Interactions of the products of oxidative polymerization of hydroquinone as catalyzed by birnessite with Fe (hydr)oxides – an implication of the reactive pathway for humic substance formation

Abstract

Polyphenol polymerization (PP) catalyzed by MnO₂ is an important pathway in humification processes in soils. Due to a lack of aliphatic carbons, the products of PP are considered to be humic substance-like materials (HQs), which are subject to adsorption by Fe (hydr)oxide upon their formation. However, the effects of the interactions of HQs with Fe (hydr)oxide in humification processes have received less attention. In this study, hydroquinone was reacted with birnessite for 1 d (HQ-1). After the removal of residual birnessite, the filtrates were incubated for another 7 and 20 days and were denoted as HQ-7, and HQ-20, respectively. The spectroscopic analyses of the HQ samples indicated that oxidative polymerization of hydroquinone occurred within 1 d. With an increase in incubation time, the molecular structure of the resultant HQ continued to change and became increasingly similar to that of natural humic acid (HA), even in the absence of birnessite. Upon adsorption of the HQs by Fe (hydr)oxide, changes in the IR absorption band indicated the complexation of HQ carboxyl groups with metal centers on the mineral surfaces. HQ-20 was preferentially adsorbed on the Fe (hydr)oxides because it contained a higher number of carboxyl groups than their counterparts with a smaller molecular weight. In addition, the steric arrangements and the distributions of the adsorption sites on Fe (hydr)oxide, closely matching the structures of larger molecules, may also contribute to the preferential adsorption. This study implies that HQs with a larger molecular size can be accumulated in soils while reacting with Fe (hydr)oxides during their formation, and the association of HQs with aliphatic carbons derived from the Maillard reaction may contribute to humic substance (HS) formation.

---

Introduction

Humic substances (HSs) are comprised of humic acid, fulvic acid and humin, the major components of soil organic matter (SOM), which are commonly found in surface soils, sediments and water.^1,2 HSs are derived from the humification processes that transform organic precursors, such as sugars, polyphenols and amino compounds, to recalcitrant organic components in soils and sediments.^3,4 In general, humic substances consist of various organic compounds that contain aliphatic moieties and oxidized aromatic components from lignin and polyphenols. The transformation from organic precursors to HSs in soil is usually driven through abiotic catalytic reactions, including the Maillard reaction and polyphenol polymerization (PP). The Maillard reaction is a reaction between amino acids and reducing sugars under ambient conditions.^5,6 Jokic et al.⁴ found that aliphatic structures could be produced during the polycondensation process between glycine and catechol while the Maillard regent (i.e. glucose) was added. On the contrary, the PP reaction results in an increase in the number of aromatic structures due to its higher degree of condensation via the oxidative polymerization of polyphenol.⁷ Because of the lack of aliphatic structures, the chemical compositions of PP products are different to those of HSs, which contain certain amounts of aliphatic molecules and recalcitrant residues of plants and algae.⁸ Thus, we define the products of PP as humic substance-like materials (HQ), which are distinct from HSs.⁹

As reported in previous studies, minerals showed the promise to promote the PP reaction in soils. For example, Jokic et al.⁴ found that clay minerals and metal oxides could promote the transformation of phenolic compounds to precursors of HSs through an oxidative polymerization reaction, including ring cleavage, decarboxylation, and/or dealkylation. Risser and Bailey¹⁰ found that manganese oxides (e.g., birnessite, δ-MnO₂), existing in many soils and sediments, exhibited stronger ability in the abiotic conversion of organic compounds compared to other metal oxides, such as those of iron, aluminum, and silicon. Zherebker et al.¹¹ also reported that Mn(IV) oxides could enhance the oxidation of phenolic compounds, such as hydroquinone, over a pH range of 4–8 in soils. Shindo and Huang¹² indicated that HQs were produced within one day in the presence of MnO₂; however, our preliminary studies showed that the newly formed HQs were unstable, because the absorbance of HQs changed continuously over a prolonged reaction time, even in the absence of MnO₂. Because HQs are important intermediates in humification processes, their interaction with soil minerals might influence the formation and stabilization of soil organic matter (SOM) and control sustainable carbon cycling. In soils, it is well known that HSs might associate or form assemblages with minerals/metal oxides which would hinder the microbial decay and decomposition of the organic materials.^13–15 Regarding HQs, however, their interactions with soil minerals have been rarely studied. Because Fe (hydr)oxides are widely distributed in soils and are associated closely with MnO₂, rgw contributions of Fe (hydr)oxides to the formation, accumulation, and conversion of HQs should be addressed more precisely to clarify the humification processes of organic matter in soils.

