Removal of pharmaceutical- containing wastewater by bio- and hydrochar adsorbents
Adsorption capacity and surface functionalities
Alexandra Charlson
Alexandra Charlson Master Thesis 60 ECTS Report passed: 20 June 2017 Supervisor: Stina Jansson Examiner: Lars Backman
I
Abstract
Water scarcity, a lack of adequate sanitation and poor-quality drinking water are some of the major problems that developing countries face. Wastewater remediation by a sustainable and low-cost adsorbent is what bio- and hydrochar can offer. In this project, the adsorption capacity of agricultural by-products transformed into bio- and hydrochar was investigated. Five different agricultural by-product feedstocks and a faecal simulant were transformed into bio- and hydrochar. Two experiments were performed to evaluate the remediation efficiency of the chars; (i) kinetic adsorption and (ii) sorption isotherms. From the kinetic adsorption experiment horse manure chars prepared by torrefaction were found to be the most versatile chars remediating all 12 pharmaceuticals sufficiently. Neither torrefaction nor hydrothermal carbonization was observed to be the best carbonization technique, in concern to removal efficiency, for all feedstocks, however 230 oC was the best temperature for both techniques. Sorption isotherms revealed which chars have the largest maximum capacities (KF) and are more favorable to adsorption (n). Two pharmaceuticals were analyzed, atenolol (polar) and carbamazepine (non-polar). Atenolol was removed with the highest efficiency by a torrefaction horse manure char and carbamazepine had the best removal efficiency by a hydrothermal carbonization char made from Raphia Farinifera seed capsules. The surface characteristics of the bio- and hydrochar samples were analysed by diffuse reflectance Fourier transformation spectroscopy, X-ray photoelectron spectroscopy and Brunauer-Emmett-Teller surface analysis. The diffuse reflectance Fourier transformation spectroscopy results reveled that functional groups were present in different quantities for the different chars and that there were no trends for functional group quantities in respect to carbonization technique and temperature. General observations from the X-ray photoelectron spectroscopy data showed that with increasing temperatures the amount of oxygen and trace elements diminish. Less trace elements were found on the surface of hydrothermal carbonized chars. Rice husk and horse manure hydrochars were found to have the highest surface area in the series (12 – 49 m2/g). The remaining chars had small surface areas (<2.5 m2/g). The results suggest that most agricultural by-products either transformed into bio- and hydrochar are suitable for wastewater treatment. Surface characteristics show that oxygen rich adsorbents have a higher affinity for adsorption. These results are promising as the low-cost and sustainable production of char adsorbents could help clean water for better sanitation worldwide.
Keywords: Biochar, hydrochar, wastewater remediation, kinetic adsorption, sorption isotherm
II
III
List of abbreviations
AGA Aktiebolaget Gasaccumulator
BET Brunauer-Emmett-Teller
CAS Chemical Abstracts Service
CEC Carbon exchange capacity
DL Detection limit
DRIFTS Diffuse reflectance infrared Fourier transform spectroscopy EDS Energy-dispersive X-ray spectroscopy
FESEM Field Emission Scanning Electron Microscope HDBSD High definition backscattered electrons HTC Hydrothermal carbonization
KF Freundlich affinity constant
LOD Limit of detection
logP Logarithm of polarity logS Logarithm of solubility
MS Mass spectrometry
n Freundlich exponential constant
NSAID Non-steroidal anti-inflammatory drug
OSPE-LC-MS/MS Online solid phase liquid chromatography tandem mass spectrometry
PAH Polycyclic aromatic hydrocarbons pKa Acid dissociation constant
PZC Point of zero charge
QL Limit of quantification
RSD Relative standard deviation SEM Scanning electron microscopy
SPE Solid phase extraction
UNDP United Nations Development Programme
WHO World Health Organization
WWTP Wastewater treatment plant XPS X-ray photoelectron spectroscopy
IV
V
Table of contents
Abstract ... I
1. Introduction ... 1
1.2 Aim of the diploma work ... 2
2. Popular scientific summary including social and ethical aspects ... 2
2.1 Popular scientific summary ... 2
2.2 Social and ethical aspects ... 3
3. Survey of the field ... 4
3.1 Carbonization techniques ... 4
3.2 Characterization of char ... 4
3.3 Remediation of organic contaminants in wastewater ... 5
3.4 Analysed pharmaceuticals ... 6
3.5 Char feedstocks ... 8
4. Experimental ... 9
4.1 Human fecal simulant ... 9
4.2 Preparation of bio- and hydrochar ... 9
4.3 Sorption tests ... 10
4.3.1 Rotating bed reactor ... 10
4.3.2 Kinetic adsorption ... 11
4.3.3 Sorption isotherm ... 11
4.3.4 Analysis of contaminates with online SPE LC-MS-MS ... 11
4.3.5 Data analysis ... 12
4.4 Characterization of adsorbents ... 13
4.4.1 DRIFTS ... 13
4.4.2 XPS ... 14
4.4.3 BET surface area and pore volume ... 14
4.4.4 SEM ... 14
5. Results and Discussion ... 14
5.1 Sorption tests ... 14
5.1.1 Rotating bed reactor ... 14
5.1.2 Kinetic adsorption ... 15
5.1.2.1 Ampicillin ... 17
5.1.2.2 Atenolol ... 17
5.1.2.3 Bezafibrate ... 18
5.1.2.4 Carbamazepine ... 18
5.1.2.5 Ciprofloxacin ... 18
5.1.2.6 Diclofenac ... 18
5.1.2.7 Erythromycin ... 18
5.1.2.8 Ketoprofen ... 18
5.1.2.9 Metronidazole ... 19
5.1.2.10 Nalidixic acid ... 19
5.1.2.11 Sulfametazole ... 19
5.1.2.12 Trimethoprim ... 19
5.1.2.13 Total relative removal efficiency ... 19
5.1.3 Sorption isotherm ... 20
5.1.3.1 Sorption isotherm relative removal efficiencies ... 23
5.2 Characterization of adsorbents ... 24
5.2.1 DRIFTS ... 25
5.2.2 XPS ... 26
VI
5.2.3 BET surface area and pore volume ... 28
5.2.4 SEM ... 29
5.3 Surface functionalities influence on adsorption capacity ... 31
5.3.1 Kinetic adsorption ... 31
5.3.2 Sorption isotherm ... 31
6. Conclusions ... 31
7. Outlook ... 32
Acknowledgement ... 33
References ... 33
Appendix ... 1
Appendix 1 – Kinetic adsorption ... 1
Appendix 2 – Sorption isotherm plots ... 1
Appendix 3 – Diffuse reflectance Fourier transform spectra ... 1
Appendix 4 – X-ray photoelectron spectroscopy spectra ... 1
Appendix 5 – Energy-dispersive X-ray spectroscopy ... 1
1
1. Introduction
As the human population increases the gap between water supply and demand widens.
