Crystal structure of 2,5-di­hy­droxy­terephthalic acid from powder diffraction data

The crystal structure of anhydrous 2,5-dhy­droxy­terephthalic acid, C₈H₆O₆, was solved and refined using laboratory X-ray powder diffraction data, and optimized using density functional techniques. The published structure of 2,5-di­hydroxy­terephthalic acid dihydrate was also optimized. The carb­oxy­lic acid groups form strong hydrogen bonds, which form centrosymmetric rings with graph set R²₂(8). These hydrogen bonds link the mol­ecules into chains along [011]. There is an intra­molecular O—H⋯O hydrogen bond between the hydroxyl group and the carbonyl group of the carb­oxy­lic acid. The hydrogen bonding in the dihydrate is very different. Although the intra­molecular hy­droxy/carb­oxy­lic acid hydrogen bond is present, the water mol­ecule acts as an acceptor to the carb­oxy­lic acid and a donor to two other oxygen atoms. The carb­oxy­lic acid groups do not inter­act with each other directly.

1. Chemical context

2,5-Di­hydroxy­terephthalate (C₈H₄O₆^2–; dhtp) is of current inter­est as a linker in metal–organic frameworks (MOFs). It can add extra functionality to alter adsorption and catalytic properties. In an attempt to replicate the ionothermal preparation of the Co-dhtp MOF Co₂(dobdc)-ST (Azbell et al., 2022), an unexpected product was obtained, namely anhydrous 2,5-dhy­droxy­terephthalic acid, C₈H₆O₆, (I).

The crystal structures of three Co-dhtp MOFs have been reported: Cambridge Structural Database refcodes FEGBEB (Gen, 2017), VOFJIM (Rosnes et al., 2019) and VOFJIM01 (Ayoub et al., 2019). The calculated powder patterns of these three compounds, which have been given the name CPO-27-Co, indicate that they have the same structure (Fig. 1).

Figure 1 Calculated (using Mercury; Macrae et al., 2020) powder diffraction patterns (Cu Kα radiation) for CPO-27-Co [FEGBEB (Gen, 2017), VOFJIM (Rosnes et al., 2019) and VOFJIM01 (Ayoub et al., 2019)]. The differences in peak positions result from the different temperatures of the diffraction studies. Image generated using JADE Pro (MDI, 2022).

2. Structural commentary

Compound (I) crystallizes in the triclinic space group P with half a mol­ecule in the asymmetric unit. The root-mean-square Cartesian displacements of the non-H atoms in the Rietveld-refined and CRYSTAL17-optimized structures is 0.053 Å (Fig. 2). The good agreement provides strong evidence that the structure is correct (van de Streek & Neumann, 2014). The CRYSTAL17 and VASP-optimized structures are essentially identical (r.m.s. displacement = 0.031 Å). This discussion concentrates on the CRYSTAL17-optimized structure. The full mol­ecule (with atom numbering) is illustrated in Fig. 3 and a view of the packing down the a-axis direction is shown in Fig. 4.

Figure 2 Comparison of the Rietveld-refined (red) and VASP-optimized (blue) structures of anhydrous 2,5-di­hydroxy­terephthalic acid. The r.m.s. Cartesian displacement is 0.053 Å. Image generated using Mercury (Macrae et al., 2020).

Figure 3 The full 2,5-di­hydroxy­terephthalic acid mol­ecule, with the atom numbering. The atoms are represented by 50% probability spheroids. Image generated using Mercury (Macrae et al., 2020). Symmetry code: (a) 1 − x, 1 − y, 1 − z.

Figure 4 The crystal structure of anhydrous 2,5-di­hydroxy­terephthalic acid, viewed down the a-axis. Image generated using DIAMOND (Crystal Impact, 2022).

All of the bond distances, angles, and torsion angles fall within the normal ranges indicated by a Mercury Mogul geometry check (Macrae et al., 2020). The plane of the phenyl ring lies approximately on the (98) Miller plane. The peak profiles are dominated by anisotropic microstrain broadening: the average microstrain is 8362 ppm.