In humic acid extracted from volcanic soil enriched with Fe (hydr)oxides, Chen et al.¹⁶ and Huang et al.¹⁷ found that larger molecular weights (MWs) of aromatic groups dominated the compositions of the humic acids. However, humic acids extracted from peat soil with fewer Fe (hydr)oxides present showed different distributions of MWs and aromaticity.¹⁸ That is, the variances in the chemical compositions of SOM are related to not only the origins of the source SOM,^19,20 but also the presence of specific soil minerals with high affinities to SOM. Because the PP reaction is considered to be a major abiotic pathway in the conversion of organic precursors to HSs, the association of PP products obtained at different stages with soil minerals may be helpful in preserving labile substances against microbial decay and benefitting subsequent humification. While previous studies gauged the magnitude of the PP processes, a complementary mechanistic study that provides explanations of the molecular structures and chemical compositions of the HQs is still missing. Besides, to the best of our understanding, exactly how the Fe (hydr)oxides are involved and how they affect the humification of SOM through abiotic PP processes is still unclear. Thus, we aimed to determine the products of oxidative polymerization of hydroquinone with and without catalysis by birnessite and investigate the adsorption kinetics of the HQs on Fe (hydr)oxides upon their formation. We speculated that the interactions of the HQs with Fe (hydr)oxides, such as ferrihydrite and goethite, may contribute towards diversities in the chemical compositions of SOM.

Experimental

Preparation of humic substance-like materials

The oxidative polymerization of hydroquinone by birnessite (δ-MnO₂) was carried out according to the procedures of Shindo and Huang¹² and Jokic et al.⁴ Birnessite was prepared using the method of McKenzie.²¹ Sodium acetate (0.1 M, 125 mL) containing 10 mM hydroquinone (Sigma, purity ≥ 99%) at a pH of 6 in a 500 mL flask was mixed with 125 mg birnessite. The chemicals and reagents used in the study are all of analytical grade. The mixture suspension was incubated in an oscillating shaker in a water bath at 25 °C for 24 h and filtered using a 0.2 μm pore-size membrane filter. The solids on the filter were collected and analyzed using an X-ray diffractometer (PANalytical X’Pert Pro MRD) with Cu-Kα radiation. The XRD patterns were recorded for values of 2θ from 2 to 80° with a scan rate of 2° min⁻¹. The filtrate was subsequently allowed to age under ambient conditions for 1, 7 and 20 days, respectively. Then, the suspensions were collected, and acidified to pH 1.0 using 6 M HCl to obtain the precipitates. The precipitates were centrifuged at 2683g for 30 min and transferred to dialysis tubing with a molecular weight (MW) cutoff of 3500 Da and dialyzed against deionized water until the electrical conductivity (EC) of the dialysate was <50 μS cm⁻¹. The precipitates of the humic substance-like material products, referred to as HQ-1, HQ-7, ad HQ-20, respectively, were then lyophilized and stored prior to use.

Sample collection and extraction of HA from volcanic soil

Volcanic soil collected from the Yangming Mountain area (25N, 121°32′50′′E) represents a sample of soil that has been likely formed under the influence of Fe (hydr)oxides because the soil is enriched with Fe (hydr)oxides and contains larger MW HSs in the SOM. The Yangming mountain soil is classified as Andisol.²² The physical and chemical properties of the volcanic soil can be referred to in the reports of Huang et al.²³ and Chang et al.²⁴ The major minerals in the volcanic soils include kaolinite, illite, quartz, gibbsite, chlorite, and high contents of Fe/Al non-crystalline minerals. The total organic carbon is 156 g kg⁻¹.

Extraction of the HA followed the standard procedure of the International Humic Substance Society (IHSS), as outlined by Swift.²⁵ Briefly, an aliquot (30 g, air-dried, passed through a 2 mm sieve) was treated with HCl (300 mL, 1 M) in a 500 mL centrifuge bottle. After shaking for 1 h, the suspension was centrifuged at 2683g for 20 min and the supernatant was decanted. The residue was re-suspended in NaOH (300 mL, 0.1 M, in a N₂ atmosphere) on a shaker for 24 h (25 °C). Following centrifugation, the supernatant was collected and acidified with 6 M HCl (to pH 1–2) to obtain the HA fraction.

Sample characterization

UV-visible spectroscopy. Prior to being acidified, the soluble HQ samples, formed in the presence or absence of birnessite, were examined using a UV-visible spectrophotometer (Varian, Cary-50) over a wavelength range of 400–700 nm. The readings of absorbance within the specific wavelength range were commonly used to examine the extent of humification of the HAs.^12,26

Elemental compositions. The elemental contents (C, H, O, and N) of the HQ and HA precipitates were investigated using a Heraeus CHNOS Rapid F002 Elemental Analyzer.

Fourier transform infrared spectroscopy (FT-IR)

FT-IR spectra of the original hydroquinone, HQ-1, HQ-7, HQ-20 and the HA samples were obtained using a Thermo-Nicolet Nexus FT-IR spectrometer in the range 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹. The samples were purged with dry N₂ gas for at least 10 min to remove atmospheric CO₂ and moisture prior to analysis. A precise quantity (1 mg) of each sample was mixed with 200 mg KBr powder, and the mixture was ground and compressed into a translucent sample disk. FT-IR spectra were obtained by the co-addition of 64 individual scans.

