Supplementary Materials for

Discrimination of xylene isomers in a stacked coordination polymer

Liangying Li et al.

Corresponding authors: Zongbi Bao, [email protected]; Jing Li, [email protected]

Science 377, 335 (2022)

DOI: 10.1126/science.abj7659

The PDF file includes:

Materials and Methods

Figs. S1 to S67
Tables S1 to S6
References

Other Supplementary Material for this manuscript includes the following:

Data S1
 Materials and Methods
Materials
2,5-dihydroxy-1,4-benzoquinone (H2dhbq, 98%) was purchased from Alfa-Aesar Co. (USA).
Manganese acetate (Mn(CH3COO2)4H2O, 99%) was obtained from Sigma-Aldrich Co. (USA).
5 N2 (99.999%), He (99.999%), and CO2 (99.999%) were purchased from Jingong Co., Ltd (China).
Para-xylene (PX, anhydrous, 99%), meta-xylene (MX, 99%), and ortho-xylene (OX, 99%) were
purchased from Sigma-Aldrich Co. (USA). 1,4-Diethylbenzene (PDEB, >99.5%) was provided by
Macklin. All chemicals and gases were used without further purification.

10 Methods
Synthesis of Mn-dhbq
The Mn-dhbq samples were synthesized following the reported method with some
modifications, in which manganese sulfate was replaced by manganese acetate in order to
accelerate the rate of reaction (29, 30). H2dhbq (0.21 g, 1.5 mmol) and Mn(CH3COO2)4H2O, (0.37
15 g, 1.5 mmol) were mixed in deionized water (10 mL) and stirred overnight at room temperature.
After that, the Mn-dhbq samples were centrifuged and washed with deionized water several times
until the aqueous solution became clear. The sample was activated at 423 K for 24 h under
ultrahigh vacuum before single-component xylene vapor adsorption.

20 Scale-up synthesis of Mn-dhbq

H2dhbq (21.0 g, 150 mmol) and Mn(CH3COO2)4H2O, (36.5 g, 150 mmol) were added
into 1000 mL deionized water in a round-bottomed flask. The mixture was stirred at ambient
temperature and pressure for 48 h. Then, the product was collected and washed with deionized
water, followed by centrifuged until the aqueous solution was clear. Actually, the amount of the
25 ligand and metal salt can be increased relevantly as increasing the stirred time since the mole ratio
of the ligand (H2dhbq) and metal salt (Mn(CH3COO2)4H2O) keep the same.

Synthesis of Mn-dhbq molded pellets

In the process of molding, hydroxypropyl cellulose (HPC) was used as the binder. The as-
30 synthesized Mn-dhbq sample was dried at ambient conditions before being molded. A certain
amount of Mn-dhbq powder sample was mixed with 5 wt% of HPC as a binder material, and proper
amount of deionized water was added and mixed uniformly to form a paste (for example, 1.0 g
Mn-dhbq powder, ~0.6 mL deionized water, 50 mg HPC). The paste was partially dried at ambient
temperature and was formulated into pellets by hands. The diameter of the pellets was controlled
35 at 1~2 mm and the weight of each pellet was about 3~6 mg. The as-molded Mn-dhbq pellets were
placed on petri dishes and semi-dried at ambient temperature for about 12 h. The dried pellets were
then activated at 423 K for 24 h under ultrahigh vacuum before single-component xylene vapor
adsorption and breakthrough tests.

40 Determination of BET surface area

The BET surface areas of as-synthesized Mn-dhbq, Mn-dhbq after treatment under different
conditions, and Mn-dhbq molded pellets were characterized by N2 adsorption-desorption

2
 isotherms at 77 K. The adsorption-desorption isotherms were both obtained on a Micromeritics
ASAP 3Flex adsorption analyzer (Micromertics Instrument Corp., USA). About 120 mg samples
were degassed at 423 K for 24 h under ultrahigh vacuum before tests.

5 Characterizations of Mn-dhbq
Powder X-ray diffraction (PXRD) patterns of the samples were recorded with a Rigaku
Ultima-IV automated diffraction system (Rigaku Corp., Japan) using Cu K radiation ( = 1.5406
Å). The data were collected at ambient temperature in a 2θ range of 8-37.5o with a scan speed of
1o/min. The operating power was 40 kV/40 mA. In-situ X-ray diffraction (XRD) analysis of
10 powder samples was carried out on a Nano-inXider system (Xenocs, Sassenage, France) using Cu
Kα emission radiation (λ = 0.154 nm). The sample-to-detector distance is 79 mm resulting in the
effective range of the diffraction angle 2θ of 5–60°. The diffraction data was collected on a Pilatus
100 K semiconductor pixel detectors (Dectris, Swiss). Thermogravimetric analyses (TGA) of
samples were performed using the TA Instrument Q5000IR thermal gravimetric analyzer with N2
15 flow at 25 mL/min. About 4-6 mg of sample was loaded onto a platinum sample pan and heated
from ambient temperature to 973 K under N2 flow with a rate of 10 K/min.

Vapor kinetic adsorption measurements

Xylene adsorption kinetic experiments were performed on a homemade gravimetric
20 adsorption unit modified from a TGA Q50 thermogravimetric analyzer (TA Instruments).
Ultrahigh pure N2 (99.999%) was used as a carrier gas passing through a bubbler filled with single-
component liquid xylene. The partial pressure of hydrocarbon was controlled by adjusting the
blend ratio of pure N2 and N2 saturated with xylene vapor. Adsorbed amount was monitored by
the weight change of the sample and the adsorbed time for Mn-dhbq sample was monitored by the
25 computer. About 20 mg of Mn-dhbq samples were heated at 423 K under N2 flow for 2 h to remove
any residual water molecules. After cooling down to the adsorption temperature (303, 333, 363,
and 393 K), another N2 flow passing through a xylene bubbler connected to a condenser maintained
at 291 K was mixed with the pure N2 stream and then introduced to the adsorption chamber.
The pore diffusivity of PX was measured on a BELSORP-max II (BEL Japan, Inc.) automated
30 volumetric sorption analyzer and fitted automatically with BEL-Master software. About 100 mg
sample was degassed at 423 K under ultrahigh vacuum for 24 h before the experiment. The kinetic
adsorption characteristics were measured when the adsorption equilibrium data were collected.
After the adsorbate gas was introduced into the adsorption system at a given dose, the changes of
gas pressure and xylene uptake with time were recorded and converted into the transient
35 normalized uptake as a function of time.

Vapor adsorption measurements

The samples of Mn-dhbq and Mn-dhbq pellets were degassed at 423 K for 24 h under
ultrahigh vacuum prior to single-component xylene vapor adsorption measurements. The single-
40 component isotherms of para-xylene, meta-xylene, and ortho-xylene were measured at 303, 333,
363, and 393 K on a BELSORP-max II (BEL Japan, Inc.) equipped with a vapor dosing bottle.
Each sample tube was subsequently immersed in a temperature-controlled heating mantle that
surrounded most of the sample tube. The manifold of the instrument itself including the vapor
dosing bottle was heated to 323 K and kept at this temperature for all these single-component
45 xylene measurements.

3
 Multicomponent xylene vapor-phase breakthrough experiments
In a typical experiment, about 2.2-2.3 g of activated Mn-dhbq molded pellet sample was
packed into a stainless steel HPLC column (10 mm I.D.  50 mm). The column packing was
5 conducted in glove-box filled with Ar. A stream of N2 flow (20 mL/min) was introduced into the
column to further purge the pellets at 393 K for 24 h prior to measuring the vapor breakthrough
experiments. A mixture of ternary para-xylene, meta-xylene, and ortho-xylene was loaded into a
glass bubbler that connected to the sample column and N2 cylinder, respectively. N2 (99.999%)
was flowed through the bubbler at a rate of 100 mL/min, which was controlled by a mass flow
10 controller. The composition of the ternary components in the bubbler was adjusted until an
equimolar mixture was achieved in the vapor phase as detected by the gas chromatograph (GC).
Typically, the volume of the liquid ternary xylene isomers was about 17, 18 and 23 mL for p-para-,
meta-, and ortho-xylene, respectively. The xylene mixture from the bubbler was carried by N2 flow
through the sample column to the GC with a rate of 80 mL/min. The temperature of the sample
15 column was controlled through an oil bath at 303, 333, 363, and 393 K, respectively, to evaluate
the xylene isomers separation performance on the Mn-dhbq molded pellet samples at various
temperatures. The binary mixture of para-/meta-xylene, para-/ortho-xylene, and orhto-/meta-
xylene was also introduced with a constant flow rate of 80 mL/min and the sample column was
controlled through an oil bath at 303, 333, 363 and 393 K, respectively. The bubbler containing
20 binary xylene isomers was controlled at specific temperature to make the partial pressure of each
xylene component equal to that of the ternary mixtures. The outlet vapor from the sample column
was monitored using a GC-2010 Pro (SHIMADZU) gas chromatography with a flame ionization
detector (FID). The vapor mixture was separated by a capillary column (PEG-20M, 0.32  60
m) at 343 K. After every breakthrough experiment, the pelleted sample column was activated and
25 regenerated with a N2 flow at 423 K for 72-96 h.

