Perfluorocarbons: Life sciences and biomedical uses Dedicated to the memory of Professor Guy Ourisson, a true RENAISSANCE man.


  • Marie Pierre Krafft,

    Corresponding author
    1. Systèmes Organisés Fluorés à Finalités Thérapeutiques (SOFFT), Institut Charles Sadron (CNRS), 6 rue Boussingault., 6700 Strasbourg Cedex, France
    • Systèmes Organisés Fluorés à Finalités Thérapeutiques (SOFFT), Institut Charles Sadron (CNRS), 6 rue Boussingault., 6700 Strasbourg Cedex, France
    Search for more papers by this author
  • Jean G. Riess

    1. MRI Institute, University of California at San Diego, Medical Center, 410 Dickinson St, San Diego, California 92103-1990
    Search for more papers by this author


Perfluorocarbons are primarily characterized by outstanding chemical and biological inertness, and intense hydrophobic and lipophobic effects. The latter effects provide a powerful noncovalent, labile binding interaction that can promote selective self- assembly. Perfluoro compounds do not mimic nature, yet they can offer abiotic building blocks for the de novo design of functional biopolymers and alternative solutions to physiologically vital issues. They offer new tags useful for molecular recognition, selective sorting, and templated binding (e.g., selective peptide and nucleic acid pairing). They also stabilize membranes and provide micro- and nanocompartmented fluorous environments. Perfluorocarbons provide inert, apolar carrier fluids for lab-on-a-chip experiments and assays using microfluidic technologies. Low water solubility, combined with high vapor pressure, allows stabilization of injectable microbubbles that serve as contrast agents for diagnostic ultrasound imaging. High gas solubilities are the basis for an abiotic means for intravascular oxygen delivery. Other biomedical applications of fluorocarbons include lung surfactant replacement and ophthalmologic aids. Diverse colloids with fluorocarbon phases and/or shells are being investigated for molecular imaging using ultrasound or magnetic resonance, and for targeted drug delivery. Highly fluorinated polymers provide a range of inert materials (e.g., fluorosilicons, expanded polytetrafluoroethylene) for contact lenses, reconstructive surgery (e.g., vascular grafts), and other devices. © 2007 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 45: 1185–1198, 2007.


Perfluorochemicals (PFCs) are known for a range of unmatched specific properties.1, 2 Their exceptional chemical and biological inertness constitutes the most essential common basis for their biomedical potential. PFCs' outstanding stability and inertness relate to the strength of the C[BOND]F bond, the strongest single bond found in organic chemistry, and to the dense, protective, and repellent electron sheath that coats F-chains (CnF2n+1; the italized prefixal symbol F- conventionally stands for perfluoro). PFCs are also characterized by a distinctive contrast between the strength of their intramolecular C[BOND]C and C[BOND]F bonds and the weakness of their intermolecular interactions. The low polarizability of the fluorine atom results indeed into low van der Waals forces. Consequently, liquid PFCs display very low intermolecular cohesiveness, a feature they share with gases. Low cohesiveness is reflected by the highest vapor pressures (relative to molecular weight), lowest surface tensions, and lowest water solubilities of all liquids, and an exceptional propensity for liquid PFCs to dissolve gases. Concurrently, higher F-chain rigidity, when compared with hydrocarbon chains, translates into higher melting points than for the corresponding hydrocarbon compounds, resulting in a narrower liquid phase domain. Extreme hydrophobicity, supplemented by substantial lipophobicity, causes PFCs to segregate among themselves rather than to meddle with other material, including membrane lipids, a feature that certainly contributes to their biological inertness.2–4

The powerful noncovalent and reversible “superhydrophobic” interactions engendered by F-moieties enrich the toolbox of those dealing with fluorinated colloids and interfaces5 or developing supramolecular constructs and practicing complex constitutional dynamic chemistry.6 For biochemists seeking alternatives to hydrogen bonding for nucleic acid pairing or selective protein interactions, F-moieties provide an abiotic (or abiologic) implement. In this respect, it is noteworthy that fluorine appears to display little capacity to act as a hydrogen bond acceptor.

Further PFC-specific or PFC-enhanced properties potentially useful in biomedical applications include high specific gravity, fluidity, and spreading coefficient, high contact angle with water, translating into low friction materials, low refractive index, absence of protons, and a concentration of 19F nuclei that provides a valuable NMR probe.

The exceptional biological inertness of PFCs is illustrated by the marketing approval by the US Food and Drug Administration of the ingestion of F-octyl bromide, C8F17Br, in liter-size doses for use as a contrast agent for bowel delineation by magnetic resonance imaging (MRI) [Fig. 1(a)]. Likewise, filling the lungs of a patient with a liquid PFC can be done without serious untoward effects [Fig. 1(b)]. PFCs administered parenterally, in the form of emulsions, are not metabolized, but excreted unchanged in the expired air. Fluorinated surfactants, although they are usually much more surface active than hydrogenated analogs, are generally much less hemolytic, if at all.3

Figure 1.

(a) X-ray image of the bowel of an adult patient filled with F-octyl bromide (the bromine atom makes the compound radioopaque, and hence, clear on the radiograph). Courtesy Alliance Pharmaceutical Corp. (b) Lungs of an infant patient filled with the same material. (Reproduced from ref. 7 with permission from Lippincott Williams & Wilkins.)