The Bravais–Friedel–Donnay–Harker (Bravais, 1866; Friedel, 1907; Donnay & Harker, 1937) morphology suggests that we might expect platy (with {001} as the major faces) morphology for this crystal. A 4th order spherical harmonics preferred orientation model was included in the refinement. The refined texture index was 1.059 (2), indicating that preferred orientation was small for this capillary specimen. In flat plate specimens examined in Bragg–Brentano geometry using Cu radiation, the preferred orientation tended to be higher.

3. Supra­molecular features

In the extended structure of (I), the carb­oxy­lic acid groups form strong O3—H4⋯O5 hydrogen bonds, which form centrosymmetric loops with graph set R₂²(8) (Etter, 1990; Bernstein et al., 1995; Shields et al., 2000). These hydrogen bonds link the mol­ecules into chains propagating along [011] (Table 1; Fig. 5). There is an intra­molecular O1—H2⋯O5 hydrogen bond between the hydroxyl group and the carbonyl group of the carboxyl acid. A C—H⋯O hydrogen bond also contributes to the lattice energy. The Mercury aromatics analyser indicates one strong inter­action with a centroid–centroid distance of 4.26 Å, and a moderate one at 5.59 Å.

Table 1
Hydrogen-bond geometry (Å, °) for (I)

D—H⋯A	D—H	H⋯A	D⋯A	D—H⋯A
O3—H4⋯O5^i	1.00	1.69	2.689	174
O1—H2⋯O5	0.99	1.68	2.567	147
Symmetry code: (i) .

Figure 5 The hydrogen bonds in the structure of anhydrous 2,5-di­hydroxy­terephthalic acid. Image generated using Mercury (Macrae et al., 2020).

The hydrogen bonding in the dihydrate DUSJUX (Cheng et al., 2010) is very different (Table 2; Fig. 6). Although the intra­molecular hy­droxy–carb­oxy­lic acid O—H⋯O hydrogen bond is present, the water mol­ecule acts as an acceptor to the carb­oxy­lic acid and a donor to two other oxygen atoms. The carb­oxy­lic acid groups do not inter­act with each other directly.

Table 2
Hydrogen-bond geometry (Å, °) for DUSJUX

D—H⋯A	D—H	H⋯A	D⋯A	D—H⋯A
O2—H2⋯O4	1.07	1.43	2.500	178
O1—H1⋯O3^i	1.01	1.64	2.562	149
O4—H4⋯O3^ii	0.99	1.78	2.736	161
O4—H5⋯O1^iii	0.99	1.82	2.794	169
Symmetry codes: (i) ; (ii) ; (iii) .

Figure 6 The hydrogen bonds in the structure of 2–5-di­hydroxy­terephthalic acid dihydrate DUSJUX. Image generated using Mercury (Macrae et al., 2020).

The CRYSTAL17 (Dovesi et al., 2018) calculations suggest that DUSJUX is 28.5 kcal mol⁻¹ lower in energy than the sum of anhdyrous 2,5-di­hydroxy­terephthalic acid and two water mol­ecules. The corresponding VASP (Kresse & Furthmüller, 1996) calculations indicate that DUSJUX is 114.0 kcal mol⁻¹ more stable. As chemists, we would like to attribute the `extra' energy to the formation of additional hydrogen bonds. Rammohan & Kaduk (2018) developed (for citrates using earlier versions of CRYSTAL) a correlation between the energy of an O—H⋯O hydrogen bond and the Mulliken overlap population between the H and the O acceptor: E (kcal mol⁻¹) = 54.7(overlap)^1/2. Using this correlation to estimate the energies of the individual hydrogen bonds, we calculate that DUSJUX is 59.6 kcal mol⁻¹ lower in energy than the sum of the anhydrous mol­ecule and two water mol­ecules – a value between the two DFT calculations. While the disagreements indicate that the absolute energy calculated for a hydrogen bond may be more uncertain than we would like, the Mulliken overlap population is certainly a guide to whether a hydrogen bond is stronger or weaker than another, and to whether a (geometrically possible) hydrogen bond is real or not.