Thus, the reuse of, in many cases, untreated wastewater is used to irrigate arable land in developing countries. The water carries with it harmful pathogens as well as organic and inorganic contaminants which may pose a risk for human health and the environment1,2. An environmentally friendly and cost-efficient way of treating the untreated wastewater, or the land that the wastewater is applied to, from these unwanted contaminants is by remediation with bio- and/or hydrochar. It is used as an adsorbent that captures contaminants and makes the practice of using untreated wastewater for irrigation purposes safer.
Bio- and hydrochar are two hot topics within remediation research as they are renewable, cheap and sustainable adsorbents3. Biochar is a solid material obtained from the dry thermochemical conversion of biomass in an oxygen-limited environment4. Whereas, hydrochar is wet slurry which is produced from a wet thermochemical process5. In this thesis, the name biochar will be used for chars produced by torrefaction and hydrochar will be used for chars produced by hydrothermal carbonization (HTC).
This project is tied to a larger VR (vetenskapsrådet – the Swedish research council) project where the aim is to capture and immobilize pollutants in wastewater in Africa using biochar from local crops. A small-scale decentralized wastewater treatment solution (DEWATS)6 is one of the main removal systems in the project. A DEWATS can be built and placed anywhere and is thus a useful alternative to a conventional waste water treatment plant (WWTP). Biochar will be used as an adsorbent in the system that captures pollutants from the influent to ensure a cleaner effluent. To ensure that the system is sustainable and cost-efficient local crop residues will be transformed into biochar and used to remediate the wastewater from pollutants.
The Bremen Overseas Research and Development association (BORDA) are the developers behind the cost-efficient and low-maintenance DEWATS systems6. The DEWATS systems are designed for densely populated urban areas where there is no or little wastewater treatment. As an uncontrolled discharge of wastewater is a dangerous for both human health and the environment the DEWATS system is a reliable and efficient option to traditional WWTPs6. The design of a DEWATS system is dependent on the location of the system and what kind of wastewater is to be treated. In many cases DEWATS systems use wetlands to remediate various types of wastewaters as it is a cost-efficient and low-maintenance remediator. Biochars can be added on or in the media of the wetland to enhance the removal of pollutants. In Figure 1 a picture of a DEWATS located in Newlands (north of Durban), South Africa can be observed7. The DEWATS system treats wastewater from approximately 80 low- to medium income households.
Figure 1. A decentralized wastewater treatment solution (DEWATS) system at the Newlands Mashu research site, north of Durban, South Africa7. In the picture the wetland portion of the DEWATS system can be seen.
2 Wastewater contains a multitude of pollutants including organic compounds, metals and other inorganic compounds. Prior research has found that different surface characteristics on the char are necessary for the removal of different compounds. For example, more polar compounds are favourably adsorbed by char rich in oxygen functional groups whereas more non-polar compounds are more readily adsorbed by carbon rich chars8. Different carbonization techniques, temperatures, retention times and feedstocks cause bio- and hydrochar to have different surface characteristics from each other5. Many kinds of feedstocks can be transformed into char. These includes agricultural by-products, such as peels, husks, manure and so on, that otherwise would have been discarded. Depending on the type of biomass used and the char production method the characteristics of the biochar vary. This makes it optimal for the adsorption of some contaminants but not others8,9,10,11.
This project focuses on the remediation of 13 pharmaceuticals from wastewater with bio- and hydrochar. The surface functionalities and trace elements found on the surface of the chars were characterised to see if there are any trends between removal ability and surface functionalities.
1.2 Aim of the diploma work
This project focuses on the remediation of selected pharmaceuticals from wastewater by bio- and hydrochar. The surface functionalities of the produced chars were characterised. The aim is to display the functionalities of char, i.e. how well it can capture various pollutants and at which concentrations it is saturated. The results will be used to determine useful bio- and hydrochar preparation techniques and suggestable feedstocks for the application of char to a DEWATS system in Africa. Topics that were investigated throughout the project were:
i. The adsorbent capacity of various chars.
ii. Are the adsorbent capacities different for bio- and hydrochars?
iii. How do the characteristics vary between bio- and hydrochar?
iv. Which characteristics are optimal for adsorbing organic contaminants, in specific pharmaceuticals?
v. Which surface characteristics influence the adsorption of polar and non-polar contaminants?
vi. Is human excrement transformed into hydrochar a good adsorbent?
vii. Which feedstocks transformed into char are good remediators?
2. Popular scientific summary including social and ethical aspects
2.1 Popular scientific summary
Water scarcity is one of the major issues that many developing are faced with today. To reduce this burden untreated wastewater is applied to arable land. The use of this type of water is both a blessing and a curse as it is a plentiful source of water but also carries with it harmful pathogens as well as organic and inorganic pollutants. These contaminants may pose a risk for human health and the environment1,2. An environmentally friendly and cost-efficient way of treating the untreated wastewater, or the land that the wastewater is applied to, from these unwanted contaminants is by remediation with bio- and/or hydrochar. It is used as an adsorbent that captures contaminants and makes the practice of using untreated wastewater for irrigation purposes safer.