PFCs can be synthesized industrially by substitution of hydrogen atoms by fluorine atoms in an organic compound using electrochemical fluorination, reaction with heavy metal fluorides, or direct action of molecular difluorine. If not properly controlled, these highly exothermal processes tend, however, to yield complex, often ill-defined mixtures. Alternately, PFCs can be obtained by assembling small building blocks that are already fluorinated. Telomerization of tetrafluoroethylene, the precursor of Teflon®, provides very pure materials, as separation of successive terms of a homologous series of F-compounds is very easy. Among the PFCs most investigated for biomedical applications, one can list F-propane, F-butane, F-hexane, F-decalin, F-octyl bromide, F-1,8-dichlorooctane, as well as fluorinated polymers such as expanded polytetrafluoroethylene (ePTFE). Mixed (F-alkyl)alkane diblock compounds, CnF2n+1Cm H2m+1, constitute further valuable components for the engineering of fluorinated colloidal systems. A range of F-alkylated surfactants, with a large variety of polar head groups, has also been synthesized.8 In terms of surface activity, F-surfactants are both more effective and more efficient than their hydrogenated counterparts. Small amounts of such materials allow reducing surface tensions to values that cannot be attained otherwise. The increasing number of pharmaceuticals and agrochemicals that contain individual fluorine atoms or trifluoromethyl groups is out of the scope of this short highlight.9, 10 Only a limited number of selected examples and literature sources are provided.

Inordinately strong hydrophobic and lipophobic effects constitute a powerful driving force for segregation and self-organization of fluorinated amphiphiles. These effects promote phase separation and ordering among fluorinated components and, within such components, of the fluorinated moieties. Highly fluorinated materials thus offer unique components and tools for the engineering of stable self-assembled supramolecular and colloidal systems.2, 11 These systems can be two or three dimensional and usually display several nanocompartmented phases. Reversible temperature-dependent partitioning of PFC-tagged molecules between a PFC phase and an organic solvent is the basis for “fluorous” synthesis, catalysis, and separation technologies.12


Nature (or chance and necessity) chose hydrogen, the most frequently encountered atom on Earth, rather than fluorine as a partner of carbon to engender life as we know it on our planet. Fluorine, although the most abundant of the halogens, ranking 13th in order of frequency of the elements in Earth's crust (carbon ranks 14th), is seldom found in natural compounds. Only a handful of (usually toxic) mono or difluorinated compounds are known. Enzymatic processes exist that form and break the C[BOND]F bonds present in these compounds.13 On the other hand, no natural perfluorocompounds or moieties have, as yet, been found. Consequently, nature did not bother to develop the enzymatic machinery that would allow for their generation and recycling. Nature, thereby, left a wide field open for chemists to develop their own bioinspired (or not) substitute solutions for physiological needs, and beyond.

A priori, the question of how far exogenous PFC (rather than “natural” hydrocarbon) chemistry could provide access to, control of, or interference with living constructs and processes may appear science-fictional and futile. It is certainly intriguing, and hence, heuristic. The large body of literature that exists on investigations of PFCs in relation to life sciences and biomedical issues indicates that such questioning is not devoid of interest. Highly fluorinated compounds and moieties can actually allow unique functional modifications of natural compounds, offer surrogate means of molecular recognition and selection, as well as provide unique microenvironments unknown to nature.

Designing Abiotic Self-Assembling and Selection Processes

Improved selectivity, stability, and performance of therapeutic proteins are important goals for the pharmaceutical industry. Introduction of F-alkyl or aryl groups into peptides and proteins would, a priori, be expected to substantially modify the in vivo behavior of these molecules, their interactions with receptors, and their pharmacological characteristics. Remarkably, the examples given later indicate that fluorinated proteins can retain the structure and activity of their native model. For those engaged in artificial protein and nucleic acid engineering, the introduction of highly fluorinated fragments provides a novel, complementary means for controlling, sorting, and assembling amino acids. It allows templated biosynthesis of peptides from fluorinated amino acids, stabilizing protein folding, achieving selective protein–protein recognition and assembly, DNA recognition and binding, and the modulation of biological processes. The stability of peptides is generally increased when natural amino acids are substituted by fluorinated analogues. Such stabilization augmented with the number of hexafluoroleucine residues introduced.14 Native-like structure was preserved, but the peptides had a more structured backbone and less fluid hydrophobic core. Substitution of four leucine residues by trifluoroleucines in the leucine zipper peptide GCN4-p1d led to a substantial gain in thermal stability and resistance to chemical denaturation of the dimeric coiled-coil structure.15 The DNA binding behavior of the similarly fluorinated construct GCN4-bZip remained identical in terms of affinity and specificity to that of the wild-type peptide. These examples indicate that some protein domains can tolerate extensive fluorination without loss of function. The coenzyme nicotinamide adenine dinucleotide, when fitted with a perfluoropolyether side-chain, actually displayed augmented coenzyme activity in horse liver alcohol dehydrogenase-catalyzed oxidation/reduction reactions, the fluorinated coenzyme being dissolved in a fluorous solvent or in liquid CO2.16

Incorporation of fluorinated amino acids (e.g., tri- or hexafluoroleucine, hexafluorovaline, etc.) creates a fluorinated patch within peptides that constitutes a new type of highly selective peptide–peptide hydrophobic interaction motif. Disproportionation, under equilibrium conditions, of a heterodimer composed of a highly fluorinated peptide and its nonfluorinated analogue led almost exclusively to the homodimers.17 Phase separation (self-sorting) of such fluorous patches appropriately placed onto hydrophobic folds direct helix–helix self-assembly (pairing or oligomerization) within micelles.18 (Fig. 2). Interactions among fluorous patches, coupled with hydrogen bonds, can be superior and more efficient at directing oligomerization of transmembrane helices within phospholipid bilayer membranes than those based on natural aliphatic interfaces.19 This is a fine example of constitutional dynamic synthesis, where the self-assembly of the fluorine-patched elements into complex functional biopolymers is dictated by adaptation to the environment. Systematic studies of the influence of amino-acid side-chain fluorination on protein folding and stability have been initiated.20

Figure 2.

