Low-Flammable Parahydrogen-Polarized MRI Contrast Agents

Abstract

Many MRI contrast agents formed with the parahydrogen-induced polarization (PHIP) technique exhibit biocompatible profiles. In the context of respiratory imaging with inhalable molecular contrast agents, the development of nonflammable contrast agents would nonetheless be highly beneficial for the biomedical translation of this sensitive, high-throughput and affordable hyperpolarization technique. Pointing the way in this direction, we assess the hydrogenation kinetics, the polarization levels and the lifetimes of PHIP hyperpolarized products (acids, ethers and esters) at various degrees of fluorine substitution. The results highlight important trends as a function of molecular structure that are instrumental to design new, safe contrast agents for in vivo imaging applications of the PHIP technique, with an emphasis on the highly volatile group of ethers used as inhalable anesthetics.

Graphical Abstract

To design low-flammable NMR/MRI inhalable contrast agents with the parahydrogen-induced polarization technique, trends are derived from kinetic, polarization, and relaxation measurements for a number of functionalized hydrocarbons with various degrees of fluorine substitution.

Introduction

NMR and MRI represent a set of powerful techniques to provide diagnostic information through molecular characterization, yet their signal intensity is directly proportional to the nuclear spin polarization (P) that is only about 10⁻⁵ for protons in physiologically and clinically relevant conditions.^[1] Hyperpolarization techniques aim at increasing P through the manipulation of the population difference between nuclear spin levels, resulting in several orders of magnitude gains in sensitivity.^[2–5] One of them, Parahydrogen-Induced Polarization (PHIP) converts the singlet spin-order of parahydrogen (p-H₂) into a large P via the pairwise addition of p-H₂ to an unsaturated substrate.^[6–8] The high magnetization of the hyperpolarized (HP) pool of p-H₂-derived protons induces the occurrence of RASER (Radiofrequency Amplification by Stimulated Emission Radiation), a quantum effect in which the radiofrequency modes of the inductive detector of an NMR spectrometer are enslaved by the proton nuclear spin evolution. Proton RASER allows for determining J-coupling constants and chemical shift differences while surpassing the nominal spectroscopic resolution and uncovering nonlinear effects in coupled spin oscillators, e.g., frequency combs, line collapse, chaos, etc.^[9–12]

PHIP and other hyperpolarization techniques such as dissolution Dynamic Nuclear Polarization (d-DNP)^[13] and Spin-Exchange Optical Pumping (SEOP)^[14] have attracted much attention for their potential for in vivo applications.^[15–23] Hyperpolarized (HP) compounds can indeed be employed as injectable or inhalable contrast agents.^{[24, 25]} Currently, in vivo applications of hyperpolarization methods use ¹³C and ¹²⁹Xe nuclei because of the greater lifetimes of their HP states compared to those of protons (>10 s for ¹³C and ¹²⁹Xe vs. ~1–2 s for ¹H).^{[17, 26–28]} However, despite major successes of HP ¹³C and ¹²⁹Xe contrast agents in a number of biomedical studies and clinical trials, ¹³C and ¹²⁹Xe detection is not available on clinical MRI scanners. As a result, a costly (~$0.5M) and clunky hardware upgrades and custom MRI sequences must be implemented on a clinical MRI scanner to enable these functionalities. All-in-all, even though the ¹²⁹Xe and ¹³C-based contrast agents are ready for prime time biomedical applications, the MRI scanner availability is clearly a substantial translational barrier.

