- Email: [email protected]
Hydrocarbon adsorption performance and regeneration stability of diphenyldichlorosilane coated zeolite and its application in permeable reactive barriers: Column studies
Microporous and Mesoporous Materials xxx (xxxx) xxx
Contents lists available at ScienceDirect
Microporous and Mesoporous Materials journal homepage: http://www.elsevier.com/locate/micromeso
Hydrocarbon adsorption performance and regeneration stability of diphenyldichlorosilane coated zeolite and its application in permeable reactive barriers: Column studies Junchao Ma, Yong Wang, Geoffrey W. Stevens, Kathryn A. Mumford * Particulate Fluids Processing Centre, Department of Chemical and Biomolecular Engineering, The University of Melbourne, VIC, 3010, Australia
A R T I C L E I N F O
A B S T R A C T
Keywords: DPDSCI coated zeolite Toluene adsorption Regeneration Column tests Axial dispersion Transport modelling
The adsorption behaviour and long-term stability of diphenoldichlorosilane (DPDSCI) coated zeolite was investigated in a column study by studying the toluene adsorption performance and material regeneration at various temperatures. Computer modelling (CXTFIT and MATLAB) was adapted to explore the adsorption mechanism, including the calculation of the axial dispersion coefficient and maximum adsorption capacity. The results indicate that DPDSCI coated zeolite presented a higher adsorption capacity towards toluene than that found in previous batch experiments. Under high flow rates, temperature had little influence on the adsorption behaviour. Regeneration tests prove that this material may be cleaned and reused at least three times without significant reduction in adsorption effectiveness. Axial dispersion studies show that transitional flow occurs during the adsorption process and is dominated by diffusion processes. Good agreement between experimental data and model prediction are achieved with an R2 value between 0.97 and 0.99 obtained for all tests. Overall, the column tests demonstrate DPDSCI coated zeolite may be suitable for the remediation of sites contaminated with hydrocarbons, even in cold regions and variable water flows.
1. Introduction Worldwide, accidental spills; unsatisfactory disposal of industrial chemicals; poor agricultural practices and mining activities has resulted in water, especially groundwater being seriously contaminated by pol lutants, including heavy metals, hydrocarbons and pesticides amongst others [1]. As humans increase their presence across the globe, this not only occurs in temperate climates, but also polar regions [2]. Therefore, it is necessary to develop suitable technologies for contaminated site remediation in both temperate and cold climates. Previous researchers [3] have found that the Permeable Reactive Barrier (PRB) technologies are suitable for application in situ [4]. The key factors that influence PRB performance include; the characteristics of the reactive media such as porosity and particle size as well as contaminant type, nutrient avail ability and microbial activity [5,6]. Previous studies have applied Granular Activated Carbon (GAC) or Surfactant Modified Zeolite (SMZ) as principal reactive materials for laboratory or field scale column experiments towards the removal of hydrocarbon contaminants, such as BTEX, phenanthrene or catechol
under temperate temperatures [7–12]. As these are commercial prod ucts, they are relatively easy to access and are cost competitive [7]. Evaluation of the work conducted reveals that for most applications, GAC exhibits superior adsorption behaviour as compared to SMZ [7,13]. This is likely due to the porous granular structure and surface functional groups of GAC. In contrast, although SMZ does not exhibit as strong as adsorption potential for hydrocarbons, the base material, zeolite has a high cation exchange capacity and so may be used to remove metal pollutants [14]. This indicates that SMZ may be able to adsorb both hydrocarbon and heavy metals. As most column tests present in literature are conducted at room temperatures, their use in low temperatures need interpretation [4,15]. The results from a large scale installation in the Antarctic indicate that GAC is too fragile to withstand repeated freeze-thaw cycles and breaks down, thereby changing the flow path through the PRB itself [16]. Also, it was found that the surfactant coating of SMZ may leach after several regenerations, as the surfactant is only held by electrostatic bonds on the zeolite surface, thereby influencing the surface capacity of SMZ for hy drocarbons [7,17]. The limitations of both GAC and SMZ indicate that a
* Corresponding author. E-mail addresses: [email protected] (J. Ma), [email protected] (Y. Wang), [email protected] (G.W. Stevens), mumfordk@ unimelb.edu.au (K.A. Mumford). https://doi.org/10.1016/j.micromeso.2019.109843 Received 23 January 2018; Received in revised form 18 October 2019; Accepted 28 October 2019 Available online 31 October 2019 1387-1811/© 2019 Elsevier Inc. All rights reserved.
Please cite this article as: Junchao Ma, Microporous and Mesoporous Materials, https://doi.org/10.1016/j.micromeso.2019.109843
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
novel composite material that possesses the high hydrocarbon adsorp tion ability of GAC, ion capture ability and rigid structure of SMZ, and a stable coating is still required. Based on these requirements, chlorosilane is introduced as a coating for the zeolite [18]. It has been reported that chlorosliane may be grafted to the surface of zeolite via a series of chemical reactions [18]. This surface coating improves the hydrocarbon adsorption capacity without damage to the ion capture ability [19]. Regeneration tests showed that this coating is highly stable due to the presence of covalent bonds [19]. Among the common chlorosilanes available, diphenoldichlorosilane (DPDSCI) was found to exhibit a better affinity towards hydrocarbon contaminants, especially the BTEX and aromatic compounds, as they have similar structures [18]. Tests investigating temperature depen dence, showed that DPDSCI coated zeolite performed well even under cold temperatures i.e. 4 � C [20]. Although chloroslianes are highly reactive materials and may hydrolyse and liberate hydrochloride acid when exposed to moisture [21], the hazards are minimized during the modification process by using acetone and water to wash the excessing DPDSCI and oven drying. As the DPDSCI is covalently bonded on the surface of zeolite, it is not be easily washed off, and this does not pollute environment. Therefore, as it is efficient to adsorb hydrocarbons and safe to the environment, we extend upon our previous work and examine the performance of the material with column studies. To test the long-term stability of the material, previous studies have used air sparging or wet air oxidation [7,8] to remove contaminants on the surface of sorbents. However, according to the batch experiments of DPDSCI coated zeolite conducted previously [20], it was found that water may have similar results for toluene. In addition, as it is intended that the material will be regenerated via microbial activity, a water regeneration method is applied in this study. To appropriately simulate the various field conditions likely to be experienced, it is desirable to carry out adsorption and regeneration column tests at both temperate and cold temperatures. To investigate the flow characteristics through the column and explore the adsorption mechanisms, axial dispersion and reactive transport modelling column tests are conducted, and modelled using standard methods [13]. This will enable the evaluation of the likely performance of the material in the field.
2.3. Analysis of toluene in solution High performance liquid chromatography (HPLC) (Agilent 1200) was used to measure concentrations of toluene in the influent (C0) and effluent (C) water samples taken throughout the column tests. The analysis method was the same as that used previously and described elsewhere [20]. 2.4. Adsorption column tests Columns tests were conducted in a vertical glass column with a height and inner diameter of 129 mm and 28 mm respectively. The materials were packed as follows: 32 mm height of glass beads at the base, followed by 32 mm height of DPDSCI coated zeolite and then 62 mm height of glass beads at the top. All materials were separated by Nylon mesh (200 μm). The function of glass beads in the column was to support the zeolite and fill the rest of the column. As they were inert to the toluene, they would not affect the adsorption results. 50 mg/L toluene solution was introduced in an up flow direction at 4.5 rpm pump speed (equivalent to flow rates of 11.24 PV/h). A single pore volume (PV) is defined as the volume of toluene solution present within the section containing modified zeolite in the column per unit time, which is commonly used in column studies [7,8,13]. The effluent was collected at the end of the column and analysis conducted via HPLC. Tests were conducted at both 4 � C and 20 � C in the incubator (Thermoline Scientific TLM-590) and sampled in triplicate. 2.5. Regeneration tests Regeneration tests were conducted to detect whether the material could be used multiple times without significant decrease in hydrocar bon adsorption ability. The column set up was the same as for the toluene adsorption test. After the column was saturated with toluene, the influent was replaced with water until all the toluene solution was ejected. The column after the first washing cycle was termed 1st re generated. This column was then saturated again with toluene, and subsequently washed and termed 2nd regenerated. This was repeated once more and the final material termed 3rd regenerated. Tests were conducted at both 4 � C and 20 � C and sampled in triplicate.
