C A R B O N 7 7 ( 2 0 1 4 ) 6 2 7 – 6 3 6

.sc ienced i rec t .com

Avai lab le a t www

ScienceDirect

journal homepage: www.elsevier .com/ locate /carbon

A novel porous carbon derived from hydrothermalcarbon for efficient adsorption of tetracycline

http://dx.doi.org/10.1016/j.carbon.2014.05.0670008-6223/� 2014 Elsevier Ltd. All rights reserved.

* Corresponding author: Fax: +86 21 65642297.E-mail address: zhangsc@fudan.edu.cn (S. Zhang).

Xiangdong Zhu, Yuchen Liu, Chao Zhou, Gang Luo, Shicheng Zhang *, Jianmin Chen

Shanghai Key Laboratory of Atmospheric Particle Pollution and Prevention (LAP3), Department of Environmental Science and Engineering,

Fudan University, Shanghai 200433, China

A R T I C L E I N F O

Article history:

Received 1 January 2014

Accepted 23 May 2014

Available online 6 June 2014

A B S T R A C T

Increasing attention is being paid to hydrothermal carbonization (HTC) of waste biomass,

due to energy shortages, environmental crises and developing customer demands. How-

ever, most research has been dedicated to the production of bio-oil, with few studies focus-

ing on the application of hydrothermal carbon (hydrochar), a solid residue from HTC of

biomass. In this study, a novel porous carbon (PC) was prepared from hydrochar, via pyro-

lysis at different temperatures (300–700 �C), the characteristics of PC as well as tetracycline

(TC) adsorption behavior were investigated. The hydrochar and PC samples showed a

remarkable range of surface properties, as characterized by Boehm titration, the Fourier

transform infrared spectra and nuclear magnetic resonance spectra. The changes in char-

acteristics suggested that the PC samples produced at high activation temperature

(500–700 �C) were well carbonized and exhibited a high surface area (>270 m2/g). Linear

relationships were obtained between Freundlich adsorptive capacity (KF) and elemental

atomic ratios, surface area and pore volume. The high adsorption capacity of PC samples

can be attributed to its low polarity and high aromaticity, surface area and pore volume.

The molecular variations among the hydrochar and PC samples translated into differences

in their ability to adsorb TC.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

In the last few years, the hydrothermal carbonization (HTC) of

waste biomass has received increasing attention for several

reasons: (1) the precursors are readily renewable and cheap,

(2) it is a simple and environmentally friendly (‘‘green’’) pro-

cess, (3) bio-oil, chemicals and carbonaceous solids can be

obtained simultaneously [1]. The advantage of HTC over the

pyrolysis process of biomass is that it can convert wet input

material into carbonaceous solids at relatively high yields [2].

Much attention has been focused on obtaining bio-oil from

the HTC process; studies on the application of carbonaceous

solids are very scarce. The carbonaceous solids, (i.e., hydro-

thermal carbon, hydrochar), have a less aromatic structure

and thermal recalcitrance, a low surface area and poor poros-

ity, hindering the effective exploitation of hydrochar for

environmental and agricultural applications [2–5]. As a conse-

quence, a post-activation method is required to increase the

surface area and porosity of hydrochar.

The easiest possibility for the development of the surface

area of hydrochar is thermal treatment under an inert atmo-

sphere, where small organic molecules are removed to gener-

ate microporosity, thereby reaching a specific surface area are

comparable to conventional carbon materials [3].

Hydrochar-derived black carbon (i.e., porous carbon, PC)

can be considered structurally similar to pyrochar and

628 C A R B O N 7 7 ( 2 0 1 4 ) 6 2 7 – 6 3 6

activated carbon, consisting primarily of short stacks of

graphite sheets with O-containing groups rimmed on the

edge to form connected microporous networks [6]. Biomass-

derived pyrochar shows an extraordinary strong affinity for

hydrophobic and hydrophilic organic contaminants removal,

due to its large specific surface area, pore volume and abun-

dant functional groups [6–8]. Since activation of waste hydro-

char with thermal treatment can change the surface area as

well as the amounts and types of functional groups, its

adsorption capability can be affected.

Tetracycline (TC) is widely used in aquaculture and veteri-

nary medicines to improve growth rates and feed efficiencies

[9,10]. TC is excreted through feces and urine as un-metabo-

lized matter, and the most negative effect is the development

of multi-resistant bacterial strains that can no longer be trea-

ted with presently known drugs [10]. Due to its potential risk,

it is of great importance to explore efficient and cost-effective

treatment technologies for TC removal. TC is amphoteric

compound and contains multiple functional groups such as

phenol and alcohol, which are active and can induce complex

reactions with black carbon.

It has been reported that various adsorbents, including

black carbon, can remove TC by means of surface adsorption,

metal bridging, H-bonding and p–p interaction between TC

and the corresponding structural components of the adsor-

bents [6,7,11]. However, no relevant studies have been per-

formed to test the adsorption behavior of TC on waste

hydrochar-derived PC.

In this work, PC samples were produced at different tem-

peratures from real hydrochar that was obtained from a

pilot-scale plant. To our knowledge, there is no research that

discusses the characterization and adsorption behaviors of

waste hydrochar-derived PC in the literature. The objective

of this study was, therefore, the investigation of the effects

of activation temperature on the characterization of waste

hydrochar-derived PC and on the performance of PC in

adsorbing TC from aqueous solutions. The materials pro-

duced in this study were then characterized in terms of

porosity, structure, composition and functionality. It is

expected that this work will enhance the production value

of waste biomass HTC and promote the development of

HTC in industry.