The unveiling of long-lived singlet states (LLS) by M. H. Levitt and colleagues, with T_LLS values significantly greater than corresponding spin-lattice relaxation time constant T₁, has revived the interest in imaging of proton-hyperpolarized contrast agents.^[29–32] Interestingly, we have recently demonstrated the existence of LLS in HP propane and HP diethyl ether in the gas phase using homogeneous and heterogeneous (HET) PHIP^[33–37]— HET-PHIP is particularly promising since it allows production of pure catalyst-free HP hydrocarbon gas. Both compounds (propane and diethyl ether) with anaesthetic features showed promising enough polarization levels and lifetimes (T_LLS of 3–4 s at physiologically relevant conditions) to envision testing them as inhalable contrast agents in our up-and-coming preclinical laboratory. It should be stressed that this gain in the HP state lifetime is substantial – for example, HP propane would retain less than 15% of its potency during 2-s long inhalation if stored using T₁ versus retaining over 50% of its potency during 2-s long inhalation if stored in accord to T₁ under physiologically relevant conditions.^[33–37] The key benefits of proton-hyperpolarized inhalable contrast agents is their low cost of precursors, low cost of hyperpolarization process and high production throughput. Moreover, unlike HP ¹²⁹Xe, proton-hyperpolarized contrast agents can be readily detected on already installed clinical MRI scanners thereby making this new technology readily available to the doctors and patients.

Pulmonary functional imaging is a substantial unmet medical need: there is currently no widespread clinical imaging modality to perform high-resolution functional lung imaging: Computed Tomography (CT), conventional MRI, and X-ray can only provide structural images of dense tissues—informing about pathologies like tumors and pneumonia—but yielding little or no information about lung ventilation, perfusion, alveoli size, gas-exchange efficiency, etc. Deadly diseases such as chronic obstructive pulmonary disease (COPD), Asthma, Constrictive Bronchiolitis, Lung Injury, and Pulmonary Fibrosis affect >300 million people worldwide and cause ~3 million annual deaths (5% of all deaths worldwide).^[38] These diseases do not have any clinical imaging marker, especially in the early disease stages when intervention can be potentially curative. This state of affairs is in contrast with that of cancer imaging, which includes a wide range of imaging modalities with their own merits, such as MRI, CT, ultrasound, mammography, Positron Emission Tomography (PET), and others—collectively enabling early diagnosis via population screening and monitoring response to treatment. Furthermore, CT scans (2D and 3D X-ray) expose the body to ionizing radiation, and thus cannot be performed frequently due to increased risk associated with cancer-inducing radiation. On the other hand, MRI involves no ionizing radiation, and is effectively non-invasive. Inhalable HP contrast agents can provide 3D functional lung imaging information on single breath hold^{[17, 21]}—thus, the use of proton-hyperpolarized contrast agents^[23] has a potential to revolutionize pulmonary health care.

In our recent work, we have demonstrated the feasibility of clinical-scale production of HP propane gas and its high-resolution (0.3×0.3 mm² pixel size).^[39] The feasibility of pulmonary imaging even when using non-hyperpolarized propane has been demonstrated by Hane and co-workers in rats.^[40] Feasibility of HP diethyl ether MRI has also been demonstrated.^[41]

Although diethyl ether is approved for medical use in some countries including Russia, the development of a new, gas-phase, inhalable proton-hyperpolarized contrast agent demands to address flammability concerns specifically. For example, the use of the first commercially developed anesthetic (diethyl ether) has been phased out in many developed countries and WHO removed it from the list of essential medications after 2003.^[42] In this realm, we report here on the chemical conversion and proton polarization build-up, maxima and lifetimes of PHIP products fluorinated to various extents, revealing trends as a function of structure and bonding patterns that are essential to tackle with the development of efficient and low-flammable PHIP contrast agents. Fluorinated hydrocarbons hyperpolarized via the PHIP technique have been reported previously, showing possibility for inducing polarization transfer from ¹H to ¹⁹F via magnetic field cycling^[43–45] or for taking advantage of the lipophilic propensity of fluorocarbons in in vivo imaging,^[46] both examples illustrating valuable advantages in using HP fluorinated compounds other than non-flammability. However, the previous works have been focused on materials intended for the use in the liquid state. As the medical field of anaesthesiology employs nowadays a group of highly volatile, halogenated, and non-flammable ethers as inhalable anaesthetics, such as isoflurane CF₃CHClOCHF₂, sevoflurane (CF₃)₂CHOCH₂CH₂F, and desflurane CF₃CHFOCHF₂, this work examines alternative scenarios for PHIP, which could eventually provide novel in vivo molecular contrast agents combining inhalable, anaesthetic and non-flammable properties with the high detection sensitivity and simplicity of the PHIP hyperpolarization technique.