High performance size exclusion chromatography (HPSEC)

HPSEC was used to investigate the changes in MW distributions of each HQ sample compared to that of HA. The HQ and HA samples were dissolved in 0.1 M KOH prior to HPSEC analyses. The SEC system consisted of a high-pressure liquid chromatography pump (Varian, ProStar 210) and a UV detector (ProStar 320-UV-Vis Detector). A Phenomenex protein was used in the SEC column (BioSep-SEC-s2000). The mobile phase consisted of 2 mM phosphate and 10 mM NaCl solutions at pH 7.0. The flow rate was maintained at 1 mL min⁻¹ and the samples were detected at 254 nm. This technique was also used to examine the preferential adsorption of specific portions of HQ samples upon reaction with Fe (hydr)oxides.

Gas chromatography mass spectrometry (GC-MS)

A GC-MS technique was used to identify the molecular variations of HQ1 and HQ20 samples during the PP reaction and the results were compared with that of hydroquinone, a precursor of the PP reaction. GC-MS was performed with an Agilent 7890A gas chromatograph and an Agilent 5975C mass selective detector. A HP-5 silica column (30 m × 0.32 mm, 0.25 μm film thickness) was used, and operated with the following oven temperature program: a temperature of 150 °C was held for 6 min, followed by a rise at a rate of 3 °C min⁻¹ to 300 °C, at which the temperature was held for 15 min. Helium was the carrier gas at 1 mL min⁻¹. The splitless injection mode was utilized, mass spectra (2.35 scan s⁻¹) were recorded under electron ionization at 70 eV, and the compounds were assigned by comparing with the NIST library.

Adsorption of HQs and HA onto Fe (hydr)oxides

Fe (hydr)oxide preparation. Two-line ferrihydrite was synthesized by neutralizing a 0.1 M Fe(NO₃)₃ solution with 1 M NaOH. Well-crystallized goethite (α-FeOOH) was prepared by rapidly adding 5 M KOH to 1 M Fe(NO₃)₃ under constant stirring. The suspensions were immediately diluted to a volume of 2 L with deionized water, transferred to a polyethylene flask and incubated in an oven at 70 °C for 60 h. Each synthesized mineral was washed thoroughly using deionized water and dialyzed against deionized water until free of Cl⁻. The samples were stored as suspensions under refrigeration. Details of the procedures are outlined by Schwertmann and Cornell.²⁷ The random powder samples of ferrihydrite and goethite were analyzed using an X-ray diffractometer (PANalytical X’Pert Pro MRD).

Preparation of HQ/HA stock solutions. Stock solutions of HA and HQ samples were prepared by dissolving 2.0 g HA and 1.0 g HQs in a 1 L solution with 0.1 M KOH under continuous stirring overnight. After passing through a 0.2 μm pore-size membrane filter, the filtrate was stored in a refrigerator for further use. The C contents of the stock solutions were verified using a multi-N/C total organic carbon (TOC) analyzer (Analytik Jena, 2100).

Adsorption kinetics. Stocks of the ferrihydrite and goethite suspensions were dispersed in an ultrasonic bath for 30 min and stirred constantly for 1 h. An appropriate volume of the suspension was extracted from each sample, transferred to a 50 mL centrifuge tube, and maintained at pH 4.0 using KOH (0.01 M) or HCl. An appropriate quantity of HA and HQ stock solution was added to the suspension to bring the final volume to 40 mL and HA and HQ concentrations to 30 and 50 mg C L⁻¹ for goethite and ferrihydrite, respectively. The final suspension densities for ferrihydrite and goethite were 0.25 and 0.62 g L⁻¹, respectively. The adsorption experiments were carried out on a rotary shaker, and the tubes were removed periodically. The suspensions were passed through 0.2 μm pore-size membrane filters prior to being analyzed colorimetrically at 254 nm. Changes in the absorbance of HA or HQs after interaction with the solids were attributed to adsorption. A linear calibration curve of the absorbance measured by UV/Vis versus the C mass determined by TOC was obtained, and thus, the absorbance could be readily converted to the C mass. HA and HQ adsorption was expressed in units of mass of C per gram of adsorbent.

Results and discussion

UV-visible spectroscopy

The time-dependent absorbance changes in the products of PP reactions were examined via wavelength scans of the filtrate from 400 to 700 nm (Fig. 1), which is a rapid method used to determine the degree of humification.¹² The absorbance increased slightly with incubation time, indicating that the oxidative polymerization of hydroquinone occurred slowly in the absence of Mn oxides (Fig. 1a). Addition of Mn oxide could accelerate the oxidative polymerization of hydroquinone, as evidenced by a rapid increase in the maximum absorbance at 400 nm (Fig. 1b). These results are consistent with the reports of Shindo and Huang,¹² showing that the colored (browning) reaction occurred within 1 d of incubation. In this study, however, a continuous increase in absorbance was observed upon increasing the incubation time up to 20 days after filtering out MnO₂ (Fig. 1b), demonstrating that the polymerization reactions proceeded even after the oxidant (i.e., birnessite) was removed (Fig. 1b).

Fig. 1 Effects of incubation time on the absorbance (400–700 nm) of filtrates obtained from the reaction of hydroquinone (a) without and (b) with birnessite at pH 6.