2. Materials and methods

2.1. Thermal conversion of waste hydrochar

The waste hydrochar was produced from Salix psammophila of

HTC at 300 �C [12]. A more detailed procedure for the prepara-

tion of hydrochar is available in Supporting Information. The

waste hydrochar was activated via thermal conversion under

nitrogen gas (N2) conditions at 300, 400, 500, 600 and 700 �C. In

brief, the hydrochar samples of 5 g were packed into a cera-

mic pot and then pyrolyzed at different temperatures in a

box-type resistance furnace (SLQ1100-30, Shengli Co., Ltd.,

Shanghai) under N2 flow of 1 L min�1 for 4 h at a heating rate

of 4 �C min�1. The yield of the PC samples were recorded and

then milled to pass through a 0.25 mm sieve (60 mesh) prior

to further analyses. The activated samples are hereafter

referred to as P300, P400, P500, P600 and P700, where the suffix

number represents the activation temperature.

2.2. Characterization of samples

The elemental (C, H, N) analyses were performed with an

Elemental Analyzer Vario EL 3 instrument. Ash content was

measured by heating the samples at 200 �C for 1 h and then

at 500 �C for an additional 4 h under an air atmosphere [13].

The pH of samples was measured in a suspension of 1:10

sample/deionized water using a combination electrode. The

suspension was shaken for 1 h before measurement [14].

The oxygenated acidic groups and basic components of sam-

ples were determined using the Boehm’s titration method

[15]. A more detailed account of the Boehm’s titration is avail-

able in Supporting Information.

The thermogravimetry (TG) and derivative thermogravi-

metric (DTG) of the hydrochar sample were analyzed by a

thermo-gravimetric analyzer (Perkin Elmer, USA) under an

N2 atmosphere, by heating the sample from room tempera-

ture to 800 �C at a rate of 10 �C min�1. The Brunauer–

Emmett–Teller (BET) surface area of the samples was deter-

mined with a N2 adsorption–desorption isotherm measured

at 77 K using a Quantasorb SI instrument (Quantachrone,

USA). The morphology of samples was examined by scanning

electron microscopy (SEM) using a Philips (XL300) microscope.

The functional groups and surface properties of samples were

examined through solid state 13C nuclear magnetic resonance

(NMR) spectra with cross polarization magic angle spinning

(Bruker DSX 300) and Fourier transform-infrared (FT-IR, Nexus

470) techniques. The samples were also characterized by

X-ray diffraction (XRD) and X-ray photoelectron spectrometer

(XPS) techniques; and, the binding energies for the high-

resolution spectra were calibrated by setting C to 1 s at

284.6 eV.

2.3. Batch sorption experiment

Tetracycline (TC) was selected as the model compound.

Adsorption isotherms were obtained at the concentration

range of 5–50 mg L�1 TC, and the background solution was

0.02 M sodium chloride (NaCl) [6]. To initiate the experiments,

a 60 mL amber glass vial with 40 mL sorption solution

received a weighted amount of sorbent (0.04 g) and was then

shaken at 150 rpm for 4 days at 30 �C. After reaching sorption

equilibrium, the suspensions were filtered through a 0.45 lm

polytetrafluoroethylene (PTFE) membrane filter, and aliquots

of the filtrate were analyzed by UV–vis spectroscopy [16].

The absorbance of TC was measured at 360 nm.

3. Results and discussion

3.1. Characterizations of hydrochar and porous carbonsamples

3.1.1. TG-DTG analysis of hydrochar sampleThe pyrolysis characteristics, (i.e. both TG and DTG curves)

are shown in Fig. S1. Weight loss of the hydrochar sample

occurred between 130 and 700 �C, and the amount of solid

C A R B O N 7 7 ( 2 0 1 4 ) 6 2 7 – 6 3 6 629

residue left was around �60% (wt). In the DTG curve, there

were five major weight loss peaks. The decomposition peak

at 238 �C may be attributed to the loss of residual hemicellu-

lose [17], however, the other four peaks may be assigned to

different boiling point products derived from the decomposi-

tion of cellulose, hemicellulose and lignin during the HTC

process. In addition, the residual lignin contributed to the

wide and flat DTG peak, due to its slow carbonization process.

Thus, the activation temperatures at 300, 400, 500, 600 and

700 �C for the prepared PC samples corresponded to the

decomposition degree of different components within the

hydrochar.

3.1.2. Proximate analysis and elemental compositionsTable 1 displays the numerical results for proximate and ele-

mental analysis of hydrochar and PC samples. Yields declined

gradually from 300 to 700 �C, due to the slow decomposition

of the components of hydrochar. The final yield of PC sample

was 61.6%, which was consistent with the TG measurement.

The ash content increased slightly up to 500 �C and stabilized

at �23% at higher temperature. An acidic pH for the hydro-

char was observed, due to the solubilization of inorganic frac-

tion and acidic products gathered in the solid residue [18].

The pH of the PC samples increased with the increasing tem-

perature, suggesting that higher activation temperature led to

the higher pH of the PC due to the separating of alkali salts

from organic component and basic organic oxides [19,20].

This was important for the employment of porous carbon,

because the natural alkalinity of char can resist soil acidity

and improve crop growth and yield [13].

The relative element contents showed the gradual losses

of H and O, which resulted from the condensation reactions

for raw hydrochar. The content of N was relatively stable,

but showed a maximum (1.60%) at 500 �C, suggesting enrich-

ment of N-containing compounds [21]. The molar ratios of

elements were calculated to estimate the aromaticity (H/C)

and polarity (O/C, O + N/C) of the samples and a typical van

Krevelen diagram, which is shown in Fig. S2.

