The goal of this project is to use molecular modeling and experimental results to design inexpensive, environmentally benign, CO₂-soluble compounds that can decrease the mobility of CO₂ at typical enhanced oil recovery (EOR) reservoir conditions.

University of Pittsburgh, Pittsburgh, PA 
Yale University, New Haven, CT

The research group previously formulated the only known CO₂ thickener, a (fluoroacrylate-styrene) random copolymer, but this proof-of-concept compound was expensive and environmentally unacceptable because it was fluorous. They then identified the most CO₂-soluble, high-molecular-weight, conventional polymer composed solely of carbon, hydrogen, and oxygen: poly(vinyl acetate), or PVAc. PVAc could not dissolve at pressures below the minimum miscibility pressure (MMP), however. The current research effort, therefore, was directed at using molecular modeling and experimental tools to design polymers that are far more CO₂-soluble than PVAc. The subsequent goal was to incorporate this polymer into a thickening agent that will dissolve in CO₂ below the MMP and generate a two- to ten-fold decrease in CO₂ mobility at concentrations of 0.01–1.0 percent by weight. Although most of the thickeners envisioned are copolymers, researchers will also evaluated several small hydrogen-bonding agents and surfactants with oligomeric (very short polymer) tails that form viscosity-enhancing structures in solution , and novel CO₂ soluble surfactants that may be able to generate foams in situ as they mix with reservoir brine (without the need for the injection of alternating slugs of water).

About 2 billion standard cubic feet of CO₂ is injected in domestic oil reservoirs every day in dozens of CO₂flooding projects that produce more than 200,000 incremental barrels of oil per day. However, CO₂ flood performance still has room for improvement because of the low viscosity of dense CO₂. For example, about 8,000 standard cubic feet of CO₂ injection is required for each barrel of oil recovered. Further, large slugs of water must be injected alternately with slugs of CO₂ (water-after-gas or WAG) in an attempt to suppress CO₂“fingering.” This project’s objective is to identify an inexpensive CO₂ thickener composed primarily of carbon, hydrogen, and oxygen. Such a thickener could delay breakthrough of CO₂, reduce the amount of CO₂ required to recover a barrel of oil, and double or triple the oil production rate. Lowering the cost of carbon dioxide enhanced oil recovery would widen the possibility of its application, increasing the volume of oil producible from domestic oil reservoirs.

Previously, researchers used molecular models to design three new types of polymers that may exhibit a high degree of solubility in CO₂. They then synthesized several low-molecular-weight versions of these polymers. These new polymers included poly(3-acetoxy oxetane), or PAO, polyvinyl methoxy methyl ether, or PVMME, and polyvinyl methoxy ethyl ether, or PVMEE. Project performers were able to synthesize and characterize low-molecular weight versions of PAO, PVMME and PVMEE. These polymers proved to be CO₂-soluble, but they were not as soluble in CO₂ as polyvinyl acetate, PVAc, which the researchers had previously identified as the most CO₂ soluble inexpensive polymer. Recently, polymer scientists at GE have agreed to assist the University of Pittsburgh team in the synthesis of high molecular weight PAO at no cost to this project. This high purity sample was also less CO₂-soluble than PVAc. The researchers also identified two non-fluorous polymers that were soluble in dense CO₂ over a wide range of molecular weight, but both were less CO₂ soluble than PVAc. Poly (1-O-(vinyloxy)ethyl-2,3,4,6-tetra-O-acetyl-ß-D-glucopyranoside has been determined to be the second-most CO₂-soluble hyrocarobn-based polymer, and amorphous poly(lactic acid) is the third most CO₂-soluble hydrocarbon-based polymer. p

The researchers successfully designed the first non-fluorous CO₂ thickener by incorporating an associating group into PVAc. PVAc was selected as the “base polymer” for the copolymeric thickener because it remains the most CO₂-soluble, non-fluorous, high molecular weight polymer that has yet been identified. Because associating groups are CO₂-phobic, this copolymer was less CO₂-soluble than PVAc. Therefore, this compound was designed as a “proof-of-concept” thickener that would demonstrate the ability to thicken CO₂ using a non-fluorous copolymer, albeit at pressures well above the MMP. The benzoyl monomer contains a pendant aromatic ring (just like the highly effective fluorinated “polyFAST” thickener previously designed by the researchers) that enables intermolecular associations to occur via intermolecularp- p stacking of the aromatic rings. Therefore a 95mol% vinyl acetate - 5mol% benzoyl copolymer was synthesized. This copolymer was capable of increasing the viscosity of CO₂ by 70% at 298K, 9500 psia and a concentration of 2wt% .