FT-IR spectroscopy

Infrared spectra of the reaction products (i.e., HQ-1, HQ-7, HQ-20), HA from volcanic soil, and pure hydroquinone, are shown in Fig. 2a. Compared with the spectra of the HQs (Fig. 2b), hydroquinone exhibited more characteristic absorption bands in the fingerprint region. The bands at 3220, 3030 and 1517/1475 cm⁻¹ were assigned to OH, C–H and C=C stretching, respectively, on the single aromatic ring. The bands at 1209/1096 and 827/760 cm⁻¹ were attributed to C–O stretching and C–H out-of-plane bending, respectively.²⁸

Fig. 2 (a) FT-IR spectra of HQ samples obtained after reactions for 1, 7 and 20 days, denoted as HQ-1, HQ-7 and HQ-20, respectively. HA from Yangming Mountain is provided for comparison. (b) Enlargement of the 600–2000 cm⁻¹ region.

A comparison of the absorption bands of hydroquinone with those from HQ-1 revealed that the intensities of the major hydroquinone bands became weaker or disappeared as the result of the addition of Mn oxide, which gave rise to the oxidative polymerization reaction (Fig. 2a). The broad band at 3420 cm⁻¹ was assigned to OH stretching, and the bands at 1720 and 1609 cm⁻¹ were attributed to the asymmetric stretching of COOH and the stretching of the H-bonded C=O of the carbonyl groups, respectively (Fig. 2a). The presence of these two bands in the spectrum of HQ-1 indicated that the OH group of hydroquinone was oxidized by Mn oxide to COOH and C=O groups. Besides, the hydroquinone conjugated C=C modes at 1475 and 1517 cm⁻¹ shifted to 1450 and 1500 cm⁻¹, respectively, upon oxidation (Fig. 2b). These changes in the spectra of the major absorption peaks indicated the structural differences between hydroquinone and HQ-1.^4,15

HQ-7 and HQ-20 exhibited analogous spectra; nonetheless, a significant increase in the band intensity at 1720 cm⁻¹, attributed to the asymmetric C=O stretching of COOH, was observed when the incubation period was extended to 20 days (Fig. 2b). An increase in the band intensity from the –COOH groups indicated that the oxidative reaction continued, most likely because of the oxidation of OH or ether groups in the compounds with phenolic/alcoholic or aryl ether groups. Thus, the band at 1250 cm⁻¹, assigned to the C–O stretching of aryl ethers, gradually decreased (Fig. 2b). In addition to the enhancement of carboxylic groups, the bands at 1550, 1450 and 1050 cm⁻¹, attributed to C–C stretching, disappeared when the incubation time was extended (Fig. 2b). The occurrence of aromatic ring cleavage and re-polymerization of the aliphatic fragments may have caused the disappearance of these characteristic bands of the HQ-20 sample. The spectra for HQ-20 and HA showed distinct differences at 2905, 2849 and 1050 cm⁻¹, corresponding to C–C/C–H stretching. These bands were present only for the HA sample, suggesting that the products synthesized from HQs lacked aliphatic carbons and differed significantly in chemical composition from the HAs, which contained both aromatic and aliphatic domains.²³ Furthermore, elemental analyses of the HQ samples showed lower H/C atomic ratio values than HA, indicating that HA was more aliphatic than the HQ samples, which was consistent with the FTIR results (Table 1).

Table 1 Elemental composition (%) of HQ samples derived from the browning reaction and comparisons with HA

Adsorbent	Proportion (%) of total weight				H/C^a
	N	C	H	O
a Atomic ratio.
HA	2.5	54.9	4.2	38.2	0.9
Hydroquinone	0.1	65.4	5.3	28.2	—
HQ-1	0.1	54.8	3.4	40.2	0.7
HQ-7	0.1	56.6	3.3	40.1	0.7
HQ-20	0.1	54.9	3.4	41.5	0.8

---

HPSEC investigation of HQ samples

HPSEC was used to further investigate the effect of incubation time on the changes in the MW of HQs. The HPSEC elution curves of HQ-1, HQ-7, HQ-20, and HA are shown in Fig. 3. Two main and distinct distributions could be observed for the HA sample, and a wider multi-modality was observed for all the HQ samples (Fig. 3). HQ-1 exhibited a longer retention time, indicating the presence of small molecules. Oxidative polymerization during the extended incubation time led to a shift in the HQ absorption peak to a shorter retention time, particularly for HQ-20 (Fig. 3). This suggested that HQ-20 contained molecules that were larger than those in HQ-1 and HQ-7. Although the HPSEC exhibited the intrinsic limitation of being unable to determine precisely the MWs of HQ/HA, due to the lack of a suitable standard with similar structures to those of HQ/HA, the technique may be still feasible to describe qualitatively the changes in the apparent MWs of HQ/HA. Combined with the corresponding FTIR results, the results of the HPSEC analysis further demonstrated that the oxidative polymerization of hydroquinone occurred with a longer incubation time.

Fig. 3 HPSEC elution curves of multi-modality of HQ-1, HQ-7, HQ-20 and HA.