As expected, raw hydrochar followed the typical dehydra-

tion reaction as a function of activation temperature up to

700 �C. Lower molar ratios were obtained in P700 sample, indi-

cating its high aromaticity and low polarity compared to the

raw hydrochar, due to the formation of aromatic structures

with a higher degree of carbonization of organic matter and

the removal of polar components [14].

Table 1 – Yields, pH, ash contents, elemental compositions and aprepared under various temperatures.

Sample Yielda (%) pHb Asha (%) Cc (%) Hc

HC 100 5.2 15.2 66.3 5.21P300 89.5 6.7 14.7 72.1 4.12P400 81.0 7.6 15.8 72.9 3.58P500 71.1 7.9 22.8 80.7 3.21P600 66.2 8.5 22.9 92.5 2.98P700 61.6 11.8 23.7 94.1 2.54a Yields and ash contents are on a water-free basis.b The ratios of char and deionized water are 1:10 (wt/wt).c Elemental compositions and atomic ratios are on a water- and ash-free

oxygen to carbon. (O + N)/C: atomic ratio of sum of oxygen and nitrogen t

3.1.3. Surface area and morphologyAs shown in Table 2, the surface area of hydrochar was extre-

mely low (7.16 m2/g). This can be related to pore blockage, due

to the presence of organic matter that was not transferred to

the liquid phase during HTC process or due to compounds

from liquid phase that migrated to the surface of the hydro-

char [22]. The surface area of the PC samples gradually

increased from 7.16 to 288 m2/g up to 500 �C and became sta-

ble at higher temperatures.

The N2 adsorption–desorption isotherms and pore size dis-

tribution are exhibited in Figs. S3 and S4, respectively. Accord-

ing to the IUPAC (International Union of Pure and Applied

Chemistry) classification [23], the N2 isotherm for hydrochar

is of type III (macroporous adsorption with weak adsorbate–

adsorbent interaction), suggesting that the raw hydrochar

possessed less developed porous structure. However, the N2

isotherms for the P300 and P400 samples were of type I and

IV (micropore adsorption and appearance of adsorption hys-

teresis, respectively), which is indicative of microporous and

mesporous structures. Interestingly, the N2 isotherms for

P500, P600 and P700 evolved to type I (micropore adsorption),

reflecting the evolution of microporous structure. As shown

in Fig. S4, it is evident that mesopores of about 4 nm in diam-

eter increased considerably when the activation temperature

was higher; and, the total pore volume of the PC increased

from 0.055 to 0.242 cm3/g accordingly with increasing activa-

tion temperature (Table 2).

The morphologies of the hydrochar and PC samples are

illustrated in Fig. 1. The raw hydrochar appeared to be a

sphere-like structure (�15 lm in diameter), which was caused

by the degradation of the cellulose component during the

HTC process and its subsequent precipitation and growth as

spheres [24]. In addition, there were some rudimentary pores

(�100 nm in diameter) within the structure, due to the change

of composition owing to thermal decomposition [25]. In all PC

samples, it could be observed that, the sphere was broken and

split into flat structure. Furthermore, the sponge-like struc-

ture confirmed the existence of cavities, and the development

of which is dependent on the activation temperature.

3.1.4. Surface acidity and basicityTable 3 presents the oxygen-containing functional groups and

total alkalines (including all mineral salt and basic groups,

such as ketones, pyrones and chromens) from the Boehm

titration. The data indicated that three groups of organic acids

tomic ratios of waste hydrochar (HC) and porous carbon (PC)

(%) Nc (%) Oc (%) H/Cc O/Cc (O + N)/Cc

1.26 27.2 0.942 0.307 0.3241.54 22.3 0.685 0.232 0.2501.54 22.0 0.590 0.226 0.2441.60 14.5 0.477 0.134 0.1511.49 3.00 0.387 0.024 0.0381.40 1.92 0.324 0.015 0.028

basis. H/C: atomic ratio of hydrogen to carbon. O/C: atomic ratio of

o carbon.

HC HC

P700

P500P400

Fig. 1 – Scanning electron micrographs (SEM) of raw hydroxchar (HC) and selected PC samples.

Table 2 – Surface area, pore size and pore volume parameters for waste hydrochar (HC) and porous carbon prepared undervarious temperatures.

Sample Surface areaa

(m2/g)Pore sizeb (d, nm) Vmic

c (cm3/g) Vmesd (cm3/g) Vt

e (cm3/g) Vmic/Vt

HC 7.16 30.5 0.004 0.051 0.055 0.07P300 80.3 3.73 0.018 0.064 0.082 0.22P400 142 3.73 0.036 0.124 0.160 0.23P500 288 3.99 0.124 0.088 0.212 0.58P600 270 4.01 0.117 0.098 0.215 0.54P700 316 4.03 0.136 0.106 0.242 0.56a Measured using N2 adsorption with the Brunauer–Emmett–Teller (BET) method.b Pore size in diameter calculated by the desorption data using Barrett–Joyner–Halenda (BJH) method.c Micropore volume calculated using the t-plot method.d Mesopore volume calculated as the difference between the Vt and Vmic.e Total pore volume determined at P/P0 = 0.99.

630 C A R B O N 7 7 ( 2 0 1 4 ) 6 2 7 – 6 3 6

including carboxylic, lactonic and phenolic decreased with

increasing activation temperature, due to the removal of com-

ponents, however, the carbonylic group sharply decreased

before 500 �C and then gradually increased. Interestingly, total

acidity and basicity exhibited opposite trends. The total basi-

city increased with increasing activation temperature, due to

the increasing contents of ash and basic functional groups,

such as ketones, pyrones, and chromenes [13,26]. The total

basicity slightly decreased at 700 �C, due to the highly carbon-

ized sample. The results of Boehm titration illustrated that

the low-temperature PC sample were more organic in nature

[15].