Stability tests of Mn-dhbq

After being centrifuged and washed with deionized water for several times, the pure as-
synthesized Mn-dhbq powder samples were activated at 423 K for 24 h to remove all the water
30 molecules. The activated Mn-dhbq powder samples, about 0.3 g for each batch, were immersed in
10 mL deionized water and pure para-xylene solution, respectively, and then heated to 393 K in a
Teflon-lined reactor for 24 h. After cooling to ambient temperature, these two samples were
filtered and dried at room temperature. Additionally, the as-synthesized Mn-dhbq samples were
immersed in deionized water and exposed to air (relative humidity ~30%) for 7 days. All these
35 treated samples were dried at ambient temperature before PXRD measurements. The BET surface
areas of these samples were determined by the N2 adsorption-desorption isotherms at 77 K.
To better understand the structure changes of activated Mn-dhbq when it was exposed to
different xylene isomers, 0.1 g of the activated Mn-dhbq powder samples for each batch were
backfilled with single-component xylene vapor for about 24 h at 303, 333, 363, and 393 K,
40 respectively. The backfill xylene vapor was controlled by a Quantachrome Autosorb-iQ2 gas
sorption analyzer (USA) equipped with a vapor dosing bottle. The vapor dosing bottle was
controlled at 323 K to produce the vapor xylene. Then these samples were filtered and dried at
room temperature prior to PXRD tests.
The temperature-variable in-situ powder X-ray diffraction measurements were performed to
45 investigate thermal stability of Mn-dhbq. The activated samples were heated up to 433, 453, and

4
 473 K with a heating rate of 30 K/min, respectively, and kept at that temperature for various times.
The in-situ XRD data were collected at set temperature in the range of 5-60° (2θ) with a step size
of 0.02o.

5 Recyclability tests
To compare recyclability of Mn-dhbq and Cu-metallocycle, we performed consecutive PX
adsorption-desorption cycles. About 500 mg of freshly activated Mn-dhbq or Cu-metallocycle
samples were soaked in 10 mL PX liquid at 393 K for 1 h as the adsorption step. After that, the
samples were collected by filtering at the adsorption temperature and were allowed to dry on filter
10 paper for 3-4 hours in order to remove residual xylene adhered to the surface of samples. Then the
PX-adsorbed samples were held in a tube furnace under nitrogen atmosphere and heated from
room temperature to 473 K and kept for 30 minutes to fully desorb the PX. The adsorption-
desorption steps were repeated ten times. The mass fraction of PX adsorbed in each cycle was
measured by thermogravimetry by taking a very small portion of the PX-adsorbed samples. the
15 quantity of the PX desorbed was calculated as a difference between the mass of adsorbent before
and after the thermal treatment.

Nuclear Magnetic Resonance (NMR) of xylene-adsorbed Mn-dhbq

About 60 mg samples of pristine Mn-dhbq (the weight of activated sample was about 50 mg)
20 were placed into a sample tube of Quantachrome Autosorb-iQ2 gas sorption analyzer and
evacuated at 423 K for 24 h prior to xylene vapor backfilled. The liquid binary or ternary of xylene
mixtures were put into a vapor dosing bottle and the temperature was controlled at 323 K during
the backfilling process. To make sure the vapor concentration of each xylene isomer identical, for
each batch measurement, PX, 3.4 mL; MX, 3.6 mL; OX, 4.6 mL, was added into the vapor dosing
25 bottle. The sample tube was controlled at 333 K to measure the adsorption capacity of equimolar
binary vapor of MX/OX mixtures. And for the equimolar binary/ternary vapor of PX/OX, PX/MX,
and PX/MX/OX mixtures, the tested sample tube was controlled under the constant temperature
of 363 K. The total pressure (P0) was set as 10 torr, and the relative pressure was set to P/P0 = 0.6
for the backfill process. After being backfilled with the specific xylene vapor, each sample was
30 taken out from the sample tube and transferred into vial. In order to completely exchange the
adsorbed xylene isomers, 2 mL 36~38 wt% HCl was added into each batch vial to decompose all
these tested samples, and then these samples were soaked in 2 mL CDCl3 for 12 h, enabling the
corresponding xylene isomers were completely exchanged by CDCl3. The lower organic layers of
each samples were collected to measure 1H NMR spectra. It should be noted that the NMR spectra
35 of each other exhibit slightly shift due to the slightly different concentration of xylene isomers and
H+. To make it easier to compare all the spectra, the relevant peaks of xylene isomers were fixed
by setting the residue peak of TMS at 0 ppm.
To verify the feasibility, NMR spectra of pure unary and equimolar binary or ternary of xylene
isomers were first collected. About 50 mg activated Mn-dhbq sample was soaked in 2 mL 36~38
40 wt% HCl, and then 5 μL of each single xylene isomer was added into the vial and 2 mL CDCl 3
was used to make the xylene isomers exchanged from the inorganic solvent.

Gas chromatography analysis of liquid-phase xylene adsorption in Mn-dhbq

To accurately quantify the uptake of individual xylenes in the Mn-dhbq, the xylene-loaded
45 samples collected from the static solid-liquid extraction experiments was destroyed by

5
 concentrated hydrochloric acid (36~38 wt%) followed by CHCl3 extraction rather than directly
extracted by organic solvent because of strong interaction between xylenes and the framework.
The binary/ternary equimolar mixtures of xylene isomers were prepared by combining PX, MX
and/or OX in appropriate quantities weighed accurately using an analytical balance with a
5 precision of 0.1 mg. In a typical experiment, about 200 mg of freshly activated Mn-dhbq samples
were soaked in 6 g (~7 mL) of the binary or ternary mixtures and the sample was allowed to
equilibrate at a temperature of 303, 333, 363 and 393 K, respectively. After 1 h, the solid
adsorbents were collected by filtering at the adsorption temperature and were allowed to dry on
filter paper for 3-4 hours in order to remove xylene adhered to the surface of samples, after which
10 the xylene adsorbed Mn-dhbq samples were decomposed by 2 mL concentrated hydrochloric acid.
The destroyed Mn-dhbq samples were then immersed in 2 mL CHCl3 for 12 h, enabling the xylene
isomers to be completely extracted by CHCl3. The lower organic layer was collected for GC
measurements. The peak areas of individual xylene isomer shown in chromatograms were used to
calculate the selectivity.
15 In order to mimic the actual xylene mixture solutions under industrial conditions, 6 g of
quaternary mixture of xylene isomers (PX/MX/OX/EB = 22/22/50/6, w/w/w/w) was used to
evaluate the separation performance of Mn-dhbq, and treated with 200 mg of the activated Mn-
dhbq samples in a 10 mL glass spawn bottle and kept at 363 and 393 K for 1 h, respectively. After
that, the samples were collected by filtering at the adsorption temperature. Finally, the air-dried
20 Mn-dhbq samples was treated using the same procedure described above for the selectivity
measurements of xylene isomers.
To compare the performance of Cu-metallocycle and Mn-dhbq, the same experiments were
carried out on Cu-metallocycle under exactly the same conditions.

25 Adsorption heats of xylenes on Mn-dhbq

Adsorption heats of xylene isomers on Mn-dhbq were measured by differential scanning
calorimetry (DSC) on STD 650 (TA Instruments, Inc.). Runs at the thermobalance were performed
by feeding a flow stream of xylene diluted with nitrogen onto about 10 mg of evacuated sample at
303 K. Prior to measurements, the baseline was monitored under dry nitrogen flow at the same
30 temperature of 290 K, and then a nitrogen gas flow was introduced by bubbling the carrier gas in
a saturator containing single-component liquid xylene at a given temperature and the DSC signal
were recorded to obtain the heat of adsorption.

Adsorption selectivity calculation

35 The adsorption selectivity for xylenes separation, Sij, is defined by the following equation:
𝑥𝑖
⁄𝑦
𝑆𝑖𝑗 = 𝑥 𝑖
𝑗
⁄𝑦𝑗

where xi, xj are the equilibrated adsorption capacity of component i and j in the adsorbed phase,
40 respectively; and yi, yj are the molar fraction of component i and j in the vapor phase or liquid
phase. For equimolar mixtures, the selectivity can be simplified as

6
 𝑥𝑖
𝑆𝑖𝑗 =
𝑥𝑗
The ratio of xi/ xj can be derived from the integrated area ratio of corresponding methyl groups or
methylene group of individual C8 species in 1H-NMR spectra or the peak areas in GC
chromatograms. The calculated values of selectivity coefficients are summarized in Table S2 and
5 S5, respectively. It is worth noting that for the liquid-phase adsorption selectivities of the
quaternary mixture of PX/MX/OX/EB (22/22/250/6), the yi/yj must be determined by their GC
peak areas of individual component in the equilibrium solution.