GC-MS investigation of HQ samples

To investigate the products of the oxidative polymerization of hydroquinone, HQ-1, HQ-20, and hydroquinone samples were selected and examined using GC-MS, and the results are shown in Fig. 4 and 5. The appearance of larger organic fragments, such as the peaks at 8.912, 11.766 and 12.757 min with major m/z (mass-to-charge) ratios of 253–342 in the spectra after 1 day of reaction between hydroquinone and birnessite suggested the occurrence of oxidative polymerization reactions (Fig. 4). The spectra of HQ-1 showed some small MS fragments at 3.250, 4.717 and 5.982 min (Fig. 4 and Table 2), which may be the oxidative products or the derivatives of oxidative products of hydroquinone, such as para-benzoquinone, forming during the mass analyses. However, these fragments disappeared for HQ-20, indicating a continuous polymerization or re-arrangement of the structures of HQ-1 with prolonged incubation time, even though the birnessite was removed after 1 day of the reaction (Fig. 5). Although the structures of HAs are more complex than those of the HQ samples,^29–31 the fragmentation patterns in the mass spectra of the HQ-20 sample were similar to those of natural HA (comparing Fig. 5c with Fig. 5d). This suggested that the progressive polymerization of HQ-1 would produce larger molecular structures which possessed the ionization properties of HA, i.e., resisting the formation of charged molecules or molecule fragments upon mass analysis. Thus, the major fragments of 8.912, 11.766 and 12.757 min in the mass spectra of HQ-20 and HA samples exhibited relatively weak peak intensities compared with those of HQ-1 (Fig. 4). The results corresponded with the report of Gao et al.,³² who proposed that bisphenol A would form dimer products via hydroxylation and dealkylation in the presence of MnO₂.

Fig. 4 Total ion chromatogram of the GC-MS spectra of hydroquinone and HQ samples obtained at 1 and 20 days, denoted as HQ-1 and HQ-20, respectively. HA is shown for comparison.

Fig. 5 3D profile of GC-MS spectra of hydroquinone and HQ samples obtained at 1 and 20 days, denoted as HQ-1 and HQ-20, respectively. HA is shown for comparison.

Table 2 GC-MS peaks of HQ samples and HA identified by NIST library

Peak time (min)	Compounds	Major (m/z)
a Peak of compound from hydroquinone.b Peak of compound from HQ-1.c Peak of compound from HQ-20.d Peak of compound from HA.
1.469	Acetaldehyde^a	44
3.250	Methyl-phenylindole^b, 1-benzopyrylium, 2-phenyl^b	207
4.717	Benzo-phenothiazine-dioxide^b	281
5.982	Salicylic acid^b	282
	4-Hydroxymandelic acid^b, ethyl ester^b	340
6.905	Hydroquinone^a, 1,4-benzenediol^a	110
7.368	3,5,7-Trimethoxy-2-(4-methoxyphenyl)-/benzopyran^b	342
	10-Ethyl-8-phenyl^b, tetrahydro-dimethoxybenzo^b	341
8.912/8.900	Benzoic acid/hydroxyl-methylphenyl^b^,^c^,^d	296
	Benzofuran^b^,^c^,^d	310
	Siloxane/nonamethylpentasiloxane^d	340
11.766/11.760	Dihydrobenzo(E)pyrene^b^,^c^,^d	254
	Quinolinone/3-hydroxy-4-(3-hydroxyphenol)-^d	253
	Cyclopentadienebutanenitrile^d	253
	Androstan^d	346
12.757/12.753	14-Ethenyl-7,8-dihydro^b^,^c^,^d	327
	Naphtho-cyclodeca-biphenylene^d	328
14.028^b	2-Methoxybenzoylformic acid^b, pentamethyl phenyl-^b	135
15.480^b	Pentamethyl phenyl-^b	135

---

Accordingly, the results of the FT-IR (Fig. 2), HPSEC (Fig. 3), and GC-MS (Fig. 4 and 5) spectra/curves indicated that in the absence of birnessite, long term incubation allowed the further conversion of HQ-1 to form compounds (i.e., HQ-20) with larger MWs and distinct chemical properties. Additionally, the XRD patterns of MnO₂ treated with hydroquinone revealed that the characteristic peaks (7.15, 3.60, 2.44 Å) of birnessite disappeared gradually upon increasing the reaction time (Fig. 6), accompanied with the appearances of peaks of MnCO₃ (3.07, 2.76, 2.47, 2.36 Å). This result indicated that birnessite was transformed (reduced) to rhodochrosite upon hydroquinone oxidation. Because Fe-containing minerals, such as ferrihydrite and goethite, associate intimately with Mn oxides and are widely distributed in soils, their interactions with HQs should be investigated to clarify the possible role of Fe (hydr)oxides involved in PP-related humification processes.

Fig. 6 XRD patterns of the solid residues obtained from the reactions of hydroquinone and birnessite at different reaction times.