Table 3 – Contents of oxygen-containing functional groups and total alkaline contents for waste hydrochar (HC) and porouscarbon prepared under various temperatures.

Sample Carboxylica Lactonicb Phenolicc Carbonylicd Total aciditye Total basicityf

HC 1.22 0.12 0.50 0.60 2.44 0.22P300 0.25 0.86 BDL 0.14 1.25 0.37P400 0.06 0.36 0.12 0.06 0.60 1.14P500 0.06 0.06 0.02 0.14 0.28 1.20P600 BDLg 0.10 BDL 0.16 0.26 1.22P700 BDL 0.04 0.02 0.20 0.26 0.85a The contents of carboxylic (mmol/g) calculated as the consumption of NaHCO3.b The contents of lactonic (mmol/g) calculated as the difference in the consumption between Na2CO3 and NaHCO3.c The contents of phenolic (mmol/g) calculated as the difference in the consumption between NaOH and Na2CO3.d The contents of carbonylic (mmol/g) calculated as the difference in the consumption between C2H5ONa and NaOH.e The total acidity (mmol/g) calculated as the consumption of C2H5ONa.f The total basicity (mmol/g) calculated as the consumption of HCl.g Below detectable level.

C A R B O N 7 7 ( 2 0 1 4 ) 6 2 7 – 6 3 6 631

3.1.5. FTIR analysisThe FTIR spectra and spectroscopic assignment of hydrochar

and PC samples are shown in Fig. 2 and Table S1, respectively.

No noteworthy changes occurred at 300 �C. After heating to

400 �C, the band intensities at 2920 and 2845 cm�1 (aliphatic

C-H stretching) were completely disappeared, due to removal

of low boiling points components (e.g. paraffin) of hydrochar.

A similar trend was observed for the band at 1314 cm�1,

which is an indicator of carboxyl or calcium oxalate. A new

band at 1408 cm�1 (CO32�) represented the presence of calcite

[27], which was resulted from the decomposition of calcium

oxalate at high temperatures [28].

Heating to 500 �C resulted in more substantial chemical

transformations. The polar group (band at 1700 cm�1) and

indicator of residual lignin (aromatic C@C vibration at

1442 cm�1) gradually decreased. Greater aromatic C@C

stretching vibrations (1600 cm�1) was observed; and, further

evidence for aromatic C was provided in the appearance of

the bands at 1021, 877, 790 and 746 cm�1 [29]. Interestingly,

these bands were dramatically diminished at 600 �C and com-

pletely disappeared at 700 �C. In addition, the functional

group was almost invisible at 700 �C, which closely resembled

FTIR spectra of graphitic domains of amorphous carbon

[21,30]. Hence, an increasing degree of condensation was

observed with the increasing heating temperature.

4000 3500 3000 2000 1500 1000 500

HC

P300

P400P500

P600P700

OHaliphatic C-H C=O

aromatic C=C

C=C

aromatic C-HC-H

Wavenumber (cm-1)

Abs

orba

nce

(arb

itary

uni

ts)

CO32-

-COOH

a

Fig. 2 – FTIR spectra (a) and X-ray diffraction patterns (b) of raw

ranging from 300 �C to 700 �C.

3.1.6. XRD analysisAs shown in Fig. 2, the XRD patterns of hydrochar and PC

samples had different peaks, indicating the presence of min-

eral crystals. It was noticed that the major crystalline phases

in hydrochar sample were whewellite (calcium oxalate, Ca(C2-

O4)Æ2H2O) and low-temperature quartz (SiO2). These are the

most frequent and abundant minerals in plants, indicating

that hydrochar was rich in calcium and silica [31,32].

As the heating temperature increased to 400 �C, a new

peak was observed and can be attributed to calcite (calcium

carbonate, CaCO3), which was originated from the heating

transformation of whewellite. The formation of calcite con-

tributed to the alkalinity of the studied PC samples, as shown

by their high pH values (Table 1). With further heating, the

peaks assigned to quartz and calcite (i.e. 26.4� and 29.2�,respectively) progressively increased in intensity and then

significantly decreased at 700 �C, indicating that decomposi-

tion of calcite was occurred at high temperature. In addition,

the XRD results confirmed the band changes of FTIR at

1408 and 1314 cm�1 (CO32�, carboxyl or calcium oxalate,

respectively).

3.1.7. XPS analysisTo obtain more detailed information regarding surface func-

tionality presented at the outer surface of the hydrochar

10 20 30 40 50

♦♦♦♦♥♦

♦♦

♥

••

••

•

♥

♥•♥• •• ♥♥♥♥

♥♥

♥♥

P700

P600

P500

P400

P300

••••

•

♥♦♦♦ ♦♦ ♦♦

♦

2θ

Diff

ract

ion

Inte

nsity

(arb

itary

uni

ts)

b ♦ whewellite • calcite ♥ quartz

♦ ♥

•

HC

hydrochar (HC) and PC samples generated at temperatures

632 C A R B O N 7 7 ( 2 0 1 4 ) 6 2 7 – 6 3 6

and PC sample, XPS was performed on representative sam-

ples (Fig. 3 and Table 4). The high resolution C 1(s) photoelec-

tron spectrum for the raw hydrochar sample included four

common signals, which can be attributed to the aliphatic/aro-

matic carbon groups (C1, C–Hx, and C–C), hydroxyl group (C2,

C–O–H), carbonyl group (C3, C@O), and carboxylic/ester/lac-

tone group (C4, O@C–O) [18]. Thermal treatment caused a

reduction from 80.0% to 75.8% in the C1 peak, indicating the

loss of volatile compounds and a resultant increase of surface

hydrophobicity [33]. However, the increased relative intensity

of the C1 peak was observed for calcination at more than

500 �C, suggesting a resultant increase in degree of aromatics

and condensation. The new C5 peak can be attributed to the

presence of carbonate produced from the degradation of whe-

wellite, as confirmed by XRD result. Increasing the activation

290 288 286 284 282 280

4000

8000

12000

16000

20000

C1s (4)C1s (3)

C1s (2)

Inte

nsity

(a.u

.)