Researchers also synthesized the first non-fluorous CO₂-soluble ionic surfactant by using short PVAc “tails” and tested its ability to form stable foams when mixed with water. Such foams, which would be formed as the CO₂+surfactant solution was injected into a waterflooded reservoir, could be useful in the reduction of CO₂mobility. Further, the introduction of the surfactant with the CO₂ could preclude the need for the injection of alternating slugs of water. Unfortunately, the resultant foams formed with the CO₂-soluble ionic surfactant were not as stable as foams formed with conventional, water-soluble surfactants and required pressures far above typical MMP values to attain dissolution of the surfactant. Subsequent efforts in this area are being directed at the design of non-fluorous, nonionic, commercially available surfactants that generate stable foams because nonionics can dissolve at substantially lower pressures than ionic surfactants. Specific CO₂-philic tails identified as promising include branched alkanes, branched alkylphenyls, poly(butylenes glycol), and oligomers of vinyl acetate; the most obvious nonionic hydrophile is an oligomer of poly(ethylene oxide).

The researchers also successfully designed and synthesized small molecules that were intended to dissolve in CO₂ and then associate in solution via hydrogen bonding, thereby forming viscosity-enhancing macromolecules. Although the first CO₂ soluble hydrogen bonding bisurea compound was synthesized during the course of this research, it did not thickened the dense CO₂. The researchers also synthesized the first non-fluorous CO₂-soluble dendrimer. Although this dendrimer contained hydrogen bonding groups, steric hindrance of the CO₂-philic “arms” apparently prevented any intermolecular viscosity-enhancing associations from occurring.

The project has resulted in:

• The first successful use of computational chemistry methods based on quantum chemistry to design polymers to dissolve in CO₂ from; acetoxy oxetane, vinyl methoxy methyl ether and vinyl methoxy ethyl ether.
• Researchers identified three monomer groups that exhibit stronger interactions with CO₂ than vinyl acetate, which allowed for the possibility that polymers based on these novel monomers may be more CO₂ soluble than PVAc. (Poly vinyl acetate, PVAc, is an amorphous polymer rich in acetate groups that is the most CO₂-soluble polymer based on C, H and O that has yet been identified.)
• Synthesis of the monomer required for poly(3-acetoxy oxetane), or PAO, was completed.
• Development of a technique for the production of high purity sample of poly(3-acetoxy oxetane), thanks to the assistance of GE Global researchers. Synthesis of novel monomers and successful polymerization of then to generate poly vinyl methoxy methyl ether, or PVMME, and poly vinyl methoxy ethyl ether, PVMEE, was completed.
• Synthesis of the monomer 1-O-(vinyloxy)ethyl-2,3,4,6-tetra-O-acetyl-ß-D-glucopyranoside and the subsequent polymerization yielded poly(1-O-(vinyloxy)ethyl-2,3,4,6-tetra-O-acetyl-ß-D-glucopyranoside) P(AcGIcVE). Although not specifically designed by molecular modeling, this polymer was designed to be amorphous (crystalline polymers are difficult to dissolve in CO₂) and rich in acetate groups. This polymerwas determined to be the second most CO₂-soluble polymer that has yet to be identified (second only to PVAc).
• Polylactic acid, PLA, was also investigated because (like PVAc) it is very rich in oxygen content. Researchers studied the phase behavior of different molecular weight (12K-128K) of amorphous, acid-terminated PLA at a concentration of 5wt%. The pressure required to dissolve PLA ranges between 120 – 140 MPa at 298K which is more than twice that required to dissolve PVAc in dense CO₂. PLA is the third most CO₂-soluble high molecular weight polymer, following PVAc and P(AcGIcVE).
• The researchers successfully designed and synthesized the first CO₂ soluble dendrimer. This compound (C76H106N4O50) was extremely soluble in CO₂ at modest pressures, but it did not induce a viscosity change probably due to the steric hindrance of the four acetylated arms, which inhibited the internmolecular associations required for viscosity enhancement . The dendrimer was soluble in CO₂ for 5 wt% at 50 MPa but did not result in viscosity enhancement.
• The synthesis of non-fluorous hydrogen bonding compounds, containing two CO₂ philic tails (rich in acetate groups) separated by bis-urea group, was completed. The hydrogen bonding compound dissolved at 1wt% in CO₂ at 65 MPa. No significant viscosity enhancement occurred, however, even though some degree of association occurred (as evidenced by the formation of a solid micro-fibrillar foam when the CO₂ was released). Researchers synthesized the first non-fluorous CO₂-soluble ionic surfactant by using short PVAc “tails”. The foam formed when the CO₂+surfactant solution was mixed with brine was very modest compared to foams formed by mixing water+surfactant solutions with CO₂.
• The researchers have established that tri-tert-butylbenzene, tritertbutylphenol and acetylated monosaccharides and disaccharides are solids that melt in dense CO₂ and exhibit incredibly high solubility. This observation, combined with recent reports in the literature that highly branched (e.g. containing many tert-butyl groups) hydrocarbon tails can be used in the design of CO₂ soluble surfactants, has provided the group with another promising candidate for CO₂-philic tails for surfactants or other types of small thickeners. The t-butyl groups have not been effective in enhancing the solubility of polymers, however.
• The researchers successfully designed the first non-fluorous CO₂ thickener by incorporating an associating group into PVAc. PVAc was selected as the “base polymer” for the copolymeric thickener because it remains the most CO₂-soluble, non-fluorous, high molecular weight polymer that has yet been identified. Because associating groups are CO₂-phobic, this copolymeric thickener was less CO₂-soluble than PVAc. Therefore, this compound was designed as a “proof-of-concept” thickener that would demonstrate the ability to thicken CO₂ using a non-fluorous copolymer, albeit at pressures well above the MMP. The benzoyl monomer contains a pendant aromatic ring (just like the highly effective fluorinated “polyFAST” thickener previously designed by the researchers) that enables intermolecular associations to occur via intermolecular p- p stacking of the aromatic rings. Therefore a 95mol% vinyl acetate - 5mol% benzoyl copolymer was synthesized. This copolymer was capable of increasing the viscosity of CO₂ by 70% at 298K, 9500 psia and a concentration of 2wt%.
• The basic strategy for synthesizing CO₂-soluble nonionic surfactants has been developed. These surfactants would not be direct thickeners; rather they would form CO₂-in-brine emulsions as the CO₂+surfactant solution is injected into the sandstone or limestone formation targeted for EOR. The CO₂-phile should be a branched alkane, a branched alkyl phenyl group, an oligomer of vinyl acetate, or an oligomer of polypropylene oxide or butylenes oxide. The hydrophile should be an oligomer of ethylene oxide.
• This strategy for designing CO₂-soluble surfactants capable of generating CO₂-in-brine emulsions or foams in-situ (detailed in the prior result) is proving to be successful in a current IAES research project.