Interactions of HQs and of HA with Fe (hydr)oxides

The adsorption kinetics of HA and HQs on ferrihydrite and goethite are shown in Fig. 7. Adsorption proceeded rapidly and reached a plateau within 20 min. The ferrihydrite and goethite demonstrated a similar adsorption order toward HQs and HA, i.e. HA > HQ-20 > HQ-7 > HQ-1. The adsorption of HQs and HA on ferrihydrite and goethite followed the second order kinetic model with rate constants between 5.8 × 10⁻² and 2 × 10⁻³ L mg⁻¹ min⁻¹ and between 2 × 10⁻⁴ and 7 × 10⁻⁵ L mg⁻¹ min⁻¹, respectively (Table 3). Ferrihydrite exhibited greater adsorption rates (particularly for HQ-20) and capacities than goethite due to its higher surface area and reactivity.

Fig. 7 Adsorption kinetics of HA, HQ-20, HQ-7, and HQ-1 on (a) ferrihydrite and (b) goethite with suspension densities of 0.25 and 0.62 g L⁻¹, respectively, at pH 4 and at 25 °C.

Table 3 First and second order regressions for adsorption kinetics

	First-order		Second-order
	k (min^−1)	R^2	k (L mg^−1 min^−1)	R^2
Ferrihydrite
HQ-1	2.3 × 10^−2	0.90	2 × 10^−3	0.92
HQ-7	5.3 × 10^−2	0.86	2.4 × 10^−2	0.92
HQ-20	5.8 × 10^−2	0.89	3 × 10^−2	0.97
HA	7.6 × 10^−2	0.70	5.8 × 10^−2	0.83
Goethite
HQ-1	1.2 × 10^−3	0.79	7 × 10^−5	0.80
HQ-7	1.4 × 10^−3	0.88	7 × 10^−5	0.88
HQ-20	1.6 × 10^−3	0.82	8 × 10^−5	0.82
HA	3.0 × 10^−3	0.84	2 × 10^−4	0.84

---

The time-dependent changes in the MW distributions of HQ-20 upon interaction with ferrihydrite were selected to investigate changes in the molecular sizes of HQ-20 whilst in contact with Fe (hydr)oxides, and the results are shown in Fig. 8. A remarkable drop and disproportionate decrease in the presence of larger molecules occurred within a short retention time, particularly within the first minute (Fig. 8), which was attributed to the initial rapid adsorption of HQ-20 onto ferrihydrite (Fig. 7a). These results suggested that the adsorption of the larger organic molecules of HQ-20 on Fe (hydr)oxides was favorable. Due to a greater degree of polymerization, the steric arrangements and distributions of the adsorption sites on Fe (hydr)oxide may closely match the structures of the larger hydrophobic molecules of HQ-20, leading to preferential adsorption. The phenomenon of preferential adsorption of specific organic moieties enriched with carboxyl groups on soil minerals was also observed by Balcke et al.,³³ Hur and Schlautman,³⁴ and Meier et al.³⁵ Because HQ-20 contained more carboxyl groups than the other HQ samples, the preferential adsorption of HQ-20 on Fe (hydr)oxides was expected. FT-IR spectra of the ferrihydrite before and after interaction with HQ-20 showed that, upon the adsorption of HQ-20 on ferrihydrite, the band at 1720 cm⁻¹, assigned to the asymmetric C=O stretching vibration of COOH, disappeared (Fig. 9). In contrast, the band at 1609 cm⁻¹, attributed to the stretching of H-bonded C=O, weakened and shifted to 1575 cm⁻¹, indicating complexation with Fe-oxides. These band changes and shifts are consistent with the results of Gu et al.³⁶ and Kaiser and Guggenberger¹⁵ who illustrated that a characteristic absorption of the carboxylate band occurred when this group interacted with Fe-oxides. Gu et al.^36,37 and Parfitt et al.³⁸ suggested that the band shift from 1609 to 1575 cm⁻¹ was due to complexation of carboxyl groups with metals on the mineral surfaces via a ligand exchange reaction.

Fig. 8 Time-dependent changes of multi-modality of HQ-20 upon interaction with ferrihydrite.

Fig. 9 FT-IR spectra of HQ-20, ferrihydrite, and HQ-20-bearing ferrihydrite.

Following the rapid interaction of HQ-20 with ferrihydrite, the adsorption rate of HQ-20 decreased slowly (Fig. 7a). Although the adsorption kinetics of the HQs on Fe (hydr)oxide suggested that an apparent equilibrium could be reached within 30 min of reaction (Fig. 7), the curve intensity of the HPSEC spectra changed continuously (Fig. 8). For example, during the stage of slow adsorption, the larger molecules in HQ-20 exhibited preferential adsorption by the ferrihydrite, as indicated by the gradual shift of the adsorption peaks toward a longer retention time in the HPSEC spectra when the reaction progressed (Fig. 8). The result suggested that the adsorption of HQ-20 may proceed via structural rearrangement of the adsorption sites, which allowed the adsorption of additional HQ-20 through molecule–molecule interactions or molecular diffusion into the interior of the adsorbent.