Bonding Energy (ev)

HC

C1s (1)

290 288 286 284 282 280

4000

8000

12000

16000

20000

C1s (5)C1s (4)

C1s (3)C1s (2)In

tens

ity (a

.u.)

Bonding Energy (eV)

P500

C1s (1)

Fig. 3 – High-resolution XPS scans of C1 (s) photoelectron en

Table 4 – Experimental C1(s), binding energy (eV) and chemicalsample.

Sample C1 C2

sp2-Graphitic orC–C/C–Hx (%)

C–O (%)

HC 284.6/80.0 285.7/10.7P400 284.6/75.8 286.1/7.55P500 284.6/77.8 286.1/9.76P700 284.6/78.7 286.1/8.51a Below detectable level.

temperature to 700 �C led to a loss in the residual carbonate

group.

3.1.8. 13C CP NMR analysisThe changes of structure of the hydrochar and PC samples

could be further elucidated by the solid state 13C NMR spectra.

As shown in Fig. 4, the 13C NMR spectrum of hydrochar sam-

ple contained a range of C types. The obvious peaks at 168

and 30 ppm was assigned to the carboxyl group and long-

chain aliphatic structures, and other small peaks at 154 and

14 ppm were represented the presence of the phenolic group

and short-chain aliphatic structure. In addition, the small sig-

nals at 146 and 56 ppm were assigned to residual lignin and

represented O-substituted aromatic C and methoxyl C [34];

and, the peak at 75 ppm could be attributed to cellulose [35],

292 290 288 286 284 282 280

4000

8000

12000

16000

20000

C1s (5)C1s (4)

C1s (3)C1s (2)

Inte

nsity

(a.u

.)

Bonding Energy (eV)

P400

C1s (1)

290 288 286 284 282 280

4000

8000

12000

16000

20000

C1s (5)C1s (4)

C1s (3)C1s (2)In

tens

ity (a

.u.)

Bonding Energy (eV)

P700

C1s (1)

velope for raw hydrochar (HC) and selected PC samples.

state assignments for raw hydrochar (HC) and selected PC

C3 C4 C5

C@O (%) O@C–O (%) CO32� (%)

287.0/4.59 288.9/4.69 BDLa

287.0/8.98 288.9/4.89 289.5/2.76287.0/5.50 288.5/4.27 289.0/2.67287.0/9.23 288.1/1.65 289.3/1.91

∗∗

∗

P600∗

127

∗

∗

∗

127

P500∗

300 250 200 150 100 50 0 -50

∗∗

P700

127∗

ppm

∗

14

HC

168

30

14675

∗154

56

300 250 200 150 100 50 0 -50

∗

∗

∗

ppm

127

P400153∗

∗

P300

168

∗

30

128

14147

Fig. 4 – Stolid-state 13C NMR spectra of raw hydrochar (HC)

and PC samples generated at temperatures ranging from

300 �C to 700 �C (Asterisks mean spinning side bands).

C A R B O N 7 7 ( 2 0 1 4 ) 6 2 7 – 6 3 6 633

which indicated unconverted components of biomass during

the HTC process. The 13C NMR spectrum revealed that the

hydrochar sample was less aromatic in nature.

After heating, there was a clear change in the structure of

the PC samples. In the NMR spectrum of the P300 sample, the

signals at 154 ppm (phenolic group), 75 ppm (cellulose) and

56 ppm (lignin) completely disappeared; and, a new peak

became evident at 128 ppm, which can be assigned to

C- and H- substituted aromatic C, indicating the breakdown

of residual lignocellulose and HTC products. When the heat-

ing temperature increased to 400 �C, only a peak at 127 ppm

(aromatic C) remained in the PC sample, indicating the evolu-

tion of aromaticity. However, the peak at 153 ppm again

appeared, due to the transformation of lignin and cellulose.

At the higher activation temperatures (500, 600 and 700 �C),

spectra with only aromatic C were observed, due to its

increasing degree of carbonization. This also reflected that

0 5 10 15 20 25 30 35

0

5

10

15

20

25

30

HC

P300

P400P500

P600

Amou

nt S

orbe

d (q

e, m

g/g)

Equilibrium Concentration (Ce, mg/L)

P700a

Fig. 5 – Adsorption isotherms of TC for raw hydrochar (HC) and P

700 �C in aqueous solution.

the development of aromatic structure was greatly dependent

on the activation temperature. From the series of spectra,

spinning sidebands associated with the aryl peak (marked

with an asterisk) was clear, which occurred as the rate of

magic angle spinning became insufficient to overcome the

chemical shift anisotropy [36]. Overall, the NMR-derived char-

acteristics of the PC samples were in keeping with the data

from other characterization techniques (such as element

analysis, Boehm titration, FTIR).