2. Materials and methods
2.6. Reactive transport modelling test
2.1. Adsorption materials
To develop models that describe the material performance, the entire glass column (129 mm height, 28 mm inner diameter) was packed with DPDSCI coated zeolite. The column was pre-saturated with MilliQ water and agitated such that no air bubbles were present. 50 mg/L toluene solution was then introduced to the saturated column at time zero in an up flow direction at various flow rates. The pump speeds were set 2.4, 3.5, 4.5 rpm respectively (equivalent to flow rates of 1.52, 2.31 and 2.81 V/h). The effluent was collected at the column outlet and HPLC analysis conducted to determine the toluene concentration. Tests were conducted at both 4 � C and 20 � C and sampled in triplicate.
The raw zeolite material used was natural clinoptilolite zeolite (Castle Mountain Zeolite, Quirindi, N.S.W., Australia) and has been previously well characterised [20]. Prior to use, the zeolite was sieved with an 8–16 US mesh sieve. The Diphenyldichlorosilane modification process followed the method developed previously [20]. After modifi cation, the product was sieved again using same mesh sieve. The density of natural zeolite and modified zeolite were 2.112 g/cm3 and 2.103 g/cm3, respectively. Inert ballotini glass beads (Potters Industries Inc.25–35 US sieve) were used in the adsorption tests to hold the reac tive materials in place. Before use, they were pre-treated with MilliQ water and dried at 150 � C for 6 h.
2.7. Axial dispersion tests
2.2. Reagents
Using a similar method to that previously described [13], axial dispersion tests were conducted by running 500 mg/L NaCl solution through the saturated column (using the same configuration as the reactive transport modelling test) in an up flow direction at a known flow rate. This was continued until the effluent concentration was the same as influent, after which it was replaced with a deionised water feed which was denoted time zero. The column was packed as per the transport modelling tests and flow rates investigated were 1.52, 2.31 and 2.81 PV/h respectively. The effluent samples were collected every 2 min and a conductivity probe and meter (Crison GLP 31) was used to measure the Naþ concentration via use of a previously prepared
Except where stated, chemicals used in the adsorption tests were all received and used without further purification. All solutions were made with MilliQ water. Toluene (Ajax) was selected to evaluate the hydro carbon sorption properties of the material. Sodium Chloride (Chemsupply, purity 99.7%) was used as the tracer in the axial dispersion tests to determine the axial dispersion coefficient.
2
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
calibration curve. Tests were conducted duplicate at same flow rate in 4 � C and 20 � C.
CXTFIT code was developed by the U.S. Agriculture Department for determining contaminant transport parameters in laboratory column tests and model non-reactive and reactive transport of solutes through porous media.
3. Modelling
3.1.2. Reynolds number The Reynolds number is a dimensionless parameter that shows the ratio of inertial to viscous forces during flow. The conventional Reynolds number is used to distinguish laminar flow and turbulent flow. For fluid flow in the porous media, energy losses from the fluid due to viscous and form drags would influence the predication of flow type [22]. Therefore, the modified Reynolds number (Re’) [22] is used to determine the flow within in porous media as follows:
3.1. Flow characteristics Several parameters are used to describe the flow characteristics through a column, including the axial dispersion coefficient, Reynolds and P�eclet numbers. 3.1.1. Axial dispersion coefficient Dl In a saturated, homogeneous, isotropic media system with steadystate uniform flow, the advection-dispersion equation (ADE) for the transport of a non-reactive solute in the axial (x-axis) direction (Fig. 1) [13] can be expressed as:
∂C ∂2 C ¼ Dl 2 ∂t ∂x
v
∂C ∂x
Re’ ¼
Sv ¼
(1)
(3)
nÞSv μ
6 d
(4)
where ρ and μ are water density (kg/m3) and viscosity (Pa⋅s), U0 is su perficial fluid velocity in the column (m/s), and n is the material porosity. Sv is specific surface area per unit volume which is determined by d, the particle diameter (m). After modification, typical packed bed Reynolds numbers [22] are updated to: Laminar flow (Re ’ � 0:01); Transitional flow (0:01 < Re ’ � 2); Turbulent flow (2 < Re ’).
where Cis the tracer concentration in bulk solution (mg/L), Dl is the axial dispersion coefficient (m2/s), x is the distance taken along the flow line (m), v is the average pore velocity (m/s), which is the real average speed of the solution passing through a porous medium [22]. Inverse modelling (Fig. 1) was carried out to determine the required parameters by fitting a solution of the ADE (Eq. (1)) to an axial dispersion test result. With the continuous tracer input into a semiinfinite length column with an initial solute concentration:Cðx; 0Þ ¼ 0 � � � at x � 0, the boundary conditions are: Cð0; tÞ ¼ C0 at t � 0, and dc ¼ dt � 0 at t � 0. � � � �� � � � C 1 x vt vx x þ vt ¼ erfc pffiffiffiffiffiffi þ exp erfc pffiffiffiffiffiffi C0 2 Dl 2 Dl t 2 Dl t
ρU 0 ð1
3.1.3. P�eclet number The P� eclet number represents the relative effect of advective solute transport compared to dispersive transport. The modified P�eclet number (Pe’) for flow in porous media is:
x¼∞
Pe’ ¼ Re’ � Sc ¼
(2)
Sc ¼
where erfc is the complimentary error function. The parameters Dl and v were determined by Eq. (2) using the computer program CXTFIT [13].
ð1
U0 nÞSv Dl
μ ρDl
(5) (6)
where Sc is the Schmidt number, which represents the characteristics of simultaneous momentum and mass diffusion convection processes. At relatively high P� eclet numbers, advection dominates the transport process whereas diffusion/dispersion dominate the transport process at relatively low P� eclet numbers [23]. 3.2. Reactive transport in the column The loss or gain of solute within an elemental volume of porous media can occur from chemical reactions that take place within the pore water or because of transfer of solute to or from the solid phase [24]. The loss of solute from solution due to adsorption can be described by including an additional term in Eq. (1) to describe the effect of the rate of change in solute concentration in the solid phase. This is known as the advection-dispersion-reaction solute transport equation (ADRE) [13]:
∂C ∂2 C ¼ Dl 2 ∂t ∂x
v
∂C ∂x
ρb ∂q n ∂t
(7)
where ρb is the bulk density of the DPDSCI coated zeolite column and q is
the toluene concentration on adsorption materials (mg/g). The term ρnb ∂∂qt is the rate of change in solution concentration due to sorption processes. Using results obtained from previous batch experiments [20], the Langmuir sorption isotherm is used to describe the relationship between the fluid and solid phase solute concentration for the toluene-modified zeolite material system. Therefore in this study, the ‘q’ in equation (6) is equal to qe in Langmuir model: qe ¼ Fig. 1. Schematic diagram of column design. 3
qmax Kl Ce 1 þ Kl Ce
(8)
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
Fig. 2. The toluene adsorption curves of unamended zeolite and fresh DPDSCIzeolite at 20 � C and 4 � C. Initial toluene concentration (C0) is 50 mg/L and the flow rate is 11.24 PV/h.
Fig. 3. The regeneration tests of DPDSCI-zeolite at 20 � C. Initial toluene con centration (C0) is 50 mg/L and the flow rate is 11.24 PV/h.
where Ce is the concentration of contaminants in solution at equilib rium (mg/L), qe is the adsorption capacity at equilibrium (mg/g). qmax , is the maximum adsorption capacity (mg/g) of the adsorbent and Kl is the equilibrium constant (L/mg). Langmuir equations for toluene sorption on DPDSCI coated zeolite have been determined previously [20] as: q ¼ 0:02289Ce = ð1 þ 0:028Ce Þ q ¼ 0:01694Ce = ð1 þ 0:027Ce Þ at 20 � C and 4 � C respectively. MATLAB, a general mathematical computer program, is used to solve the partial differential equations mentioned. The initial concentration is:Cðx; 0Þ ¼ 0 atx � 0, the boundary conditions are: Cð0; tÞ ¼ C0 at t � 0, � � � and dc ¼ 0 at t � 0. dt � x¼∞
Fig. 4. The regeneration tests of DPDSCI-zeolite at 4 � C. Initial toluene con centration (C0) is 50 mg/L and the flow rate is 11.24 PV/h.
4. Results and discussion 4.1. Adsorption performance
proved that this modified zeolite could be used for hydrocarbon capture over a wide temperature range.