The Antoine equation

10 For each batch of experiments, the saturated pressure of each xylene isomers was calculated
by the following equation:
𝐵
log(𝑃) = 𝐴 −
𝑇+𝐶
Where P is the pressure of pure component, mmHg. A, B and C were constants and the
physical property data can be found in various manuals. T was the temperature, oC. The values of
15 A, B and C were displayed in Table S6.

MicroED Mn-dhbq sample preparation

The as-synthesized Mn-dhbq samples were activated at 423 K under high vacuum for 24 h
prior to be dispersed in dried ethanol. About 40 mg activated Mn-dhbq samples were placed in a
20 glass vial containing 20 mL of dried ethanol. Similarly, the p-xylene-adsorbed Mn-dhbq samples
were also dispersed in pure p-xylene solution. Then, for each experiment, about 2 μL of the
supernatant was applied onto the surface of grid, and dried at room temperature for a few minutes.
The grid was subsequently placed into auto-loader cartridges under liquid nitrogen and loaded into
a Talos F200C cryo-TEM operated at 200 kV and equipped with a CETA 4 k × 4 k camera, which
25 was used for diffraction data collection.

Rietveld refinement details

To confirm the crystal structure of the as-synthesized sample, the Rietveld refinement was
performed based on the powder X-ray diffraction data by using the Fullprof software. Final
30 refinement on the positions of the thermal factors, occupancies, and lattice parameters, background,
and profiles converges with satisfactory R-factors. Besides, the Rietveld refinement was also
introduced for the activated and xylene-loaded Mn-dhbq in order to determine the structural
validity. Unfortunately, until recently our results have been uniformly disappointing.

35 DFT molecular simulation

All density functional theory (DFT) calculations were carried out using the DMol3 as
implemented in the Materials Studio package (35). DFT calculations were performed under the
generalized gradient approximation (GGA) with perdew-burke-ernzerhof (PBE) (36) and double
numerical pluspolarization (DNP) basis set (37). A semi-empirical dispersion correction was
40 included in the calculation to account for Van der Waals interactions. The boundary positions of
the model were saturated with hydrogen atoms. We have tried to resolve the crystal structure of
the activated sample with many technologies, but it is a long way to go. The activated Mn-dhbq

7
 framework contains stacks of 1D chains with high concentration of open metal sites and with no
hydrogen bonds, and we speculate favorable adsorption binding sites should be located among the
1D chains. Thus, we chose a segment of the 1D chain as the unit cluster to calculate the static
binding energy and adsorption sites between the xylene isomers and 1D chains. The convergence
5 tolerances with a fine quality (energy, gradient and displacement convergence were 2×10-4 Ha,
4×10-3 Ha Å-1, and 5×10-3 Å, respectively) were employed in the calculations. Each xylene isomer
was introduced to different locations of the chain to simulate the binding sites. The static binding
energy was calculated by the equation:

10 𝐸𝐵 = 𝐸(𝑎𝑑𝑠𝑜𝑟𝑏𝑒𝑛𝑡 + 𝑔𝑢𝑒𝑠𝑡) − 𝐸(𝑎𝑑𝑠𝑜𝑟𝑏𝑒𝑛𝑡) − 𝐸(𝑔𝑢𝑒𝑠𝑡)

8
 Fig. S1.
Microscope image of the as-synthesized Mn-dhbq crystal samples.

9
 Fig. S2.
Rietveld refinement of the powder X-ray diffraction data. Experimental (red, circles), calculated
5 (black, line), and the difference (blue, line below observed and calculated patterns) powder X-ray
diffraction profiles for the as-synthesized Mn-dhbq, measured at room temperature. The vertical
bars (green) indicate the calculated positions of Bragg peaks.

10
 Fig. S3.
The comparison of thermogravimetric (TG) curves of the as-synthesized and activated Mn-dhbq
samples under nitrogen atmosphere with a heat rate of 10 oC/min.
5

11
 150

Gas uptake (cm /g)

120

3 90
Activated at 323 K
60 Activated at 373 K
Activated at 423 K
Activated at 473 K
30

0
0.0 0.2 0.4 0.6 0.8 1.0
Relative pressure (p/p0)
0.30

0.25
dV/dw (cm3/gÅ)

0.20
Activated at 323 K
Activated at 373 K
0.15 Activated at 423 K
Activated at 473 K
0.10

0.05

0.00
3 6 9 12 15 18 21 24
Pore width (Å)
Fig. S4.
5 The N2 adsorption isotherms at 77 K and the pore distribution of Mn-dhbq samples activated at
various conditions. The pore distribution was determined by using the pore geometry of cylinder
model.

12
Fig. S5.
The color change of the Mn-dhbq sample before (left) and after (right) activation.

13
 Fig. S6.
The PXRD patterns of the pristine Mn-dhbq and samples under different conditions.
5

14
 As synthesized
Activated
PX-loaded
Intensity Re-activated

10 15 20 25 30 35 40
2 ()
Fig. S7.
The comparison of PXRD patterns of the as-synthesized, activated, PX-loaded, and reactivated
Mn-dhbq samples.
5

15
 Fig. S8.
The photograph of as-synthesized molded pellets of Mn-dhbq samples.
5

16
 120
Mn-dhbq pellets-adsorption
Desorption
90

N2 uptake (cm /g)

3 0.20

60 0.15

dV/dw (cm /(g·Å))

0.10

3
30 0.05

0.00
2.5 5.0 7.5 10.0 12.5 15.0
Pore width (Å)
0
0.0 0.2 0.4 0.6 0.8 1.0
Relative pressure (p/p0)
BET Report

BET surface area: 292.6512 ± 0.3556 m²/g

Slope: 0.014871 ± 0.000018 g/cm³ STP
Y-intercept: 0.000002 ± 0.000000 g/cm³ STP
C: 7600.351641
Qm: 67.2363 cm³/g STP
Correlation coefficient: 0.9999993
Molecular cross-sectional area: 0.1620 nm²

Fig. S9.
The N2 adsorption-desorption isotherms and the pore distribution of Mn-dhbq pellet samples at 77
5 K.

17
 200 303 K

Uptake amount (mg/g) 150

100 PX
MX
OX
50

0
0 300 600 900 1200
Pressure (Pa)
Fig. S10.
Single-component adsorption isotherms of p-xylene (PX), m-xylene (MX), and o-xylene (OX)
vapors on a Mn-dhbq sample at 303 K.
5

18
 200
333 K

Uptake amount (mg/g) 150

100 PX
OX
MX
50

0
0 300 600 900 1200
Pressure (Pa)
Fig. S11.
Single-component adsorption isotherms of p-xylene (PX), m-xylene (MX), and o-xylene (OX)
vapors on a Mn-dhbq sample at 333 K.
5

19
 160 363 K

Uptake amount (mg/g) 120

PX
MX
80 OX

40

0
0 300 600 900 1200
Pressure (Pa)
Fig. S12.
Single-component adsorption isotherms of p-xylene (PX), m-xylene (MX), and o-xylene (OX)
vapors on a Mn-dhbq sample at 363 K.
5

20
 160
393 K
Uptake amount (mg/g)
120

80 PX
MX
OX

40

0
0 300 600 900 1200
Pressure (Pa)
Fig. S13.
Single-component adsorption isotherms of p-xylene (PX), m-xylene (MX), and o-xylene (OX)
5 vapors on a Mn-dhbq sample at 393 K.

21
5 Fig. S14.
Single-component adsorption isotherms of p-xylene vapors on powder and molded pellet samples
of Mn-dhbq and Mn-dhbq at 303 K and 333 K.

22
 Fig. S15.
Single-component adsorption isotherms of o-xylene vapors on powder and molded pellet samples
5 of Mn-dhbq and Mn-dhbq at 303 K and 333 K.

23
 Fig. S16.
Single-component adsorption isotherms of m-xylene vapors on powder and molded pellet samples
5 of Mn-dhbq and Mn-dhbq at 303 K and 333 K.

24
 Fig. S17.
Magnified 1H NMR spectrum for the methyl groups of xylene isomer by decomposing Mn-dhbq
sample after adsorption of pure p-xylene at ambient conditions.
5

25
 Fig. S18.
Magnified 1H NMR spectrum for the methyl groups of xylene isomer by decomposing Mn-dhbq
sample after adsorption of pure m-xylene at ambient conditions.
5

26
 Fig. S19.
Magnified 1H NMR spectrum for the methyl groups of xylene isomer by decomposing Mn-dhbq
sample after adsorption of pure o-xylene at ambient conditions.
5

27
 Fig. S20.
Magnified 1H NMR spectrum for the methyl groups of xylene isomers by decomposing Mn-dhbq
sample after adsorption of a mixture of two xylene isomers (p-xylene : m-xylene = 1:1, liquid v/v)
5 at ambient conditions.

