Cavity Size Effect on Host-Guest Property of Tiara-like Structural M_n(SR)_2n Nanoclusters Probed by NMR Spectroscopy

Abstract

The lack of detect technology hinders the understanding of host-guest (H-G) chemical properties for thiolate-protected tiara-like structural nanoclusters (M_n(SR)_2n). In this work, NMR spectroscopy is demonstrated as a powerful tool to probe the H-G structure of M_n(SR)_2n both experimentally and theoretically. A low-field shifting and wide chemical shift (CS) signal of the H nucleus in CH₂Cl₂ is observed in the NMR spectrum of the mixture of CH₂Cl₂ and Pd₈(PET)₁₆ (PET is 2-phenylethanethiol), agreeing with the theoretical results that a deshielding area appears in the central cavity of Pd₈(SR)₁₆. All M_n(SR)_2n own similar nucleus-independent chemical shift maps and deshielding cavities, which means that the H nucleus in small molecules trapped by M_n(SR)_2n should have consistent low-field shifted CSs. However, such a phenomenon was only observed in the NMR spectrum of the mixed solution of Pd₈(SR)₁₆ and CH₂Cl₂, indicating that Pd₈(SR)₁₆ is the only one in the series of Pd_n(SR)_2n (n = 4~16) analogues that can capture a CH₂Cl₂, the H-G properties of M_n(SR)_2n are highly dependent on their cavity sizes, and a guest molecule only inserts into the matching cavity of M_n(SR)_2n. We anticipate that the realization of such convenient probe strategy will give a deeper understanding of the H-G properties of M_n(SR)_2n.

1. Introduction

A unique tiara-like framework endows thiolate-protected group 10 transition-metal nanoclusters (M_n(SR)_2n (M is Ni, Pd, or Pt), an inorganic analogy of metallocrowns, with excellent physical and chemical properties, such as nonlinear absorption, stability, photoluminescence, catalytic performance, and host-guest (H-G) chemical properties [1,2,3,4,5,6,7,8]. M_n(SR)_2n were considered potential host molecules since Ni₆(SC₂H₅)₁₂ was structurally determined [9]. The H-G structure of M_n(SR)_2n was first experimentally determined in 2002 by the reporting of a tetrahydrofuran that was captured by Ni₁₁(SC₆H₅)₂₂ [10]. Then, several G@M_n(SR)_2n (G denotes the guest molecule) were reported, including benzene and toluene molecules inserted into Ni₁₀(SR)₂₀ [3,11], a guest Ag⁺ ion accommodated in Pt₆(SR)₁₂ [12], and Pd₈(SR)₁₆ or Pt₈(SR)₁₆ encapsulating a small molecule, such as CH₂Cl₂, CH₂Br₂, CH₂ClCH₂Cl, CH₃I, and I₂ [13,14]. To the best of our knowledge, single crystal X-ray diffraction (SCXRD) is the only technique to obtain the H-G structures of G@M_n(SR)_2n.

As we know, the size of the cavity in a host molecule plays a very prominent role in the formation of the H-G structure. Until now, M_n(SR)_2n with various metal atoms, including Ni_n(SR)_2n (n = 4~6, 8~12) [3,10,15,16], Pd_n(SR)_2n (n = 4~20) [17], and Pt_n(SR)_2n (n = 5~13) [1], were synthesized and isolated, but their large, high-quality single crystals, especially for the larger sized ones, were difficult to grow [18]. This is a huge obstacle to correlate their cavity sizes to H-G chemical properties using SCXRD. Therefore, it is highly desirable to develop an alternative analytical method for probing the H-G structure of M_n(SR)_2n without a single crystal sample. The NMR technique was used to detect the H-G properties for a large number of materials [19,20,21,22,23,24]. For example, Hu et al. confirmed that the stereoselective, guest-driven, self-assembly of either Λ4- or Δ4- type Eu4L4 cages has been realized via chiral induction with R/S-BINOL or R/S-SPOL templates by NMR [20]. Schäfer et al. revealed the existence of a complex dynamic equilibria of oligomers that are formed by the host with bidentate guests based on diffusion NMR spectroscopy [21]. Mi et al. used solid-state NMR to ensure that the ratio of two host and guest molecules is 1:1:2 in interheteromacrocyclic hosts charge transfer crystals [23]. However, none has been reported for M_n(SR)_2n thus far, which may be due to the complexity of their guest molecules, including the metal ion (Ag⁺), the small organic molecules, and the reductive inorganic molecule (I₂). In this work with Pd₈(SR)₁₆ and CH₂Cl₂ as model host and guest molecules (Scheme 1), NMR spectroscopy is shown to be a highly effective approach to confirm the H-G structure of M_n(SR)_2n both experimentally and theoretically. Furthermore, we use NMR spectroscopy to study the size-dependent characteristics of host Pd_n(SR)_2n NCs capturing guest molecules. In the future, we plan to use NMR spectroscopy for the detection of M_n(SR)_2n capturing other molecules and expand their applications based on their H-G properties.