Results and Discussion

The hydrogenation reactions studied here are shown in Scheme 1. The fluorinated molecules include:

Scheme 1.

PHIP reactions reported in the present study. Pairwise addition of parahydrogen in CD₃OD is found efficient for all unsaturated substrates, at the exception of 17FPDe, which dissolves poorly in methanol, and 10FPVE, for which the reaction with H₂ leads to the formation of hydrofluoric acid (HF).

• Vinyl trifluoroacetate (3FVA) leading to ethyl trifluoroacetate (3FEA);

• Trifluoroethyl acrylate (3FEAcr), hexafluoroisopropyl acrylate (6FPA), and dodecafluoroheptyl acrylate (12FHA), leading to trifluoroethyl propionate (3FEP), hexafluoroisopropyl propionate (6FPP), and dodecafluoroheptyl propionate (12FHP), respectively;

• Trifluoroethyl vinyl ether (3FEVE), the first halogenated hydrocarbon anesthetic to be produced (a.k.a. fluroxene), leading to trifluoroethyl ethyl ether (3FDE);

• Trifluoromethyl acrylic acid (3FMAA) leading to 2-trifluoromethyl propanoic acid (3FMPA).

The results are compared to those obtained with corresponding non-fluorinated molecules: vinyl acetate (VA) leading to ethyl acetate (EA), hydroxyethyl acrylate (HEA) to hydroxyethyl propionate (HEP), ethyl vinyl ether (EVE) to diethyl ether (DE), and methacrylic acid (MAA) to 2-methyl propanoic acid (MPA). Note that two other fluorinated precursors were tentatively hydrogenated: 1H,1H,2H-Perfluoro-1-decene (17FPDe) and perfluoro propyl vinyl ether (10FPVE). For alkene 17FPDe, the chemical conversion toward alkane 17FPDa (~15–20% after 30 s reaction) and the ¹H polarization of 17FPDa (~0.1% at 10 s reaction) were low and 17FPDe was poorly dissolved in solvent methanol-d4, aborting further experiments with this molecule. For ether 10FPVE, the hydrogenation reaction was found hazardous due to the formation of hydrofluoric acid (HF) seen as glass NMR tube melting.

All measurements were performed with standard high-throughput 5 mm NMR tubes filled under argon atmosphere. The solutions contained ~ 40 mM of substrates and ~4 mM of rhodium catalyst dissolved in CD₃OD, at the exception of the EVE + p-H₂ → DE reaction for which 186 mM of EVE were used in a recent study.^[37] The samples were pressurized at 8 bar with p-H₂ (>99%) and heated at 80 °C for 30 s. Note that all studied compounds are volatile liquids under normal pressure—as a result, if the depressurization of the HP liquid would render the expansion of HP proton-hyperpolarized gas. However, this depressurization would lead to unwanted polarization losses and make the comparison and quantitative studies of reaction kinetics rather challenging. As a result, the NMR detection of proton-hyperpolarized compounds was performed in the liquid state in 5 mm NMR tubes.

The hydrogenation reaction was achieved by bubbling p-H₂ though the solution at a flow rate of 150 standard cubic centimeter per minute (sccm) in the Earth’s magnetic field (50 μT), corresponding to ALTADENA conditions^[47]—the sample transfer took less than 2 seconds from cessation of p-H₂ flow and spectrum acquisition unless noted otherwise. The chemical conversion was evaluated for each sample from the thermal spectra acquired before and after the reaction, as well as the residual concentration of the products in CD₃OD (liquid fraction) in the case of highly volatile ethers. ¹H polarization levels were corrected accordingly (see Supporting Information for details).