The aim of this project was to analyse how well various agricultural by-products converted into bio- and hydrochar adsorb contaminants from a simulated wastewater spiked with selected pharmaceuticals. Agricultural by-products that otherwise would have been discarded were transform into char though torrefaction, a dry carbonisation technique, and HTC, a wet carbonisation technique. The surface chemistry of the bio- and hydrochars were analysed by several techniques to determine which functional
3 groups, such as alcohols, were present, how big the surface area was and if there were any trace elements, such as iron, present on the surface.
Selected pharmaceuticals that are commonly used in Africa and found in African rivers were subject to two experiments with the bio- and hydrochar. In the kinetic experiment a constant concentration of the chars was used and the removal efficiencies were analysed after certain time intervals between 0 and 120 minutes. In the sorption isotherm experiment a set time of 48 hours was used and different concentration of two pharmaceuticals, one polar and one non-polar, was analysed. The results from both experiments revealed that surface characteristics, the type of feedstock used and retention time are important factors that determine how well a char can remediate the wastewater. In this study horse manure produced at 230 oC by torrefaction proved to be the best adsorbent as it could capture a majority of the pharmaceuticals in the kinetic experiment and removed 99 % of the initial concentration in the sorption isotherm experiment.
As the results found that all the feedstocks produced chars favourable to adsorb most pharmaceuticals it gives hope that the results found in this project can hopefully be applied to large scale production of biochar in African countries for the remediation of contaminated wastewater. One can also aspire that the results can be applied globally as biochar is a sustainable and cost-efficient adsorbent that is made from unwanted agricultural and/or industrial by-products.
2.2 Social and ethical aspects
The use of untreated wastewater is both a blessing and a curse as it is a much-needed source of water and brings wanted nutrients to the land, but on the other hand it also brings with it harmful contaminants such as pathogens, heavy metals and organic pollutants. Uptake of these pollutants may pose a threat to human health as well as causing damage to the environment1,2. The main application of wastewater, treated or untreated, is to arable land. In 2013 it was reported that 20 million hectares of arable land were irrigated by wastewater1. In developing countries though there are set regulations for the application of wastewater they are not always met due to a lack of funds or technology. Therefore, wastewater in its untreated state is commonly used to water crops.
The United Nations Development Programme (UNDP) wrote in their Human Development Report from 2006 that ‘over 1 billion people are denied the right to clean water and 2.6 billion people lack access to adequate sanitation’. It was also reported that ‘1.8 million children die from diarrhoea and other diseases caused by unclean water and poor sanitation’12. Conventional WWTPs are not designed for the remediation of pharmaceuticals or their metabolites, thus WWTPs are commonly pointed out as major point of emission for pharmaceutically active compounds13. Research by Kasprzyk-Horden and co-workers found that carbamazepine was not removed by conventional WWTPs in Wales14. They also recorded, among others, the removal percentages for the following pharmaceuticals: atenolol 10 – 55 %, bezafibrate 67 %, metronidazole 23 %, erythromycin 50 %, sulfamethoxazole 70 % and diclofenac 33 %. In this study chloramphenicol was the only antibiotic that was totally removed (100 %) in a conventional WWTP14.
It is important to remediate pharmaceuticals from wastewater due to antibiotic resistance and the possible uptake of pharmaceuticals by humans, animals and/or plants. The World Health Organization (WHO) state that ‘antibiotic resistance is one of the biggest threats to global health today’15. Gao and colleagues studied the bioaccumulation of antibiotics in fish and found elevated concentrations in fish muscles compared to their surrounding water making them conclude that it is plausible that fish accumulate certain antibiotics16. A paper published in 2004 revealed that diclofenac (a NSAID) was the cause of a declining, then listed as critically endangered, vulture population in Pakistan17. Diclofenac has also been found to accumulate in rainbow trout18. The environmental fate of many pharmaceuticals is still unknown and there is a possible risk for adverse effects. As human welfare improves so does the use of pharmaceuticals, and therefore the residues of pharmaceuticals increase in wastewater.
4 Thus, it is of the upmost importance to resolve these issues as access to clean water and sanitation is a human right.
3. Survey of the field
3.1 Carbonization techniques
There are many carbonization techniques used to produce char, e.g. pyrolysis, gasification, torrefaction, and HTC3,5,8,10. This project will only focus on torrefaction and HTC. Torrefaction is a mild pyrolysis technique with a low temperature ranging from 200 - 300 oC19. The residence time, for lab scale production, is generally 30 minutes to several hours5. The yield of a solid product is approximately 60 – 90 % and a mix of volatile gas5,20. As the temperature used during torrefaction is low compared to those used during conventional pyrolysis (300 – 650 oC)5 the solid products physicochemical properties are described to be in-between those of a raw biomass and a biochar5. HTC can be referred to as a wet torrefaction method that transforms wet biomass into a char slurry3,5,8. The biomass is placed in a confined system, water is used as a solvent and a self-generated pressure is formed21. Typically, temperatures range from 180 – 260 oC and the pressure is between 2 – 6 MPa. For lab scale production the residence time generally varies between 5 minutes to 12 hours5. Once the carbonization step is complete the solid hydrochar is separated (from the char slurry) by filtration. An advantage with HTC compared to more common carbonization techniques is that wet biomass can be applied directly into the HTC system5.
3.2 Characterization of char
The characteristics of biochar i.e. carbon rich, porous with oxygen functional groups and aromatic surfaces allows it to adsorb both organic and inorganic contaminants9,10. The physicochemical properties of biochar are governed by which biomass is used, the carbonisation technique and the reaction temperature. The residence time has less impact on the surface characteristics5,8,22. Parameters that govern the adsorption capacity of char are the inorganic fraction, textural characteristics and morphology, surface functionalities, aromatic fraction, pH and cation exchange capacity (CEC)8,10. An increase in temperature and residence time tends to increase the carbon percentage while decreasing the oxygen percentage of the char. Hydrogen-containing functional groups also decrease with temperature which causes the surface to be less polar and thus hydrophilic due to the higher extent of carbonization3,8,10. These biochars seem to be more efficient at adsorbing inorganic contaminants whereas chars made at lower temperatures (which have a higher surface area and more stable carbon oxygen complexes) are more suitable for organic contaminants8,9.