2. Materials and Methods

2.1. Materials

All reagents were commercially available and used as received. Palladium nitrate, sodium trifluoroacetate, and triethylamine were purchased from Sinopharm Chemical Reagent Co., Ltd, Shanghai, China. 2-phenylethanethiol, deuterochloroform, and trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene] malononitrile (DCTB) were purchased from Adamas, Shanghai, China.

2.2. Synthesis and Purification of Pd_n(PET)_2n

The synthesis of thiolate-protected palladium nanoclusters followed that of a previous report [17]. We dissolved 0.2 g Pd(NO₃)₂·2H₂O in 12 mL acetonitrile. After stirring for 30 min, 210 μL 2-phenylethanethiol was added into the above solution. Then, the reaction mixture was stirred for another 15 min, and 0.5 mL triethylamine was rapidly added. The reaction was allowed to continue under constant stirring for 5 h. After the reaction stopped, the yellow precipitate was washed several times with methanol and water, and then collected by centrifugation. The crude product was dissolved in CH₂Cl₂, and the precipitate consisting of insoluble Pd-thiolates was removed. The as-obtained soluble product was further isolated via thin-layer chromatography. Each Pd_n(PET)_2n NC was purified by TLC for a minimum of three times.

2.3. Characterization

Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) was collected by an autoflex speed TOF/TOF mass spectrometer (Bruker, Billerica, MA, USA) in reverse positive mode with DCTB used as the matrix. The ¹H NMR and 2D COSY of the mixture of Pd_n(PET)_2n (5 ≤ n ≤ 16) and CH₂Cl₂ were recorded on a Bruker 400 MHz spectrometer (Billerica, MA, USA) at room temperature. Approximately 10 mg of each nanocluster was dissolved in 1 mL CDCl₃ containing 0.1 μL of CH₂Cl₂ (approximately 1.6 × 10⁻³ mmol) (tetramethylsilane was used as internal standards).

2.4. Calculational Details

The geometrical structure of Pd₈(SCH₃)₁₆, simplified by replacing the -CH₂CO₂CH₃ groups of the experimentally obtained Pd₈(SCH₂CO₂CH₃)₁₆ molecule with methyl groups, was used for the theoretical calculation. All theoretical calculations were carried out by the Density Functional Theory (DFT) method using the Gaussian 16 program (Revision B01) [25]. The selected functional was the hybrid-GGA functional B97-2 [26], which reaches rather constantly the lowest standard deviations among several common functionals [27]. The employed basis sets were the pcSseg-1 (for C, H, and S atoms), which are developed and optimized for NMR calculations [28], and the Stuttgart–Dresden double-ζ (SDD) with an effective core potential (ECP) (for Pd atoms) [29].

The nucleus-independent chemical shift (NICS) maps of the studied Pd₈(SCH₃)₁₆ toroid structure were carried out by calculating the isotropic shielding value (δ_(r)) and its ZZ component (δ^zz_(r)) of 132,528 points in the space of 6.348 Å (12 Bohr) extension of the molecular coordinate [30]. These δ_(r) and δ^zz_(r) values were used to generate the maps to denote the shielded (negative) and deshielded (positive) areas of the system by using the Multiwfn program (Revision 3.6) [31].