We note that no HP resonances was observed before p-H₂ bubbling. Moreover, once p-H₂ bubbling stopes, the chemical reaction is effectively ceased under our conditions.^[48]

¹H NMR spectra were measured with a benchtop NMR spectrometer (Spinsolve Carbon 60, Magritek, 61 MHz / 1.4 T) and polarization levels were calibrated against reference spectra of neat ¹³C-labeled ethyl acetate (Figure 1a,b). RASER activity, which prevents from measuring polarization levels just as the reaction is completed, is only briefly observed with 40 mM solutions after full polarization build-up for the most strongly hyperpolarized molecules, i.e. ethyl acetate and propionates. Therefore, a delay of only 5 seconds in the Earth’s magnetic field before inserting the samples in the bore of the spectrometer was necessary to avoid RASER. The polarization levels were back calculated to account for the corresponding relaxation. Significant RASER activity can however be easily observed at higher concentrations, that is even for HP molecules with relatively modest polarization levels of a few %, such as shown in Figure 1c,d for 3FDE where a two-mode RASER is induced without any radiofrequency excitation pulse.

Figure 1.

a) ¹H NMR spectrum of HP 3FDE in CD₃OD solution acquired using 8° flip angle. Polarization of 3.2% was measured after correction for evaporation and the Earth’s field relaxation. b) Corresponding NMR spectrum of neat thermally polarized ethyl-acetate-1-¹³C. c) ALTADENA RASER signal of HP 3FDE recorded without excitation pulse. d) Fourier transform (FT) spectra of the RASER signal for the regions outlined by purple and orange boxes in panel c (H_A / H_B two-mode RASER).

Figure 2 shows the ¹H polarization levels P_H measured as a function of reaction (bubbling) time t and fitted to:

$$P_H\left(t\right)= \frac{P_{max}}{T_{cat}/T_{LLS}-1}\left(exp\left(-\frac{t-t_0}{T_{cat}}\right)-exp\left(-\frac{t-t_0}{T_{LLS}}\right)\right)$$ Eq. 1

in order to derive P_max, the theoretical maximum polarization neglecting relaxation, and T_cat, the time constant for the catalytic reaction (or polarization build-up), while leaving t₀ fixed to 1 s and T_LLS fixed to the H_A-H_B averaged values measured independently in the Earth’s magnetic field (~50 μT, see Figure S2). Figure 3 compiles P_max, T_cat, T_LLS, and T₁ (measured at 1.4 T, see Figure S2) values for all compounds shown in Figure 2, at the exception of 3FEA for which the measured P_H were too low.

Figure 2.

PHIP kinetics for hydrogenation reactions in CD₃OD at 8 bar and 80 °C. All reactions were studied with ~ 40 mM solutions, at the exception of the EVE + p-H₂ → DE reaction ([EVE] = 186 mM). Solid circles correspond to polarization of H_A protons (methine CH group of 2-methylpropanoic acid (MPA) and 2-trifluoromethylpropanoic acid (3FMPA), methylene CH₂ group in all other cases) and empty squares correspond to polarization of H_B protons (methyl CH₃ group). H_A and H_B polarizations were averaged in the fit. (a) ¹H polarization of ethyl acetate and ethyl trifluoroacetate (EA and 3FEA) as a function of reaction time. (b) ¹H polarization of propionates (HEP, 3FEP, 6FPP and 12FHEP) as a function of reaction time. (c) ¹H polarization of diethyl ether and 2,2,2-trifluoroethyl ethyl ether (DE and 3FDE) as a function of reaction time. Insert represents the liquid fraction of EVE / DE and 3FEVE / 3FDE, which decreases with reaction time due to evaporation (d) ¹H polarization of 2- MPA and 3FMPA as a function of reaction time.

Figure 3.

(a) Maximum polarization without relaxation P_max. (b) Build-up time T_cat. (c) Earth’s field relaxation time constant T_LLS. (d) High field (1.4 T) relaxation time constant T₁. H_A and H_B polarizations are averaged to derive P_max (a), T_cat (b) and T_LLS (c). H_A notation corresponds to HP protons of CH group (in case of MPA and 3FMPA) or CH₂ group (in all other cases) and H_B notation correspond to HP protons of CH₃ group.

