Two major current approaches to removing phenol group compounds (e.g., phenol, cresols, etc.) from gas systems are adsorption using activated carbon and absorption using solvents and resins. These suffer from multiple limitations, including high cost and limited capacity to remove phenol at lower concentrations. Thus, it is required to identify novel additives or combinations of additives (formulated as gels, pastes, and solids at room temperature) that would be able to effectively remove phenol group compounds from gas systems.
Hydroxypropyl-β-cyclodextrin (HPβCD) has been identified and proposed as suitable liquid absorbent for phenol and cresols present in a waste gas stream. HPβCD forms inclusion complexes with phenol, o-, m-, and p-cresols at room temperature. These complexes are reversible and hence HPβCD can be regenerated within 30 to 60 min and reused. HPβCD is safe for humans and relatively cheap.
THE OPTIMAL TOLUENE ABSORBENT
Cyclodextrins (CDs): Chemical Structure
CDs are produced via partial degradation of starch, the cleaved glucose units of which is then enzymatically coupled generating crystalline, homogenous toroidal chiral structures of various molecular size: α-CD (6 glucose units), β-CD (7 glucose units), and γ-CD (8 glucose units). The surface of the toroidal structure is hydrophilic, whereas its cavity is hydrophobic (apolar). Thus, CDs can include (encapsulate) other hydrophobic (apolar) molecules of appropriate dimensions within their cavities and bind them through various types of interactions (London dispersion forces, hydrogen bonding, or dipole-dipole interactions).
Volatile organic compounds (VOCs) encapsulation in CDs represents a feasible and efficient tool to retain and when required to modulate the release of the encapsulated volatiles. CDs, termed “hosts”, form inclusion complexes with the “guests” (i.e., VOCs). To quantitatively analyze the binding strength between the CD-host and VOC-guest the formation constant (Kf) of each inclusion complex is determined1. If there is a 1:1 stoichiometry between the host and the guest, then Kf can be expressed as follows:
Kf=[inclusion complex]/[Host][Guest] or Kf=[CD/G]/[CD][G]
, where CD is cyclodextrin (host), G is VOC (guest), and CD/G is their complex. It is clear that the higher the Kf value the higher the propensity to form an inclusion complex between the guest and the host.
X-ray structures demonstrate rigidity and inflexibility for the β-CD and γ-CD inflexible structures as well as stability in a battery of aqueous and organic solvents, whereas bonds of glucose units within α-CD are weaker and stretchable. Cyclodextrins are stable within a 3–14 pH range.
CDs: Physical properties
Cyclobond™ handbook2 provides the main physical characteristics of CDs (Table, below):
| CD | Glucose unit per cycle (number) | Mw (g/mol) | Cavity diameter (nm) | Water solubility (g/100 ml) |
|---|---|---|---|---|
| α | 6 | 972 | 0.57 | 14.5 |
| β | 7 | 1135 | 0.78 | 1.85* |
| γ | 8 | 1297 | 0.95 | 23.2 |
*If position 2 or 3 hydroxyl groups of the β-CD are ethoxylated or propoxylated the solubility increases to 45g/100mL.
The major message of the Table above is that β-CD is very poorly soluble in water, whereas when ethoxylated or propylated the solubility of the respective β-CD modifications increase significantly. Thus, the solubility of the hydroxypropyl-β-CD (HPβCD) is approximately 45 grams (g) per 100 mL of water.