Based on these results, we developed a model to describe the humification processes with various scenarios under different environmental conditions. The findings and results in this study have brought new insights into humification processes in soils, which have not been well understood before. The possible fates of HQs produced from the oxidative polymerization of hydroquinone in the presence of MnO₂ in a soil profile influenced by Fe (hydr)oxides are described in Fig. 10. Condition I illustrates the outcome of strong leaching conditions, or in soils with limited Fe (hydr)oxides or other adsorptive minerals, allowing the small and polar fractions of HQs, formed in the early stages of the oxidative polymerization of hydroquinone, to be distributed and moved rapidly throughout the environment. These hydroquinone–quinone intermediates may be leached out of the soil profile or interact with environmental contaminants, such as Cr(VI) and chlorophenol,^16,23 and thus the polymerization of quinones, leading to the formation of the precursor HS, becomes inhibited. In contrast, weak leaching conditions, illustrated for condition II, cause the small fragments of HQs to be further polymerized and become associated with aliphatic compounds derived from the Maillard reaction in soil solutions. Because Fe (hydr)oxides preferentially adsorb larger MW organic compounds, the biotic decompositions of quinones become inhibited, and thus, larger MW HSs can be accumulated in the SOM. Humic acid derived from Taiwanese volcanic soil, where the larger MWs of aromatic groups dominate the compositions of the humic acids, appears to resemble condition II.¹⁸ Finally, condition III describes what happens with a limited amount of Fe (hydr)oxides in soils. In this case, organic compounds derived from quinones or the Maillard reaction may be accessible and be partially degraded by microorganisms upon their formation, and thus, the MWs and aromaticity of humic acids are more evenly distributed in the soils.¹⁸

Fig. 10 The possible fate of hydroquinone in soil profiles developed under different environmental conditions with various Fe (hydr)oxide contents.

However, the current scenario of describing the possible fates of HQs in the soil profiles is more appropriately applied when HQs are produced with discrete or coated MnO_x prior to entering the soil pores with various amounts of Fe (hydr)oxides. On the other hand, the co-existence of Fe (hydr)oxides and MnO_x, either in discrete or associated phases may interfere with each other while interacting with hydroquinone. Besides, the reactions are greatly affected by the crystalline properties and the abundances (ratios) of Fe (hydr)oxides and MnO_x, as well as the exposed proportions of activated sites if the two minerals are strongly associated. Considering the complexity of a system with many minerals present, the influences of the mineral components on the fate of hydroquinone and the subsequent humification processes under various environmental conditions merit alternative investigation.

Conclusions

The oxidative polymerization of hydroquinone catalyzed by birnessite can lead to the production of HQ within 24 h. FT-IR, GC-MS and HPSEC analysis further proved that the oxidative polymerization proceeded continuously in the absence of birnessite up to 20 days. A significant increase in band intensity at 1720 cm⁻¹ in the FTIR spectra resulted from the polymerization, which is attributed to the asymmetric C=O stretching of COOH. The results were most likely caused by the oxidation of hydroxyl or ether groups in functionalities such as phenolic/alcoholic or aryl ether groups. In addition, the GC-MS and HPSEC profiles indicated that larger molecules were formed after longer incubation times. The compositions of the HQs differed from those of natural HA; however, the kinetic results suggested that HQ/HA adsorption by either ferrihydrite or goethite followed a similar reaction mechanism. The steric arrangement and distribution of the absorption sites on Fe (hydr)oxide, which closely match the structures of larger molecules, may have led to the preferential adsorption of higher MW HQs/HA by these two adsorbents. In addition, larger molecular moieties in HQ/HA, with abundant carboxylic groups, the major adsorptive sites, contributed to preferential adsorption. This study addresses that in the absence of MnO₂, a slow process is involved in the conversion of simple aromatic compounds toward larger organic moieties in PP reactions. This conversion may enhance the adsorption of PP products by soil metal oxides, thereby preventing degradation by microbial attack, and thus, the aromatic-enriched C products from PP reactions can be sequestrated in soils with Fe (hydr)oxides. Thus, understanding the PP products at different stages and their interactions with soil minerals is helpful in clarifying the role of PP reactions in the production of HS in soils.

Acknowledgements

The authors are deeply grateful for to the late P. M. Huang (University of Saskatchewan, Canada) for inspiration and M. H. B. Hayes (University of Limerick, Ireland) for invaluable suggestions which improved the manuscript. Thanks are also due to Yi-Chi Pao for her assistance in sample preparation. The work was financially supported by the National Science Council, ROC under project No. 101-2313-B-005-047-MY3 and 101-2621-M-005-005 and, in part, by the Ministry of Education, ROC under the Aim for Top University (ATU) plan.