4. Database survey

A connectivity search in the Cambridge Structural Database [CSD version 5.43 June 2022 (Groom et al., 2016); ConQuest 2022.2.0 (Bruno et al., 2002)] of a 2,5-di­hydroxy­terephthalate fragment with the elements C, H, and O only yielded the structure of the dihydrate (Cheng et al., 2010; DUSJUX), as well as two esters. The dihydrate was also obtained accidentally during the synthesis of metal–organic coordination polymers. Removing the chemistry limitation yielded 249 entries, many of which are metal–organic frameworks. A search of the powder pattern against the Powder Diffraction File (Gates-Rector & Blanton, 2019) yielded no hits. Not even the usual accidental matches were obtained; this pattern evidently occupies a very different portion of `diffraction space'.

5. Synthesis and crystallization

Cobalt(II) chloride hexa­hydrate (1.78 g, 7.50 mmol) and 2,5-di­hydroxy­terephthalic acid (1.00 g, 5.05 mmol) were crushed together with mortar and pestle and added to a 10 ml round-bottom flask. The flask was connected to a Schlenk line and placed in a glass bowl of sand on top of a hot plate. The hot plate was heated to 443 K for approximately 18 h and the round-bottom flask was under vacuum. After being removed from the heat and allowed to cool, the remaining solid was transferred to a Pyrex container with aceto­nitrile (50 ml) and placed in a vacuum oven at 343 K for 24 h. After removal from the oven, the solution was deca­nted and replaced with aceto­nitrile (50 ml). This wash procedure was done a total of three times. The remaining solid was then added to 100 ml of methanol at 343 K for 24 h and deca­nted, this wash was done two times. The remaining solid was then added to a vacuum oven at 423 K for 24 h. The remaining solid was then added to a scintillation vial wrapped with Parafilm for storage.

6. Refinement

Crystal data, data collection and structure refinement details are summarized in Table 3. A portion of the sample was blended with 11.51% < 1 micron diamond powder (Alfa Aesar) inter­nal standard in a mortar and pestle until the color was uniform. The X-ray powder diffraction pattern was measured from a 0.7 mm diameter static capillary specimen on a PANalytical Empyrean diffractometer using Mo Kα radiation. The pattern was measured from 1.0–50.0° 2θ in 0.0083560° steps, counting for 4 sec/step.

Table 3
Experimental details

	(I)	DUSJUX (DFT)
Crystal data
Chemical formula	C8H6O6	C8H6O6·2(H2O)
Mr	198.08	--
Crystal system, space group	Triclinic, P	Monoclinic, P21/c
Temperature (K)	302	--
a, b, c (Å)	4.2947 (5), 5.6089 (5), 8.2331 (19)	5.18830, 17.54500, 5.49900
α, β, γ (°)	93.612 (4), 102.219 (4), 96.7621 (14)	90, 103.03, 90
V (Å^3)	191.69 (1)	487.68
Z	1	2
Radiation type	Mo Kα1,2, λ = 0.70932, 0.71361 Å	--
Specimen shape, size (mm)	Cylinder, 12 × 0.7	--
Data collection
Diffractometer	PANalytical Empyrean
Specimen mounting	Glass capillary
Data collection mode	Transmission
Data collection method	Step
θ values (°)	2θmin = 1.002 2θmax = 49.991 2θstep = 0.008
Refinement
R factors and goodness of fit	Rp = 0.034, Rwp = 0.042, Rexp = 0.019, χ^2 = 5.148
No. of parameters	53
No. of restraints	18
(Δ/σ)max	2.635
The same symmetry and lattice parameters were used for the DFT calculations as for the powder diffraction study for (I). Computer program: GSAS-II (Toby & Von Dreele, 2013).