The removal of toluene from simulated waste gas stream by using various CDs
Blach et al. examined the absorption efficiency of toluene from gas and the potential for regeneration by using liquid absorbents based on CDs. The liquid absorbents they used were aqueous solutions of α-CD, β-CD (and modifications: HPβCD , randomly methylated-β-cyclodextrin [RAMEB], low methylated-β-cyclodextrin [CRYSMEB], sulfobutyl-ether-β-cyclodextrin [4-SBEβCD]), and γ-CD. Specifically, it was demonstrated that all CD-derivatives tested were able to decrease the Henry’s law constant of toluene. Henry’s law describes the distribution of a volatile contaminant between the aqueous and gas phase. The dimensionless and static Henry’s law constant (Hc), or the gas–liquid partition coefficient, is the ratio of a compound’s concentration in the gas phase to its concentration in the aqueous phase at equilibrium. Blach et al. showed the reduction of toluene’s volatility up to 95%, depending on CD chemical structure and concentration. β-CD and its modifications bound toluene the strongest. Absorption experiments demonstrated liquid absorbents based on β-CD can be up to 250-fold more efficient than water3.
Absorption capacity as a function of CD concentration, temperature, and pH
For HPβCD, the highest dynamic absorption capacity was recorded for 10% HPβCD: 0.284 mg of toluene was bound by each gram of liquid absorbent (10% HPβCD) at toluene concentration of 500 mg/m3. The tests were performed in an experimental device (to mimic industrial absorption processes, generally based on a scrubber, a dynamic bubbling set-up has been developed at laboratory scale).
Dynamic absorption of toluene was typically performed at +20°C. However, when the temperature was increased to +60°C, the capacity of 10% HPβCD to absorb toluene decreased 20-fold. Highly basic pH (12) had no influence on toluene’s absorption by 10% HPβCD.
Absorption capacity as a function of toluene concentration in the waste gas stream
The amount of toluene trapped in 10% HPβCD increased linearly with the concentration in the waste gas stream. Specifically, at the toluene’s concentration of 2,000 mg/m3 the absorption capacity of the 10% HPβCD liquid was approximately 1.5 mg/g and when the toluene’s concentration was 6,000 mg/m3 the absorption capacity of the 10% HPβCD liquid increased to approximately 4.5 mg/g. Importantly, the experimental data obtained by Blach and colleagues showed no saturation. Thus, at toluene’s concentrations above 6,000 mg/m3 the absorption capacity of the 10% HPβCD liquid is expected to increase above 4.5 mg/g.
Regeneration of the absorbent
Air stripping in combination with heat was used for the regeneration of the 10% HPβCD liquid absorbent. First, the absorption of toluene was performed at +20°C. Second, the desorption was performed at +20°C, +40°C and +60°C. It was found that the time required for removing toluene decreased with increasing temperature. After the release of toluene from 10% HPβCD, the aqueous 10% HPβCD solution could be used to absorb toluene again. The regeneration procedure took approximately 30 min for air stripping at +40°C and +60°C.
Safety for humans and cost
Numerous clinical studies demonstrated that HPβCD is well tolerated and safe in most of the human patients receiving HPβCD at daily oral doses of 4 – 8 g for at least 2 weeks4. Daily oral doses of 16 – 24 g given for 14 days to human volunteers increased incidences of soft stools and/or diarrhea. Thus, HPβCD is considered to be non-toxic (at least for 14 days) if the daily oral dose does not exceed 16 g.
No measurable effect on kidney function was found in an intravenous dosing study with single doses of HPβCD up to 3 g, which were well-tolerated by all human volunteers5. No adverse effects were reported after a one week intravenous study, when patients received a single HPβCD dose of 1 g.
Following intravenous injection, rapid biphasic decline of unchanged HPβCD in human plasma was observed with single doses of HPβCD up to 3 g (for both women and men). The elimination of HPβCD proceeded almost in its entirety through kidneys without any sign of tubular reabsorption. After oral administration, HPβCD was not detected in plasma or urine, which suggested no absorption from the gastrointestinal tract and low oral bioavailability6.
Thus, at described doses and specified routes of administration, the HPβCD is completely safe for humans. The price for HPβCD varies between USD $13.14 and USD $40.0 per 1 kg.
Solvents
The removal of toluene from simulated waste gas stream by using various solvents
Recently, similar experiments to those performed by Blach et al. were performed using the same experimental device and essentially the same formulas for Hc determination, but with acetic acid, benzyl alcohol, and propylene glycol were used as absorbents7.