References

1. M. H. B. Hayes and C. E. Clapp, Soil Sci., 2001, 166, 723–737.
2. A. Piccolo, Soil Sci., 2001, 166, 810–832.
3. C. Liu and P. M. Huang, Org. Geochem., 2002, 33, 295–305.
4. A. Jokic, M. C. Wang, C. Liu, A. I. Frenkel and P. M. Huang, Org. Geochem., 2004, 35, 747–762.
5. J. I. Hedges, in Humic substances and their role in the environment, ed. F. H. Frimmel and R. F. Christman, John Wiley and Sons, Chichester, UK, 1988, pp. 45–58.
6. J. Burdon, Soil Sci., 2001, 166, 752–769.
7. F. J. Stevenson, Humus Chemistry-Genesis, Composition, Reactions, John Wiley and Sons, New York, NY, 2nd edn, 1994.
8. S. Derenne and C. Largeau, Soil Sci., 2001, 166, 833–847.
9. M. H. B. Hayes and R. S. Swift, in The Chemistry of Soil Constituents, ed. D. J. Greenland and M. H. B. Hayes, Wiley, Chichester, 1978, pp. 179–320.
10. J. A. Risser and G. W. Bailey, Soil Sci. Soc. Am. J., 1992, 56, 82–88.
11. A. Y. Zherebker, D. Airapetyan, A. I. Konstantinov, Y. I. Kostyukevich, A. S. Kononikhin, I. A. Popov, K. V. Zaitsev, E. N. Nikolaev and I. V. Perminova, Analyst, 2015, 140, 4708–4719.
12. H. Shindo and P. M. Huang, Nature, 1982, 298, 363–365.
13. R. G. Keil, D. B. Montlucon, F. G. Prahl and J. I. Hedges, Nature, 1994, 370, 549–552.
14. J. A. Baldock and J. O. Skjemstad, Org. Geochem., 2000, 31, 697–710.
15. K. Kaiser and G. Guggenberger, Org. Geochem., 2000, 31, 711–725.
16. S. Y. Chen, S. W. Huang, P. N. Chiang, J. C. Liu, W. H. Kuan, J. H. Huang, J. T. Hung, Y. M. Tzou, C. C. Chen and M. K. Wang, J. Hazard. Mater., 2011, 197, 337–344.
17. S. W. Huang, P. N. Chiang, J. C. Liu, J. T. Hung, W. H. Kuan, Y. M. Tzou, S. L. Wang, J. H. Huang, C. C. Chen, M. K. Wang and R. H. Loeppert, Chemosphere, 2012, 87, 587–594.
18. Y. M. Tzou, S. L. Wang, J. C. Liu, Y. Y. Huang and J. H. Chen, J. Hazard. Mater., 2008, 152, 812–819.
19. B. S. Xing, Environ. Pollut., 2001, 111, 303–309.
20. B. S. Xing and J. J. Pignatello, Environ. Sci. Technol., 1997, 31, 792–799.
21. R. M. McKenzie, Mineral. Mag., 1971, 38, 493–502.
22. Soil Survey, USDANRCS Agric. Handbook No. 436. US Gov. Print, Office, Washington, DC, 1999.
23. Y. Y. Huang, S. L. Wang, J. C. Liu, Y. M. Tzou, R. R. Chang and J. H. Chen, Chemosphere, 2008, 70, 1218–1227.
24. R. R. Chang, R. Mylotte, M. H. B. Hayes, R. McInerney and Y. M. Tzou, Naturwissenschaften, 2014, 101, 197–209.
25. R. S. Swift, Soil Science Society of America, in Methods of soil analysis. Part 3. Chemical methods, ed. D. L. Sparks, et al., American Society of Agronomy, Madison, WI, 1996, pp. 1018–1020.
26. Y. Chen, N. Senesi and M. Schnitzer, Soil Sci. Soc. Am. J., 1977, 41, 352–358.
27. U. Schwertmann and R. M. Cornell, Iron oxides in the laboratory: preparation and characterization, VCH, Weinheim, 1991.
28. B. Sherine, A. J. A. Nasser and S. Rajendran, Int. Trans. J. Eng., Manage., Appl. Sci. Technol., 2010, 2, 341–357.
29. C. Xiaoli, T. Y. Shimaoka, G. Qiang and Z. Youcai, Waste Manag., 2008, 28, 896–903.
30. X. Q. Lu, J. V. Hanna and W. D. Johnson, Appl. Geochem., 2000, 15, 1019–1033.
31. M. A. Olivella, J. C. d. Rio, J. Palacios, M. A. Vairavamurthy and F. X. C. d. l. Heras, J. Anal. Appl. Pyrolysis, 2002, 63, 59–68.
32. N. Gao, Z. Yu, J. Hong, P. Peng and W. Huang, Soil Sci., 2011, 176, 265–272.
33. G. U. Balcke, N. A. Kulikova, S. Hesse, F.-D. Kopinke, I. V. Perminova and F. H. Frimmel, Soil Sci. Soc. Am. J., 2002, 66, 1805–1812.
34. J. Hur and M. A. Schlautman, J. Colloid Interface Sci., 2003, 264, 313–321.
35. M. Meier, K. Namjesnik-Dejanovic, P. A. Maurice, Y.-P. Chin and G. R. Aiken, Chem. Geol., 1999, 157, 275–284.
36. B. Gu, J. Schmltt, Z. Chen, L. Liang and F. McCarthy, Environ. Sci. Technol., 1994, 28, 38–46.
37. B. Gu, J. Schmitt, Z. Chen, L. Liang and J. F. McCarthy, Geochim. Cosmochim. Acta, 1995, 59, 219–229.
38. R. L. Parfitt, A. R. Fraser and V. C. Farmer, Soil Sci., 1977, 28, 289–296.

---