Schematic view of the programmed two-step dynamic self-assembly of fluorinated protein segments in lipid micelles. First, the hydrophobic peptides partition into the micelles, forming α-helices. Then, the strings of superhydrophobic hexafluoroleucine residues seek each other, causing self-association into dimers and higher order aggregates. Fluorine is green, while the backbone of the α-helices is purple. (Reproduced from ref. 18 with permission from National Academy of Sciences.)

Remarkably, incorporation of fluorinated amino acids into proteins can also be accomplished in vivo. This supposes that the fluorinated amino acid analogues are recognized by the appropriate amino acyl-tRNA synthetase enzyme with efficiency similar to that of the natural amino acid. The proliferase response elicited by a fluorinated analogue (a trifluoroisoleucine derivative) of murine interleukin-2 produced in an appropriate E. coli strain was nearly as high as that of the authentic cytokine, indicating folding into an authentic, native structure.21

Where nucleic acids are concerned, the enhanced hydrophobicity of abiotic polyfluorinated aromatic bases (e.g., tetrafluorobenzene or tetrafluoroindole deoxyribose derivatives) was exploited as an alternative to “natural” (Watson and Crick) hydrogen bonding to achieve selective and stable nucleic-acid base pairing in duplex DNA.22 DNA replication was examined using polyfluorinated-nucleotide analogues as substrates. A DNA polymerase active site was able to process the polyfluorinated base pairs more effectively than the analogous hydrocarbon pairs, demonstrating hydrophobic selectivity of polyfluorinated bases for other polyfluorinated bases.23

Incorporation of non-natural polyfluorinated bases, capable of enhanced (or “hyper”-) hydrophobic interactions, expands the genetic alphabet beyond the natural one used in replication, thus illustrating the emerging role of F-compounds in developing chemistry, including constitutional dynamic “bio”chemistry that lies well “beyond” nature.

Creating Novel “Abiotic” Environments—“Hypernonpolar” Compartments for Segregation and Confinement

Living organisms are composed of specialized functional subunits that allow exchange processes, including selective transport, and hence, communication. Their structure, therefore, relies heavily on self-assembly of amphiphilic molecules (e.g., phospholipids) into multiple compartments possessing a panoply of specific characters. At the molecular level, functional proteins also exhibit diverse regions offering distinct environments. Substantial efforts are being devoted to producing micrometer-sized and fluid cell-like compartments that mimic some vital cell functions. For example, a model system for in vitro directed evolution has been designed that uses the aqueous compartments of water-in-oil emulsions to allow gene selection and link genotype and phenotype at the molecular level.24F-alkylated compounds, because of their exceptional aptitude at developing both hydrophobic and lipophobic effects, are highly effective in generating stable, organized, and compartmented molecular systems with controlled complexity. “Multicompartment micelles”25 have, for example, been prepared by micellar terpolymerization of acrylamide with both hydrocarbon- and fluorocarbon-polymerizable surfactants.26 Other multicompartmented systems have been obtained by self-assembly in water of linear27 or star-shaped [Fig. 3(a)]28 triblock copolymers bearing a highly fluorinated arm, a polystyrene arm, and a polymorpholinium or poly(ethylene oxide) arm.

Figure 3.

(a) Schematic representation and cryoTEM image of a multicompartmented micelle obtained from self-assembly in water of a star terpolymer having a perfluoropolyether arm (green), a polystyrene arm (red), and a poly(oxyethylene) arm (blue). (Reproduced from ref. 28 with permission from American Association for the Advancement of Science.) (b) Self-assembled rigid cage with a fluid fluorocarbon interior. The cage is made of 12 palladium(II) ions (yellow dots) and 24 linkers with a fluorocarbon chain (green dots); the fluorocarbon chains-lined cavity within the cage is depicted in light green. (Reproduced from ref. 29 with permission from American Association for the Advancement of Science.)

Exquisite space control at the molecular level is exemplified by the engineering of a spherical shell through spontaneous assemblage of 12 palladium(II) ions with 24 bent F-alkylated pyridine-based ligands [Fig. 3(b)].29 The disordered F-alkyl chains are tethered to the concave side of the rigid shell, hence providing a fluid inner, nanodroplet-like, fluorinated (fluorous) environment whose size is determined by the length of the F-chains. Selective incorporation (cryptation) of PFC molecules was demonstrated. Fluorinated dendrimers also offer precisely shaped nanosize fluorous domains or corona.30, 31 Highly organized helical pyramidal columns were formed by self-assembly of semifluorinated dendrons attached to electron–donor groups.32 These columns subsequently self-organize into large supramolecular pyramidal liquid crystals.

Further examples of controllable spherical or tubular micro- and nanosized fluorinated environments are found in self-assembled vesicles, tubules, ribbons, as well as in a variety of microbubbles and microdroplets, as in aerosols, emulsions, microemulsions, reverse emulsions, gel-emulsions, and multiple emulsions with highly fluorinated phase or membrane components.3, 11 2-D constructs with fluorinated microcompartments include monolayer films, bilayer membranes, and interfaces, as found in Langmuir or Langmuir–Blodgett films or within surface micelles33 and 3-D colloids. Figure 4 illustrates the formation of layered hydrophilic, lipophilic, and fluorophilic shells within the bilayer membrane of fluorinated vesicles.3 These vesicles were obtained either from lipids having F-alkylated chains or from combinations of standard phospholipids with fluorocarbon/hydrocarbon diblocks. Segregation of hydrogenated and fluorinated domains within Langmuir monolayers has led to both laterally and vertically phase-separated zones.34

Figure 4.