28
 Fig. S21.
Magnified 1H NMR spectrum for the methyl groups of xylene isomers by decomposing Mn-dhbq
sample after adsorption of a mixture of two xylene isomers (m-xylene : o-xylene = 1:1, liquid v/v)
5 at ambient conditions.

29
 Fig. S22.
Magnified 1H NMR spectrum for the methyl groups of xylene isomers by decomposing Mn-dhbq
sample after adsorption of a mixture of two xylene isomers (p-xylene : o-xylene = 1:1, liquid v/v)
5 at ambient conditions.

30
 Fig. S23.
Magnified 1H NMR spectrum for the methyl groups of xylene isomers by decomposing Mn-dhbq
sample after adsorption of a mixture of three xylene isomers (p-xylene : m-xylene : o-xylene =
5 1:1:1, liquid v/v/v) at ambient conditions.

31
 Fig. S24.
Magnified 1H NMR spectrum for the methyl groups of xylene isomers by Mn-dhbq after being
loaded with a mixture of two xylene isomers (p-xylene : m-xylene = 1:1, vapor pressure) at 363 K.
5 The vapor loaded sample was decomposed by 36~38 wt% 2 mL HCl solution and then exchanged
by using 2 mL CDCl3.

32
 Fig. S25.
Magnified 1H NMR spectrum for the methyl groups of xylene isomers by Mn-dhbq after being
loaded with a mixture of two xylene isomers (p-xylene : o-xylene = 1:1, vapor pressure) at 363 K.
5 The vapor loaded sample was decomposed by 36~38 wt% 2 mL HCl solution and then exchanged
by using 2 mL CDCl3.

33
 Fig. S26.
Magnified 1H NMR spectrum for the methyl groups of xylene isomers by Mn-dhbq after being
loaded with a mixture of two xylene isomers (m-xylene : o-xylene = 1:1, vapor pressure) at 333 K.
5 The vapor loaded sample was decomposed by 36~38 wt% 2 mL HCl solution and then exchanged
by using 2 mL CDCl3.

34
 Fig. S27.
Magnified 1H NMR spectrum for the methyl groups of xylene isomers by Mn-dhbq after being
loaded with a mixture of two xylene isomers (p-xylene : m-xylene = 1:1, vapor pressure) at 303 K.
5 The vapor loaded sample was decomposed by 36~38 wt% 2 mL HCl solution and then exchanged
by using 2 mL CDCl3.

35
 Fig. S28.
Magnified 1H NMR spectrum for the methyl groups of xylene isomers by Mn-dhbq after being
loaded with a mixture of three xylene isomers (p-xylene : m-xylene : o-xylene = 1:1:1, vapor
5 pressure) at 363 K. The vapor loaded sample was decomposed by 36~38 wt% 2 mL HCl solution
and then exchanged by using 2 mL CDCl3.

36
 (a)

(b)

(c)

(d)

Fig. S29.
Gas chromatograms used to quantify the selectivity coefficients of activated Mn-dhbq powders
after being soaked in the equimolar liquid mixtures of (a) PX/MX, (b) PX/OX, (c) MX/OX, and
5 (d) PX/MX/OX at 303 K, respectively. Note that minor amount of PX is present in the
commercially pure OX and MX, thus a small peak of PX is present in the cases involving MX/OX
mixtures.

37
 (a)

(b)

(c)

(d)

Fig. S30.
Gas chromatograms used to quantify the selectivity coefficients of activated Mn-dhbq powders
after being soaked in the equimolar liquid mixtures of (a) PX/MX, (b) PX/OX, (c) MX/OX, and
5 (d) PX/MX/OX at 333 K, respectively. Note that minor amount of PX is present in the
commercially pure OX and MX, thus a small peak of PX is present in the cases involving MX/OX
mixtures.

38
 Fig. S31.
Gas chromatograms used to quantify the selectivity coefficients of activated Mn-dhbq powders
after being soaked in the equimolar liquid mixtures of (a) PX/MX, (b) PX/OX, (c) MX/OX, (d)
5 PX/MX/OX, and (e) a quaternary mixture of PX/MX/OX/EB with molar ratio of 22:22:50:6 at
363 K, respectively. Note that minor amount of PX is present in the commercially pure OX and
MX, thus a small peak of PX is present in the cases involving MX/OX mixtures.

39
 Fig. S32.
Gas chromatograms used to quantify the selectivity coefficients of activated Mn-dhbq powders
after being soaked in the equimolar liquid mixtures of (a) PX/MX, (b) PX/OX, (c) MX/OX, (d)
5 PX/MX/OX, and (e) a quaternary mixture of PX/MX/OX/EB with molar ratio of 22:22:50:6 at
393 K, respectively. Note that minor amount of PX is present in the commercially pure OX and
MX, thus a small peak of PX is present in the cases involving MX/OX mixtures.

40
 (a)

(b)

(c)

(d)

Fig. S33.
Gas chromatograms used to quantify the selectivity coefficients of activated Cu-metallocycle
5 crystals after being soaked in the equimolar liquid mixtures of (a) PX/MX, (b) PX/OX, (c)
MX/OX, and (d) PX/MX/OX at 303 K, respectively. Note that minor amount of PX is present in
the commercially pure OX and MX, thus a small peak of PX is present in the cases involving
MX/OX mixtures.

10

41
 (a)

(b)

(c)

(d)

Fig. S34.
Gas chromatograms used to quantify the selectivity coefficients of activated Cu-metallocycle
5 crystals after being soaked in the equimolar liquid mixtures of (a) PX/MX, (b) PX/OX, (c)
MX/OX, and (d) PX/MX/OX at 333 K, respectively. Note that minor amount of PX is present in
the commercially pure OX and MX, thus a small peak of PX is present in the cases involving
MX/OX mixtures.

42
 Fig. S35.
Gas chromatograms used to quantify the selectivity coefficients of activated Cu-metallocycle
crystals after being soaked in the equimolar liquid mixtures of (a) PX/MX, (b) PX/OX, (c)
5 MX/OX, (d) PX/MX/OX, and (e) a quaternary mixture of PX/MX/OX/EB with molar ratio of
22:22:50:6 at 363 K, respectively. Note that minor amount of PX is present in the commercially
pure OX and MX, thus a small peak of PX is present in the cases involving MX/OX mixtures.

43
 Fig. S36.
Gas chromatograms used to quantify the selectivity coefficients of activated Cu-metallocycle
crystals after being soaked in the equimolar liquid mixtures of (a) PX/MX, (b) PX/OX, (c)
5 MX/OX, (d) PX/MX/OX, and (e) a quaternary mixture of PX/MX/OX/EB with molar ratio of
22:22:50:6 at 393 K, respectively. Note that minor amount of PX is present in the commercially
pure OX and MX, thus a small peak of PX is present in the cases involving MX/OX mixtures.

44
 (a) (b)

(c) (d)

Fig. S37.
Two-component vapor-phase breakthrough measurements for an equimolar mixture of p-xylene
(orange) and o-xylene (blue) vapor on Mn-dhbq molded pellets at (a) 303, (b) 333, (c) 363, and (d)
5 393 K. The temperature of the composition of these two components in the bubbler was controlled
at 313 K via water bath.

45
 (a) (b)

(c) (d)

Fig. S38.
Two-component vapor-phase breakthrough measurements for an equimolar mixture of p-xylene
(orange) and m-xylene (purple) vapor on Mn-dhbq molded pellets at 303, 333, 363, and 393 K.
5 The temperature of the composition of these two components in the bubbler was controlled at 303
K via water bath.

46
 (a) (b)

(c) (d)

Fig. S39.
Two-component vapor-phase breakthrough measurements for an equimolar mixture of m-xylene
(purple) and o-xylene (blue) vapor on Mn-dhbq molded pellets at 303, 333, 363, and 393 K. The
5 temperature of the composition of these two components in the bubbler was controlled at 313 K
via water bath.

47
 (a) (b)

(c) (d)

Fig. S40.
Multicomponent vapor-phase breakthrough measurements for an equimolar mixture of p-xylene
(orange), m-xylene (purple), and o-xylene (blue) vapor on Mn-dhbq molded pellets at 303, 333,
5 363, and 393 K. The temperature of the composition of these three components in the bubbler was
controlled at 323 K via water bath.

48
 Fig. S41.
Adsorption kinetic curves of p-xylene (PX), m-xylene (MX), and o-xylene (OX) on Mn-dhbq
samples at 303 K. The saturated vapor pressures of p-xylene, m-xylene, and o-xylene are 5.8, 5.4,
5 and 4.3 torr, respectively. And the tested vapor pressures of p-xylene, m-xylene, and o-xylene are
5.3, 5.0, and 4.0 torr, respectively.

49
 Fig. S42.
Adsorption kinetic curves of p-xylene (PX), m-xylene (MX), and o-xylene (OX) on Mn-dhbq
samples at 333 K. The saturated vapor pressures of p-xylene, m-xylene, and o-xylene are 5.8, 5.4,
5 and 4.3 torr, respectively. And the tested vapor pressures of p-xylene, m-xylene, and o-xylene are
5.3, 5.0, and 4.0 torr, respectively.