The adsorption tests of toluene onto DPDSCI coated zeolite at 20 � C and 4 � C, and natural zeolite at 20 � C were conducted in the column containing glass beads and zeolite and the results are presented in Fig. 2. It can be observed that the chlorosilane coating greatly improved the toluene adsorption ability of the natural zeolite. Natural zeolite adsor bed minimal quantities of toluene (less than 5% of the toluene present was removed), while DPDSCI coated zeolite showed distinct capture ability before 30 PV of influent had been injected, adsorbing 60% of toluene at the early stages. The increasing adsorption capacity is a result of surface modification where Van de Waals interactions are enhanced via further hydrophobic bonding and π-π electron stacking interactions between aromatic moieties [20]. Meanwhile, temperature did not appear to have significant influence on the adsorption performance at the flowrates investigated. This reiterates the results obtained in the previous batch study [20]. Furthermore, the trend of toluene adsorption whereby the concentration sharply increases, before stabilising and becoming saturated, is similar to that of SMZ [7]. From this, it may be speculated that if other hydrocarbons sorb onto SMZ well, they may also sorb well onto DPDSCI coated zeolite. Overall, the adsorption tests
4.2. Regeneration test The regeneration tests of DPDSCI coated zeolite at 20 � C and 4 � C were conducted in a column containing glass beads and zeolite. The results of long term stability tests of DPDSCI coated zeolite at the tem peratures investigated are shown on Fig. 3 (20 � C) and Fig. 4 (4 � C). Though slight difference may exist at the initial period, the whole adsorption trend is similar for all regeneration tests at both 20 � C and 4 � C. This indicates that there is no significant reduction in the sorption capacity of DPDSCI coated zeolite after repeated adsorption and washing processes. Similar phenomena was found in the investigation of the regeneration of SMZ contaminated with BTEX [7]. The positive outcomes of this series of tests indicate that this material may be promising in field applications. With biotechnology increasingly being used in the remediation of hydrocarbon contaminated sites [2], continuous adsorption and elution of contaminants on the surface of materials is of benefit when combined with biofilm technology [25]. The
4
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
Fig. 5. The effect of flow rate on axial dispersion at 20 � C. The lines present the CXTFIT model fit.
Fig. 7. The relationship between modified Reynolds number (Re’) and axial dispersion coefficient (Dl) of full packed column.
comparison of the curves presented in Figs. 5 and 6, at different tem peratures, the time at which the Naþ concentration began decreasing, and the time water fully saturated the column under 4 � C were only slightly slower than that at 20 � C, indicating temperature has little in fluence on axial dispersion process. The modified Reynolds numbers calculated, refer from Table 1, are generally around 0.01, indicating that transitional flow occurs in the packed column [22]. Meanwhile, the modified P�eclet number (Table 1) shows relatively low numbers (around 0.01 to 0.02), indicating diffu sion, especially molecular diffusion dominate the whole solute transport process [23,27]. It is also observed that the modified P� eclet numbers show little dependence on modified Reynolds numbers, which is consistent with other studies regarding liquids dispersion through packed beds [28]. The relationship between the modified Reynolds number and the axial dispersion coefficient of the full packed column is shown in Fig. 7. It can be observed that at the same temperatures, the modified Reynolds number has a linear correlation with the axial dispersion coefficient for all flow rates. For the two temperatures, the curves have similar slopes (1.2 � 10 4 and 1.3 � 10 4), but different intercepts (9.3 � 10 7 and 3.7 � 10 7). It may be speculated that for a given column, temperature may slightly affect the intercept of regression equation between Re’ and Dl. Meanwhile, the packed column design, such as column parameters, material characteristics, might influence the slope of regression. This may be proven via comparison with other column test results. Arora investigated the adsorption performance of GAC in the same column
Fig. 6. The effect of flow rate on axial dispersion at 4 � C. The lines present the CXTFIT model fit.
contaminants adsorbed to the surface of this material can be regarded as a food source for microbial communities [25]. Due to the release of hydrocarbons from the material, the food source may be rapidly pro vided and the sufficient carbon supply would benefit microbial breeding and contaminant removal [26]. 4.3. Axial dispersion coefficient The axial dispersion tests were used to understand the flow charac teristics within a column containing DPDSCI coated zeolite. The effluent concentration trends in Figs. 5 and 6, show that at higher flowrates contaminant breakthrough occurs earlier. The corresponding axial dispersion coefficient under different flow rates and temperatures mimicked this result by being larger at high flow rates (Table 1). Upon
Table 1 The flow parameters of axial dispersion test at both 20 � C and 4 � C. Temperature (� C) 20 4
Pump Flow rate (PV/h) 1.52 2.31 2.81 1.52 2.31 2.81
Dl (m2/s) 8.12 � 10 1.40 � 10 2.56 � 10 4.61 � 10 7.95 � 10 1.39 � 10
7 6 6 7 7 6
Re’
Pe’
0.0130 0.0209 0.0296 0.0060 0.0095 0.0130
0.0161 0.0151 0.0116 0.0203 0.0189 0.0147
Fig. 8. The observed experimental points and fitted toluene adsorption curves of DPDSCI-zeolite at 20 � C under various flow rates. 5
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
adsorption curves (Fig. 9) are not influenced by the flow rate, being very close to one another. This may be because the maximum adsorption capacity is lower at lower temperature [20](Table 2) and so, the ca pacity is reached quicker at all flowrates investigated. Similar behaviour has also been observed for the adsorption of copper by Na-clinoptilolite [29]. Through the evaluation of toluene adsorption onto DPDSCI coated zeolite, it could be concluded that: 1) in temperate climates, the per formance of the adsorption material may be impacted by environmental factors such flow rate, while in cold regions this seems to have less impact; 2) the ADRE equation could adequately predict the performance of DPDSCI coated zeolite under various temperature and flow rate conditions, which could be applied in the future to simulate the toluene or hydrocarbon adsorption process by this material in the field. This observation has also been reported for other sorbents on other aromatic pollutants [9].
Fig. 9. The observed experimental points and fitted toluene adsorption curves of DPDSCI-zeolite at 4 � C under various flow rates.
5. Conclusions The toluene adsorption efficiency, longevity and mechanism of the DPDSCI coated zeolite in a packed column was studied by measuring both inlet and exit concentrations of toluene as a function of injected volume and fitting the results to ADRE. According to the adsorption tests and regeneration tests, compared with unamended zeolite, the adsorp tion capacity and long term stability of DPDSCI coated zeolite were not impacted by temperature or flow rate. These characteristics make it a promising material for future remediation of hydrocarbon contaminated sites at various temperatures. To investigate the adsorption and trans port mechanism, axial dispersion tests and reactive transport modelling were conducted. The modified Reynolds number and modified P�eclet number were obtained from axial dispersion tests. These showed that transitional flow occurred during the adsorption process and diffusion may dominate the flow. The good agreement between experiment data and modelling prediction indicates the ADRE equation with Langmuir equilibrium isotherm could simulate the toluene adsorption perfor mance of DPDSCI coated zeolite well. The maximum capacity calculated from modelling present behaviour of this material exceeds that deter mined in batch experiments. Overall, the column tests of DPDSCI coated zeolite prove this mate rial could be suitable for use in hydrocarbon contaminants site reme diation even at cold temperature for a long period. Performance modelling shows that the behaviour of this material could be adequately predicted at various temperatures and flow rate conditions.