For 3FVA / 3FEA, it is clear that the presence of the CF₃ group on the carboxylic side induces deleterious effects: the ¹H polarization reaches a maximum of about 0.1% and the chemical conversion is less than 40% after 10 s of reaction, whereas in the non-fluorinated case (VA / EA) the chemical conversion is complete after 10 s of reaction with P_max = 25% and T_cat ~ 1 s. These results are in sharp contrast with those obtained for the series of acrylates / propionates. We find that the replacement of –CH₂OH by –CF₃ or even –(CF₂)₅CF₂H does not affect much of the speed and specificity of the catalytic pairwise addition reaction, with P_max on the order of 20%, T_cat ~ 1 s and full chemical conversion after 10 s of reaction. In the case of the branched 6FPP, P_max and T_cat are equal to 14% and ~1 s, respectively, and the chemical conversion is also completed after 10 s of p-H₂ supply. The negative inductive effect of the electron-accepting fluorinated group, which affects the electron density at the CC double bond through the conjugated π-system involving the CO carbonyl group, is thus only deleterious at short distance; in 3FEP, 6FPP and 12FHP the fluorinated groups and the conjugated systems are separated by an aliphatic carbon so that little effects on the hydrogenation reaction outcomes are observed. On the other hand, expectedly, the larger the carbon framework is the shorter is the relaxation constant (both T_LLS in the strong coupling regime and T₁ in the weak coupling regime), with a clear 3FEP > 6FPP > 12FHP trend.

In the case of highly volatile ethers EVE / DE, substituting the methyl group with –CF₃ induces a decrease of polarization by a factor of ~2.5. However, the hydrogenation reaction remains fast with 100% conversion reached after 10 s of reaction, and neither the relaxation nor the buildup time are affected by the presence of F atoms. Non-specific interactions in the catalytic hydrogenation reaction are therefore more pronounced for HP ethers than HP propionates, but remain far from deleterious.

Methacrylic acid represents a test case for F substitution on a methyl group that forms a branch adjacent to the double bond hosting the pairwise addition of p-H₂. For both MPA and 3FMPA hydrogenated products, the chemical conversion is found similar, i.e., 79% and 75% after 10 s of reaction, respectively, and the buildup time constants are significantly longer (7 to 9 s) than for the other compounds studied here. Interestingly, the effective P_H of 3FMPA is found about 3 times higher than the one of MPA, even though HP MPA and 3FMPA show similar relaxation constants at both Earth’s field and high field. Note that, methine (H_A) and methyl (H_B) groups exhibit a pronounced difference in T₁ at high field, with values > 30 s for methine and < 10 s for methyl. This is also manifested in the strong coupling regime (at Earth’s magnetic field), with methyl group polarization values inferior by about 20% when compared to methine group values. This effect is not due to the presence of F atoms in the vicinity of the HP protons, as it is observed similarly for MPA and 3FMPA.

We now turn the discussion toward the designing of an effective, non-flammable and inhalable PHIP contrast agent. As the flammability limits for most of the molecules studied here (PHIP substrates and products) are yet unknown, we invoke the empirical model developed by Kondo et al. to predict the lower (L) and upper (U) flammability limits (in vol%) of partially fluorinated and perfluorinated compounds.^{[49, 50]} This model is based on experimental values measured for 74 alkanes, alkenes, ethers, and esters following the ASHRAE criteria.^[51] The experimental values of L and U can be predicted with an average relative deviation of 10% and 36%, respectively, as shown in Figure 4 for a selection of 42 benchmark molecules along with the PHIP substrates and hydrogenated products. Although the predicted upper flammability limit of non-fluorinated ethers such as EVE and DE are underestimated, this model reproduces fairly well the subtle differences in partially fluorinated ethers and esters that are closely related to the molecules studied here.

Figure 4.