50
 Fig. S43.
Adsorption kinetic curves of p-xylene (PX), m-xylene (MX), and o-xylene (OX) on Mn-dhbq
samples at 363 K. The saturated vapor pressures of p-xylene, m-xylene, and o-xylene are 5.8, 5.4,
5 and 4.3 torr, respectively. And the tested vapor pressures of p-xylene, m-xylene, and o-xylene are
5.3, 5.0, and 4.0 torr, respectively.

51
 Fig. S44.
Adsorption kinetic curves of p-xylene (PX), m-xylene (MX), and o-xylene (OX) on Mn-dhbq
samples at 393 K. The saturated vapor pressures of p-xylene, m-xylene, and o-xylene are 5.8, 5.4,
5 and 4.3 torr, respectively. And the tested vapor pressures of p-xylene, m-xylene, and o-xylene are
5.3, 5.0, and 4.0 torr, respectively.

52
 Fig. S45.
The adsorption kinetic curves along with the calculated diffusion rate of PX in the Mn-dhbq at 303
5 and 393 K, fitted automatically with BEL-Master software according to the Crank theory (33).

53
 Fig. S46.
The adsorption kinetic curves along with the calculated diffusion rate of PX in the Cu-metallocycle
5 at 303 and 393 K, fitted automatically with BEL-Master software according to the Crank theory
(33).

54
 Activated
PX-vapor-303 K
PX-vapor-333 K
Intensity PX-vapor-363 K
PX-vapor-393 K

10 15 20 25 30 35 40
2 ()
Fig. S47.
The comparison of PXRD patterns of the activated Mn-dhbq sample and those after being loaded
5 with p-xylene at different temperatures.

55
 Activated
MX-vapor-303 K
MX-vapor-333 K
Intensity MX-vapor-363 K
MX-vapor-393 K

10 15 20 25 30 35 40
2 ()
Fig. S48.
The comparison of PXRD patterns of the activated Mn-dhbq sample and those after being loaded
5 with m-xylene at different temperatures.

56
 Fig. S49.
The comparison of PXRD patterns of the activated Mn-dhbq sample and those after being loaded
5 with o-xylene at different temperatures.

57
 Fig. S50.
The comparison of PXRD patterns of the activated Mn-dhbq sample and those after being loaded
with an equimolar mixture of p-/m-/o-xylene at different temperatures.
5

58
 (a) (b)

(c) (d)

Fig. S51.
The activated Mn-dhbq crystals yield high resolution diffraction patterns. (a) and (b) A
magnification search of the grid showed some broken within crystals. (c) and (d) Electron
diffraction patterns collected from different nanocrystals showed elongated and messy diffraction
5 spots.

59
 (a) (b)

Fig. S52.
The p-xylene-adsorbed Mn-dhbq crystals yield high resolution diffraction patterns. (a) A
magnification search of the grid showed some broken within crystals. (b) Electron diffraction
5 patterns collected from nanocrystal showed elongated and messy diffraction spots.

60
 Fig. S53.
Heat flow and heats of adsorption for xylene isomers in Mn-dhbq by DSC measurements. Note
that the adsorption heat of o-xylene is significantly lower than the theoretically predicted value
5 due to the large error as the adsorbed amount of o-xylene is small due to its low partial pressure
(1.2 kPa) generated at 290 K.

61
 Fig. S54.
5 The TG profile of the PX loaded Mn-dhbq sample.

62
 Fig. S55.
The TG profile of the MX loaded Mn-dhbq sample.
5

63
Fig. S56.
The TG profile of OX loaded Mn-dhbq sample.

64
Fig. S57.
The comparison of the PXRD patterns of Mn-dhbq materials with different treatments.

65
 Soaked in H2O for 1 week
150 Pristine sample-adsorption 150
Desorption
Desorption
N2 uptake (cm /g)

N2 uptake (cm /g)

120 120
3

3
0.25 0.5
90 0.20 90 0.4

dV/dw (cm /(g·Å))

dV/dw (cm /(g·Å))
0.15 0.3
Soaked in water for 1 week

60 3 60

3
0.10 0.2
0.05 0.1

30 0.00
30 0.0
2.5 5.0 7.5 10.0 12.5 15.0
2.5 5.0 7.5 10.0 12.5 15.0
Pore width (Å)
Pore width (Å)
0 0
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Relative pressure (p/p0) Relative pressure (p/p0)

150 Exposed to air for 1 week 150 Soaked in 393 K H2O for 24 h
Desorption Desorption
N2 uptake (cm /g)

N2 uptake (cm /g)

120 120
3

3
0.5
0.4
90 90 0.4

dV/dw (cm /(g·Å))

0.3
dV/dw (cm /(g·Å))

0.3

3
0.2
60 60 0.2
3

0.1 0.1

30 0.0
30 0.0

2.5 5.0 7.5 10.0 12.5 15.0 2.5 5.0 7.5 10.0 12.5 15.0
Pore width (Å) Pore width (Å)
0 0
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Relative pressure (p/p0) Relative pressure (p/p0)

150 Soaked in 393 K PX for 24 h 150 Soaked in 393 K H2O for 1 week
Desorption Desorption
N2 uptake (cm /g)

N2 uptake (cm /g)

120 120
3

0.5 0.5

90 0.4 90 0.4
dV/dw (cm /(g·Å))
dV/dw (cm /(g·Å))

0.3 0.3
3
3

60 0.2 60 0.2

0.1 0.1

30 0.0 30 0.0
2.5 5.0 7.5 10.0 12.5 15.0 2.5 5.0 7.5 10.0 12.5 15.0
Pore width (Å) Pore width (Å)
0 0
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Relative pressure (p/p0) Relative pressure (p/p0)

Fig. S58.
5 The comparison of N2 adsorption isotherms and the pore distribution of treated Mn-dhbq samples
at 77 K.

66
 Fig. S59.
5 PXRD patterns for the activated Mn-dhbq samples at 433, 453 and 473 K for various time
period. The results confirm that the crystallinity of Mn-dhbq is well retained at all temperatures
433 - 473 K (160 - 200 C).

67
 Fig. S60.
5 PXRD patterns for the activated Cu-metallocycle samples at 433, 453 and 473 K for various time
period. The results show that structure started to degrade at 453 K and lost its crystallinity totally
at 473 K.

68
 Fig. S61.
5 Uptake amount of p-xylene for the 20 consecutive adsorption cycles of Mn-dhbq (Adsorption
condition: 393 K, p-xylene vapor pressure 5.3 torr; Regeneration condition: 423 K, nitrogen flow
for 2 h for each batch of measurement).

69
 Fig. S62.
P-xylene adsorption-desorption recyclability test on Mn-dhbq and the uptake amount of p-xylene
for the 20 consecutive adsorption cycles of Mn-dhbq remains nearly identical (Adsorption
5 condition: 393 K, p-xylene vapor pressure 5.3 torr; Regeneration condition: 423 K, nitrogen flow
for 2 h for each batch of measurement).

70
 Fig. S63.
The PX adsorption-desorption recyclability tests on Mn-dhbq and Cu-metallocycle samples. The
results show that Mn-dhbq well retains the adsorption capacity after ten cycles and Cu-
5 metallocycle essentially lost its adsorption capacity after the first cycle.

71
 (a)

(b)

Fig. S64.
The PXRD patterns of (a) Mn-dhbq and (b) Cu-metallocycle after PX adsorption-desorption cycles
5 compared with those of the simulated, as-synthesized and activated samples. The result confirms
that Mn-dhbq can maintain its crystallinity after ten cycles while Cu-metallocycle lost its
crystallinity after the first cycle.

72
 Fig. S65.
Gas chromatogram of elution solution of 1,4-diethylbenzene, which was used to extract PX at 433
K from Mn-dhbq after adsorption of a quaternary mixture of PX/MX/OX/EB (22:22:50:6) at 393
5 K for 1h.

73
 Fig. S66.
Comparison of PXRD patterns of simulated (red), as-synthesized (orange), activated (gray), scale-
up (blue), and molded pellets (green) of Mn-dhbq samples.
5

74
 Fig. S67.
5 The comparison of PXRD patterns of activated Mn-dhbq powder and Mn-dhbq molded pellet
samples after being loaded with p-xylene at different temperature.