conditions at both 20 � C and 4 � C [13]. Based on her data, the re lationships between modified Reynolds number and axial dispersion coefficient at 20 � C and 4 � C were Dl ¼ 1.1 � 10 5Re’ þ 1.8 � 10 6 and Dl ¼ 1.5 � 10 5Re’ – 3.2 � 10 8, respectively. The different intercepts result from the influence of temperature. The different magnitude of regression slope between DPDSCI-zeolite and GAC column is due to the different materials characteristics. 4.4. Reactive transport modelling To explore the mechanism of toluene adsorption, a theoretical model (ADRE) was applied to simulate and predict the transport process. As previously discussed the Langmuir isotherm was used to predict toluene adsorption onto DPDSCI coated zeolite. As shown in Figs. 8 and 9, the predicted curves are in good agreement with experimental data. The comparison of Langmuir parameters in batch and column tests are shown in Table 2. The equilibrium constant (Kl) and maximum adsorption capacity (qmax) calculated from column data were found to be larger than those observed from batch experiments at both 20 � C and 4 � C, indicating even greater adsorption occurs in laboratory columns. Similar discrepancies have been found previously including in the work of Simpson and Bowman, 2009 and Woinarski et al., 2006 [29,30]. At 20 � C, the adsorption performance of the column containing DPDSCI coated zeolite under various flow rates are presented in Fig. 8. At lower flow rates, later breakthrough occurs, which is in accordance with the data presented in Table 2. The reason for this differences is at room temperature, the breakthrough curve for a given adsorbent with same column parameters mainly depends on the contact time, which is dominated by the fluid superficial velocity [31]. With slower influent passing, the functional groups on sorbents surface may have more op portunity to interact with toluene in the water, which extends the time to reach breakthrough and saturation points [9,17]. At 4 � C, nearly all
Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Table 2 Summary of experimental conditions, breakthrough point (C/C0 ¼ 0.10) and saturation point (C/C0 ¼ 0.99) of column experiments, toluene adsorption capacity at both batch and column test after calculation. Temperature (� C) 20 4
Flow rate (PV/h) 1.52 2.31 2.81 1.52 2.31 2.81
Breakthrough point (PV) 2.90 2.09 1.27 1.99 1.74 1.55
Saturation Point (PV) 11.92 8.82 6.78 7.18 6.62 6.50
6
Kl (10
2
L/mg)
qmax (10
1
mg/g)
Batch
Column
Batch
Column
2.81
6.55
8.15
16.58
2.73
7.75
6.21
7.06
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
Acknowledgements
[13] M. Arora, I. Snape, G.W. Stevens, Toluene sorption by granular activated carbon and its use in cold regions permeable reactive barrier: fixed bed studies, Cold Reg. Sci. Technol. 69 (2011) 59–63. [14] S. Wang, Y. Peng, Natural zeolites as effective adsorbents in water and wastewater treatment, Chem. Eng. J. 156 (2010) 11–24. [15] G. Hornig, K. Northcott, I. Snape, G. Stevens, Assessment of sorbent materials for treatment of hydrocarbon contaminated ground water in cold regions, Cold Reg. Sci. Technol. 53 (2008) 83–91. [16] K.A. Mumford, J.L. Rayner, I. Snape, G.W. Stevens, Hydraulic performance of a permeable reactive barrier at Casey Station, Antarctica, Chemosphere 117 (2014) 223–231. [17] J. Schick, P. Caullet, J.-L. Paillaud, J. Patarin, C. Mangold-Callarec, Nitrate sorption from water on a surfactant-modified zeolite. Fixed-bed column experiments, Microporous Mesoporous Mater. 142 (2011) 549–556. [18] P. Huttenloch, K.E. Roehl, K. Czurda, Sorption of nonpolar aromatic contaminants by chlorosilane surface modified natural minerals, Environ. Sci. Technol. 35 (2001) 4260–4264. [19] K.A. Northcott, J. Bacus, N. Taya, Y. Komatsu, J.M. Perera, G.W. Stevens, Synthesis and characterization of hydrophobic zeolite for the treatment of hydrocarbon contaminated ground water, J. Hazard Mater. 183 (2010) 434–440. [20] J. Ma, G.W. Stevens, K.A. Mumford, The effect of temperature on hydrocarbon adsorption by diphenyldichlorosilane coated zeolite and its application in permeable reactive barriers in cold regions, Cold Reg. Sci. Technol. 145 (2018) 169–176. [21] P.A. Jean, R.H. Gallavan, G.B. Kolesar, W.H. Siddiqui, J.A. Oxley, R.G. Meeks, Chlorosilane acute inhalation toxicity and development of an LC50 prediction model, Inhal. Toxicol. 18 (2008) 515–522. [22] R.G. Holdich, Fundamentals of Particle Technology, Midland Information Technology and Publishing, 2002. [23] C.D. Shackelford, Critical concepts for column testing, J. Geotech. Eng. 120 (1994) 1804–1828. [24] R.A. Freeze, J.A. Cherry, Groundwater, Prentice-Hall, Englewood Cliffs, NJ, 1979. [25] B.L. Freidman, S.L. Gras, I. Snape, G.W. Stevens, K.A. Mumford, The performance of ammonium exchanged zeolite for the biodegradation of petroleum hydrocarbons migrating in soil water, J. Hazard Mater. 313 (2016) 272–282. [26] K.A. Mumford, S.M. Powell, J.L. Rayner, G. Hince, I. Snape, G.W. Stevens, Evaluation of a permeable reactive barrier to capture and degrade hydrocarbon contaminants, Environ. Sci. Pollut. Control Ser. 22 (2015) 12298–12308. [27] D. Yu, K. Jackson, T. Harmon, Dispersion and diffusion in porous media under supercritical conditions, Chem. Eng. Sci. 54 (1999) 357–367. [28] M. Edwards, J. Richardson, Gas dispersion in packed beds, Chem. Eng. Sci. 23 (1968) 109–123. [29] A.Z. Woinarski, G.W. Stevens, I. Snape, A natural zeolite permeable reactive barrier to treat heavy-metal contaminated waters in Antarctica, Process Saf. Environ. Prot. 84 (2006) 109–116. [30] J.A. Simpson, R.S. Bowman, Nonequilibrium sorption and transport of volatile petroleum hydrocarbons in surfactant-modified zeolite, J. Contam. Hydrol. 108 (2009) 1–11. [31] D. Wang, E. McLaughlin, R. Pfeffer, Y.S. Lin, Aqueous phase adsorption of toluene in a packed and fluidized bed of hydrophobic aerogels, Chem. Eng. J. 168 (2011) 1201–1208.
The authors greatly acknowledge the support from Qiang Sun in the modelling of the column performance. This study was supported by the Particulate Fluids Processing Centre, a Special Research Centre of the Australia Research Council. The Australian zeolite was supplied by the Australian Antarctic Division. Thanks to University of Melbourne and Chinese Scholarship Council for scholarship support. References [1] R. Thiruvenkatachari, S. Vigneswaran, R. Naidu, Permeable reactive barrier for groundwater remediation, J. Ind. Eng. Chem. 14 (2008) 145–156. [2] H.E. De Jesus, R.S. Peixoto, Bioremediation in Antarctic soils, J. Pet Environ. Biotechnol. 06 (2015). [3] I. Snape, C.E. Morris, C.M. Cole, The use of permeable reactive barriers to control contaminant dispersal during site remediation in Antarctica, Cold Reg. Sci. Technol. 32 (2001) 157–174. [4] F. Obiri-Nyarko, S.J. Grajales-Mesa, G. Malina, An overview of permeable reactive barriers for in situ sustainable groundwater remediation, Chemosphere 111 (2014) 243–259. [5] M.M. Scherer, S. Richter, R.L. Valentine, P.J.J. Alvarez, Chemistry and microbiology of permeable reactive barriers for in situ groundwater clean up, Crit. Rev. Microbiol. 26 (2000) 221–264. [6] K.A. Mumford, J.L. Rayner, I. Snape, S.C. Stark, G.W. Stevens, D.B. Gore, Design, installation and preliminary testing of a permeable reactive barrier for diesel fuel remediation at Casey Station, Antarctica, Cold Reg. Sci. Technol. 96 (2013) 96–107. [7] J. Michael Ranck, R.S. Bowman, J.L. Weeber, L.E. Katz, E.J. Sullivan, BTEX removal from produced water using surfactant-modified zeolite, J. Environ. Eng. 131 (2005) 434–442. [8] C.R. Altare, R.S. Bowman, L.E. Katz, K.A. Kinney, E.J. Sullivan, Regeneration and long-term stability of surfactant-modified zeolite for removal of volatile organic compounds from produced water, Microporous Mesoporous Mater. 105 (2007) 305–316. [9] D. Richard, M.d.L. Delgado Nú~ nez, D. Schweich, Adsorption of complex phenolic compounds on active charcoal: breakthrough curves, Chem. Eng. J. 158 (2010) 213–219. [10] S.T.M.L.D. Senevirathna, S. Tanaka, S. Fujii, C. Kunacheva, H. Harada, B.H.A.K. T. Ariyadasa, B.R. Shivakoti, Adsorption of perfluorooctane sulfonate (n-PFOS) onto non ion-exchange polymers and granular activated carbon: batch and column test, Desalination 260 (2010) 29–33. [11] D. Chatzopoulos, A. Varma, Aqueou-phase adsorption and desorption of toluene in activated carbon fixed beds: experiments and model, Chem. Eng. Sci. 50 (1995) 127–141. [12] R. Murillo, T. Garcı ́a, E. Ayl� on, M.S. Call� en, M.V. Navarro, J.M. L� opez, A. M. Mastral, Adsorption of phenanthrene on activated carbons: breakthrough curve modeling, Carbon 42 (2004) 2009–2017.