Summary 
The objective of this research was to use molecular modeling techniques, coupled with the researchers’ prior experimental results, to design, synthesize, and evaluate inexpensive, non-fluorous CO₂ thickening agents. The first type of thickener to be considered was associating polymers.

First, a highly CO₂-philic, hydrocarbon-based monomers were identified. Polymers or oligomers (small polymers) of this monomer must exhibit high CO₂ solubility at EOR MMP conditions. Second, the molecular weight of a homopolymer of the CO₂-phile must be increased as much as possible without causing the polymer to become insoluble in CO₂. Finally, a small concentration of a CO₂-phobic moiety that promotes viscosity-enhancing macromolecular interactions while not substantially diminishing CO₂ solubility must be incorporated into the polymer. Researchers at Yale University assisted the University of Pittsburgh team in the design and synthesis of these novel monomers.

Yale was solely responsible for the synthesis of a second type of thickener: small, hydrogen-bonding compounds. These molecules have a core that contains one or more hydrogen-bonding groups, particularly, a urea group. Non-fluorous, CO₂-philic functional groups were attached to the hydrogen-bonding core of the compound to impart CO₂ stability and macromolecular stability to the linear “stack” of these compounds. Although we successfully designed these compounds, they did not impart a discernible viscosity increase to the CO₂. Upon depressuriozation, a microfibrous network of these solids remained indicating that the compounds did indeed associate, but apparently not enough to cause a viscosity increase.