After correcting the peak positions using the known diamond peak positions, the pattern was indexed using JADE Pro (MDI, 2022) on a primitive triclinic cell with a = 4.26420, b = 5.58601, c = 8.17902 Å, α = 93.53, β = 12.13, γ = 96.78° and V = 188 ų. Since the volume corresponded to one mol­ecule of 2,5-di­hydroxy­terephthalic acid, the space group was assumed to be P, with half a mol­ecule in the asymmetric unit. A reduced cell search of the CSD yielded no hits. Preliminary indexing attempts using the default peak list from a pattern collected using Cu radiation were unsuccessful (monoclinic cells with no reasonable structures), until closer examination of the pattern revealed that the peak at 21.6° (9.7° Mo) was actually a doublet, and that there was an additional peak at 22.0° (9.9° Mo). Including these two additional peaks yielded the triclinic cell.

The 2,5-di­hydroxy­terephthalic acid mol­ecule was extracted from the DUSJUX structure using Materials Studio (Dassault Systèmes, 2021), and saved as a .mol2 file. The crystal structure was solved using Monte Carlo simulated annealing techniques as implemented in EXPO2014 (Altomare et al., 2013), using a whole mol­ecule as the fragment. Since the mol­ecule occupies a center of symmetry, the two halves overlapped partially. The overlapping atoms were averaged manually using Materials Studio to obtain the asymmetric unit.

Rietveld refinement was carried out using GSAS-II (Toby & Von Dreele, 2013). All non-H bond distances and angles were subjected to restraints, based on a Mercury Mogul geometry check (Sykes et al., 2011; Bruno et al., 2004). A planar restraint was applied to the benzene ring. The Mogul average and standard deviation for each qu­antity were used as the restraint parameters. The restraints contributed 1.9% to the final χ². The hydrogen atoms were included in calculated positions, which were recalculated during the refinement using Materials Studio (Dassault Systèmes, 2021). The U_iso of the heavy atoms were grouped by chemical similarity. The U_iso for the H atoms were fixed at 1.3× the U_iso of the heavy atoms to which they are attached. The peak profiles were described using the generalized microstrain model. The background was modeled using a four-term shifted Chebyshev polynomial, along with a peak at 12.05° to model the scattering from the glass capillary and any amorphous component. The final refinements yielded the residuals reported in Table 1. The largest errors in the difference plot (Fig. 7) are small, and are in the shapes of the peaks.

Figure 7 The Rietveld plot for the refinement of anhydrous 2,5-di­hydroxy­terephthalic acid. The blue crosses represent the observed data points, and the green line is the calculated pattern. The cyan curve is the normalized error plot, and the red line is the background curve. The row of tick marks indicates the calculated reflection positions. The vertical scale has been multiplied by a factor of 10× for 2θ > 20.5°. The row of red tick marks indicate the positions of the diamond internal standard peaks.

The crystal structure (as well as that of DUSJUX and an isolated water mol­ecule) was optimized using VASP (Kresse & Furthmüller, 1996) (fixed experimental unit cells) through the MedeA graphical inter­face (Materials Design, 2016). The calculations were carried out on 16 2.4 GHz processors (each with 4 Gb RAM) of a 64-processor HP Proliant DL580 Generation 7 Linux cluster at North Central College. The calculations used the GGA-PBE functional, a plane wave cutoff energy of 400.0 eV, and a k-point spacing of 0.5 Å⁻¹ leading to a 4 × 3 × 2 mesh. The structures were also optimized (fixed experimental cells) and population analyses were carried out using CRYSTAL17 (Dovesi et al., 2018). The basis sets for the H, C, N, and O atoms in the calculations were those of Gatti et al. (1994). The calculations were run on a 3.5 GHz PC using 8 k-points and the B3LYP functional.