In the calculation, an external magnetic field (B^ext) of 1.0 T was applied perpendicular to the molecular plane, the induced magnetic field (B^ind, in ppm units) over the space was related with the shielding tensor σ and the external magnetic field B^ext, as was given elsewhere:

$${\mathit{B}}_{\left(r\right)}^{\mathit{i}\mathit{n}\mathit{d}}=-{\mathit{\sigma }}_{\left(r\right)}{\mathit{B}}^{ext}$$

The isotropic shielding value δ_(r) (in units of ppm) can be determined by the definition:

$${\delta }_{\left(r\right)}=\mathrm{NICS}=-\frac{1}{3}\mathrm{Tr}{\mathit{\sigma }}_{\left(r\right)}$$

where Trσ_(r) represents the trace of the nuclear shielding tensor.

3. Results and Discussion

The host Pd₈(SR)₁₆ molecule was synthesized utilizing a procedure from a previous report [17]. In previous reports, a toluene molecule inserted into the Ni₁₀S₂₀ framework of Ni₁₀(StBu)₁₀(SR)₁₀ with 2-ethylthioethanethiol, 2-(2-mercaptoethyl)pyridine or 2-aminoethanethiol as perpendicular thiols [3,11], indicated that the H-G properties of M_n(SR)_2n were not dependent to their thiolate ligands. To avoid the formation of an intramolecular H-G structure by the thiolate carbon chain entering the Pd₈(SR)_16′s central cavity [16], 2-phenylethanethiol (PET), a thiol owning a large carbon tail (the outer diameter of phenyl (~5.3 Å) is larger than the cavity size of the Pd₈S₁₆ framework (~5.1 Å) [13]), was used as the ligand when we conducted the synthesis. The Pd₈(PET)₁₆ was synthesized and purified via a previous route [17]. Its chemical composition was identified by MALDI-TOF-MS and NMR, as shown in Figure 1 and Figure S1. The only peak centered at 3062.8 Da in the MS spectrum, and the perfect agreement between the experimental and simulated isotopic patterns indicated that the Pd₈(PET)₁₆ sample was of high purity (Figure S1). In the ¹H NMR and COSY spectra (Figure 1), the three intriguing triplets at 2.6–3.2 ppm were assigned to the protons of the methylene groups (1 and 2 in Figure 1) within 2-phenylethanethiolate, and the peaks at 6.8–7.3 ppm corresponded to the phenyl groups (3–5 in Figure 1). The wide blank span at 3.4–6.8 ppm in the ¹H NMR spectrum would be an ideal window to observe the CS shift of the H nuclei within a guest molecule.

Comparing four previously reported organic guest molecules of Pd₈(SR)₁₆, CH₂Cl₂, a cheap and easily available solvent, was optimized for the experiment. The reasons are as follows: First, its CS at 5.3 ppm [25] was in the middle of the observation window of Pd₈(PET)₁₆ and not disturbed by 2-phenylethanethiolate protected metal nanoclusters, e.g., Ni₆(PET)₁₂ [32], a small size M_n(SR)_2n; and second, unlike other guest molecules disordered in the central cavity of Pd₈(SR)₁₆ [13], there was only one configuration in which one C-Cl bond was arranged vertically and another one was nearly horizontal when CH₂Cl₂ was encapsulated by Pd₈(SR)₁₆ [13], as shown in Scheme 1. It should be noted that CDCl₃ was used as the solvent to characterize ¹H NMR because it cannot insert into Pd₈(SR)₁₆ [13]; this would be an effective strategy to avoid the competition between solvent and guest molecules in tests.

