3+/BDT) decreases as incubation time increases in the presence of EDTA-4Na. The relationship between ln (Ct/C0) and time (t) reveals a linear correlation [ln (Ct/C0) = −0.139t + 0.063, R2 = 0.995] (Fig. 5C), where C0 and Ct refer to the loading concentration of Al3+ in BP/Al3+/BDT at an immersion time of 0 and t, respectively. The above analysis indicates that EDTA-4Na is a suitable chelator for the removal of Al3+ from the BP/Al3+/BDT surface. Al3+ ions on the BP/Al3+/BDT surface can also be removed by other chelating agents, such as sodium citrate (SC) and glutathione (GSH) (fig. S7, D to F), thus indicating its great potential for the application in biomedical-related fields.

Fig. 5 Deprotection of BP/Al3+/BDT.

(A) Schematic illustration of Al3+ ion and BDT removal by EDTA-4Na. (B) Photoluminescence (PL) emission spectra of Al3+ residue on BP/Al3+/BDT after EDTA-4Na treatment. (C) Plot of ln (Ct/C0) as a function of EDTA-4Na treatment time. (D and E) HR-XPS spectra of P 2p, Al 2p, and S 2p for BP, BP/Al3+/BDT, and deprotected BP/Al3+/BDT. (F) Plots of water contact angles and zeta potentials of BP as measured at each protective-deprotective cycle. (G) Polarizing microscope images of bulk BP (0 and 7 days) and bulk deprotected BP/Al3+/BDT (0 and 7 days). (H) Variation of PO43− concentration in solutions of BP and deprotected BP/Al3+/BDT with varying ambient exposure durations. (I) Stability of deprotected BP/Al3+/BDT with a varying residual amount of Al3+ ion on the BP surface. (J) TEM images of BP, BP/Al3+/BDT, and deprotected BP/Al3+/BDT after HAuCl4 (aqueous solution) treatment.

The hydrophobic molecules (BDT) were also removed together with Al3+. The deprotected BP/Al3+/BDT produces a hydrophilic surface with a negative zeta potential (fig. S7G), which is similar to the original BP. The P 2p binding energy of deprotected BP/Al3+/BDT (129.7 and 130.8 eV) is same as that of the original BP (129.6 and 130.7 eV) (Fig. 5D), indicating a resumed electron density on the BP surface. Furthermore, no Al and S signals were found in Al 2p and S 2p. XPS spectra of deprotected BP/Al3+/BDT (Fig. 5E) show the complete removal of Al3+ and BDT from BP/Al3+/BDT. The deprotective process does not affect the BP lattice structure (fig. S7H), Raman spectra (fig. S7I), conductivity (fig. S7J), and inherent photothermal conversion efficiency (fig. S7K).

Our protective-deprotective process achieves the reversible regulation of the BP reactivity. Figure 5F illustrates the plots of water contact angles and zeta potentials of BP measured at each interval of the protective-deprotective process cycles. In the five-cycled protective-deprotective process, the surface properties of BP fluctuate between hydrophilicity and hydrophobicity, and the corresponding zeta potentials of BP exhibit excellent reversibility.

The recovery of surface electron density and surface properties of deprotected BP/Al3+/BDT is supposed to have the same reactivity (such as degradation upon ambient exposure) as the as-prepared BP. As expected, the deprotected BP/Al3+/BDT displays the same degradation behavior as the original BP (Fig. 5G), which completely converts to PO43− after 7 days (Fig. 5H). In addition, the residue amount of Al3+ ions on deprotected BP/Al3+/BDT surface can be rationally tuned via different immersing times of BP/Al3+/BDT in EDTA-4Na aqueous solution. The reactivity of deprotected BP/Al3+/BDT highly depends on the residue amount of surface Al3+ ions (Fig. 5I and fig. S8), demonstrating the efficient approach for regulating the reactivity of BP. Beyond the degradability, the deprotected BP can also be used for further functionalization. As shown in Fig. 5J, the deprotected BP can react with HAuCl4 aqueous solution, and Au nanoparticle-functionalized BP is achieved. We also prove that Al3+-based BP reactivity regulation can be extended to other metal ions such as Fe3+, Zn2+, and lanthanide metal ions (fig. S9). Notably, some metal ions, typically Fe3+ with relatively high oxidizability, enable the oxidation of BP when the normal protective process is applied. For these cases, Fe3+ is linked to the BDT molecule to form the Fe3+-BDT complex before functionalization on the BP surface to yield BP/Fe3+/BDT (see note S4 for details). The slight modification for the protection process can reduce the induced oxidation by metal ions in high valance state, further extending the scope of the developed protective strategy.