Separated nanometer-thick domains within the bilayer membrane of fluorinated vesicles obtained either from perfluoroalkylated phospholipids (a) or from an association of standard phospholipids with (perfluoroalkyl)alkyl diblocks (b). The central hydrophobic and lipophobic fluorous core (green) is flanked by two lipophilic shells (magenta), then by the hydrophilic outer layers of polar heads (blue).2 [Color figure can be viewed in the online issue, which is available at]

Such micro- and nanometer-size environments can serve multiple purposes. They can provide microreactors and matrices for polymerization.35 A fluorosurfactant-stabilized water-in-PFC microemulsion was used to study the influence of confinement on water dynamics.36 They should allow isolation and investigation of isolated single proteins. It is noteworthy that high enzymatic activity (e.g., lipase-catalyzed alcoholysis) can be preserved and even enhanced in a fluorous medium.37

Fluorocarbons also play an essential role in the development of miniaturized “lab-on-a-chip” systems. The approach involves manipulation of discrete droplets of reagents confined and precisely positioned in microchannels.38 Microfluidic technologies allow handling, transferring, mixing, and separating numerous reactants on the nanoliter scale. PFCs provide the continuous, immiscible, inert carrier fluid that separates and carries the various reactants.39 They also offer their unique surface properties, allowing flow and friction control, and facilitating droplet size and stability control. Microfluidic cartridges preloaded with nanoliter plugs of reagents separated by gas bubbles40 within a PFC carrier fluid39 have been fabricated. The preloaded capillaries are then coupled with microfluidic chips that allow merging of reactant droplets (Fig. 5). Such systems may provide a convenient alternative to well plates for screening.39 PTFE capillaries can be used to facilitate fluid handling and to allow high-pressure reactions. Applications involve immunoassays and enzymatic assays, screening and optimization of chemical reactions, and protein crystallization conditions. Crystals grown in plugs inside a microcapillary can be analyzed in situ by X-ray diffraction.41

Figure 5.

Schematic view of PFC-implemented microfluidics for chemists. Capillary cartridges preloaded with plugs of reagents separated by gas bubbles within a fluorocarbon carrier fluid are fitted onto a microfluidic chip where the substrate stream merges with the reagent plugs. (Reproduced from ref. 39 with permission from Elsevier.)


Several highly fluorinated molecular, macromolecular, or supramolecular materials are being used or investigated for biomedical uses.1, 42 Commercially available PFC-based products include the injectable PFC-stabilized, micrometer-size gas bubbles that serve as contrast agents for ultrasound imaging; contact lenses and various materials for ophthalmologic applications; external patches for signal enhancement and homogeneity resolution in MRI; and a panoply of polymers for surgical aids, vessel grafts, and other devices for reconstructive surgery. Further products are in an advanced stage of development, in particular, the submicrometer-size emulsions that are under clinical evaluation for parenteral oxygen delivery (the so-called “blood substitutes”). Molecular imaging and targeted drug and gene delivery are also very active fields of investigation for fluorinated particles. Lung surfactant replacement preparations, cell culture media, tissue and organ preservation media, and more compliant polymers are also being prospected, as well as novel anti-inflammatory and neuroprotective techniques.

A note of caution is here appropriate in these increasingly environmentally conscious times concerning possible environmental issues that may be associated with the use of F-compounds. PFCs are not metabolized. Those injected in the vasculature are eventually excreted unmodified in the expired air. Ideally, microorganisms should be discovered or developed, which would digest PFCs into harmless metabolites. However, our understanding of the biochemistry of the C[BOND]F bond is still rudimentary and only concerns isolated C[BOND]F bonds. Alternatively, recovery of used PFCs could be envisaged. Meanwhile, only low-tonnage medical applications, as is the case for essentially all the products presently commercialized or under development, seem sustainable. The environment protection agencies appear, at this time, to consider that the risk associated with these products, in the required amounts, is insignificant. Fluorinated surfactants pose different problems: the bioaccumulation and toxicity of some of them, including F-alkyl acids such as F-decanoic acid, or of their metabolites, are well documented.43 In other cases, including F-alkylated phospholipids, acute toxicity appears to be very low, but prolonged retention in the organism (which increases rapidly when the F-chain exceeds eight carbon atoms) may limit the length of the F-chain used, the dose administered, or the frequency of their use.3 Some components that may be acceptable for occasional use, for example perioperatively, may not be tolerable for chronic daily use (e.g., delivery of antiasthma drugs through metered-dose inhalers), when the rate of accumulation exceeds the rate of excretion.

Injectable Sound Reflectors—Advances in Diagnosis

PFCs have played a key role in the development, over the past decade, of the injectable micrometer-size bubbles that are used for contrast ultrasound imaging.44–46 Gas bubbles are ideal sound reflectors. However, micrometer-size air or nitrogen bubbles, when injected in the circulation, dissolve rapidly in the blood under the combined action of arterial pressure and Laplace pressure. Low-solubility PFC gases, when introduced in such a microbubble, provide osmotic stabilization against these forces (Fig. 6), thus prolonging the microbubbles' intravascular persistence and making effective contrast radiological examination possible. Additionally, the interaction between bubble contrast agents and sound waves generates harmonics. Adequate manipulation of sound waves and filtration of the incident frequency allow generation of an image that is specific to the microbubbles, and hence, of only the tissues that contain the agent. This provides exquisite visualization of the cavities of the heart, blood vessels, and capillary beds, and can considerably improve the contrast between healthy and pathologic tissues.

Figure 6.