7
Contents lists available at ScienceDirect
Microporous and Mesoporous Materials journal homepage: http://www.elsevier.com/locate/micromeso
Hydrocarbon adsorption performance and regeneration stability of diphenyldichlorosilane coated zeolite and its application in permeable reactive barriers: Column studies Junchao Ma, Yong Wang, Geoffrey W. Stevens, Kathryn A. Mumford * Particulate Fluids Processing Centre, Department of Chemical and Biomolecular Engineering, The University of Melbourne, VIC, 3010, Australia
A R T I C L E I N F O
A B S T R A C T
Keywords: DPDSCI coated zeolite Toluene adsorption Regeneration Column tests Axial dispersion Transport modelling
The adsorption behaviour and long-term stability of diphenoldichlorosilane (DPDSCI) coated zeolite was investigated in a column study by studying the toluene adsorption performance and material regeneration at various temperatures. Computer modelling (CXTFIT and MATLAB) was adapted to explore the adsorption mechanism, including the calculation of the axial dispersion coefficient and maximum adsorption capacity. The results indicate that DPDSCI coated zeolite presented a higher adsorption capacity towards toluene than that found in previous batch experiments. Under high flow rates, temperature had little influence on the adsorption behaviour. Regeneration tests prove that this material may be cleaned and reused at least three times without significant reduction in adsorption effectiveness. Axial dispersion studies show that transitional flow occurs during the adsorption process and is dominated by diffusion processes. Good agreement between experimental data and model prediction are achieved with an R2 value between 0.97 and 0.99 obtained for all tests. Overall, the column tests demonstrate DPDSCI coated zeolite may be suitable for the remediation of sites contaminated with hydrocarbons, even in cold regions and variable water flows.
1. Introduction Worldwide, accidental spills; unsatisfactory disposal of industrial chemicals; poor agricultural practices and mining activities has resulted in water, especially groundwater being seriously contaminated by pol lutants, including heavy metals, hydrocarbons and pesticides amongst others [1]. As humans increase their presence across the globe, this not only occurs in temperate climates, but also polar regions [2]. Therefore, it is necessary to develop suitable technologies for contaminated site remediation in both temperate and cold climates. Previous researchers [3] have found that the Permeable Reactive Barrier (PRB) technologies are suitable for application in situ [4]. The key factors that influence PRB performance include; the characteristics of the reactive media such as porosity and particle size as well as contaminant type, nutrient avail ability and microbial activity [5,6]. Previous studies have applied Granular Activated Carbon (GAC) or Surfactant Modified Zeolite (SMZ) as principal reactive materials for laboratory or field scale column experiments towards the removal of hydrocarbon contaminants, such as BTEX, phenanthrene or catechol
under temperate temperatures [7–12]. As these are commercial prod ucts, they are relatively easy to access and are cost competitive [7]. Evaluation of the work conducted reveals that for most applications, GAC exhibits superior adsorption behaviour as compared to SMZ [7,13]. This is likely due to the porous granular structure and surface functional groups of GAC. In contrast, although SMZ does not exhibit as strong as adsorption potential for hydrocarbons, the base material, zeolite has a high cation exchange capacity and so may be used to remove metal pollutants [14]. This indicates that SMZ may be able to adsorb both hydrocarbon and heavy metals. As most column tests present in literature are conducted at room temperatures, their use in low temperatures need interpretation [4,15]. The results from a large scale installation in the Antarctic indicate that GAC is too fragile to withstand repeated freeze-thaw cycles and breaks down, thereby changing the flow path through the PRB itself [16]. Also, it was found that the surfactant coating of SMZ may leach after several regenerations, as the surfactant is only held by electrostatic bonds on the zeolite surface, thereby influencing the surface capacity of SMZ for hy drocarbons [7,17]. The limitations of both GAC and SMZ indicate that a
* Corresponding author. E-mail addresses: [email protected] (J. Ma), [email protected] (Y. Wang), [email protected] (G.W. Stevens), mumfordk@ unimelb.edu.au (K.A. Mumford). https://doi.org/10.1016/j.micromeso.2019.109843 Received 23 January 2018; Received in revised form 18 October 2019; Accepted 28 October 2019 Available online 31 October 2019 1387-1811/© 2019 Elsevier Inc. All rights reserved.
Please cite this article as: Junchao Ma, Microporous and Mesoporous Materials, https://doi.org/10.1016/j.micromeso.2019.109843
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
novel composite material that possesses the high hydrocarbon adsorp tion ability of GAC, ion capture ability and rigid structure of SMZ, and a stable coating is still required. Based on these requirements, chlorosilane is introduced as a coating for the zeolite [18]. It has been reported that chlorosliane may be grafted to the surface of zeolite via a series of chemical reactions [18]. This surface coating improves the hydrocarbon adsorption capacity without damage to the ion capture ability [19]. Regeneration tests showed that this coating is highly stable due to the presence of covalent bonds [19]. Among the common chlorosilanes available, diphenoldichlorosilane (DPDSCI) was found to exhibit a better affinity towards hydrocarbon contaminants, especially the BTEX and aromatic compounds, as they have similar structures [18]. Tests investigating temperature depen dence, showed that DPDSCI coated zeolite performed well even under cold temperatures i.e. 4 � C [20]. Although chloroslianes are highly reactive materials and may hydrolyse and liberate hydrochloride acid when exposed to moisture [21], the hazards are minimized during the modification process by using acetone and water to wash the excessing DPDSCI and oven drying. As the DPDSCI is covalently bonded on the surface of zeolite, it is not be easily washed off, and this does not pollute environment. Therefore, as it is efficient to adsorb hydrocarbons and safe to the environment, we extend upon our previous work and examine the performance of the material with column studies. To test the long-term stability of the material, previous studies have used air sparging or wet air oxidation [7,8] to remove contaminants on the surface of sorbents. However, according to the batch experiments of DPDSCI coated zeolite conducted previously [20], it was found that water may have similar results for toluene. In addition, as it is intended that the material will be regenerated via microbial activity, a water regeneration method is applied in this study. To appropriately simulate the various field conditions likely to be experienced, it is desirable to carry out adsorption and regeneration column tests at both temperate and cold temperatures. To investigate the flow characteristics through the column and explore the adsorption mechanisms, axial dispersion and reactive transport modelling column tests are conducted, and modelled using standard methods [13]. This will enable the evaluation of the likely performance of the material in the field.
2.3. Analysis of toluene in solution High performance liquid chromatography (HPLC) (Agilent 1200) was used to measure concentrations of toluene in the influent (C0) and effluent (C) water samples taken throughout the column tests. The analysis method was the same as that used previously and described elsewhere [20]. 2.4. Adsorption column tests Columns tests were conducted in a vertical glass column with a height and inner diameter of 129 mm and 28 mm respectively. The materials were packed as follows: 32 mm height of glass beads at the base, followed by 32 mm height of DPDSCI coated zeolite and then 62 mm height of glass beads at the top. All materials were separated by Nylon mesh (200 μm). The function of glass beads in the column was to support the zeolite and fill the rest of the column. As they were inert to the toluene, they would not affect the adsorption results. 50 mg/L toluene solution was introduced in an up flow direction at 4.5 rpm pump speed (equivalent to flow rates of 11.24 PV/h). A single pore volume (PV) is defined as the volume of toluene solution present within the section containing modified zeolite in the column per unit time, which is commonly used in column studies [7,8,13]. The effluent was collected at the end of the column and analysis conducted via HPLC. Tests were conducted at both 4 � C and 20 � C in the incubator (Thermoline Scientific TLM-590) and sampled in triplicate. 2.5. Regeneration tests Regeneration tests were conducted to detect whether the material could be used multiple times without significant decrease in hydrocar bon adsorption ability. The column set up was the same as for the toluene adsorption test. After the column was saturated with toluene, the influent was replaced with water until all the toluene solution was ejected. The column after the first washing cycle was termed 1st re generated. This column was then saturated again with toluene, and subsequently washed and termed 2nd regenerated. This was repeated once more and the final material termed 3rd regenerated. Tests were conducted at both 4 � C and 20 � C and sampled in triplicate.