Supporting information

Computing details

2,5-Dihydroxybenzene-1,4-dicarboxylic acid (I)

Crystal data

C8H6O6	β = 102.219 (4)°
Mr = 198.08	γ = 96.7621 (14)°
Triclinic, P1	V = 191.69 (1) Å^3
Hall symbol: -P 1	Z = 1
a = 4.2947 (5) Å	Dx = 1.716 Mg m^−3
b = 5.6089 (5) Å	T = 302 K
c = 8.2331 (19) Å	cylinder, 12 × 0.7 mm
α = 93.612 (4)°

Data collection

PANalytical Empyrean diffractometer	Data collection mode: transmission
Specimen mounting: glass capillary	Scan method: step

Refinement

18 restraints

Preferred orientation correction: Simple spherical harmonic correction Order = 4 Coefficients: 0:0:C(2,-2) = 0.246(11); 0:0:C(2,-1) = -0.018(11); 0:0:C(2,0) = -0.313(16); 0:0:C(2,1) = 0.217(13); 0:0:C(2,2) = -0.192(9); 0:0:C(4,-4) = -0.146(17); 0:0:C(4,-3) = 0.073(19); 0:0:C(4,-2) = -0.052(16); 0:0:C(4,-1) = 0.083(18); 0:0:C(4,0) = -0.058(17); 0:0:C(4,1) = -0.006(18); 0:0:C(4,2) = -0.196(23); 0:0:C(4,3) = 0.071(16); 0:0:C(4,4) = 0.108(25)

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
C10	0.492 (2)	0.6406 (16)	0.6392 (11)	0.0323 (10)*
C6	0.6818 (18)	0.4506 (18)	0.6571 (9)	0.0323 (10)*
C7	0.6946 (18)	0.3106 (14)	0.5107 (12)	0.0323 (10)*
C9	0.477 (2)	0.7892 (16)	0.7918 (9)	0.0553 (15)*
O1	0.8599 (12)	0.4082 (9)	0.8120 (6)	0.0323 (10)*
O3	0.2791 (12)	0.9640 (12)	0.7718 (7)	0.0553 (15)*
O5	0.6452 (16)	0.7744 (11)	0.9376 (9)	0.0553 (15)*
H8	0.84185	0.17135	0.52928	0.0420 (14)*
H2	0.83782	0.54288	0.88998	0.0420 (14)*
H4	0.30280	1.07255	0.87663	0.0719 (19)*

Geometric parameters (Å, º)

C10—C6	1.411 (5)	C9—O3	1.365 (5)
C10—C7^i	1.384 (5)	C9—O5	1.277 (5)
C10—C9	1.482 (6)	O1—C6	1.391 (5)
C6—C10	1.411 (5)	O1—H2	0.987 (5)
C6—C7	1.412 (6)	O3—C9	1.365 (5)
C6—O1	1.391 (5)	O3—H4	1.004 (6)
C7—C10^i	1.384 (5)	O5—C9	1.277 (5)
C7—C6	1.412 (6)	H8—C7	1.060 (8)
C7—H8	1.060 (8)	H2—O1	0.987 (5)
C9—C10	1.482 (6)	H4—O3	1.004 (6)
C6—C10—C7^i	124.3 (7)	C10^i—C7—H8	126.7 (10)
C6—C10—C9	118.1 (9)	C6—C7—H8	115.1 (10)
C7^i—C10—C9	117.5 (9)	C10—C9—O3	116.8 (7)
C10—C6—C7	117.4 (6)	C10—C9—O5	125.0 (9)
C10—C6—O1	121.6 (9)	O3—C9—O5	118.1 (7)
C7—C6—O1	121.0 (10)	C6—O1—H2	105.5 (6)
C10^i—C7—C6	118.1 (7)	C9—O3—H4	112.7 (5)

Symmetry code: (i) −x+1, −y+1, −z+1.