3.2. Adsorption of tetracycline

3.2.1. Adsorption isothermsThe adsorption isotherms of TC for the raw hydrochar and PC

samples are presented in Fig. 5. The adsorption data were

fitted to the Freundlich model, Qe ¼ KFCne , where Qe (mg/g)

and Ce (mg/L) are the adsorbed and aqueous-phase

concentrations, respectively, at adsorption equilibrium; KF

(mg1�nLn/g) is the Freundlich constant indicating adsorptive

capacity; n (unitless) is the Freundlich linearity index

[6,14,15]. The fitting parameters are summarized in Table S2,

which shows that the Freundlich model reasonably fit all

adsorption data (R2 > 0.94).

Notably, the n values were <1 in all cases, thereby suggest-

ing nonlinearity in the isotherms [14]. The Freundlich adsorp-

tive capacity (KF) increased largely with increasing activation

temperatures, reflecting the high adsorption capacity of the

PC sample produced at 700 �C. The increase from HC to the

P300 sample was modest. In particular, KF increased dramat-

ically from the P300 to P400 samples, possibly due to the con-

siderable increase in surface area, which may have enhanced

the diffusion of TC into the well-developed pores [37].

3.2.2. Correlation between sample properties and TCadsorptionIt has been well documented in the literature that TC adsorp-

tion on carbonaceous material is driven by a combination of

pore-filling effect, nonspecific van der Waals forces and spe-

cific p–p electron–donor–acceptor (EDA) interaction with the

graphite surfaces mechanism [6,38]. Hence, the adsorption

behavior of carbonaceous material for TC may depend on

its various properties. To gain a direct comparison of the

-0.6 -0.3 0.0 0.3 0.6 0.9 1.2 1.5

-0.6

-0.3

0.0

0.3

0.6

0.9

1.2

1.5

P700

P600 P500 P400

P300

log

qe

log Ce

HC

b

C samples generated at temperatures ranging from 300 �C to

634 C A R B O N 7 7 ( 2 0 1 4 ) 6 2 7 – 6 3 6

adsorption capacity between different samples, the KF and

Qe20 (taken from the adsorption isotherm at a constant TC

equilibrium concentration of 20 mg/L) were plotted against

the various properties of the samples, including the H/C,

O/C and (O + N)/C atomic ratios, surface area, total pore vol-

ume and micropore volume (Fig. 6 and Fig. S5). Thus, under

a given activation condition, the adsorption parameters could

be used to estimate the activation temperature of hydrochar

[8].

As shown in Fig. 6, a linear decrease in the KF value as a

function of H/C, O/C, (O + N)/C atomic ratios was obtained,

whereas a linear increase in the KF value as a function of the

surface area, total pore volume and micropore volume was

achieved (resulting from the pore-filling effect). Correspond-

0.2 0.4 0.6 0.8 1.0

0

2

4

6

8

K F (m

g1-n Ln /g

)

P700

P600

P500

P400

P300

H/C Atomic Ratio

y=-13.88x+12.32R2=0.94

HC

0.0 0.1 0.2 0.3

0

2

4

6

8

K F (m

g1-n Ln /g

)

P700

P600

P500

P400

(O+N)/C Atomic Ratio

HC

P300

y=-26.11x+8.95R2=0.96

0.05 0.10 0.15 0.20 0.25

0

2

4

6

8

K F (m

g1-n Ln /g

)

P700P600

P500

P400

P300

Total Pore Volume (cm3/g)

HC

y=40.37x-2.06R2=0.92

Fig. 6 – Correlation between Freundlich adsorptive capacity (KF)

atomic ratio, surface area, the total pore volume and the microp

temperatures ranging from 300 �C to 700 �C.

ingly, the polarity (O/C and (O + N)/C atomic ratios) was the

best for correlation with the KF value (R2 = 0.96). It should be

emphasized that TC is bulky molecule and, as expected,

induced a marked size-exclusion effect when adsorbing on

carbonaceous material [6]. Accordingly, the micropore volume

was the worst for correlation with the KF value (R2 = 0.87).

To further elucidate the TC adsorption, the Qe20 value was

also correlated with the properties of the samples, as shown

in Fig. S5. A negative correlation between the Qe20 values

and aromaticity index (H/C ratios) was observed, suggesting

that EDA interaction could occur between the PC sample

and the TC molecule. TC molecules are strong p-acceptors,

due to their ketone functional groups; and, a high-

temperature PC sample can act as a p-donor, due to graphitic

0.0 0.1 0.2 0.3

0

2

4

6

8

K F (m

g1-n Ln /g

)

P700

P600

P500

P400

P300

O/C Atomic Ratio

y=-26.53x+8.59R2=0.96

HC

0 75 150 225 300

0

2

4

6

8P700

P600

P500

P400

P300

K F (m

g1-n Ln /g

)

Surface Area (BET, m2/g)

HC

y=0.02x-0.04R2=0.91

0.00 0.04 0.08 0.12

0

2

4

6

8

K F (m

g1-n Ln /g

)

P700P600

P500P400

P300

Micropore Volume (cm3/g)

HC

y=-50.77x+0.76R2=0.87

and the H/C atomic ratio, the O/C atomic ratio, the (O + N)/C

ore volume of raw hydrochar and PC samples generated at

C A R B O N 7 7 ( 2 0 1 4 ) 6 2 7 – 6 3 6 635

carbon. However, the large amount of carboxyl groups in raw

hydrochar and low-temperature PC samples, such as P300,

also can act as p-acceptors [6,39]. Consequently, a stronger

EDA interaction occurred between the TC molecule (p-accep-

tor) and a high-temperature PC sample (p-donor) than with

low-temperature PC samples.

A negative correlation between Qe20 values and polarity

index (O/C atomic ratios) was also observed, which indicated

that the TC adsorption increased with a decrease in the polar-

ity of the PC samples. Hence, higher polar functional groups

in a PC sample may have inhibited TC adsorption, due to

the formation of larger and denser water molecule clusters

as a result of hydrogen bonding and hydrophobic interaction

playing an important role during the TC adsorption [39].