2. Materials and methods
2.6. Reactive transport modelling test
2.1. Adsorption materials
To develop models that describe the material performance, the entire glass column (129 mm height, 28 mm inner diameter) was packed with DPDSCI coated zeolite. The column was pre-saturated with MilliQ water and agitated such that no air bubbles were present. 50 mg/L toluene solution was then introduced to the saturated column at time zero in an up flow direction at various flow rates. The pump speeds were set 2.4, 3.5, 4.5 rpm respectively (equivalent to flow rates of 1.52, 2.31 and 2.81 V/h). The effluent was collected at the column outlet and HPLC analysis conducted to determine the toluene concentration. Tests were conducted at both 4 � C and 20 � C and sampled in triplicate.
The raw zeolite material used was natural clinoptilolite zeolite (Castle Mountain Zeolite, Quirindi, N.S.W., Australia) and has been previously well characterised [20]. Prior to use, the zeolite was sieved with an 8–16 US mesh sieve. The Diphenyldichlorosilane modification process followed the method developed previously [20]. After modifi cation, the product was sieved again using same mesh sieve. The density of natural zeolite and modified zeolite were 2.112 g/cm3 and 2.103 g/cm3, respectively. Inert ballotini glass beads (Potters Industries Inc.25–35 US sieve) were used in the adsorption tests to hold the reac tive materials in place. Before use, they were pre-treated with MilliQ water and dried at 150 � C for 6 h.
2.7. Axial dispersion tests
2.2. Reagents
Using a similar method to that previously described [13], axial dispersion tests were conducted by running 500 mg/L NaCl solution through the saturated column (using the same configuration as the reactive transport modelling test) in an up flow direction at a known flow rate. This was continued until the effluent concentration was the same as influent, after which it was replaced with a deionised water feed which was denoted time zero. The column was packed as per the transport modelling tests and flow rates investigated were 1.52, 2.31 and 2.81 PV/h respectively. The effluent samples were collected every 2 min and a conductivity probe and meter (Crison GLP 31) was used to measure the Naþ concentration via use of a previously prepared
Except where stated, chemicals used in the adsorption tests were all received and used without further purification. All solutions were made with MilliQ water. Toluene (Ajax) was selected to evaluate the hydro carbon sorption properties of the material. Sodium Chloride (Chemsupply, purity 99.7%) was used as the tracer in the axial dispersion tests to determine the axial dispersion coefficient.
2
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
calibration curve. Tests were conducted duplicate at same flow rate in 4 � C and 20 � C.
CXTFIT code was developed by the U.S. Agriculture Department for determining contaminant transport parameters in laboratory column tests and model non-reactive and reactive transport of solutes through porous media.
3. Modelling
3.1.2. Reynolds number The Reynolds number is a dimensionless parameter that shows the ratio of inertial to viscous forces during flow. The conventional Reynolds number is used to distinguish laminar flow and turbulent flow. For fluid flow in the porous media, energy losses from the fluid due to viscous and form drags would influence the predication of flow type [22]. Therefore, the modified Reynolds number (Re’) [22] is used to determine the flow within in porous media as follows:
3.1. Flow characteristics Several parameters are used to describe the flow characteristics through a column, including the axial dispersion coefficient, Reynolds and P�eclet numbers. 3.1.1. Axial dispersion coefficient Dl In a saturated, homogeneous, isotropic media system with steadystate uniform flow, the advection-dispersion equation (ADE) for the transport of a non-reactive solute in the axial (x-axis) direction (Fig. 1) [13] can be expressed as:
∂C ∂2 C ¼ Dl 2 ∂t ∂x
v
∂C ∂x
Re’ ¼
Sv ¼
(1)
(3)
nÞSv μ
6 d
(4)
where ρ and μ are water density (kg/m3) and viscosity (Pa⋅s), U0 is su perficial fluid velocity in the column (m/s), and n is the material porosity. Sv is specific surface area per unit volume which is determined by d, the particle diameter (m). After modification, typical packed bed Reynolds numbers [22] are updated to: Laminar flow (Re ’ � 0:01); Transitional flow (0:01 < Re ’ � 2); Turbulent flow (2 < Re ’).
where Cis the tracer concentration in bulk solution (mg/L), Dl is the axial dispersion coefficient (m2/s), x is the distance taken along the flow line (m), v is the average pore velocity (m/s), which is the real average speed of the solution passing through a porous medium [22]. Inverse modelling (Fig. 1) was carried out to determine the required parameters by fitting a solution of the ADE (Eq. (1)) to an axial dispersion test result. With the continuous tracer input into a semiinfinite length column with an initial solute concentration:Cðx; 0Þ ¼ 0 � � � at x � 0, the boundary conditions are: Cð0; tÞ ¼ C0 at t � 0, and dc ¼ dt � 0 at t � 0. � � � �� � � � C 1 x vt vx x þ vt ¼ erfc pffiffiffiffiffiffi þ exp erfc pffiffiffiffiffiffi C0 2 Dl 2 Dl t 2 Dl t
ρU 0 ð1
3.1.3. P�eclet number The P� eclet number represents the relative effect of advective solute transport compared to dispersive transport. The modified P�eclet number (Pe’) for flow in porous media is:
x¼∞
Pe’ ¼ Re’ � Sc ¼
(2)
Sc ¼
where erfc is the complimentary error function. The parameters Dl and v were determined by Eq. (2) using the computer program CXTFIT [13].
ð1
U0 nÞSv Dl
μ ρDl
(5) (6)
where Sc is the Schmidt number, which represents the characteristics of simultaneous momentum and mass diffusion convection processes. At relatively high P� eclet numbers, advection dominates the transport process whereas diffusion/dispersion dominate the transport process at relatively low P� eclet numbers [23]. 3.2. Reactive transport in the column The loss or gain of solute within an elemental volume of porous media can occur from chemical reactions that take place within the pore water or because of transfer of solute to or from the solid phase [24]. The loss of solute from solution due to adsorption can be described by including an additional term in Eq. (1) to describe the effect of the rate of change in solute concentration in the solid phase. This is known as the advection-dispersion-reaction solute transport equation (ADRE) [13]:
∂C ∂2 C ¼ Dl 2 ∂t ∂x
v
∂C ∂x
ρb ∂q n ∂t
(7)
where ρb is the bulk density of the DPDSCI coated zeolite column and q is
the toluene concentration on adsorption materials (mg/g). The term ρnb ∂∂qt is the rate of change in solution concentration due to sorption processes. Using results obtained from previous batch experiments [20], the Langmuir sorption isotherm is used to describe the relationship between the fluid and solid phase solute concentration for the toluene-modified zeolite material system. Therefore in this study, the ‘q’ in equation (6) is equal to qe in Langmuir model: qe ¼ Fig. 1. Schematic diagram of column design. 3
qmax Kl Ce 1 þ Kl Ce
(8)
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
Fig. 2. The toluene adsorption curves of unamended zeolite and fresh DPDSCIzeolite at 20 � C and 4 � C. Initial toluene concentration (C0) is 50 mg/L and the flow rate is 11.24 PV/h.
Fig. 3. The regeneration tests of DPDSCI-zeolite at 20 � C. Initial toluene con centration (C0) is 50 mg/L and the flow rate is 11.24 PV/h.
where Ce is the concentration of contaminants in solution at equilib rium (mg/L), qe is the adsorption capacity at equilibrium (mg/g). qmax , is the maximum adsorption capacity (mg/g) of the adsorbent and Kl is the equilibrium constant (L/mg). Langmuir equations for toluene sorption on DPDSCI coated zeolite have been determined previously [20] as: q ¼ 0:02289Ce = ð1 þ 0:028Ce Þ q ¼ 0:01694Ce = ð1 þ 0:027Ce Þ at 20 � C and 4 � C respectively. MATLAB, a general mathematical computer program, is used to solve the partial differential equations mentioned. The initial concentration is:Cðx; 0Þ ¼ 0 atx � 0, the boundary conditions are: Cð0; tÞ ¼ C0 at t � 0, � � � and dc ¼ 0 at t � 0. dt � x¼∞
Fig. 4. The regeneration tests of DPDSCI-zeolite at 4 � C. Initial toluene con centration (C0) is 50 mg/L and the flow rate is 11.24 PV/h.
4. Results and discussion 4.1. Adsorption performance
proved that this modified zeolite could be used for hydrocarbon capture over a wide temperature range.