75
 Table S1. The physical properties and dimensions of xylene isomers (38, 39).

Boiling Melting Kinetic Molecular dimensional

Xylene Polarizability
point point diameter (Å) z/x
isomer  1025/cm3
(oC) (oC) (Å) x y z
p-xylene 138.38 13.2 5.8 6.62 3.81 9.15 1.38 137-149
o-xylene 144.44 -25.2 6.8 7.27 3.83 7.83 1.08 141-149
m-xylene 139.19 -47.9 6.8 8.99 3.95 7.32 1.23 142

76
 Table S2. Summary of reported adsorbents for adsorptive separation of xylene isomers.
Temperature Selectivity
Adsorbents Refs.
(K) PX/OX PX/MX MX/OX
12.5
363 37.9 34.8 This worka
Mn-dhbq (333 K)
393 66.8 25.5 63.4 This workd
AgLClO4 298 20.3 5.4 1.15 (9)a
MIL-125(Ti)-NH2 298 2.2 3.0 1.03 (40)b
MIL-140B 323 1.8 1.6 - (41)b
ZIF-8 398 3.9 1.6 2.4 (42)b
MCF-50 453 1.6 1.3 1.3 (43)c
[Ce(HTCPB)] 298 5.7 4.6 1.2 (44)d
EtP5 298 1.9 2.1 1.1 (8)d
EtP6 298 14.3 10.2 1/3.8 (8)d
BaX nanosize 423 2.82 7.19 - (45)b
KX nanosize 423 2.43 5.36 - (46)b
ZSM-5 (H/Li/Na/K) 403 3.9-16.8 4.0-25.0 - (47)b
MOF-monoclinic 393 4.6 2.5 - (48)b
MFI zeolite 298 104 - - (7)d
Zn-MOF 298 50.0 11.8 4.3 (27)a
Cu-metallocycle 293 51.6 65.7 37.5 (12)d
AZO-Cage 298 15.6 10.9 9.7 (13)a
OX-selective adsorbents
SAMM-3-Cu-OTf 298 1/23.1 - 1/6.1 (49)a
sql-1-Co-NCS 293 1/9.6 1.5 1/7.5 (26)d
MIL-101(Cr) 408 1/1.6 1/1.1 1/1.5 (50, 51)c
UiO-66 313 1/2.4 - 1/1.8 (52, 53)b
HKUST-1 398 1/1.2 1/1.1 1/1.1 (54)b
Zn(BDC)(Dabco)0.5 448 1/1.88 1/1.25 1/1.12 (55)b
CPO-27-Ni 398 1/3.3 1/2.0 1/1.7 (54)b
Co2(dobdc) 306 1/3.9 1/1.6 1/2.5 (14)d
[Ni(NCS)2(ppp)4] 295 1/40.5 1/12.7 1/34.2 (11)a
MIL-47(V) 343 1/1.4 2.5 1/2 (15, 56)b
MIL-53(Fe/Al) 293 1/3.5-1/2.5 - 1/2.7-1/1.6 (25, 57-59)b
CAU-13 298 1/1.5 1.3 1/1.9 (60)d
MFM-300(In/V/Fe) 293 1/3.0-1/1.3 1/3.7-1/3.5 1.1-2.3 (17)b
ZU-61 398 1/2.6 1/2.9 - (28)b
For each material, only the highest/best values of selectivity were selected, regardless of whether the experiments were conducted
with adsorbate in liquid or vapor phase.
a Calculated from vapor-phase sorption experiments. b Calculated from experimental breakthrough curves.
c Calculated from chromatographic separation data. d Calculated from batch solid-liquid sorption experiments.
5

77
Table S3. Crystallographic data of as-synthesized Mn-dhbq compound obtained from the
Rietveld refinement based on the powder X-ray diffraction data.
As-synthesized
Formula C12H12O12Mn2 (2[Mn(C6H2O4)·2H2O])
Formula weight 458.091
Crystal system 3D Monoclinic-B
Space group C2/m
a/Å 6.9401(9)
b/Å 7.7563(11)
c/Å 8.1148(9)
/ o 90
/ o 113.855(11)
/ o 90
3
Volume/Å 399.503(96)
calc g/cm3 1.904
Gof 1.354
Radiation CuK

78
 Table S4. The comparison of BET surface areas and mean pore size of as-synthesized and
treated Mn(II)-dhbq samples.

BET surface areas Mean pore size a

Conditions
(m2/g) (Å)
Activated at 323 K 408.8 ± 0.6 5.4
Activated at 373 K 416.2 ± 0.5 5.4
Activated at 423 K 428.9 ± 0.6 5.6
Activated at 473 K 434.3 ± 1.5 5.4
Soaked in H2O for 1 week 436.5 ± 0.4 5.5
Exposed to air for 1 week 421.2 ± 0.5 5.6
Boiled in H2O for 24 h 466.6 ± 1.8 5.1
Boiled in H2O for 1 week 458.0 ± 2.3 5.4
Boiled in 393 K PX for 24 h 439.4 ± 1.2 5.4
Mn-dhbq pellets 292.7±0.4 5.3
a
The mean pore sizes of treated Mn(II)-dhbq samples are calculated by the Horvath-Kawazoe
5 (H-K) model.

79
Table S5. Liquid-phase adsorption selectivities for the equimolar binary/ternary and quaternary
xylene mixtures on Mn-dhbq and Cu-metallocycle.

Selectivity
Temperature Xylenes
Mn-dhbq Cu-metallocycle
PX/MX 18.2 150.3
303 K PX/OX 76.9 443.3
(30 C) MX/OX 47.2 5.2
PX/MX/OX 153.8/6.8/1 600.0/1.5/1
PX/MX 18.0 277.1
333 K PX/OX 74.5 263.2
(60 C) MX/OX 50.7 27.1
PX/MX/OX 114.2/5.4/1 158.5/1.1/1
PX/MX 16.2 70.8
PX/OX 65.0 73.3
363 K MX/OX 42.3 24.3
(90 C) PX/MX/OX 78.4/3.8/1 45.4/1.4/1
PX/MX/OX/EB 153.1/7.1/1/23.0 86.3/2.4/1/4.9
(22/22/50/6)
PX/MX 25.5 30.4
PX/OX 66.8 41.3
393 K MX/OX 63.4 12.8
(120 C) PX/MX/OX 84.6/2.5/1 19.4/1.6/1
PX/MX/OX/EB 171.1/4.9/1/18.4 61.7/4.2/1/5.8
(22/22/50/6)

80
Table S6. The values of A, B and C for xylene isomers from the Antoine equation.

Compound A B C
o-xylene 6.99891 1474.679 213.686
m-xylene 7.00908 1462.266 215.105
p-xylene 6.99052 1453.430 215.307