To verify this concept, we used the established protective strategy for tuning the reactivity of BP in practical application (e.g., solar vapor generation). BP, BP/Al3+/BDT, and deprotected BP were deposited on hydrophilic poly(vinylidene fluoride) (PVDF) to fabricate BP/PVDF, BP/Al3+/BDT/PVDF, and the deprotected BP film, respectively (fig. S10, A and B; see also experimental details in the Supplementary Materials). BP/PVDF, BP/Al3+/BDT/PVDF, and deprotected BP show similar H2O evaporation rates (fig. S10, C and D), suggesting the same photothermal conversion property of these samples. However, after five cycles, the evaporation rates for the BP/PVDF film gradually decreased (fig. S10E). Further characterization revealed the significant degradation of the BP/PVDF film (fig. S10F), and the BP content in the BP/PVDF film dropped significantly (fig. S10G). In contrast, for the BP/Al3+/BDT/PVDF film, after five cycles, the evaporation rates did not change. No such degradation of BP/Al3+/BDT/PVDF film was observed (fig. S10F), and the BP content in the sample exhibited almost no changes (fig. S10G) after five cycles. The result demonstrates the low reactivity (to oxygen/water) of the protected BP during the solar vapor generation. For the deprotected BP film, its structure (fig. S10F) and BP content changed significantly, and therefore, similar degradation behavior to that of BP/PVDF film was observed, as expected. Analyzed together, these results suggest the feasibility of using the developed protective strategy for efficient regulation of the reactivity of BP for practical application.

DISCUSSION

We have demonstrated a protective chemistry–based approach for the rational regulation of BP reactivity. The protective step includes the binding of Al3+ ions with lone pair electrons of BP to decrease the surface electron density of BP. This protection is enhanced by the functionalization of BDT on the BP/Al3+ surface via Al─S bonds to form a dense hydrophobic layer. The decreased electron density of BP and the dense hydrophobic layer greatly decrease the reactivity of BP. By treating it with chelator (EDTA-4Na), Al3+ ions together with the hydrophobic layer can be removed from BP, leading to the reverting of the high electron density on the BP surface. The deprotected BP shows high reactivity in its original state as demonstrated by its ambient degradation and further functionalization with metal nanoparticles. The developed protective chemistry–based process gives a tunable approach for manipulating the reactivity of BP, which is hardly achievable by conventional functionalization (including covalent and noncovalent functionalization). Exploration of this protective strategy leads to dexterity in regulating reactivity of nanomaterials, which is of great importance for facilitating future nanomaterial-based multistep programmable functionalization and applications.

MATERIALS AND METHODS

Preparation of BP/Al3+/BDT

Bulk BP and BP nanosheets were prepared via a methodology highlighted in a previous study (see the Supplementary Materials for details) (21). For the preparation of BP/Al3+/BDT, BP nanosheets (0.02 mg) were added to AlCl3 (0.5 M, 20 ml) ethanol solution under stirring. After incubation for 2 hours, the precipitated product was separated from the supernatant by centrifugation and washed with ethanol and deionized water several times. Then, the precipitated product was added to BDT (5 mM) ethanol solution and incubated for 24 hours. The mixture was centrifuged at 12,000 rpm for 20 min, and the precipitated BP/Al3+/BDT was collected for use in subsequent experiments.

Deprotection of BP/Al3+/BDT

BP/Al3+/BDT (0.02 mg) was added to the EDTA-4Na (5 mM, 20 ml) aqueous solution under vigorous stirring for 20 min. Then, the mixture was centrifuged at 12,000 rpm for 20 min to collect the deprotected BP.

Electron density computational methods

The electron transfer effect for binding of BP with Al3+/-SH was investigated by DFT through calculating the NBO charge in optimized combination geometry and analyzing the electron density difference. The calculations were performed using the Gaussian 09 program (43) with the B3LYP functional. Lanl2DZ pseudopotential basis sets were used for aluminum, and 6-31g(d,p) basis sets were used for other elements. Molecular vibration frequencies were checked after geometry optimization to ensure no imaginary frequency. Monolayer BP consisting of 64 P atoms terminated by hydrogen was optimized, after which Al3+ and SH were combined in sequence. NBO charge of P, Al, and S was analyzed before and after combination. The plots of electron density differences of each complex were analyzed using Multiwfn software (44) to describe the electron transfer process. The electron density differences are defined as followsΔρ(BP/Al3+)=ρ(BP/Al3+complex)ρ(BP)ρ(Al3+)Δρ(BP/Al3+/SH)=ρ(BP/Al3+/SH complex)ρ(BP/Al3+)ρ(SH)

SUPPLEMENTARY MATERIALS

Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/6/46/eabb4359/DC1

https://creativecommons.org/licenses/by-nc/4.0/

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial license, which permits use, distribution, and reproduction in any medium, so long as the resultant use is not for commercial advantage and provided the original work is properly cited.

REFERENCES AND NOTES

Acknowledgments: We are grateful for the technical support from H. Wang, R. Yu, Y. Yang, and L. Yang from the Department of Physics and College of Materials, Xiamen University. Funding: This study was financially supported by the National Natural Science Foundation of China (21771154), the Shenzhen Fundamental Research Programs (JCYJ20190809161013453), the Natural Science Foundation of Fujian Province of China (2018J01019 and 2018J05025), and the Fundamental Research Funds for the Central Universities (20720180019 and 20720180016). This research was also supported by the Singapore National Research Foundation Investigatorship (NRF-NRFI2018-03). Author contributions: J.X., J.W., and Y.Z. conceived the idea and supervised the project. X.L. performed the experiments and collected the data. L.X. performed the TEM and analyzed the results. W.L. performed the Raman measurements. X.L., J.X., J.W., and Y.Z. analyzed the data and cowrote the paper. C.Z. and Q.X. discussed the results and commented on the paper. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Additional data related to this paper may be requested from the authors.

Stay Connected to Science Advances

Navigate This Article