The adsorption tests of toluene onto DPDSCI coated zeolite at 20 � C and 4 � C, and natural zeolite at 20 � C were conducted in the column containing glass beads and zeolite and the results are presented in Fig. 2. It can be observed that the chlorosilane coating greatly improved the toluene adsorption ability of the natural zeolite. Natural zeolite adsor bed minimal quantities of toluene (less than 5% of the toluene present was removed), while DPDSCI coated zeolite showed distinct capture ability before 30 PV of influent had been injected, adsorbing 60% of toluene at the early stages. The increasing adsorption capacity is a result of surface modification where Van de Waals interactions are enhanced via further hydrophobic bonding and π-π electron stacking interactions between aromatic moieties [20]. Meanwhile, temperature did not appear to have significant influence on the adsorption performance at the flowrates investigated. This reiterates the results obtained in the previous batch study [20]. Furthermore, the trend of toluene adsorption whereby the concentration sharply increases, before stabilising and becoming saturated, is similar to that of SMZ [7]. From this, it may be speculated that if other hydrocarbons sorb onto SMZ well, they may also sorb well onto DPDSCI coated zeolite. Overall, the adsorption tests
4.2. Regeneration test The regeneration tests of DPDSCI coated zeolite at 20 � C and 4 � C were conducted in a column containing glass beads and zeolite. The results of long term stability tests of DPDSCI coated zeolite at the tem peratures investigated are shown on Fig. 3 (20 � C) and Fig. 4 (4 � C). Though slight difference may exist at the initial period, the whole adsorption trend is similar for all regeneration tests at both 20 � C and 4 � C. This indicates that there is no significant reduction in the sorption capacity of DPDSCI coated zeolite after repeated adsorption and washing processes. Similar phenomena was found in the investigation of the regeneration of SMZ contaminated with BTEX [7]. The positive outcomes of this series of tests indicate that this material may be promising in field applications. With biotechnology increasingly being used in the remediation of hydrocarbon contaminated sites [2], continuous adsorption and elution of contaminants on the surface of materials is of benefit when combined with biofilm technology [25]. The
4
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
Fig. 5. The effect of flow rate on axial dispersion at 20 � C. The lines present the CXTFIT model fit.
Fig. 7. The relationship between modified Reynolds number (Re’) and axial dispersion coefficient (Dl) of full packed column.
comparison of the curves presented in Figs. 5 and 6, at different tem peratures, the time at which the Naþ concentration began decreasing, and the time water fully saturated the column under 4 � C were only slightly slower than that at 20 � C, indicating temperature has little in fluence on axial dispersion process. The modified Reynolds numbers calculated, refer from Table 1, are generally around 0.01, indicating that transitional flow occurs in the packed column [22]. Meanwhile, the modified P�eclet number (Table 1) shows relatively low numbers (around 0.01 to 0.02), indicating diffu sion, especially molecular diffusion dominate the whole solute transport process [23,27]. It is also observed that the modified P� eclet numbers show little dependence on modified Reynolds numbers, which is consistent with other studies regarding liquids dispersion through packed beds [28]. The relationship between the modified Reynolds number and the axial dispersion coefficient of the full packed column is shown in Fig. 7. It can be observed that at the same temperatures, the modified Reynolds number has a linear correlation with the axial dispersion coefficient for all flow rates. For the two temperatures, the curves have similar slopes (1.2 � 10 4 and 1.3 � 10 4), but different intercepts (9.3 � 10 7 and 3.7 � 10 7). It may be speculated that for a given column, temperature may slightly affect the intercept of regression equation between Re’ and Dl. Meanwhile, the packed column design, such as column parameters, material characteristics, might influence the slope of regression. This may be proven via comparison with other column test results. Arora investigated the adsorption performance of GAC in the same column
Fig. 6. The effect of flow rate on axial dispersion at 4 � C. The lines present the CXTFIT model fit.
contaminants adsorbed to the surface of this material can be regarded as a food source for microbial communities [25]. Due to the release of hydrocarbons from the material, the food source may be rapidly pro vided and the sufficient carbon supply would benefit microbial breeding and contaminant removal [26]. 4.3. Axial dispersion coefficient The axial dispersion tests were used to understand the flow charac teristics within a column containing DPDSCI coated zeolite. The effluent concentration trends in Figs. 5 and 6, show that at higher flowrates contaminant breakthrough occurs earlier. The corresponding axial dispersion coefficient under different flow rates and temperatures mimicked this result by being larger at high flow rates (Table 1). Upon
Table 1 The flow parameters of axial dispersion test at both 20 � C and 4 � C. Temperature (� C) 20 4
Pump Flow rate (PV/h) 1.52 2.31 2.81 1.52 2.31 2.81
Dl (m2/s) 8.12 � 10 1.40 � 10 2.56 � 10 4.61 � 10 7.95 � 10 1.39 � 10
7 6 6 7 7 6
Re’
Pe’
0.0130 0.0209 0.0296 0.0060 0.0095 0.0130
0.0161 0.0151 0.0116 0.0203 0.0189 0.0147
Fig. 8. The observed experimental points and fitted toluene adsorption curves of DPDSCI-zeolite at 20 � C under various flow rates. 5
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
adsorption curves (Fig. 9) are not influenced by the flow rate, being very close to one another. This may be because the maximum adsorption capacity is lower at lower temperature [20](Table 2) and so, the ca pacity is reached quicker at all flowrates investigated. Similar behaviour has also been observed for the adsorption of copper by Na-clinoptilolite [29]. Through the evaluation of toluene adsorption onto DPDSCI coated zeolite, it could be concluded that: 1) in temperate climates, the per formance of the adsorption material may be impacted by environmental factors such flow rate, while in cold regions this seems to have less impact; 2) the ADRE equation could adequately predict the performance of DPDSCI coated zeolite under various temperature and flow rate conditions, which could be applied in the future to simulate the toluene or hydrocarbon adsorption process by this material in the field. This observation has also been reported for other sorbents on other aromatic pollutants [9].
Fig. 9. The observed experimental points and fitted toluene adsorption curves of DPDSCI-zeolite at 4 � C under various flow rates.
5. Conclusions The toluene adsorption efficiency, longevity and mechanism of the DPDSCI coated zeolite in a packed column was studied by measuring both inlet and exit concentrations of toluene as a function of injected volume and fitting the results to ADRE. According to the adsorption tests and regeneration tests, compared with unamended zeolite, the adsorp tion capacity and long term stability of DPDSCI coated zeolite were not impacted by temperature or flow rate. These characteristics make it a promising material for future remediation of hydrocarbon contaminated sites at various temperatures. To investigate the adsorption and trans port mechanism, axial dispersion tests and reactive transport modelling were conducted. The modified Reynolds number and modified P�eclet number were obtained from axial dispersion tests. These showed that transitional flow occurred during the adsorption process and diffusion may dominate the flow. The good agreement between experiment data and modelling prediction indicates the ADRE equation with Langmuir equilibrium isotherm could simulate the toluene adsorption perfor mance of DPDSCI coated zeolite well. The maximum capacity calculated from modelling present behaviour of this material exceeds that deter mined in batch experiments. Overall, the column tests of DPDSCI coated zeolite prove this mate rial could be suitable for use in hydrocarbon contaminants site reme diation even at cold temperature for a long period. Performance modelling shows that the behaviour of this material could be adequately predicted at various temperatures and flow rate conditions.
conditions at both 20 � C and 4 � C [13]. Based on her data, the re lationships between modified Reynolds number and axial dispersion coefficient at 20 � C and 4 � C were Dl ¼ 1.1 � 10 5Re’ þ 1.8 � 10 6 and Dl ¼ 1.5 � 10 5Re’ – 3.2 � 10 8, respectively. The different intercepts result from the influence of temperature. The different magnitude of regression slope between DPDSCI-zeolite and GAC column is due to the different materials characteristics. 4.4. Reactive transport modelling To explore the mechanism of toluene adsorption, a theoretical model (ADRE) was applied to simulate and predict the transport process. As previously discussed the Langmuir isotherm was used to predict toluene adsorption onto DPDSCI coated zeolite. As shown in Figs. 8 and 9, the predicted curves are in good agreement with experimental data. The comparison of Langmuir parameters in batch and column tests are shown in Table 2. The equilibrium constant (Kl) and maximum adsorption capacity (qmax) calculated from column data were found to be larger than those observed from batch experiments at both 20 � C and 4 � C, indicating even greater adsorption occurs in laboratory columns. Similar discrepancies have been found previously including in the work of Simpson and Bowman, 2009 and Woinarski et al., 2006 [29,30]. At 20 � C, the adsorption performance of the column containing DPDSCI coated zeolite under various flow rates are presented in Fig. 8. At lower flow rates, later breakthrough occurs, which is in accordance with the data presented in Table 2. The reason for this differences is at room temperature, the breakthrough curve for a given adsorbent with same column parameters mainly depends on the contact time, which is dominated by the fluid superficial velocity [31]. With slower influent passing, the functional groups on sorbents surface may have more op portunity to interact with toluene in the water, which extends the time to reach breakthrough and saturation points [9,17]. At 4 � C, nearly all
Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Table 2 Summary of experimental conditions, breakthrough point (C/C0 ¼ 0.10) and saturation point (C/C0 ¼ 0.99) of column experiments, toluene adsorption capacity at both batch and column test after calculation. Temperature (� C) 20 4
Flow rate (PV/h) 1.52 2.31 2.81 1.52 2.31 2.81
Breakthrough point (PV) 2.90 2.09 1.27 1.99 1.74 1.55
Saturation Point (PV) 11.92 8.82 6.78 7.18 6.62 6.50
6
Kl (10
2
L/mg)
qmax (10
1
mg/g)