These variations in biochar characteristics make it possible to tailor make biochars specific for certain contaminants. For example, when the pH is lower than the pH for the point of zero charge (PZC) then the surface of the biochar is positively charged.
Thus, anions are readily adsorbed. On the other hand, when the pH is higher than the pHPZC then the surface is negatively charged and cations are more favourably adsorbed by the char surface. The solution pH should preferably be in the higher range for optimal cation capture10.
The surface area of char is largely dependent on which type of biomass is used.
Chars made from crop residues and wood biomass tend to have a larger surface area than those made from animal litter and solid waste (e.g. waste tire rubber23)8. Downie and co-workers stated that there is a positive correlation between the micropore volume and surface area24. This indicates that the pore volume is a key factor for the increase in surface area for chars24.
As biochar and hydrochar are made under different conditions the characteristics also vary. The production of hydrochar is initiated by hydrolysis and due to the elevated temperatures, the organic matter is subsequently dehydrated, decarboxylated and can lose functional groups. Aromatization occurs at the end of the process25. Biochar prepared at a low temperature have a graphite like structure and a lot of aromatic groups. Compared to biochar, hydrochar tends to have a more spherical morphology5,21, fewer aromatic groups and have a larger verity of alkyl moieties26. A drawback with
5 hydrochar is that it has a lower porosity and smaller surface area than biochar tends to have5,21. This can be compensated for by employing an activation technique, either physical or chemical. Physical activation is acquired by subjecting the hydrochar to elevated temperatures (>900 oC) with a controlled flow of CO2. The CO2 reacts with the carbon and removes it from the hydrochar surface, opening clogged pores and thus increasing the active surface area. KOH is commonly added to the hydrochar to enhance surface properties5,21. KOH, or other activating agents e.g. K2CO321, are added at either room temperature or at elevated temperatures (600 – 800 oC). Kambo and Dutta explain the reaction mechanism as follows “during chemical activation potassium from precursor chemical separates the lamellae of crystallites from the C structure”5. Once the hydrochar is rinsed the potassium is washed off and a more porous surface structure is gained5.
3.3 Remediation of organic contaminants in wastewater
A multitude of studies where biochar has been used for the remediation of wastewater have been performed. Some examples of remediation of organic contaminants from water are shown in Table 1. As one can see in the table there are many variations in the types of biomass used to produce char, a multitude of temperature ranges as well as different adsorption mechanisms. This illustrates how versatile biochar is for the remediation of organic contaminants.
As displayed in Table 1 there are several mechanisms proposed for the adsorption of organic contaminants. These include electrostatic interactions, hydrogen bonding, π- π interactions, hydrophobic effect and pore filling, see Figure 2. Kumar et al. (2011) suggest that electrostatic attraction is the main mechanism of adsorption of organic contaminants9.
Figure 2. Adsorption mechanisms of organic contaminants on biochar, modified from Tan et al10. The grey spheres are organic contaminants and the green spheres are metals that are attached to the biochar.
In a study by Xu et al. methyl violet dye was successfully adsorbed onto various biochars27. The chars were made from the following crop residues; canola straw, peanut straw, soybean straw and rice hull. Rice hull char was the most efficient adsorbent due to its high negative surface area. Electrostatic interaction, interaction between the contaminant and the carboxylate and phenolic hydroxyl groups as well as surface precipitation are the methods used to adsorb the dye27. The results from this study shows that the type of biomass used has a large impact of on adsorption ability.
6
Table 1. Some examples of remediation of organic contaminants from water with the aid of char. The type of biomass used and suggested mechanisms applied for the adsorption are stated.
Containment Char type Function Reference
Atrazine and
simazine Green waste
biochar (450 oC)
Adsorption and partition 28
Naphthalene and
1-nappthol Orange peel
(150 – 700 oC) Adsorption and partition 29
Fluorinated
herbicides Plant biomass Adsorption due to electrostatic interactions, H-bonding and/or non-specific London forces and π-π electron donor-acceptor interaction
30
Bisphenol A,
17α-ethinylestradiol Poultry litter and wheat straw
Pore filling, H-bonding and π-π electron donor-acceptor interaction
31
Dibromo- chloropropane
Almond shell (650 – 800 oC)
Adsorption 32
Methyl violet dye Crop residues
(350 oC) Electrostatic interaction, interaction between the containment and the carboxylate and phenolic hydroxyl groups as well as surface precipitation are the methods used to adsorb the dye
27
Naphthalene, Nitrobenzene, m-dinitrobenzene
Pine needles
(100 – 700 oC) Adsorption and partition 33
Tetracycline Rice husks (450 – 500 oC)
Adsorption due to π-π electron donor-acceptor interaction between the ring structure of the tetracycline and the graphite like structures of the biochar
34
Sulfamethazine Hardwood
litter (600 oC) Adsorption due to π-π electron donor-acceptor interaction and negative charge assisted H- bonding
35
3.4 Analysed pharmaceuticals
Figure 3. A map of South Africa, the province of KwaZulu-Natal and the Umgeni river. The map was modified from Gakuba et al. 201536. As this project aims to remediate African wastewater with char adsorbents the selected pharmaceuticals used in this project were selected from two articles (37,38) where the authors sampled the Umgeni river, in South Africa, for selected pharmaceuticals. The specific sample locations can be found in the published articles.
This study aims to investigate the adsorbent capacity of bio- and hydrochar with selected pharmaceuticals. The choice of pharmaceuticals was based on two studies in
River flow direction
Umgeni river KwaZulu-Natal South Africa
7 KwaZulu-Natal, South Africa where the researchers investigated the concentrations of selected pharmaceuticals in the Umgeni river37,38. A map of South Africa, Kwazulu- Natal and the Umgeni river can be seen in Figure 3. Specific sample sites can be found in articles published from Agunbiade et al. (2014)37 and Matongo et al. (2015)38. From these studies eight antibiotics, two non-steroidal anti-inflammatory drugs (NSAID), one anti-epileptic, one anti-hyperlipidaemia and one β-blocker were selected. The majority of these drugs are classed as essential medicine, i.e. “those that satisfy the priority health care needs of the population”, by the Department of Health in South Africa39. In Table 2 the selected pharmaceuticals CAS registration number, molar mass, polarity (logP), solubility (logS), acidic dissociation constant (pKa) and concentration found in the influent water to a WWTP in KwaZulu-Natal are shown and in Figure 4 the structures of the antibiotics and other pharmaceuticals are illustrated.