To obtain the CS changes of the CH₂Cl₂ molecules before and after entering Pd₈(PET)₁₆, the ¹H NMR spectra of pure CH₂Cl₂, pure Pd₈(PET)₁₆, and their mixture in CDCl₃ solution were compared first. In the mixture solution, the molar ratio of Pd₈(PET)₁₆ to CH₂Cl₂ was 2:1, ensuring that all CH₂Cl₂ molecules could insert into Pd₈(PET)₁₆. The ¹H NMR spectra of pure CH₂Cl₂, pure Pd₈(PET)₁₆, and their mixture in CDCl₃ solution are compared in Figure 2 and Figure S2. The pure Pd₈(PET)₁₆ and the mixture solutions have the same CS signals of the H nuclei within Pd₈(PET)₁₆, but there were two changes in the H nuclei’s CS within CH₂Cl₂ between the pure CH₂Cl₂ and the mixture solutions. One change was the CS of CH₂Cl₂ in the mixture moving toward low field (0.07 ppm). As shown in Figure 2a, unlike the CS of pure CH₂Cl₂ at 5.30 ppm [33], there was only one single, central peak at 5.37 ppm corresponding to CH₂Cl₂ in the ¹H NMR spectrum of the mixture. Another change was the CS signal of CH₂Cl₂ in the mixture solution owning a wider full width at half maximum (FWHM). As shown in Figure 2b,c, the FWHM of the mixture is approximately 0.0220 ppm, which is ~8.5 fold wider than that of pure CH₂Cl₂ (approximately 0.0026 ppm).

To verify the theory that the CS of H within CH₂Cl₂ shifting to low field was caused by CH₂Cl₂ inserted into Pd₈(PET)₁₆, we drew the NICS map of the host Pd₈(SR)₁₆ molecule to understand its magnetic response under an extra magnetic field. The structure of Pd₈(SCH₃)₁₆ (Scheme 1), simplified from the crystal structure of Pd₈(SCH₂CO₂CH₃)₁₆ [13], was used to carry out the calculation. After analyzing the crystal structures of M_n(SR)_2n [19,20,21,34], especially Pd₈(SR)₁₆ [13,14,35], with or without guest molecules (

Table 1. The structural information of Mn(SR)2n NCs with or without guest molecules.

Series	H-G Molecules	Bonding Length (Å)			Bonding Angles (°)				Ref.
		M-M	M-S	S-α-C	M-S-M	S-M-Sa	S-M-Se	M-S-α-C
Pd8(SR)16	Pd8(SCH2CH2CH3)16	3.23	2.32	1.84	88.05	97.77	82.27	107.06	[35]
	Pd8(SCH2CO2CH3)16	3.23	2.32	1.82	88.20	97.69	82.71	107.20	[13]
	CH2Cl2@Pd8(SCH2CO2CH3)16	3.24	2.32	1.82	88.43	97.66	82.71	107.07	[13]
	CH2Br2@Pd8(SCH2CO2CH3)16	3.24	2.32	1.81	88.44	97.77	82.62	107.24	[13]
	(CH2Cl)2@Pd8(SCH2CO2CH3)16	3.24	2.32	1.82	88.29	97.74	82.66	107.00	[13]
	CH3I@Pd8(SCH2CO2CH3)16	3.25	2.33	1.80	88.46	97.65	82.69	107.12	[13]
	I2@Pd8(SCH2CO2CH3)16	3.26	2.33	1.82	88.60	97.79	82.41	106.25	[14]
Pt8(SR)16	Pt8(SCH2CO2CH3)16	3.29	2.32	1.82	90.77	98.79	81.48	107.77	[13]
	CH2Cl2@Pt8(SCH2CO2CH3)16	3.29	2.33	1.82	90.37	98.04	82.11	108.22	[13]
	CH2Br2@Pt8(SCH2CO2CH3)16	3.29	2.32	1.83	90.43	98.72	81.56	107.40	[13]
	(CH2Cl)2@Pt8(SCH2CO2CH3)16	3.29	2.32	1.81	90.41	98.68	81.51	108.18	[13]
Pt6(SR)12	Pt6[S-(CH2)11-CH3]12	3.17	2.32	1.83	86.26	98.41	81.35	108.94	[12]
	Ag@Pt6[S-(CH2)11-CH3]12	3.08	2.32	1.83	83.24	97.61	82.39	109.95	[12]
Ni10-	Ni10(StBu)10(SC2H5)10	3.15	2.20	1.83	91.18	96.91	83.12	109.67	[34]
(StBu)10-	CH3C6H5@Ni10(StBu)10(etet)10	3.21	2.22	—	93.12	96.68	82.90	—	[3]
(SR)10	CH3C6H5@Ni10(StBu)10(pyet)10	3.16	2.21	1.85	91.31	97.42	83.04	109.92	[3]
	(0.5CH3C6H5)@Ni10(StBu)10(atet)10	3.17	2.21	1.85	92.06	96.96	83.07	110.02	[3]
	Ni10(StBu)10(mtet)10	3.17	2.21	1.85	91.54	97.48	82.57	110.16	[11]
	C6H6@Ni10(StBu)10(mtet)10	3.16	2.20	1.86	91.58	97.10	82.94	109.27	[11]