Gas bubbles with a polymeric or phospholipid shell membrane (a) stability enhancement design: small amounts of a fluorocarbon, here F-hexane, C6F14, provide substantial stabilization against Laplace pressure, arterial blood pressure, and oxygen consumption; (b) decay of ultrasound signal (Doppler signal intensity over time) observed in rabbits for nitrogen bubbles osmotically stabilized with F-hexane (dose: 1 mg/kg body weight) when compared with nitrogen alone. (Reproduced from ref. 47 with permission from Elsevier.)

Contrast echocardiology facilitates detection of structural and functional cardiac abnormalities (Fig. 7). Controlled destruction of the circulating microbubbles by a high-energy sound pulse allows, by monitoring the re-entry of fresh bubbles into the field of interest, assessment of tissue perfusion and blood flow abnormalities. The amount of PFC administered for an examination is on the order of a fraction of a milliliter, dispersed within some 108 bubbles, a few micrometers in diameter. Higher quality images enable patients to be diagnosed sooner and more accurately, thus increasing the physician's confidence and reducing downstream testing, and may obviously play a capital role in selection of treatment, monitoring of therapy, and patient outcome.

Figure 7.

Left ventricular cavity opacification: (a) view of the left ventricle of a patient without contrast agent; (b) the same view after a bolus injection of a microbubble contrast agent; the left ventricle (LV), endocardial border (LV/ENDO), and papillary muscle (PAP) are now precisely visualized, allowing myocardial (heart muscle) thickening to be evaluated. When watching the heart in motion, normal functioning heart muscle thickens as it contracts; abnormal functioning heart muscle moves less and does not thicken. (Reproduced from ref. 48 with permission from Blackwell Publishing.)

Further efforts are being aimed at engineering microbubbles with still longer intravascular persistence. The recent finding of a stabilization synergy between the internal PFC gas and a fluorinated shell component represents a significant advance in this respect (Fig. 8).49

Figure 8.

Synergistic stabilization of microbubbles: normalized variation of transmitted ultrasound intensity (I/I0) as a function of time at 25 °C, for microbubbles containing F-hexane and (a) of dimyristoylphosphatidylcholine (DMPC) and having a shell membrane made (b) of an F-alkylated analogue of DMPC (F-GPC). The half-lives of the microbulles with DMPC and with F-GPC shells are 5 and 70 min, respectively. (Reproduced from ref. 49.) [Color figure can be viewed in the online issue, which is available at]

Research in this field is now moving towards targeted microbubbles, capable of seeking and recognizing the specific molecular markers (i.e., unique molecular signature), rather than the anatomical signs, of a given pathology.50 Achieving such “molecular imaging,” requires binding onto the surface of the microbubbles of appropriate antibodies or other ligands capable of seeking specific antigens expressed at the surface of the diseased target cells (Fig. 9). Such agents will provide higher sensitivity and specificity than the standard microbubbles, and hence, earlier and safer assessment of pathology. The targeted pathologies are essentially those that express their antigens within the vascular lumen. Targeted microbubbles can thus allow precise localization of a thrombus (blood clot), an atherosclerotic plaque (cholesterol), an area of inflammation, or a region of active angiogenesis related to tumor growth. Appropriate bubble/sound combinations can also provide novel therapeutic tools. Blood clots can, for example, be broken up by a proper targeted bubble/focused sound wave association.52 Such combinations may also enable noninvasive brain surgery,53 as well as targeted drug delivery via transient opening of the blood–brain barrier.54

Figure 9.

Targeting strategies for microparticles (e.g., bubbles, emulsion droplets) destined for molecular imaging and drug delivery. (a) Passive targeting using the intrinsic ability of certain shell components (e.g., albumin or phosphatidylserine) to bind to receptors expressed on the target cell's surface. (b) Binding of the particle, through avidin–biotin interactions, to antibodies or other ligands that recognize specific disease-related antigens. (c) Covalent binding of such ligands, usually through a poly(ethylene glycol) (PEG) spacer, to a microparticle shell component. (d) Simultaneous binding to a targeted microparticle of stealth-providing elements (e.g., PEG strands), drugs, and markers (e.g., a Gd3+ chelate for MR contrast enhancement). (Reproduced from ref. 51 with permission from Elsevier.)

PFC-based microbubbles are not the only PFC products useful in diagnosis.42 Neat F-octyl bromide gained FDA approval for oral use as a bowel marker during MRI of the gastrointestinal tract. In this case, it is the absence of protons that creates the contrast. PFC-filed pads are available that improve magnetic homogeneity, and hence image quality, when fat-saturation techniques are used during 1H MRI. Further diagnostic tools under investigation include targeted PFC emulsions for molecular imaging of pathology.55 The targeting strategies and targeted pathologies are the same as for the microbubbles described earlier. Target detection and differentiation of pathologic from normal tissue likewise involve grafting onto the droplet's surface of ligands that specifically bind the cellular epitopes and receptors characteristic of the target pathology. While the contrast provided by isolated PFC emulsion droplets is feeble, a thin layer of contiguous droplets attached to a given tissue behaves like a continuous PFC film and provides a highly reflective interface. Incorporation of paramagnetic material, such as a gadolinium(III) complex, allows exploitation of the phenomenon in MRI.

Using the 19F nuclei, one can localize and monitor the absorption and elimination of a PFC in vivo. Analyzing the perturbation caused to the relaxation of the 19F NMR signal by the paramagnetic O2 molecule allows drawing a map of tissue oxygenation and monitoring its variations. It allows, for example, detection of highly vascularized tumor tissues. Finally, certain PFCs that are opaque to X-rays (because of the presence of a bromine atom, for example) allow imaging of the content of lymph nodes, which could prove useful for cancer staging.