Figure 4. The structures of two NSAIDs (diclofenac, and ketoprofen), one anti-epileptic (carbamazepine), one β-blocker (atenolol) and one anti-hyperlipidaemia (bezafibrate) are presented in the upper half of the figure. In the lower half, the structures of eight antibiotics screened in the South African studies are presented.
8
Table 2. Characteristics of eight antibiotics, two NSAIDs, one anti-epileptic, one anti-hyperlipidaemia and one β-blocker. Their CAS registration number and properties, including the solubility (logS), polarity (logP) and the acid dissociation constant (pKa) were gathered from Chemicalize40 and the pharmaceutical concentrations from the influent of a wastewater treatment plant (WWTP) in South Africa were taken from Agunbiade and Moodley37 and Matongo et al38.
Name Therapeutic
group
CAS number
Molar mass
(g/mol) LogS logP pKa
Influent conc. from a WWTP (µg/L) Ampicillin Antibiotic 69-53-4 349.41 -4.78 -2.00 3.24 10a Chloramphenicol Antibiotic 56-75-7 323.13 -3.24 0.88 8.69 10a Ciprofloxacin Antibiotic 85721-33-1 331.35 -2.32 -0.81 5.76 16a Erythromycin Antibiotic 114-07-8 733.93 -3.29 2.60 12.45 22a Metronidazole Antibiotic 99616-64-5 171.16 -0.41 0.46 15.41 Not detectedb Nalidixic acid Antibiotic 389-08-2 232.24 -1.98 1.01 5.95 31±3a Sulfamethoxazole Antibiotic 723-46-6 253.28 -2.25 0.79 6.16 10a Trimethoprim Antibiotic 738-70-5 290.32 -2.68 1.28 7.16a 0.13 ±2.78b Diclofenac NSAID 15307-86-5 296.15 -4.32 4.26 4.00 18a Ketoprofen NSAID 22071-15-4 254.29 -3.90 3.61 3.88 8a
Carbamazepine Anti-epileptic 298-46-4 236.27 -3.67 2.77 15.96 4.56 ±0.81b Bezafibrate Anti-hyper-
lipidaemia 41859-67-0 361.82 -4.88 3.99 3.83 8a Atenolol β-blocker 29122-68-7 266.34 -1.47 0.43 14.08 39±4a
a Concentrations analysed by Agunbiade and Moodley37.
b Concentrations analysed by Matongo et al38.
3.5 Char feedstocks
As shown in Table 1 many types of feedstocks can be used to produce char. Six types of feedstocks were used in this study and transformed into biochar and hydrochar. If these feedstocks were not transformed into char, a valuable remediation material, they would have been discarded. The feedstocks used were horse manure from a riding school in Vännäs, olive residues, tomato residues (peels and pips) and rice husks provided by the University of Alicante, Spain, Raphia farinifera seed capsules from Nigeria and a human faecal simulant prepared in Umeå. The faecal simulant was prepared with the aim to analyse whether faeces are a good adsorbent for pollutants. The feedstocks were left in their ‘natural states’ i.e. no alterations, such as grinding, were done before carbonisation. Pictures of the feedstocks pre-carbonisation are shown in Figure 5.
Figure 5. Feedstocks before being carbonised by torrefaction or hydrothermal carbonisation. Feedstocks in the top row (left to right): Horse manure, olive residues and tomato residues. Bottom row (left to right):
rice husks, Raphia Farinifera and the human faecal simulant.
As mentioned in the introduction this project is partly tied to research in KwaZulu-Natal South Africa. In this region, there are still populations which do not have access to a conventional toilet, i.e. with running water. Instead toilets may be a
9 hole in the ground which is emptied when necessary. To by-pass the need of a conventional toilet where the waste must be treated in a WWTP, or constructing more DEWATS systems, an idea is to transform the human excrement from these makeshift toilets into hydrochar. This concept would remove the need for a WWTP, reduce costs and a renewable remediation material in the form of hydrochar would be produced. A faecal simulant will be used to mimic the production of hydrochar produced from human excrement as it is not feasible to import human excrement from South Africa to Sweden. It is also unsanitary and special care would be necessary to prevent the spread of pathogens.
Although all these feedstocks are not from Africa, where the focus of this project lies, they are all large scale agricultural by-products. Using a wide verity of feedstocks that have different properties will produce chars with different surface functionalities.
From this the information from surface characterisation can show which properties are important for adsorption.
4. Experimental
The experimental part is divided into three sections: (i) production of biochar and hydrochar, (ii) sorption tests and (iii) characterization of char. In the following sections the methods and materials needed for these experiments are explained.
4.1 Human fecal simulant
As explained in section 3.5 a simulated human excrement was prepared, in Table 3 the
‘recipe’ for the faecal simulant is presented. The recipe for the human faecal simulant is from NASA and their Trash-to-gas project41. The components were mixed until a homogenous and smooth texture was acquired. Cellulose, polyethylene glycol, potassium chloride and calcium chloride used to prepare the faecal simulant were purchased from Sigma-Aldrich. Yeast and peanut oil were purchased at a local supermarket and the miso paste from a local Japanese restaurant.
Table 3. Composition of the faecal simulant41. Component Yeast Miso
paste Peanut
oil Cellulose Polyethylene
glycol Potassium
chloride Calcium
chloride Water
Mass (%) 16.5 16.5 11.0 5.5 2.7 2.2 0.5 45.1
4.2 Preparation of bio- and hydrochar
In section 1.1.5 the feedstocks used in this project are presented. All the feedstocks, apart from the human faecal simulant, were transformed into both bio- and hydrochar.
The human excrement was only transformed into hydrochar due to its large water content. In the continuation, the abbreviations presented in Table 4 will be used to identify the chars.
Table 4. Sample matrix.