Noted: StBu is 2-methyl-2-propanethiol, etet is 2-ethylthioethanethiolate, pyet is 2-(2-mercaptoethyl)pyridine, atet is 2-aminoethanethiol, mtet is methylthioethanethiolate.

), two main factors were taken into account when we used the simplified structure: First, the structural features of M_n(SR)_2n, such as the tiara-like framework, bond length and angles, and spatial arrangements of α-CH₂ (the methylene attached to the S atom), were slightly affected by their ligand species and guest molecule species; and second, for most H-G structures of G@M_n(SR)_2n, the ability of the host M_n(SR)_2n trapping guest molecules into their specific M_nS_2n toroids was independent of their thiolate ligands. So it was reasonable that we considered the nearest segments of M_n(SR)_2n to the guest molecule, including the M_nS_2n framework and the α-CH₂, when we conducted the calculation. Theoretical induced magnetic fields of 132,528 points around the Pd₈(SCH₃)₁₆ were performed to investigate the isotropic magnetic response of this area. The NICS maps for Pd₈(SCH₃)₁₆ are shown in Figure 3. As shown in Figure 3a, the δ_(r) values in the areas nearby and outside of the molecular framework of Pd₈(SCH₃)₁₆ were negative and slightly positive, indicating that these areas are shielded and slightly deshielded regions, respectively. The δ_(r) values in the central cavity of the Pd₈ ring were positive (dark blue), showing that the cavity is a deshielded area. From Figure 3b, the deshielded area could be further understood as a closed area (inside the bold solid black line) surrounded by the Pd₈(SCH₃)₁₆ molecule. This closed, deshielded area could be characterized better by the positive values of the ZZ component of δ_(r) (δ^zz_(r) values, see Figure 3c,d). Therefore, one can infer that, once a small molecule is captured by Pd₈(SCH₃)₁₆, the resonant frequency of the H nuclei within the guest molecule will move towards the low field under an external magnetic field, showing that the CS of CH₂Cl₂ in the mixture moving towards low field was caused by the CH₂Cl₂ inserting into the central cavity of Pd₈(PET)₁₆.

The wider CS signal of CH₂Cl₂ in the mixture solution was caused by CH₂Cl₂ encapsulated by Pd₈(PET)₁₆ as well. As we know, in NMR spectrum, the total observed width of the absorption line (Δν) is used to define an effective relaxation time (T) by means of the relation

$$\Delta \nu \approx T^{-1}$$

where the T is mainly determined by the spin-spin relaxation time (T₂). When low concentration CH₂Cl₂ was dissolved in CDCl₃, the CH₂Cl₂ could effectively produce spin-lattice relaxation due to its non-restricted thermal motion but generate a slow spin-spin relaxation process because CH₂Cl₂ was mainly surrounded by CDCl₃; the long distance between the two H nuclei would cause a low probability of transverse relaxation. Once a CH₂Cl₂ was encapsulated into the cavity of the Pd₈S₁₆ framework, its spin-lattice relaxation became more difficult due to its restricted thermal motion, but its spin-spin relaxation became faster (a shorter T₂) because of the fixed relative position between CH₂Cl₂ and Pd₈(PET)₁₆, and 144 hydrogen atoms in one Pd₈(PET)₁₆ molecule, especially 16 hydrogen atoms in the perpendicular α-CH₂ of eight 2-phenylethanethiolate ligands very close to CH₂Cl₂. The shorter T₂ would be the main factor of T for the mixture, which led to a wider absorption peak. Herein, both experimental and theoretical results show that ¹H NMR spectroscopy is an effective method to detect the H-G structure of CH₂Cl₂@Pd₈(SR)₁₆.