High surface areas and pore volumes of carbonaceous

material generally promote organic matter adsorption, due

to the pronounced pore-filling effect. Accordingly, a positive

correlation was obtained between the Qe20 values and the sur-

face area, with similar trends was observed for total pore vol-

ume and micropore volume. Interestingly, the total pore

volume was the best for correlation with the Qe20, rather than

the micropore volume. This can be attributed to the TC mol-

ecules exhibiting a size-exclusion effect due to their relatively

large size.

A similar result was observed in the correlation analysis of

the KF value. The properties of the PC samples influenced by

the activation temperature further affected their adsorption

behaviors. Overall, with increasing activation temperatures,

the increase of aromaticity, surface area and pore volume

and decrease of polarity of samples were observed, and ulti-

mately promoted the adsorption capacities for the removal

of TC.

4. Conclusions

The results from this study indicate that hydrochar derived

from the HTC of waste biomass can be converted into porous

carbon (PC) that shows promise as an adsorbent for the

removal of environmental antibiotics, implying that hydro-

char can potentially serve as a remediation agent.

In the present study, it was shown that the properties and

adsorption behaviors of hydrochar-derived PC depended

strongly on the formation conditions. The results indicated

increases in pH, ash contents, surface area, pore volume,

and aromaticity and a decrease in polarity of PC sample pro-

duced at high temperature. The PC samples obtained at high

temperature were well carbonized and showed a relatively

high adsorption capacity in removing TC from an aqueous

solution. Further, good correlations between adsorption

capacity and aromaticity index (H/C), polarity index (O/C,

(O + N)/C) and porosity (surface area, total pore volume and

micropore volume) were observed.

Acknowledgments

The authors are thankful for the financial support from the

Shanghai Science and Technology Committee. The authors

also thank the anonymous reviewers for fruitful suggestions.

Appendix A. Supplementary data

Supplementary data associated with this article can be found,

in the online version, at http://dx.doi.org/10.1016/j.carbon.

2014.05.067.

R E F E R E N C E S

[1] Sevilla M, Fuertes A, Mokaya R. High density hydrogenstorage in superactivated carbons from hydrothermallycarbonized renewable organic materials. Energy Environ Sci2011;4:1400–10.

[2] Libra JA, Ro KS, Kammann C, Funke A, Berge ND, Neubauer Y,et al. Hydrothermal carbonization of biomass residuals: acomparative review of the chemistry, processes andapplications of wet and dry pyrolysis. Biofuels 2011;2:71–106.

[3] Titirici M-M, White RJ, Falco C, Sevilla M. Black perspectivesfor a green future: hydrothermal carbons for environmentprotection and energy storage. Energy Environ Sci2012;5:6796–822.

[4] Falco C, Marco-Lozar J, Salinas-Torres D, Morallon E, Cazorla-Amoros D, Titirici M, et al. Tailoring the porosity ofchemically activated hydrothermal carbons: influence of theprecursor and hydrothermal carbonization temperature.Carbon 2013;62:346–55.

[5] Sun Y, Gao B, Yao Y, Fang J, Zhang M, Zhou Y, et al. Effects offeedstock type, production method, and pyrolysistemperature on biochar and hydrochar properties. Chem EngJ 2014;156:243–9.

[6] Ji L, Wan Y, Zheng S, Zhu D. Adsorption of tetracycline andsulfamethoxazole on crop residue-derived ashes: implicationfor the relative importance of black carbon to soil sorption.Environ Sci Technol 2011;45:5580–6.

[7] Jia M, Wang F, Bian Y, Jin X, Song Y, Kengara FO, et al. Effectsof pH and metal ions on oxytetracycline sorption to maize–straw-derived biochar. Bioresour Technol 2013;136:87–93.

[8] Chen B, Zhou D, Zhu L. Transitional adsorption and partitionof nonpolar and polar aromatic contaminants by biochars ofpine needles with different pyrolytic temperatures. EnvironSci Technol 2008;42:5137–43.

[9] Wang YJ, Jia DA, Sun RJ, Zhu HW, Zhou DM. Adsorption andcosorption of tetracycline and copper (II) on montmorilloniteas affected by solution pH. Environ Sci Technol2008;42:3254–9.

[10] Zhu XD, Wang YJ, Sun RJ, Zhou DM. Photocatalyticdegradation of tetracycline in aqueous solution by nanosizedTiO2. Chemosphere 2013;92:925–32.

[11] Liu P, Liu WJ, Jiang H, Chen JJ, Li WW, Yu HQ. Modification ofbio-char derived from fast pyrolysis of biomass and itsapplication in removal of tetracycline from aqueous solution.Bioresour Technol 2012;121:235–40.

[12] Toor SS, Reddy H, Deng S, Hoffmann J, Spangsmark D,Madsen LB, et al. Hydrothermal liquefaction of Spirulina andNannochloropsis salina under subcritical and supercriticalwater conditions. Bioresour Technol 2013;131:413–9.

[13] Yuan J-H, Xu R-K, Zhang H. The forms of alkalis in thebiochar produced from crop residues at differenttemperatures. Bioresour Technol 2011;102:3488–97.

[14] Ahmad M, Lee SS, Dou X, Mohan D, Sung JK, Yang JE, et al.Effects of pyrolysis temperature on soybean stover-andpeanut shell-derived biochar properties and TCE adsorptionin water. Bioresour Technol 2012;118:536–44.

[15] Chun Y, Sheng G, Chiou CT, Xing B. Compositions andsorptive properties of crop residue-derived chars. Environ SciTechnol 2004;38:4649–55.