Temperature and --- -carbonisation ---technique Feedstock and
Horse
manure Olive
residues Tomato
residues Rice
husk Raphia farinifera
Human faecal simulant
Torrefaction 230 oC HT1 OT1 TT1 RT1 FT1 -
Torrefaction 260 oC HT2 OT2 TT2 RT2 FT2 -
HTC 210 oC HH1 OH1 TH1 RH1 FH1 EH1
HTC 230 oC HH2 OH2 TH2 RH2 FH2 EH2
Biochar was prepared by placing 150 g of feedstock in a rotating oven with a temperature of 230 or 260 oC for three hours. The system had a steady flow of nitrogen (N2(g)) (AGA) gas (4 L/min). The oxygen concentrations ranged from 0 – 0.3 % and the CO2 between 11.8 – 12.0. No further treatment was done to the biochar.
The hydrochar was prepared in an Amarequip lab autoclave. A mass ratio of 1:9.4 of feedstock to milli-Q water was used. Two temperatures were selected, 210 and 230 oC, the retention time was 17 hours as the reactor took a long time to heat up. A coal-water-slurry was the final product from the reactor. The hydrochar was separated from the water by filtering it through a filter paper and thereafter it was dried in an oven
10 at 50 oC for at least 24 hours. The variations in drying-time in the oven were dependent on the moisture of the hydrochar, i.e. the chars were left in the oven until they had a constant weight.
Figure 6. Pictures of the rotating oven (left) and Amarequip lab autoclave (right) used.
After the carbonization process the yield of the chars were calculated with equation 142, where ‘m’ is the dry mass of the char and ‘mbiomass’ is the initial dry mass of the biomass before the carbonization process. The dry mass of the biomass was analysed by placing samples of biochar in an oven at 105 oC, when the sample weight was stable the dry weight was obtained.
!"#$% % = ) )*+,-.// ∗ 100 (1)
4.3 Sorption tests
4.3.1 Rotating bed reactor
The rotating bed reactor (SpinChem43) (Figure 7) is an advanced technique used to increase the mass transfer and thereby to reach equilibrium faster. The removal efficiency was evaluated by measuring the chars uptake of allura red dye (a food dye). The chemical structure of allura red can be observed in Figure 8.
The rotating bed reactor was loaded with either 1 or 2 g of bio- or hydrochar, the weight depended on the density of the char. Allura red solution (0.02g/L) was added to the reaction vessel until the reactor was covered. The reactor spun for 25 minutes at a constant speed in room temperature. Subsequently, the allura red solution was tapped from the reaction vessel, filtered through a 0.45 µm syringe filter and analysed by a RPA-II spectrometer. A calibration curve with a linear correlation ranging from 0.004 g/L to 0.02 g/L was used to analyse the absorption.
Figure 7. Rotating bed reactor from SpinChem43.
11
Figure 8. Chemical structure of allura red40.
4.3.2 Kinetic adsorption
To simulate the wastewater influent from the Umgeni river in KwaZulu-Natal, South Africa37,38 wastewater influent from Umeå WWTP was spiked with the selected pharmaceuticals (presented in Table 2). The wastewater from Umeå was spiked with the concentration found in the Umgeni river to ensure that there were equivalent or a higher concentration of the pharmaceuticals. As the concentrations of metronidazole and trimethoprim in the Umgeni river were not detected, lower than the detection limit or were detected at very low concentrations (> 1 µg/L), a concentration of 5 µg/L was used in the spiked water.
Kinetic batch adsorption tests were performed by weighing 0.06 g of char into a 15 mL falcon tube and subsequently pouring 12 mL of spiked wastewater into it. The falcon tubes were attached to a rotator under continuous rotation (50 RPM) and the samples were analysed after certain time intervals (0, 3, 7, 15, 30, 60, 90 and 120 minutes). Three replicates were performed for each char and time interval. After the set time intervals, the water was filtered through a 0.45 µm syringe filter. The equilibrium concentrations were determined by online solid phase extraction - liquid chromatography - mass spectrometry - mass spectrometry (OSPE LC-MS-MS), as explained in section 4.3.4.
4.3.3 Sorption isotherm
Atenolol, a water soluble pharmaceutical, and carbamazepine, a less soluble pharmaceutical, were selected for the adsorbance capacity evaluation. 200 mg/L of sodium azide (NaN3) (Sigma-Aldrich) and 0.01 mol/L of calcium chloride (CaCl2) (Sigma-Aldrich) were added to Milli-Q water. Sodium azide is added to inhibit microbial activity and calcium chloride water added to promote a natural water31. 0.28 g of char was weighed into a 50 mL falcon tube and 48 mL of spiked Milli-Q water was poured into the vial. The same char:water mass ratio was used in both the kinetic and sorption isotherm experiment. Nine concentrations ranging from 0 – 100 ng/ml were selected and three replicates were prepared for each unique char and pharmaceutical concentration combination. The falcon tubes were attached to a rotator with a constant speed of 50 RPM for 48 hours to ensure that equilibrium was reached. After the set time the water was centrifuged at 3700 RPM for 20 minutes and subsequently filtered through a 0.45 µm syringe filter. The equilibrium concentrations were determined by online SPE LC-MS-MS, as explained in section 4.3.4.
4.3.4 Analysis of contaminates with online SPE LC-MS-MS
Online solid phase liquid chromatography tandem mass spectrometry (OSPE-LC- MS/MS) was used to analyse the concentrations of the pharmaceuticals after the adsorption analysis was complete. The pharmaceutical concentrations were determined by a Quantiva Thermo TSQ Quantiva44. Prior to separation in a Hypersil GOLD (Thermo Scientific) C18 column45 the analytes were subject to online solid phase
12 extraction (SPE) in a pre-column. High grade acetonitrile was used during the pharmaceutical analysis with LC-MS-MS.
Kinetic (section 4.3.2.) and sorption isotherm (section 4.3.3) samples were diluted 10 times and an internal standard was added prior to analysis. Gradient elution was used to separate the analytes based on their polarity. The mobile phase composition throughout the run is displayed in Table 5. A flow rate of 0.250 mL/min and an injection volume of 1 mL was used. The same parameters were used for the kinetic and sorption isotherm samples.