Batch
Column
Batch
Column
2.81
6.55
8.15
16.58
2.73
7.75
6.21
7.06
J. Ma et al.
Microporous and Mesoporous Materials xxx (xxxx) xxx
Acknowledgements
[13] M. Arora, I. Snape, G.W. Stevens, Toluene sorption by granular activated carbon and its use in cold regions permeable reactive barrier: fixed bed studies, Cold Reg. Sci. Technol. 69 (2011) 59–63. [14] S. Wang, Y. Peng, Natural zeolites as effective adsorbents in water and wastewater treatment, Chem. Eng. J. 156 (2010) 11–24. [15] G. Hornig, K. Northcott, I. Snape, G. Stevens, Assessment of sorbent materials for treatment of hydrocarbon contaminated ground water in cold regions, Cold Reg. Sci. Technol. 53 (2008) 83–91. [16] K.A. Mumford, J.L. Rayner, I. Snape, G.W. Stevens, Hydraulic performance of a permeable reactive barrier at Casey Station, Antarctica, Chemosphere 117 (2014) 223–231. [17] J. Schick, P. Caullet, J.-L. Paillaud, J. Patarin, C. Mangold-Callarec, Nitrate sorption from water on a surfactant-modified zeolite. Fixed-bed column experiments, Microporous Mesoporous Mater. 142 (2011) 549–556. [18] P. Huttenloch, K.E. Roehl, K. Czurda, Sorption of nonpolar aromatic contaminants by chlorosilane surface modified natural minerals, Environ. Sci. Technol. 35 (2001) 4260–4264. [19] K.A. Northcott, J. Bacus, N. Taya, Y. Komatsu, J.M. Perera, G.W. Stevens, Synthesis and characterization of hydrophobic zeolite for the treatment of hydrocarbon contaminated ground water, J. Hazard Mater. 183 (2010) 434–440. [20] J. Ma, G.W. Stevens, K.A. Mumford, The effect of temperature on hydrocarbon adsorption by diphenyldichlorosilane coated zeolite and its application in permeable reactive barriers in cold regions, Cold Reg. Sci. Technol. 145 (2018) 169–176. [21] P.A. Jean, R.H. Gallavan, G.B. Kolesar, W.H. Siddiqui, J.A. Oxley, R.G. Meeks, Chlorosilane acute inhalation toxicity and development of an LC50 prediction model, Inhal. Toxicol. 18 (2008) 515–522. [22] R.G. Holdich, Fundamentals of Particle Technology, Midland Information Technology and Publishing, 2002. [23] C.D. Shackelford, Critical concepts for column testing, J. Geotech. Eng. 120 (1994) 1804–1828. [24] R.A. Freeze, J.A. Cherry, Groundwater, Prentice-Hall, Englewood Cliffs, NJ, 1979. [25] B.L. Freidman, S.L. Gras, I. Snape, G.W. Stevens, K.A. Mumford, The performance of ammonium exchanged zeolite for the biodegradation of petroleum hydrocarbons migrating in soil water, J. Hazard Mater. 313 (2016) 272–282. [26] K.A. Mumford, S.M. Powell, J.L. Rayner, G. Hince, I. Snape, G.W. Stevens, Evaluation of a permeable reactive barrier to capture and degrade hydrocarbon contaminants, Environ. Sci. Pollut. Control Ser. 22 (2015) 12298–12308. [27] D. Yu, K. Jackson, T. Harmon, Dispersion and diffusion in porous media under supercritical conditions, Chem. Eng. Sci. 54 (1999) 357–367. [28] M. Edwards, J. Richardson, Gas dispersion in packed beds, Chem. Eng. Sci. 23 (1968) 109–123. [29] A.Z. Woinarski, G.W. Stevens, I. Snape, A natural zeolite permeable reactive barrier to treat heavy-metal contaminated waters in Antarctica, Process Saf. Environ. Prot. 84 (2006) 109–116. [30] J.A. Simpson, R.S. Bowman, Nonequilibrium sorption and transport of volatile petroleum hydrocarbons in surfactant-modified zeolite, J. Contam. Hydrol. 108 (2009) 1–11. [31] D. Wang, E. McLaughlin, R. Pfeffer, Y.S. Lin, Aqueous phase adsorption of toluene in a packed and fluidized bed of hydrophobic aerogels, Chem. Eng. J. 168 (2011) 1201–1208.
The authors greatly acknowledge the support from Qiang Sun in the modelling of the column performance. This study was supported by the Particulate Fluids Processing Centre, a Special Research Centre of the Australia Research Council. The Australian zeolite was supplied by the Australian Antarctic Division. Thanks to University of Melbourne and Chinese Scholarship Council for scholarship support. References [1] R. Thiruvenkatachari, S. Vigneswaran, R. Naidu, Permeable reactive barrier for groundwater remediation, J. Ind. Eng. Chem. 14 (2008) 145–156. [2] H.E. De Jesus, R.S. Peixoto, Bioremediation in Antarctic soils, J. Pet Environ. Biotechnol. 06 (2015). [3] I. Snape, C.E. Morris, C.M. Cole, The use of permeable reactive barriers to control contaminant dispersal during site remediation in Antarctica, Cold Reg. Sci. Technol. 32 (2001) 157–174. [4] F. Obiri-Nyarko, S.J. Grajales-Mesa, G. Malina, An overview of permeable reactive barriers for in situ sustainable groundwater remediation, Chemosphere 111 (2014) 243–259. [5] M.M. Scherer, S. Richter, R.L. Valentine, P.J.J. Alvarez, Chemistry and microbiology of permeable reactive barriers for in situ groundwater clean up, Crit. Rev. Microbiol. 26 (2000) 221–264. [6] K.A. Mumford, J.L. Rayner, I. Snape, S.C. Stark, G.W. Stevens, D.B. Gore, Design, installation and preliminary testing of a permeable reactive barrier for diesel fuel remediation at Casey Station, Antarctica, Cold Reg. Sci. Technol. 96 (2013) 96–107. [7] J. Michael Ranck, R.S. Bowman, J.L. Weeber, L.E. Katz, E.J. Sullivan, BTEX removal from produced water using surfactant-modified zeolite, J. Environ. Eng. 131 (2005) 434–442. [8] C.R. Altare, R.S. Bowman, L.E. Katz, K.A. Kinney, E.J. Sullivan, Regeneration and long-term stability of surfactant-modified zeolite for removal of volatile organic compounds from produced water, Microporous Mesoporous Mater. 105 (2007) 305–316. [9] D. Richard, M.d.L. Delgado Nú~ nez, D. Schweich, Adsorption of complex phenolic compounds on active charcoal: breakthrough curves, Chem. Eng. J. 158 (2010) 213–219. [10] S.T.M.L.D. Senevirathna, S. Tanaka, S. Fujii, C. Kunacheva, H. Harada, B.H.A.K. T. Ariyadasa, B.R. Shivakoti, Adsorption of perfluorooctane sulfonate (n-PFOS) onto non ion-exchange polymers and granular activated carbon: batch and column test, Desalination 260 (2010) 29–33. [11] D. Chatzopoulos, A. Varma, Aqueou-phase adsorption and desorption of toluene in activated carbon fixed beds: experiments and model, Chem. Eng. Sci. 50 (1995) 127–141. [12] R. Murillo, T. Garcı ́a, E. Ayl� on, M.S. Call� en, M.V. Navarro, J.M. L� opez, A. M. Mastral, Adsorption of phenanthrene on activated carbons: breakthrough curve modeling, Carbon 42 (2004) 2009–2017.
7