Oxygen Delivery to Tissues—“Blood Substitutes”

Blood substitutes were initially aimed at supplying an alternative to donor blood banking and transfusion. The emergence of AIDS provided substantial impetus to their development. Considerable advances in banked blood safety progressively led to defining new therapeutic indications for injectable oxygen carriers, different from those of blood. However, blood shortages remain, in many countries, as a matter of concern, and the impact of allogeneic (donor) blood transfusion on host defenses is still poorly understood.

The fact that, after several decades of relentless research, no satisfactory oxygen carrier has reached the marketplace means that the challenge was more formidable than anticipated. Development efforts were initially directed at hemoglobin (Hb)-based products and subsequently at PFC emulsions as well.1, 56 Figure 10 schematically compares the O2 dissolution and binding mechanisms that underlie O2 transport by PFCs and by Hb. In the case of Hb, a strong, localized chemical coordination bond is established between the dioxygen molecule and the iron atom of a heme. Successive binding of four O2 molecules to the four hemes present in each Hb molecule is cooperative, and saturation occurs when all four iron atoms are coordinated. Hence, the sigmoid shape of the O2 uptake curve [Fig. 10, curve (d)]. In the case of PFCs, there is a simple physical dissolution of O2, characterized by loose, nondirectional van der Waals interactions among like materials. Indeed, liquid PFCs and gases both have very low cohesive-energy densities, as expressed by close Hildebrandt coefficients. O2 dissolution follows Henry's law, i.e., is linearly dependent upon O2 partial pressure. Because PFCs are not miscible with water, PFCs for in vivo O2 delivery are administered in the form of emulsions. Figure 10 shows that achieving full benefit of PFC emulsion administration requires that the patient breathe pure or close-to-pure oxygen. O2 release to tissues can then reach 90% of emulsion content.

Figure 10.

Oxygen transport by fluorocarbons versus Hb. (a) In the case of perfluorocarbons, O2 dissolution is characterized by loose, nondirectional van der Waals interactions. Oxygen solubility follows Henry's law, i.e., is directly proportional to the gas' partial pressure (curve c). (b) In the case of Hb, a strong, localized covalent chemical bond is established with the iron atom of a heme. Successive binding of four O2 molecules to the four hemes of Hb is cooperative, and saturation occurs when all four iron atoms are bound. Hence, the sigmoid shape of the O2 uptake curve, which levels off when the partial pressure of O2 on Earth is attained (curve d). (Riess, 2001, no. 45621.) [Color figure can be viewed in the online issue, which is available at]

The most advanced O2-carrier under development, Oxygent™, consists of a 60% weight/volume concentrated submicronic PFC emulsion, primarily F-octyl bromide (C8F17Br, perflubron), with a small percentage of F-decyl bromide (C10F21Br) added.1, 57 Selection of F-octyl bromide as the primary PFC results from a compromise between adequate emulsion stability and acceptable excretion rate. The former increases with decreasing water solubility, hence increasing molecular weight of the PFC, while the latter diminishes exponentially with increasing molecular weight. Organ retention can be mitigated by introducing a lipophilic element in the PFC, e.g., the terminal bromine atom in the case of F-octyl bromide. The slight lipophilic character facilitates transportation of the PFC by blood lipoproteins on their way to the lungs, where they are excreted. The less water-soluble added F-decyl bromide stabilizes the emulsion by reducing the rate of molecular diffusion (Ostwald ripening), which is the main cause for PFC emulsion droplet growth over time. Egg yolk phospholipids provide an effective emulsifier that is commonly used and well accepted in pharmaceuticals. The emulsion is terminally heat-sterilized, can be stored for 2 years under standard refrigeration conditions, and is ready for use.

Oxygent has passed extensive preclinical and toxicity trials.1, 58 Effective O2 delivery has been established in numerous animal models (Fig. 11) and human clinical trials, including a Phase III clinical trial conducted in general surgery patients undergoing acute normovolemic hemodilution (ANH). In this trial, use of the emulsion led to statistically significant avoidance and reduction of transfusion of banked blood.60 No significant side effects were reported as being attributable to the emulsion. The clinical data indicated that 1 g of PFC was equivalent to about 1.5 g of Hb.61 In a second Phase III trial in cardiac surgery patients undergoing cardiopulmonary bypass, in which the emulsion was used in conjunction with ANH and a second blood withdrawal procedure called intraoperative blood donation, untoward neurological and bleeding events were observed. Although the overall incidence rates of these adverse events were within clinical expectations and within the ranges reported in the literature, the rates of these events were higher in the patients that had received Oxygent when compared with the control group of patients, and thus the trial was suspended. Analysis of the clinical data statistically indicated that the imbalance in the adverse events rates was caused by excessive hemodilution in the patients receiving Oxygent, meaning that the testing protocol—and not the PFC emulsion—was responsible for the observed untoward effects.58 In parallel, extensive hemostasis and hemolysis studies investigated possible interactions between the emulsion and fluids used during surgery. However, these studies found no link between the presence of Oxygent and the observed imbalance in adverse events.62 Clinical development is now being pursued in Europe to decrease the incidence of postoperative organ failure. In China, development is directed towards reduction of exposure to allogeneic blood and blood sparing during surgery.

Figure 11.