Table 5. Gradient elution used during the separation of analytes.
Solvent
Time (min) Water (%) Acetonitrile (%)
0 – 3 10 90
3 – 6 0 Up to 100 %
6 – 8 0 100 %
8 – 10.5 90 10
4.3.5 Data analysis
The pharmaceuticals included in the kinetic adsorption samples determined by OSPE- LC-MS/MS. Internal standard (IS) quantification was used by calculating area ratios of pharmaceuticals and their assigned IS. In Table 6 the precursor mass-to-charge (m/z) ratio and assigned IS for each pharmaceutical can be observed. Pharmaceuticals that do not have their own unique IS are paired with IS that have a similar retention time.
Table 6. Precursor mass-to-charge ratio and assigned internal standard for the analysed pharmaceuticals in the kinetic adsorption experiment.
Pharmaceutical Precursor Mass-to-charge ratio Assigned internal standard
Ampicillin 382.113 Trimethoprim 13C Quan
Chloramphenicol 321.009 Carbamazepine D10 Quan
Ciprofloxacin 332.078 Ciprofloxacin D8 Qual
Erythromycin 716.374 Carbamazepine D10 Quan
Metronidazole 172.052 Trimethoprim 13C Quan
Nalidixic acid 233.061 Carbamazepine D10 Quan
Sulfamethoxazole 254.035 Sulfametazole 2H4 Qual
Trimethoprim 230.058 Trimethoprim 13C Quan
Diclofenac 295.963 Carbamazepine D10 Quan
Ketoprofen 255.078 Ketoprofen D3
Carbamazepine 237.048 Carbamazepine D10 Quan
Bezafibrate 362.096 Carbamazepine D10 Quan
Atenolol 267.200 Trimethoprim 13C Quan
Internal standard (IS) quantification was used by calculating area ratios of pharmaceuticals and their assigned IS in each sample and calibration curve. The calibration curve containing all the pharmaceuticals ranged between 1 – 2000 ng/ml.
Calibration curves for each pharmaceutical was plotted in excel. The intercept was set to zero and the k-value was later used to determine the concentration of the pharmaceuticals. To determine the concentration in ng/L, the pharmaceutical area ratio, for each bio- and hydrochar kinetic sample, was divided by their k-value. As the samples were diluted ten times pre-analysis the concentrations were multiplied ten times. A mean value for the three replicates for each char and analysis time was determined as well as a relative standard deviation (RSD). To determine the pharmaceutical adsorption by the chars the mean values were subtracted from the initial amount of contaminant. Pseudo first – and/or second order models10, 34, 46 were not employed on the kinetic tests as it brought no relative information to this project.
The internal standards were not used to quantify the sorption isotherm samples as they were not linear. Instead a calibration curve was prepared and the concentrations were extrapolated. A linear calibration curve was used for atenolol (0 – 100 ng/ml) and equation 2 was used to extrapolate the concentration of Atenolol after analysis. Where
‘Y’ is the mean absorbance, ‘m’ is the slope and ‘x’ is the concentration of atenolol.
! = )3 (2)
13 Unlike atenolol, carbamazepine did not have a linear calibration curve.
Carbamazepine had a logarithmic calibration curve. Therefore equation 3 was used to quantify the concentration of carbamazepine in the sorption isotherm samples. Where, Y’ is the mean absorbance, ‘m’ is the slope, ‘x’ is the concentration of carbamazepine and ‘b’ is the intercept.
! = ) ln 3 + 7 (3)
Equation 442 was used to calculate the adsorption capacity of the chars analysed by sorption isotherm. ‘Qe’ is the adsorption capacity (µg/g), ‘Co’ is the known initial concentration of the contaminant (µg/ml), ‘Ce’ is the equilibrium concentration (ug/ml), ‘V’ is the solution volume (mL) and ‘m’ is the dry mass of the char (g).
89 = :,− :9 ∗ < ) (4)
The Freundlich model (FM)30, 42, 47 (equation 5) was used to fit the sorption isotherm data. Where, ‘Qe’ is the adsorption capacity (µg/g), ‘KF’ is the Freundlich affinity constant ((mg/g)(mL/mg)1/n) ‘n’ is the constant that represents the intensity of adsorption (dimensionless) and ‘Ce’ is the equilibrium concentration (µg/L).
log 89 = $?@AB+DC∗ $?@:9 (5)
The detection limit (DL) and limit of quantification (QL) were calculated with equations 6 and 7, respectively. The DL and QL limits were used in for the sorption isotherm. No values under the DL are used and those under the QL are highlighted in the text.
EF = H9D/+O+P+OQG∗HIJKLMN (6)
8F = 10 ∗ EF (7)
4.4 Characterization of adsorbents
Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS), X-ray photoelectron spectroscopy (XPS), Brunauer-Emmett-Teller (BET) surface area and pore volume analysis and scanning electron microscopy (SEM) were the techniques used for the characterization of the chars. In the following sections an explanation of these techniques are presented.
4.4.1 DRIFTS
FTIR is a qualitative analytical technique used to distinguish functional groups from one another. In the wavenumber region 400 – 1800 cm-1 a unique spectrum is observed for materials, this is called the fingerprint region. From 1800 – 5200 cm-1 typical peaks for functional group bonds can be observed.
A Bruker IFS 66v/S, equipped with a Vacuum Optics Bench, was used for the DRIFTS analysis. The samples were prepared by grinding char with FTIR-grade potassium bromide (KBr)(Sigma-Aldrich) into a fine powder. Approximately 10 mg of char was placed in a mortar and then an additional 460 mg of KBr was added. The powder was grinded until it was fine and homogenous.
The Bruker IFS 66v/S was used under vacuum conditions (4 mbar) during the DRIFTS analysis. The background spectrum was set by measuring the FTIR-grade KBr.
128 interferograms were collected to gain a high signal-to-noise ratio. The spectral resolution was set to 4 cm-1 and the spectra was recorded from 400 – 5200 cm-1 where the region 400 – 2400 cm-1 was used for normalization. The spectra were baseline corrected and vector normalization was obtained by using OPUS software before the files were exported.