636 C A R B O N 7 7 ( 2 0 1 4 ) 6 2 7 – 6 3 6

[16] Wang YJ, Jia DA, Sun RJ, Zhu HW, Zhou DM. Adsorption andcosorption of tetracycline and copper(II) on montmorilloniteas affected by solution pH. Environ Sci Technol2008;42:3254–9.

[17] Yang H, Yan R, Chen H, Zheng C, Lee DH, Liang DT. In-depthinvestigation of biomass pyrolysis based on three majorcomponents: hemicellulose, cellulose and lignin. EnergyFuels 2006;20:388–93.

[18] Fuertes A, Arbestain MC, Sevilla M, Macia-Agullo J, Fiol S,Lopez R, et al. Chemical and structural properties ofcarbonaceous products obtained by pyrolysis andhydrothermal carbonisation of corn stover. Soil Res2010;48:618–26.

[19] Al-Wabel MI, Al-Omran A, El-Naggar AH, Nadeem M,Usman AR. Pyrolysis temperature induced changes incharacteristics and chemical composition of biocharproduced from conocarpus wastes. Bioresour Technol2013;131:374–9.

[20] Boehm HP. Surface oxides on carbon and their analysis: acritical assessment. Carbon 2002;40:145–9.

[21] Keiluweit M, Nico PS, Johnson MG, Kleber M. Dynamicmolecular structure of plant biomass-derived black carbon(biochar). Environ Sci Technol 2010;44:1247–53.

[22] Roman S, Valente Nabais J, Ledesma B, Gonzalez J, LaginhasC, Titirici M. Production of low-cost adsorbents with tunablesurface chemistry by conjunction of hydrothermalcarbonization and activation processes. MicroporousMesoporous Mater 2012;165:127–33.

[23] Brunauer S, Deming LS, Deming WE, Teller E. On a theory ofthe van der Waals adsorption of gases. J Am Chem Soc1940;62:1723–32.

[24] Sevilla M, Fuertes AB. The production of carbon materials byhydrothermal carbonization of cellulose. Carbon2009;47:2281–9.

[25] Liu Z, Zhang F-S. Removal of copper (II) and phenol fromaqueous solution using porous carbons derived fromhydrothermal chars. Desalination 2011;267:101–6.

[26] Ahmad M, Lee SS, Rajapaksha AU, Vithanage M, Zhang M,Cho JS, et al. Trichloroethylene adsorption by pine needlebiochars produced at various pyrolysis temperatures.Bioresour Technol 2013;143:615–22.

[27] Cao X, Ma L, Liang Y, Gao B, Harris W. Simultaneousimmobilization of lead and atrazine in contaminated soilsusing dairy-manure biochar. Environ Sci Technol2011;45:4884–9.

[28] Nanda S, Mohanty P, Pant KK, Naik S, Kozinski JA, Dalai AK.Characterization of north American lignocellulosic biomassand biochars in terms of their candidacy for alternaterenewable fuels. BioEnergy Res 2013;6:663–77.

[29] Chen Z, Chen B, Chiou CT. Fast and slow rates of naphthalenesorption to biochars produced at different temperatures.Environ Sci Technol 2012;46:11104–11.

[30] Uchimiya M, Wartelle LH, Klasson KT, Fortier CA, Lima IM.Influence of pyrolysis temperature on biochar property andfunction as a heavy metal sorbent in soil. Agric Food Chem2011;59:2501–10.

[31] Kloss S, Zehetner F, Dellantonio A, Hamid R, Ottner F, LiedtkeV, et al. Characterization of slow pyrolysis biochars: effectsof feedstocks and pyrolysis temperature on biocharproperties. Environ Qual 2012;41:990–1000.

[32] Millan A. Crystal growth shape of whewellite polymorphs:influence of structure distortions on crystal shape. CrystGrowth Des 2001;1:245–54.

[33] Yu L, Falco C, Weber J, White RJ, Howe JY, Titirici M-M.Carbohydrate-derived hydrothermal carbons: a thoroughcharacterization study. Langmuir 2012;28:12373–83.

[34] Wiedner K, Naisse C, Rumpel C, Pozzi A, Wieczorek P, GlaserB. Chemical modification of biomass residues duringhydrothermal carbonization-what makes the difference,temperature or feedstock? Organic Geochem 2012;54:91–100.

[35] Aydıncak K, Yumak Tr, Sınag A, Esen B. Synthesis andcharacterization of carbonaceous materials from saccharides(glucose and lactose) and two waste biomasses byhydrothermal carbonization. Ind Eng Chem Res2012;51:9145–52.

[36] McBeath AV, Smernik RJ, Schneider MP, Schmidt MW, PlantEL. Determination of the aromaticity and the degree ofaromatic condensation of a thermosequence of woodcharcoal using NMR. Organic Geochem 2011;42:1194–202.

[37] Kim W-K, Shim T, Kim Y-S, Hyun S, Ryu C, Park Y-K, et al.Characterization of cadmium removal from aqueous solutionby biochar produced from a giant Miscanthus at differentpyrolytic temperatures. Bioresour Technol 2013;138:266–70.

[38] Ji L, Chen W, Duan L, Zhu D. Mechanisms for strongadsorption of tetracycline to carbon nanotubes: acomparative study using activated carbon and graphite asadsorbents. Environ Sci Technol 2009;43:2322–7.

[39] Zheng H, Wang Z, Zhao J, Herbert S, Xing B. Sorption ofantibiotic sulfamethoxazole varies with biochars produced atdifferent temperatures. Environ Pollut 2013;181:60–7.