First, Villarim et al. replicated the experimental setup used by Blach et al. but using water/solvent (i.e., acetic acid, benzyl alcohol, and propylene glycol) mixtures instead of 10% HPβCD. However, the water/benzyl alcohol, water/acetic acid, and water/propylene glycol results were not as impressive as 10% HPβCD results. Specifically, only 0.35 mg of toluene was absorbed by 1 g of water/benzyl alcohol mixture.
Villarim et al. took a step further to an industrial implementation by utilizing the Terrao absorber (Terrao®; Coudekerque-Branche, France) – a direct heat and mass exchanger with compact geometric dimensions (1 m2 on the ground for 5,000 Nm3/h and 2 m high). The Terrao® exchanger has a hydraulic pumping system allowing to replace the saturated solvent with the backup solvent and has the potential to increase the turbulence of the liquid. The parameters used were: (1) the toluene concentration of 2,000 mg/m3 in the feed gas; (2) the feed gas flow rate 2,000 m3/h; and (3) 300 L of solvent. As in Figure 4, water/benzyl alcohol mixtures demonstrated higher absorption capacities than acetic acid and propylene glycol 16 water mixtures. The highest absorption capacity for toluene (22.70 g of toluene per L of solvent) was achieved with water/benzyl alcohol (20:80 wt/wt) liquid at +20°C. The density of benzyl alcohol is 1.045 g/mL (similar to water), so the resulting capacity is approximately 22,700 mg / 1,000 g = 22.7 mg/g. Thus, at toluene concentration of 2,000 mg/m3 in the feed gas, 22.7 mg of toluene is bound by every 1 g of water/benzyl alcohol liquid mixture.
It is difficult to explain (and Villarim et al. do not do that) how did authors achieved such an amplification in toluene’s binding by the water/benzyl alcohol mixture – from 0.35 mg (experimental device) to 22.7 mg (Terrao absorber) per 1 g of absorbent. The regeneration of water/benzyl alcohol mixture took 12 hours at +45°C.
Safety for humans and cost
Benzyl alcohol is classified as an antimicrobial preservative. Benzyl alcohol had been associated with: (a) neonatal “gasping syndrome”; (b) neurologic toxicity; (c) inhalation toxicity; and (d) systemic hypersensitivity8. Thus, benzyl alcohol is not safe for humans. The cost of benzyl alcohol is below USD $3.99 per liter.
10% HPβCD is the optimal absorbent for the removal of toluene from waste gas stream
In the experimental device (toluene’s concentration: 2,000 mg/m3), per every gram of liquid, 10% HPβCD binds more toluene than water/benzyl alcohol (20:80 wt/wt): 1.5 mg versus 0.35 mg. However, in the Terrao absorber (300 L of solvent), 1 g of water/benzyl alcohol (20:80 wt/wt) liquid binds 22.7 mg toluene at +20°C. Thus, the mass effect cannot be underestimated and the same jump in binding efficiency might be observed for 10% HPβCD. On the other hand, when the toluene’s concentration was 6,000 mg/m3 the absorption capacity of the 10% HPβCD liquid linearly increased to approximately 4.5 mg/g (without any signs of saturation).
If the price of HPβCD and benzyl alcohol is considered, then there might be an advantage of the former. If the price of the HPβCD is USD $13.14, then for 1 kg of 10% HPβCD only USD $1.31 is spent, whereas for 1 kg of water/benzyl alcohol (20:80 wt/wt) at least USD $3.19 should be spent – the 10% HPβCD is USD $1.88 cheaper. If the price of HPβCD is increased to USD $40.0, then water/benzyl alcohol is USD $0.79 cheaper per 1 kg. Thus, it depends, but most likely there is a parity on price here.
HPβCD and hence 10% HPβCD are safe for humans by multiple routes of administration, whereas benzyl alcohol is not safe.