Efficacy of the PFC emulsion Oxygent AF0144 in a canine model mimicking surgical blood loss after acute normovolemic hemodilution. Both the test group and the control group animals were hemodiluted (Hb brought down from ca. 14 to ca. 8 g/dL), breathed oxygen, and lost blood, hence Hb (x-axis), in a controlled manner. Mixed venous O2 tension ( equation image, the O2 tension after tissues and organs have been irrigated; y-axis) reflects the adequacy of tissue oxygenation. The difference in tissue oxygenation between treatment (a) and control groups (b) is significant. The PFC-treated dogs (1.35 g of PFC per kg body weight in this experiment) still benefit from adequate tissue oxygenation at Hb levels one-fifth of normal. (Reproduced from ref. 59 with permission from FASEB.)

Ongoing research aims at further increasing emulsion stability, controlling particle size, and prolonging intravascular persistence. Incorporation of fluorocarbon–hydrocarbon diblock compounds (e.g., C6F13C10H21) in the formulation was found to substantially increase emulsion stability and reduce droplet size.1 The diblock molecules were shown to incorporate, at least in part, into the surfactant film that coats the emulsion droplets.63 Another approach to tissue oxygenation, which consists of administering an aqueous suspension of oxygen nanobubbles stabilized by a volatile PFC, appears promising. Experimental proof of efficacy for this system includes survival of erythrocyte-depleted rats and pigs, and of pigs with potentially lethal hemorrhagic shock and with severe right-to-left shunt.64 Better understanding of the “physiology” of PFC emulsions and PFC-stabilized microbubbles is needed to optimize the conditions of use and benefit of such products for the patient.

Perfluorocarbons as Drugs and Drug-Delivery Systems

Anti-inflammatory effects of PFCs infused in the lungs have been reported repeatedly.42 PFCs have recently been shown to prevent formation of a liquid condensed (crystalline) phase in a phospholipid film (Fig. 12). This fluidization effect facilitates respreading of dimyristoylphosphatidylcholine, the main component of the lung surfactant, upon inspiration.66 Experimentation on premature rabbits demonstrated a significant increase in tidal lung volume, allowing survival of the treated animals, while controls were all dead within minutes. PFCs may thus prove useful in lung surfactant replacement compositions for neonates and possibly for the treatment of the acute respiratory distress syndrome in adults. Pulmonary infusion of cold PFCs is being investigated as a means of rapidly cooling the body and thus protecting the brain after cardiac arrest, myocardial infarction or stroke, or during neurosurgery and cardiac surgery. A PFC emulsion was found to prevent the adhesion of certain β-cell lines to cell culture plastic and promoted the formation of pseudoislets capable of insulin secretion.67

Figure 12.

Compression isotherm (π vs. A) of DPPC measured at 25 °C in an atmosphere of N2 (dashed line) or of N2 saturated with F-octyl bromide (solid line). Insets: fluorescence images of (a) the DPPC monolayer compressed at π = 15 mN/m clearly showing the crystalline domains and (b) the DPPC monolayer in contact with F-octyl bromide showing prevention of crystallization, even at high pressures (π = 30 mN/m).65

The development of appropriate drug delivery technology is gaining importance in disease management. Each drug requires its specific delivery system that optimizes efficacy and minimizes side effects (e.g., anticancer agents) through controlled release and, optimally, targeting. Many PFC- and F-surfactant-based delivery systems, including micelles, dendrimers, vesicles, microbubbles, emulsions, and other molecular or self-assembled systems, have been engineered to help ensure the delivery of fragile biopolymers (e.g., peptides, proteins, and plasmids), as well as protect normal tissue.3, 42 The stability of fluorinated vesicle membranes is generally superior to those of vesicles made from standard lipids and their permeability to drugs can be substantially lower. Incorporation of an F-alkyl chain into a molecule can modify diffusion across membranes and across the blood–brain barrier. The lung, with its large contact area, appears to provide a privileged route for delivery of such systems. Reverse water-in-PFC emulsions loaded with various drugs can, for example, be administered via metered-dose inhalers.68

Incorporation of F-chain end-groups into biodegradable polymers can help modulate biodegradation (e.g., the hydrolysis rate of polyesters) and drug release. Surface treatment of polymer capsules by PFC plasma fluorination can reduce deleterious water penetration.69 Plasma surface fluorination of polymers such as polyethylene or polypropylene results in lesser wetability and inactivation of enzymes.

PFC-based colloids, in particular micro- and nanosized emulsions and bubbles, are being investigated as site-directed drug-delivery vehicles. The position of microbubbles targeted to diseased tissues can be monitored by ultrasound imaging, and release of their drug load can eventually be triggered by a suitable ultrasound pulse. Like for other colloids, fluorinated colloids may, as needed, be rendered stimuli-dependent (pH, pressure, temperature).

Materials for Ophthalmology and Reconstructive Surgery

Various gaseous (F-propane, SF6) and liquid (F-octane, F-decalin, F-perhydrophenanthren, mixed fluorocarbon–hydrocarbon diblocks, and fluorosilicones) F-compounds are being used as intraocular tamponades for use in vitreoretinal surgery.70 Polymers with F-alkyl moieties are used in corneal inlays and implants, and in intraocular lenses.

Soft contact lenses incorporating fluorinated polymers (perfluoropolyether dimethacrylates, fluorosiloxanes) and showing increased permeability to oxygen and reduced lipophilicity, thus allowing continuous wear, are being investigated. A fluorosilicon-containing hydrogel lens is commercial.71

Polymers used in vascular grafts need to be stable in biological media and compatible with high pressure and blood flow conditions. It is essential that the grafts be nonthrombogenic (i.e., do not promote clot formation). They should ideally have viscoelasticity behavior comparable to that of native vessels and should promote arterial regeneration. Polymer characteristics should not change over time as the graft needs to remain compliant for a very long time. Surface properties are of utmost importance.