Finally, the size-dependent character of M_n(SR)_2n about their H-G properties, the issue which has never been studied before, was uncovered by ¹H NMR spectroscopy here. Previous calculated results, which showed that several analogues of Ni_n(SR)_2n own similar induced magnetic fields as Pd₈(SR)₁₆ [36], indicated that the induced magnetic fields of M_n(SR)_2n were unrelated to their sizes and metal element species, and all Pd_n(SR)_2n owned a similar isotropic magnetic response as Pd₈(SR)₁₆, which meant that the CS of the H nuclei within a guest molecule would have uniform change when the small molecule inserts into the central cavity of Pd_n(SR)_2n. To eliminate the disturbance from ligands and treated history, the series of Pd_n(PET)_2n (5 ≤ n ≤ 16) analogues were synthesized and purified via the same process [17]. In the MALDI-TOF-MS spectrum of each Pd_n(PET)_2n (5 ≤ n ≤ 16) (Figure 4a), the single molecular ion peak in each mass spectrum indicated that every cluster sample was pure. The ¹H NMR and 2D COSY spectra of CH₂Cl₂ mixed with each Pd_n(SR)_2n are shown in Figure 4b and Figures S3–S8. From the NMR data, one can easily find that the CS of CH₂Cl₂ in the mixture of CH₂Cl₂ and each Pd_n(SR)_2n (n = 7, 9~16) has a similar position at 5.3 ppm and width of FWHM as pure CH₂Cl₂, except that of the mixture of CH₂Cl₂ and Pd₈(SR)₁₆, indicating that only Pd₈(SR)₁₆ can capture a CH₂Cl₂ into its central cavity; the other sizes of Pd_n(SR)_2n cannot form a H-G structure with CH₂Cl₂. The smaller Pd_n(SR)_2n cannot capture CH₂Cl₂ perhaps due to their narrower central cavity, but the larger ones cannot capture CH₂Cl₂ perhaps due to the weak interaction force between the host and guest molecules that make it difficult to bind the guest molecules inserting into the larger cavity. Such results mean that a guest molecule only inserts into the matching cavity of M_n(SR)_2n, and 8 is the ‘magic number’ for Pd_n(SR)_2n to capture CH₂Cl₂. The unchanged CS signal of CH₂Cl₂ mixed with smaller size Pd_n(SR)_2n, such as Pd₇(SR)₁₄, also indicates that the change in the CS signal of CH₂Cl₂ mixed with Pd₈(SR)₁₆ does not originate from CH₂Cl₂ existing in the outer deshielded area of Pd₈(SR)₁₆.

4. Conclusions

In summary, ¹H NMR spectroscopy is shown to be a highly effective approach to confirm the H-G structure of M_n(SR)_2n. First, the result acquired by ¹H NMR spectroscopy shows that the CS of the H nuclei within CH₂Cl₂ shifts to low field (approximately 0.07 ppm) and widens when CH₂Cl₂ is mixed with Pd₈(PET)₁₆ in a CDCl₃ solution. Then, the deshielded effect in the inner void of the double-crown Pd₈(SR)₁₆ was clarified via the NICS procedure, which indicated that the CS of the nucleus within a small molecule will theoretically move towards low field when the guest molecule inserts into the cluster. The faster transverse relaxation process between the hydrogen nuclei of the guest CH₂Cl₂ and host Pd₈(PET)₁₆ leads to a wider CS signal of the hydrogen nuclei within CH₂Cl₂. Both the experimental and theoretical results indicated that ¹H NMR spectroscopy is a powerful tool to probe the H-G property of M_n(SR)_2n. Finally, we find that Pd₈(SR)₁₆ is the only NC in the series of Pd_n(SR)_2n analogues that can capture a CH₂Cl₂ in its central cavity, showing that the H-G properties of M_n(SR)_2n are highly dependent on their cavity size, and a guest molecule just inserts into the matching cavity of M_n(SR)_2n. This research provides an alternative strategy to study the H-G properties of thiolate-protected tiara-like structural M_n(SR)_2n nanoclusters.