The regeneration of absorbent took approximately 30 – 60 minutes for 10% HPβCD, whereas 12 hours was required for water/benzyl alcohol.
Given the facts above, 10% HPβCD liquid is the optimal absorbent for the removal of toluene.
THE PHENOL AND CRESOLS ABSORBENT
HPβCD – suitable absorbent for the removal of VOCs from waste gas stream
HPβCD will act as absorbent for phenol and cresols
The chromatogram of o-cresol, phenol, m-cresol, and p-cresol separation on CYCLOBOND I 2000 high-performance liquid chromatography (HPLC) column is given in the Figure. The CYCLOBOND I 2000 is βCD-functionalized silica gel. The fact that o-cresol, phenol, m-cresol, and p-cresol can be separated on CYCLOBOND I 2000 HPLC column means that βCD forms inclusion complexes with all these substances. If βCD forms inclusion complexes with o-cresol, phenol, m-cresol, and p-cresol, then HPβCD must form the very same inclusion complexes, though Kf value may sometimes be different . This logic is straightforward and was proven experimentally numerous times – see tables with experimental data in the articles by Kfoury et al., Szaniszló9 et al., Fourmentin10 et al., etc. Thus, similarly to toluene, o-cresol, phenol, m-cresol, and p-cresol will be bound by HPβCD, i.e., HPβCD will act as absorbent for these VOCs.
To reaffirm the logic described above two more explicit examples are brought here. First, by using a combination of affinity capillary electrophoresis, nuclear magnetic resonance titrations, and head-space gas chromatography (GC) method Lantz and colleagues measured the Kf (unit: M-1) values for HPβCD/m-cresol and HPβCD/phenol complexes as 130.5 ± 13.9 and 52.4 ± 5.6 respectively (corresponding complexes with βCD yielded 124.7 ± 27.5 and 59.6 ± 5.8)11. Second, Guo et al. determined Kf values as 94.6 ± 4.2 and 214 ± 12 for βCD/phenol and βCD/toluene complexes by using UV-vis and fluorescence spectroscopy11. Thus, 10% HPβCD liquid is suitable absorbent for o-cresol, phenol, m-cresol, and p-cresol. However, the percentage of HPβCD in aqueous solution that will be optimal for absorbing these VOCs will have to be determined experimentally.
The removal of phenol and cresols by HPβCD from waste gas
Blach et al. demonstrated that when the toluene’s concentration was 6,000 mg/m3 the absorption capacity of the 10% HPβCD liquid linearly increased to approximately 4.5 mg/g (without any signs of saturation). Thus, to bind all the toluene (6,000 mg) in 1 cubic meter (m3) 6,000/4.5=1,333.33 g of 10% HPβCD liquid will be required. The amount of solid HPβCD is hence 133.33 g. The 6,000 mg/m3 concentration of toluene corresponds to 1,600 ppm, which is 0.16%. Thus, 133.33 g of solid HPβCD will be sufficient to absorb 0.16% or 1,600 ppm of toluene. To absorb 3,000 ppm of toluene 250 g of solid HPβCD (or 2.5 L of 10% aqueous solution) will be required and to absorb 6,000 ppm of toluene 0.5 kg (or 5L of 10% solution) will be sufficient.
The typical range for phenol in the waste gas is within the 0.3 – 0.6% range or 3,000 – 6,000 ppm, which corresponds to phenol concentration 11,538.46 – 23,076.92 mg/m3. Currently, the absorption capacity of the 10% HPβCD with respect to phenol and cresols is unknown and will have to be experimentally determined.
It is instrumental to use the work by Blach et al. and the “The removal of toluene from simulated waste gas stream by using various CDs”-section (see above) as a guideline to optimize the essential parameters (HPβCD percentage, temperature, pH, etc.), absorption efficiencies and capacities, Hc and Kf constants, as well as regeneration procedure.
References