Microporous ePTFE holds a prominent place among fluorinated polymers for reconstructive surgery, especially as vessel substitutes (Fig. 13). PTFE was made microporous (standard pore size around 30 μm) by extrusion and sintering. Molded microporous ePTFE surfaces display wetting characteristics similar to those of the cardiovascular endothelial lining. Such ePTFE is highly crystalline, nonbiodegradable, and displays an electronegative luminal surface that is antithrombogenic. It has also improved cell-adhesion characteristics compared with plain PTFE. Its main disadvantage is its relatively high rigidity. ePTFE is widely used, especially as aortic and lower limb bypass grafts with excellent results.73

Figure 13.

Reconstruction of the left brachiocephalic vein using a ringed PTFE graft.72 [Color figure can be viewed in the online issue, which is available at]

Patency is, however, less satisfactory for small-diameter grafts (<6 mm). Efforts made to modify ePTFE to fill the need for small-diameter vascular grafts include carbon coating to further increase electronegativity, increasing pore sizes to encourage endothelialization, impregnation with or binding of heparin, and coating the grafts with autologous endothelial cells. Inert fluoropolymers can also be surface-modified using plasma or photochemical treatment to introduce chemical functions that allow peptide or poly(ethylene glycol) attachment. For example, a poly(vinylamine) polymer with pendant peptides specifically associated with cell adhesion (integrin-binding peptides), and fitted with F-alkyl branches for adsorption onto ePTFE, has been synthesized that facilitates the adhesion and growth of endothelial cells on ePTFE.74 In another approach, F-alkylated oligomers coupled with cell-binding peptides were blended with polyurethane. The bioactive F-alkylated species migrated to the polymer surface, providing control of cell adherence.75F-moieties can thus be used to modify polymer surfaces in view of reducing thrombogenicity, protecting a polymer from enzymatic degradation, controlling/promoting specific cell attachment, or delivering bioactive elements. More compliant polymers, having higher elasticity than ePTFE (e.g., poly (vinylidene difluoride) or poly(tetrafluoroethylene-co-hexafluoropropylene)), are being investigated.

ePTFE is also being used in vascular access devices, as for hemodialysis, and in conduits for nerve regeneration. A PFC coating can lower friction of guidewires for catheters and provide low friction antithrombogenic coating for other cardiovascular devices.

The specific properties of highly fluorinated materials are expected to continue inspire chemists and offer new research tools for the life sciences. Innovative projects that will extend biochemistry beyond natural are awaited. More effective products for diagnosis and therapy will be sought. Pharmaceutical development involves increasing complexity and heavy regulatory, financial and other constraints, and time. Managing such complexity and bringing together the required diversity of complementary skills and resources should turn the unique potential of highly fluorinated materials into useful products.


Dedicated to the memory of Professor Guy Ourisson, a true renaissance man.

Biographical Information

original image


Marie Pierre Krafft is Director of Research at the Institut Charles Sadron (Centre National de la Recherche Scientifique) in Strasbourg, France. Dr. Krafft received her Ph.D. in chemistry from the University of Nice (1989). After a postdoctoral stay in a Californian startup, she became responsible of a research group at the Laboratoire de Chimie Moléculaire in Nice. In 1997, she joined the Institut Charles Sadron, where she is now the leader of one of the research groups. She focuses on the synthesis, investigation, and applications of highly fluorinated compounds, in particular amphiphiles, as components of self-assembled molecular systems. Her present research includes the elaboration and study of multicompartmented micro- and nano-objects, nanocompartmentation being induced by fluorinated chains in organized systems such as monolayers, surface micelles, vesicles, tubules, microbubbles, and emulsions. The potential of such systems for biomedical applications (lung surfactant, tissue oxygenation, bioartificial pancreas, diagnosis) and in materials science (nanopatterned surfaces) is also explored. Dr. Krafft has published over 80 articles, including 4 book chapters, 10 invited reviews, 9 patents, and she has given numerous lectures. She sits on the Committees of the Nanosciences Program of the French National Research Agency, of the Division of Physical Chemistry, and of the French Network of Fluorine Chemistry. She coedits the Surfactant section of Current Opinion in Colloid and Interface Science.

original image


Jean G. Riess received his Doctorat-ès-Sciences degree from the University of Strasbourg, his mentor being Prof. Guy Ourisson. He then spent 2 years with Prof. John Van Wazer at Monsanto's Central Research Department in Saint-Louis, MO, learning some phosphorus and transition metal chemistry. In 1968 he became Professor at the University of Nice, France, where he founded, directed, and eventually became the honorary director of the Unité de Chimie Moléculaire (associated with the Centre National de la Recherche Scientifique). Subsequently, Prof. Riess joined a Californian start-up as a VP of Research and Development and then retired in France as a consultant. His research successively involved organic and inorganic phosphorus chemistry, transition metal chemistry, organometallics, and eventually perfluorochemicals. His present interests are in fluorocarbons, fluorinated amphiphiles, their colloid chemistry, and biomedical uses, including fluorocarbon emulsions for in vivo oxygen delivery (the so-called “blood substitutes”), fluorocarbon-based contrast agents, fluorinated self-assemblies, and drug delivery systems. Prof. Riess has published about 380 papers and holds some 25 patents. He has served on numerous councils and committees, and has chaired or cochaired international conferences on phosphorus chemistry and blood substitutes. He has won awards from the French Academy of Sciences, French Chemical Society, Alexander von Humboldt Stifftung, City of Nice, and Controlled Release Society, as well as Alliance's first Distinguished Contribution Award. He also holds a Research Associate position at the University of California, San Diego, and sits on the Board of Directors of Alliance Pharmaceutical Corporation.