14 4.4.2 XPS
The surface chemistry of the chars was analysed by XPS as it is a qualitative and quantitative method that provides information about all elements, apart from hydrogen and helium. It also gives specific information about functional groups present on the top surface layer (3 – 5 nm)48. During the XPS analysis, performed on an AXIS Ultra DLD. The lens was in hybrid mode, pass energy 160 was used for the resolution and an acquisition time of 330 seconds was applied. Five sweeps were performed with a dwell time of 60 ms. The charge nebuliser was on.
4.4.3 BET surface area and pore volume
Nitrogen adsorption is a common method used to determine the total surface area of analytes. The BET theory accounts for deviations such as multilayer adsorption during the analysis. The surface of the material is saturated with N2 (g) molecules and when the material is saturated the surface area can be calculated. A high surface area is desirable as there are more reaction sites, diffusion paths are shorter and kinetic reactions occur faster.
The BET analysis was performed with a TriStar 3000. Prior to analysis the chars were degassed, with a micromeritics SmartPrep degasser, at 120 oC with N2 (g) for approximately 2 hours. Subsequently, the chars were cooled to 77K with N2 (l) and the surface area was determined under vacuum.
4.4.4 SEM
A scanning electron microscope was used to get an optical impression of the surface morphology of the char. A Carl Zeiss Merlin Field Emission Scanning Electron Microscope (FESEM) with SmartSEM V.5.05 software was used. High definition backscattered electron (HDBSD) mode was applied. In HDBSD mode inorganic substances are lighter in colour than organic compounds. Energy-dispersive X-ray spectroscopy (EDS) was also applied to analysed the elemental composition of the char surface.
Horse manure chars prepared by torrefaction and HTC were chosen to run an SEM on to get an optical impression. Horse manure chars were chosen as they were found to be versatile chars with good remediation properties. The samples were held onto adhesive carbon tape. The surface of the chars was magnified 75 – 5000 times at different locations to get a good overview of the surface morphology.
5. Results and Discussion
5.1 Sorption tests
Three sorption experiments were performed in this study: (i) adsorption of allura red in a rotating bed reactor, (ii) adsorption of pharmaceuticals in a kinetic adsorption experiment and (iii) adsorption of atenolol and carbamazepine in sorption isotherm experiments.
5.1.1 Rotating bed reactor
The results from the adsorption of allura red dye on the chars are presented in Figure 9. The rotating bed reactor used (SpinChem)43 has a porosity of 106 µm. Therefore, chars which have a particle size smaller than the mesh size of the reactor are released from the enclosed area and deposited at the bottom of the reaction vessel. The percentage of char lost is presented as ‘loss of char’ in Figure 9. Chars made at higher temperatures tended to have a higher loss of char during the experiment as chars produced at these temperatures were finer than those produced at lower temperatures.
The highest adsorption of allura red was found for EH2, it had an adsorption of 11.7 %. Several of the chars had a negative adsorption of allura red, i.e. the finial concentration of allura red was larger than the initial concentration. There is a RSD of approximately ± 6 % as the detected absorption of allura red for HT1 and EH1 was 106
%, the highest detectable amount should be 100 %. Thus, the relative standard deviation should be considered for all the samples. No general trend in adsorption capacity
15 between the two temperatures of a carbonisation technique or between the techniques was found.
Figure 9. Adsorption of allura red and amount of char loss during the experiment.
Allura red has a logKOW of -0.55 (PubChem, 2016)49 and a logS of -4.86 (Chemicalize, 2016)40, this entails that it is quite water soluble. As the adsorption of allura red is not particularly good the char may be better suited for the adsorption of contaminants which are less water soluble.
5.1.2 Kinetic adsorption
As explained in section 3.4 wastewater from Umeå WWTP was spiked with pharmaceuticals found in the Umgeni river37,38. The pharmaceutical concentrations found in the spiked wastewater from Umeå are presented in Table 7. The average concentrations detected are based on the blank samples used for each char (66 samples in total). As one can see the standard deviation and RSD is high for most of the pharmaceuticals. The concentration of chloramphenicol in the spiked wastewater had an RSD of 170 % and is therefore invalid and was not investigated further. Matrix effects due to colloidal particles (with)in the wastewater, heterogeneous char, interference from the LC-MS-MS and human error are some factors that contribute to the divergent results.
Table 7. Pharmaceutical concentrations detected by online SPE LC-MS-MS in the spiked wastewater from Umeå wastewater treatment plant (WWTP).
Name Therapeutic
group
Influent conc. from a WWTP in the Umgeni river (µg/L)
Concentrations detected in the spiked wastewater (µg/L)
RSD (%)
Retention time (minutes)
Ampicillin Antibiotic 10a 50 ±23 47 4.95
Chloramphenicol Antibiotic 10a 2.7 + 4.6 170 5.34
Ciprofloxacin Antibiotic 16a 247 ± 49 20 4.86
Erythromycin Antibiotic 22a 734 ± 258 35 5.49
Metronidazole Antibiotic Not detectedb 21 ± 6 30 3.47
Nalidixic acid Antibiotic 31±3a 219 ± 98 45 5.58
Sulfamethoxazole Antibiotic 10a 45 ± 7 15 5.28
Trimethoprim Antibiotic 0.13 ±2.78b 32 ± 5 15 4.78
Diclofenac NSAID 18a 102 ± 12 11 6.10
Ketoprofen NSAID 8a 68 ± 8 11 5.81
Carbamazepine Anti-epileptic 4.56 ±0.81b 25 ± 3 13 5.52
Bezafibrate Anti-hyper-
lipidaemia 8a 55 ± 9 16 5.80
Atenolol β-blocker 39±4a 534 ± 95 18 3.18
a Concentrations analysed by Agunbiade and Moodley37.
b Concentrations analysed by Matongo et al38.
As the wastewater from Umeå was left in its natural state, i.e. it was not filtered or acidified, there are several matrix effects to account for. The colloidal particles in the water contribute to deviating concentrations found in the water, as seen by the, in some
-10 2030405060100
Loss of char (%) Adsorption of allura red (%)