81
References and Notes
1. M. Minceva, A. E. Rodrigues, Understanding and revamping of industrial scale SMB units for
p-xylene separation. AIChE J. 53, 138–149 (2007). doi:10.1002/aic.11062
2. G. Zhang, Y. Ding, A. Hashem, A. Fakim, N. M. Khashab, Xylene isomer separations by
intrinsically porous molecular materials. Cell. Rep. Phys. Sci. 2, 100470 (2021).
doi:10.1016/j.xcrp.2021.100470
3. D. S. Sholl, R. P. Lively, Seven chemical separations to change the world. Nature 532, 435–
437 (2016). doi:10.1038/532435a Medline
4. Y. Yang, P. Bai, X. Guo, Separation of xylene isomers: A review of recent advances in
materials. Ind. Eng. Chem. Res. 56, 14725–14753 (2017). doi:10.1021/acs.iecr.7b03127
5. L. S. Cheng, J. A. Johnson, “Adsorbents with improved mass transfer properties and their use
in the adsorptive separation of para-xylene,” US Patent 8,609,925 (2013).
6. B. Van de Voorde, B. Bueken, J. Denayer, D. De Vos, Adsorptive separation on metal-organic
frameworks in the liquid phase. Chem. Soc. Rev. 43, 5766–5788 (2014).
doi:10.1039/C4CS00006D Medline
7. G.-Q. Guo, H. Chen, Y.-C. Long, Separation of p-xylene from C8 aromatics on binder-free
hydrophobic adsorbent of MFI zeolite. I. Studies on static equilibrium. Microporous
Mesoporous Mater. 39, 149–161 (2000). doi:10.1016/S1387-1811(00)00191-8
8. K. Jie, M. Liu, Y. Zhou, M. A. Little, A. Pulido, S. Y. Chong, A. Stephenson, A. R. Hughes,
F. Sakakibara, T. Ogoshi, F. Blanc, G. M. Day, F. Huang, A. I. Cooper, Near-ideal
xylene selectivity in adaptive molecular pillar[n]arene crystals. J. Am. Chem. Soc. 140,
6921–6930 (2018). doi:10.1021/jacs.8b02621 Medline
9. N. Sun, S.-Q. Wang, R. Zou, W.-G. Cui, A. Zhang, T. Zhang, Q. Li, Z.-Z. Zhuang, Y.-H.
Zhang, J. Xu, M. J. Zaworotko, X.-H. Bu, Benchmark selectivity p-xylene separation by
a non-porous molecular solid through liquid or vapor extraction. Chem. Sci. 10, 8850–
8854 (2019). doi:10.1039/C9SC02621E Medline
10. G. W. Zhang, A.-H. Emwas, U. F. Shahul Hameed, S. T. Arold, P. Yang, A. Chen, J.-F.
Xiang, N. M. Khashab, Shape-induced selective separation of ortho-substituted benzene
isomers enabled by cucurbit[7]uril host macrocycles. Chem 6, 1082–1096 (2020).
doi:10.1016/j.chempr.2020.03.003
11. M. Lusi, L. J. Barbour, Solid-vapor sorption of xylenes: Prioritized selectivity as a means of
separating all three isomers using a single substrate. Angew. Chem. Int. Ed. 51, 3928–
3931 (2012). doi:10.1002/anie.201109084 Medline
12. M. du Plessis, V. I. Nikolayenko, L. J. Barbour, Record-setting selectivity for p-xylene by an
intrinsically porous zero-dimensional metallocycle. J. Am. Chem. Soc. 142, 4529–4533
(2020). doi:10.1021/jacs.9b11314 Medline
13. B. Moosa, L. O. Alimi, A. Shkurenko, A. Fakim, P. M. Bhatt, G. Zhang, M. Eddaoudi, N. M.
Khashab, A polymorphic azobenzene cage for energy-efficient and highly selective p-
xylene separation. Angew. Chem. Int. Ed. 59, 21367–21371 (2020).
doi:10.1002/anie.202007782 Medline
14. M. I. Gonzalez, M. T. Kapelewski, E. D. Bloch, P. J. Milner, D. A. Reed, M. R. Hudson, J.
A. Mason, G. Barin, C. M. Brown, J. R. Long, Separation of xylene isomers through
multiple metal site interactions in metal-organic frameworks. J. Am. Chem. Soc. 140,
3412–3422 (2018). doi:10.1021/jacs.7b13825 Medline
15. V. Finsy, H. Verelst, L. Alaerts, D. De Vos, P. A. Jacobs, G. V. Baron, J. F. M. Denayer,
Pore-filling-dependent selectivity effects in the vapor-phase separation of xylene isomers
on the metal–organic framework MIL-47. J. Am. Chem. Soc. 130, 7110–7118 (2008).
doi:10.1021/ja800686c Medline
16. A. Torres-Knoop, R. Krishna, D. Dubbeldam, Separating xylene isomers by commensurate
stacking of p-xylene within channels of MAF-X8. Angew. Chem. Int. Ed. 53, 7774–7778
(2014). doi:10.1002/anie.201402894 Medline
17. X. Li, J. Wang, N. Bai, X. Zhang, X. Han, I. da Silva, C. G. Morris, S. Xu, D. M. Wilary, Y.
Sun, Y. Cheng, C. A. Murray, C. C. Tang, M. D. Frogley, G. Cinque, T. Lowe, H. Zhang,
A. J. Ramirez-Cuesta, K. M. Thomas, L. W. Bolton, S. Yang, M. Schröder, Refinement
of pore size at sub-angstrom precision in robust metal-organic frameworks for separation
of xylenes. Nat. Commun. 11, 4280 (2020). doi:10.1038/s41467-020-17640-4 Medline
18. J. M. Holcroft, K. J. Hartlieb, P. Z. Moghadam, J. G. Bell, G. Barin, D. P. Ferris, E. D.
Bloch, M. M. Algaradah, M. S. Nassar, Y. Y. Botros, K. M. Thomas, J. R. Long, R. Q.
Snurr, J. F. Stoddart, Carbohydrate-mediated purification of petrochemicals. J. Am.
Chem. Soc. 137, 5706–5719 (2015). doi:10.1021/ja511878b Medline
19. L. Alaerts, C. E. A. Kirschhock, M. Maes, M. A. van der Veen, V. Finsy, A. Depla, J. A.
Martens, G. V. Baron, P. A. Jacobs, J. F. M. Denayer, D. E. De Vos, Selective adsorption
and separation of xylene isomers and ethylbenzene with the microporous vanadium(IV)
terephthalate MIL-47. Angew. Chem. Int. Ed. 46, 4293–4297 (2007).
doi:10.1002/anie.200700056 Medline
20. L. Yang, H. Liu, D. Yuan, J. Xing, Y. Xu, Z. Liu, Efficient separation of xylene isomers by a
pillar-layer metal-organic framework. ACS Appl. Mater. Interfaces 13, 41600–41608
(2021). doi:10.1021/acsami.1c10462 Medline
21. D. D. Zhou, P. Chen, C. Wang, S.-S. Wang, Y. Du, H. Yan, Z.-M. Ye, C.-T. He, R.-K.
Huang, Z.-W. Mo, N.-Y. Huang, J.-P. Zhang, Intermediate-sized molecular sieving of
styrene from larger and smaller analogues. Nat. Mater. 18, 994–998 (2019).
doi:10.1038/s41563-019-0427-z Medline
22. H. Wang, X. Dong, E. Velasco, D. H. Olson, Y. Han, J. Li, One-of-a-kind: A microporous
metal–organic framework capable of adsorptive separation of linear, mono- and di-
branched alkane isomers via temperature- and adsorbate-dependent molecular sieving.
Energy Environ. Sci. 11, 1226–1231 (2018). doi:10.1039/C8EE00459E
23. P. J. Bereciartua, Á. Cantín, A. Corma, J. L. Jordá, M. Palomino, F. Rey, S. Valencia, E. W.
Corcoran Jr., P. Kortunov, P. I. Ravikovitch, A. Burton, C. Yoon, Y. Wang, C. Paur, J.
Guzman, A. R. Bishop, G. L. Casty, Control of zeolite framework flexibility and pore
topology for separation of ethane and ethylene. Science 358, 1068–1071 (2017).
doi:10.1126/science.aao0092 Medline
24. C. Gu, N. Hosono, J.-J. Zheng, Y. Sato, S. Kusaka, S. Sakaki, S. Kitagawa, Design and
control of gas diffusion process in a nanoporous soft crystal. Science 363, 387–391
(2019). doi:10.1126/science.aar6833 Medline
25. L. Alaerts, M. Maes, L. Giebeler, P. A. Jacobs, J. A. Martens, J. F. M. Denayer, C. E. A.
Kirschhock, D. E. De Vos, Selective adsorption and separation of ortho-substituted
alkylaromatics with the microporous aluminum terephthalate MIL-53. J. Am. Chem. Soc.
130, 14170–14178 (2008). doi:10.1021/ja802761z Medline
26. S. Q. Wang, S. Mukherjee, E. Patyk-Kaźmierczak, S. Darwish, A. Bajpai, Q.-Y. Yang, M. J.
Zaworotko, Highly selective, high-capacity separation of o-xylene from C8 aromatics by
a switching adsorbent layered material. Angew. Chem. Int. Ed. 58, 6630–6634 (2019).
doi:10.1002/anie.201901198 Medline
27. S. Mukherjee, B. Joarder, B. Manna, A. V. Desai, A. K. Chaudhari, S. K. Ghosh,
Framework-flexibility driven selective sorption of p-xylene over other isomers by a
dynamic metal-organic framework. Sci. Rep. 4, 5761 (2014). doi:10.1038/srep05761
Medline
28. X. Cui, Z. Niu, C. Shan, L. Yang, J. Hu, Q. Wang, P. C. Lan, Y. Li, L. Wojtas, S. Ma, H.
Xing, Efficient separation of xylene isomers by a guest-responsive metal-organic
framework with rotational anionic sites. Nat. Commun. 11, 5456 (2020).
doi:10.1038/s41467-020-19209-7 Medline
29. S. Morikawa, T. Yamada, H. Kitagawa, Crystal structure and proton conductivity of a one-
dimensional coordination polymer, {Mn(DHBQ)(H2O)2}. Chem. Lett. 38, 654–655
(2009). doi:10.1246/cl.2009.654