The first objective of the polymer-related work in this research project was the identification of the most CO₂-philic, nonfluorous polymeric repeat unit composed solely of carbon, hydrogen, and oxygen. Ultimately it was determined that poly(vinyl acetate) was the most CO₂-soluble, hydrocarbon-based, inexpensive, benign polymer. Unfortunately, pressure required for the dissolution of even dilute amounts of this polymer greatly exceed the MMP. The second objective was to modify the CO₂-soluble polymer such that it could become a thickener. The researchers were successful in designing the first non-fluorous CO₂-thickening copolymer. The benzoyl monomer contains a pendant aromatic ring (just like the highly effective fluorinated “polyFAST” thickener previously designed by the researchers) that enables intermolecular associations to occur via intermolecular p- p stacking of the aromatic rings. Therefore a 95mol% vinyl acetate - 5mol% benzoyl copolymer was synthesized. This copolymer was capable of increasing the viscosity of CO₂ by 70% at 298K, 9500 psia and a concentration of 2wt%. This pressure is substantially above the MMP, therefore the researchers have concluded that the formulation of a non-fluorous, high molecular weight, co-polymeric CO₂ thickener that can dissolve at typical MMP conditions is not possible.

The research group has successfully designed the first CO₂-soluble ionic surfactants that are highly soluble in CO₂. The objective of these surfactants was not to thicken CO₂ directly, but to cause the in-situ generation of CO₂-in-brine mobility control foams or emulsions in the sandstone or limestone formation. The most promising surfactant has a structure similar to the widely used commercially surfactant Aerosol OT, but the hydrocarbon tails have been replaced with oligomers (short polymers) of PVAc. These non-fluorous ionic surfactants could not dissolve in CO₂ at typical MMP pressure values, however. It is the researchers’ recommendation that future CO₂ soluble surfactants intended to be used for generating mobility control foams in-situ incorporate a non-ionic ehtoxylated hydrophile and Other CO₂-philic tails found to be viable candidates for these surfactants include branched alkyl groups, branched alkyl phenyl groups, poly(propylene oxide) and poly(butylenes oxide). It should be noted that these guidelines are already yielding very promising results in an NETL IAES sponsored research program currently being conducted Enick and co-workers at Pitt and Julian Eastoe of Bristol University in the UK.

Most of the team’s work has focused on the use of chemical groups with carbon, hydrogen, and oxygen, such as the acetate group, to enhance CO₂ solubility. Researchers also have found that tert-butyl groups, which are composed solely of carbon and hydrogen, also impart CO₂ solubility to compounds. They have established this trend for small, non-polar compounds and ionic surfactants and hope to determine whether the t-butyl group can increase the CO₂ solubility of polymeric compounds. The inclusion of the simple t-butyl group into compounds may be an inexpensive and easy way to induce CO₂ solubility in small molecules such as surfactants designed for mobility control foams, but it was found to be an ineffective strategy for increasing the CO₂-solubility of polymers.

(January 2010) 
Researchers have successfully designed and synthesized several new polymers using molecular modeling and have measured their CO₂ solubility. To date PVAc remains the most CO₂ soluble polymer. High molecular weight PAO has been successfully synthesized with the assistance of GE Global Research (at no cost to this project). Although the molecular modeling predictions for this polymer were extrememly promising, this polymer was less soluble in CO₂ than PVAc and could not be made in high molecular weight forms.

The researchers successfully designed the first nono-fluorous copolymeric thickener (an analog of poly(fluoroacrylate-styrene)) composed of vinyl acetate (to enhance CO₂ solubility) and benzoyl (which contains a pendant aromatic ring to promote intermolecular interactions via p-p stacking). Poly(vinyl acetate-benzoyl) increased the viscosity of CO₂ by 70% at a concentration of 2wt% at 298 K and 9500 psia. This demonstrates for the first time that CO₂ can be thickened with a non-fluorous compound; unfortunately the requisite pressure is far above typical MMP values.

The non-flourous hydrogen bonding compounds and dendrimers made during the course of this research did exhibit CO₂-solubility. Unfortunately, neither was capable of thickening CO₂.

Because we have found that it is extraordinarily unlikely that a high molecular weight , non-fluorous, copolymeric thickener or a small hydrogen-bonding compound can dissolve in CO₂ at MMP, it now appears extremely likely that the most promising route for decreasing the mobility of CO₂ is the design of nonionic CO₂ soluble surfactants that form mobility controlling, CO₂-in-brine foams in-situ as they mix with reservoir brines. The CO₂-philic tails of the surfactant should be either a branched hydrocarbon, a branched alkyl phenyl group, poly(propylene glycol) or poly(butylenes glycol), or an oligomer of vinyl acetate, while the hydrophile should be an oligomer of poly(ethylene glycol).

This project is completed. The final report is available below under "Additional Information".

$800,000 

$221,277 (28 percent of total)

NETL - Virginia Weyland (Virginia.Weyland@netl.doe.gov or 281-494-2517)
U. of Pittsburgh - Robert Enick (rme@pitt.edu or 412-624-9649)

Final Project Report [PDF-3.58MB]