(Ia)

Crystal data

C	V = 46.12 (1) Å^3
Mr = 12.01	Z = 8
Cubic, Fd3m	Dx = 3.459 Mg m^−3
Hall symbol: -F 4vw 2vw	T = 302 K
a = 3.58625 (11) Å

Refinement

Preferred orientation correction: March-Dollase correction coef. = 1.000 axis = [0, 0, 1]

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
C1	0.12500	0.12500	0.12500	0.0159*

Geometric parameters (Å, º)

C1—C1^i	1.5529	C1—C1^iii	1.5529
C1—C1^ii	1.5529	C1—C1^iv	1.5529
C1^i—C1—C1^ii	109.471	C1^i—C1—C1^iv	109.471
C1^i—C1—C1^iii	109.471	C1^ii—C1—C1^iv	109.471
C1^ii—C1—C1^iii	109.471	C1^iii—C1—C1^iv	109.471

Symmetry codes: (i) x+1/4, y+1/4, −z; (ii) −z, x+1/4, y+1/4; (iii) y+1/4, −z, x+1/4; (iv) −x, −y, −z.

(I_DFT)

Crystal data

C8H6O6	c = 8.1976 Å
Mr = 198.08	α = 93.6590°
Triclinic, P1	β = 102.1730°
a = 4.2647 Å	γ = 96.7840°
b = 5.5912 Å	Z = 1

Data collection

h = →	l = →
k = →

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
C10	0.50725	0.65165	0.64068	0.06414*
C6	0.68695	0.46350	0.65571	0.06414*
C7	0.69097	0.31393	0.51546	0.06414*
C9	0.48403	0.80691	0.79069	0.01062*
O1	0.87656	0.42423	0.80458	0.01062*
O3	0.28942	0.97462	0.76830	0.01062*
O5	0.65017	0.78063	0.93109	0.01062*
H8	0.84185	0.17135	0.52928	0.08339*
H2	0.83782	0.54288	0.88999	0.01381*
H4	0.30280	1.07255	0.87663	0.01381*

Bond lengths (Å)

C10—C6	1.370	C7—H8	1.079
C10—C7^i	1.415	C9—O3	1.320
C10—C9	1.487	C9—O5	1.245
C6—C7	1.382	O1—H2	0.986
C6—O1	1.361	O3—H4	1.000
C7—C10^i	1.415	H4—O3	1.000

Symmetry code: (i) −x+1, −y+1, −z+1.

Hydrogen-bond geometry (Å, º)

D—H···A	D—H	H···A	D···A	D—H···A
O3—H4···O5^ii	1.00	1.69	2.689	174
O1—H2···O5	0.99	1.68	2.567	147

Symmetry code: (ii) −x+1, −y+2, −z+2.

(DUSJUX_DFT)

Crystal data

C8H6O6·2(H2O)	c = 5.49900 Å
Monoclinic, P21/c	β = 103.03°
a = 5.18830 Å	V = 487.68 Å^3
b = 17.54500 Å	Z = 2

Data collection

h = →	l = →
k = →

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Biso*/Beq
O1	0.02829	0.35100	0.84164
H1	−0.08229	0.31915	0.93268
O2	0.41842	0.59430	0.63866
H2	0.51266	0.64044	0.56603
O3	0.28123	0.68492	0.87357
C1	0.01145	0.42401	0.92007
C2	0.14692	0.48106	0.82449
H3	0.26253	0.46707	0.68792
C3	0.28585	0.61663	0.80175
C4	0.13841	0.55682	0.90233
O4	0.64658	0.70011	0.46632
H4	0.52537	0.74263	0.39890
H5	0.75145	0.68680	0.34282

Hydrogen-bond geometry (Å, º)

D—H···A	D—H	H···A	D···A	D—H···A
O2—H2···O4	1.07	1.43	2.500	178
O1—H1···O3^i	1.01	1.64	2.562	149
O4—H4···O3^ii	0.99	1.78	2.736	161
O4—H5···O1^iii	0.99	1.82	2.794	169

Symmetry codes: (i) −x, −y+1, −z+2; (ii) x, −y+3/2, z−1/2; (iii) −x+1, −y+1, −z+1.