30. T. Yamada, S. Morikawa, H. Kitagawa, Structures and proton conductivity of one-
dimensional M(dhbq)·nH2O (M = Mg, Mn, Co, Ni, and Zn, H2(dhbq) = 2,5-dihydroxy-
1,4-benzoquinone) promoted by connected hydrogen-bond networks with absorbed
water. Bull. Chem. Soc. Jpn. 83, 42–48 (2010). doi:10.1246/bcsj.20090216
31. R. W. Neuzil, “Aromatic hydrocarbon separation by adsorption,” US Patent 3,558,730
(1971).
32. D. M. Polyukhov, A. S. Poryvaev, A. S. Sukhikh, S. A. Gromilov, M. V. Fedin, Fine-tuning
window apertures in ZIF-8/67 frameworks by metal ions and temperature for high-
efficiency molecular sieving of xylenes. ACS Appl. Mater. Interfaces 13, 40830–40836
(2021). doi:10.1021/acsami.1c12166 Medline
33. J. Crank, The Mathematics of Diffusion (Oxford Univ. Press, 1975).
34. M. Minceva, A. E. Rodrigues, Adsorption of xylenes on faujasite-type zeolite: Equilibrium
and kinetics in batch adsorber. Chem. Eng. Res. Des. 82, 667–681 (2004).
doi:10.1205/026387604323142739
35. B. Delley, From molecules to solids with the DMol3 approach. J. Chem. Phys. 113, 7756–
7764 (2000). doi:10.1063/1.1316015
36. G. Kresse, J. Furthmüller, Efficient iterative schemes for ab initio total-energy calculations
using a plane-wave basis set. Phys. Rev. B 54, 11169–11186 (1996).
doi:10.1103/PhysRevB.54.11169 Medline
37. W. J. Hehre, R. Ditchfield, J. A. Pople, Self-consistent molecular orbital methods. XII.
Further extensions of gaussian-type basis sets for use in molecular orbital studies of
organic molecules. J. Chem. Phys. 56, 2257–2261 (1972). doi:10.1063/1.1677527
38. J. R. Li, R. J. Kuppler, H. C. Zhou, Selective gas adsorption and separation in metal-organic
frameworks. Chem. Soc. Rev. 38, 1477–1504 (2009). doi:10.1039/b802426j Medline
39. C. E. Webster, R. S. Drago, M. C. Zerner, Molecular dimensions for adsorptives. J. Am.
Chem. Soc. 120, 5509–5516 (1998). doi:10.1021/ja973906m
40. F. Vermoortele, M. Maes, P. Z. Moghadam, M. J. Lennox, F. Ragon, M. Boulhout, S.
Biswas, K. G. M. Laurier, I. Beurroies, R. Denoyel, M. Roeffaers, N. Stock, T. Düren, C.
Serre, D. E. De Vos, p-Xylene-selective metal–organic frameworks: A case of topology-
directed selectivity. J. Am. Chem. Soc. 133, 18526–18529 (2011). doi:10.1021/ja207287h
Medline
41. J. A. Gee, K. Zhang, S. Bhattacharyya, J. Bentley, M. Rungta, J. S. Abichandani, D. S. Sholl,
S. Nair, Computational identification and experimental evaluation of metal-organic
frameworks for xylene enrichment. J. Phys. Chem. C 120, 12075–12082 (2016).
doi:10.1021/acs.jpcc.6b03349
42. D. Peralta, G. Chaplais, J.-L. Paillaud, A. Simon-Masseron, K. Barthelet, G. D. Pirngruber,
The separation of xylene isomers by ZIF-8: A demonstration of the extraordinary
flexibility of the ZIF-8 framework. Microporous Mesoporous Mater. 173, 1–5 (2013).
doi:10.1016/j.micromeso.2013.01.012
43. J. M. Lin, C. T. He, P. Q. Liao, R. B. Lin, J. P. Zhang, Structural, energetic, and dynamic
insights into the abnormal xylene separation behavior of hierarchical porous crystal. Sci.
Rep. 5, 11537 (2015). doi:10.1038/srep11537 Medline
44. J. E. Warren, C. G. Perkins, K. E. Jelfs, P. Boldrin, P. A. Chater, G. J. Miller, T. D. Manning,
M. E. Briggs, K. C. Stylianou, J. B. Claridge, M. J. Rosseinsky, Shape selectivity by
guest-driven restructuring of a porous material. Angew. Chem. Int. Ed. 53, 4592–4596
(2014). doi:10.1002/anie.201307656 Medline
45. M. Rasouli, N. Yaghobi, F. Allahgholipour, H. Atashi, Para-xylene adsorption separation
process using nano-zeolite BaX. Chem. Eng. Res. Des. 92, 1192–1199 (2014).
doi:10.1016/j.cherd.2013.10.008
46. M. Rasouli, N. Yaghobi, S. Z. Movassaghi Gilani, H. Atashi, M. Rasouli, Influence of
monovalent alkaline metal cations on binder-free nano-zeolite X in para-xylene
separation. Chin. J. Chem. Eng. 23, 64–70 (2015). doi:10.1016/j.cjche.2014.11.005
47. M. Rasouli, N. Yaghobi, S. Chitsazan, M. H. Sayyar, Influence of monovalent cations ion-
exchange on zeolite ZSM-5 in separation of para-xylene from xylene mixture.
Microporous Mesoporous Mater. 150, 47–54 (2012).
doi:10.1016/j.micromeso.2011.09.013
48. Z. Gu, D. Jiang, H. Wang, X. Cui, X. Yan, Adsorption and separation of xylene isomers and
ethylbenzene on two Zn-terephthalate metal-organic frameworks. J. Phys. Chem. C 114,
311–316 (2010). doi:10.1021/jp9063017
49. A. M. Kałuża, S. Mukherjee, S. Q. Wang, D. J. O’Hearn, M. J. Zaworotko, [Cu(4-
phenylpyridine)4(trifluoromethanesulfonate)2], a Werner complex that exhibits high
selectivity for o-xylene. Chem. Commun. 56, 1940–1943 (2020).
doi:10.1039/C9CC09525J Medline
50. Z. Y. Gu, X. P. Yan, Metal–organic framework MIL-101 for high-resolution gas-
chromatographic separation of xylene isomers and ethylbenzene. Angew. Chem. Int. Ed.
49, 1477–1480 (2010). doi:10.1002/anie.200906560 Medline
51. P. Trens, H. Belarbi, C. Shepherd, P. Gonzalez, N. A. Ramsahye, U.-H. Lee, Y.-K. Seo, J.-S.
Chang, Adsorption and separation of xylene isomers vapors onto the chromium
terephthalate-based porous material MIL-101(Cr): An experimental and computational
study. Microporous Mesoporous Mater. 183, 17–22 (2014).
doi:10.1016/j.micromeso.2013.08.040
52. P. S. Bárcia, D. Guimarães, P. A. P. Mendes, J. A. C. Silva, V. Guillerm, H. Chevreau, C.
Serre, A. E. Rodrigues, Reverse shape selectivity in the adsorption of hexane and xylene
isomers in MOF UiO-66. Microporous Mesoporous Mater. 139, 67–73 (2011).
doi:10.1016/j.micromeso.2010.10.019
53. M. A. Moreira, J. C. Santos, A. F. P. Ferreira, J. M. Loureiro, F. Ragon, P. Horcajada, K.-E.
Shim, Y.-K. Hwang, U.-H. Lee, J.-S. Chang, C. Serre, A. E. Rodrigues, Reverse shape
selectivity in the liquid-phase adsorption of xylene isomers in zirconium terephthalate
MOF UiO-66. Langmuir 28, 5715–5723 (2012). doi:10.1021/la3004118 Medline
54. D. Peralta, K. Barthelet, J. Pérez-Pellitero, C. Chizallet, G. Chaplais, A. Simon-Masseron, G.
D. Pirngruber, Adsorption and separation of xylene isomers: CPO-27-Ni vs HKUST-1 vs
NaY. J. Phys. Chem. C 116, 21844–21855 (2012). doi:10.1021/jp306828x
55. M. P. M. Nicolau, P. S. Bárcia, J. M. Gallegos, J. A. C. Silva, A. E. Rodrigues, B. Chen,
Single- and multicomponent vapor-phase adsorption of xylene isomers and ethylbenzene
in a microporous metal–organic framework. J. Phys. Chem. C 113, 13173–13179 (2009).
doi:10.1021/jp9006747
56. J. M. Castillo, T. J. H. Vlugt, S. Calero, Molecular simulation study on the separation of
xylene isomers in MIL-47 metal–organic frameworks. J. Phys. Chem. C 113, 20869–
20874 (2009). doi:10.1021/jp908247w
57. M. Agrawal, S. Bhattacharyya, Y. Huang, K. C. Jayachandrababu, C. R. Murdock, J. A.
Bentley, A. Rivas-Cardona, M. M. Mertens, K. S. Walton, D. S. Sholl, S. Nair, Liquid-
phase multicomponent adsorption and separation of xylene mixtures by flexible MIL-53
adsorbents. J. Phys. Chem. C 122, 386–397 (2018). doi:10.1021/acs.jpcc.7b09105
58. R. El Osta, A. Carlin-Sinclair, N. Guillou, R. I. Walton, F. Vermoortele, M. Maes, D. de Vos,
F. Millange, Liquid-phase adsorption and separation of xylene isomers by the flexible
porous metal-organic framework MIL-53(Fe). Chem. Mater. 24, 2781–2791 (2012).
doi:10.1021/cm301242d
59. V. Finsy, C. E. A. Kirschhock, G. Vedts, M. Maes, L. Alaerts, D. E. De Vos, G. V. Baron, J.
F. M. Denayer, Framework breathing in the vapour-phase adsorption and separation of
xylene isomers with the metal–organic framework MIL-53. Chemistry 15, 7724–7731
(2009). doi:10.1002/chem.200802672 Medline
60. F. Niekiel, J. Lannoeye, H. Reinsch, A. S. Munn, A. Heerwig, I. Zizak, S. Kaskel, R. I.
Walton, D. de Vos, P. Llewellyn, A. Lieb, G. Maurin, N. Stock, Conformation-controlled
sorption properties and breathing of the aliphatic Al-MOF [Al(OH)(CDC)]. Inorg. Chem.
53, 4610–4620 (2014). doi:10.1021/ic500288w Medline