Fluorine substitution rate F and predicted lower (L) / upper (U) flammability limits in vol% of the PHIP substrates (regular font) and products (bold font) compared to a representative set of experimental measurements obtained with non-fluorinated and fluorinated alkanes, alkenes, ethers and esters used to model L / U.^[50] The orange line in the F rate column represents the minimal F substitution degree to allow for non-flammability (F = 0.625).

By comparing the chemical structures, F rates, and experimental and predicted flammability limits within this ensemble of molecules, one can derive the following general trends in terms of flammability:

• The larger the F substitution rate is, the less flammable the compound is. Compounds with F > 0.625 are mostly non-flammable.

• Flammability further decreases when F atoms are attached to C atoms adjacent to double bonds, whereas –C–CF₃ group has no diminishing effect in flammability other than increasing the F rate.

• In ethers, –O–CF₃ group significantly decreases flammability while –O–CF= and –O–CF₂– groups have no particular effect.

Another important dimension that must be accounted for the development of PHIP inhalable contrast agents concerns the normal boiling points (b.p.) of the hydrogenation products. Overall, perfluorination and partial fluorination tend at decreasing the boiling point, which is desirable to ensure an efficient vaporization at atmospheric pressure of the HP PHIP products. To estimate these values, we refer to the modified Joback group contribution method developed by Devotta and Pendyala.^{[52, 53]} In this empirical model originally developed for the screening of alternatives to chlorofluorocarbons (CFCs), halogens are treated differently from other functional groups—because of halogen-halogen and halogen-hydrogen nonbonding interactions—so that isomeric features are taken into account. The boiling point of 3FDE is estimated to be 15 °C lower than the one of DE (35 °C), bringing it to room temperature. Consequently, this partially fluorinated ether exhibits rather ideal properties for engineering a simple vaporizer with no condensation issues. The moderate fluorinated content of 3FDE (F = 0.3) does not however imply non-flammability.

Increasing F substitution rate while keeping intact the carbon framework and the hydrogen atoms on the hydrogenated arm of diethyl ether leads to CH₃CH₂OCF₂CF₃ (5FDE), with F = 0.5, L = 4.0, U = 15.2, and a boiling point of 19 °C. This molecule certainly lies near the non-flammability boundary, if not in the nonflammable group. We plan to test it with PHIP once its precursor CH₂CHOCF₂CF₃ (5FEVE) is available. Alternatives exist with larger systems to further reduce flammability while allowing for an effective PHIP process, as our results with propionates and acids suggest. For instance, we envision studying the following two candidates:

• CH₃CH(CF₃)OCH₂CF₃ (F = 0.5, L = 3.5 vol%, U = 12.4 vol%, and b.p. = 50.4 °C);

• CH₃CH(CF₃)OCH₂CF₂CF₃ (F = 0.57, L = 3.5 vol%, U = 12.4 vol%, b.p. = 71 °C).

These compounds are of interest especially because of the 3-fold increase in polarization measured in the case of 3FMPA, i.e. where a branched CF₃ group is attached to the methine carbon. This effect could possibly compensate for the polarization loss observed when partial fluorination is used on the non-hydrogenated arm of an ether. The potential flammability should be assessed experimentally, because the fluorine content of these systems is close to the elusive F = 0.625 non-flammability limit. Further increasing F substitution rate with larger carbon backbones would certainly guarantee non-flammability, yet at the expense of increasing the boiling point.

The feasibility imaging studies in the gas-phase is the next logical step of this work to demonstrate the feasibility of functional MRI using proton-hyperpolarized fluorinated hydrocarbons. It should be noted that future HP gas imaging studies would greatly benefit from MRI detection of long-lived spin states in the gas phase as they enable a much longer readout time window. These LLS persist in low magnetic field in the strongly coupled regime, i.e., when the spin-spin coupling is greater than the chemical shift difference of the nascent parahydrogen derived protons. In practice, such field is achieved below 0.4 T for propane and below 0.1 T for other less symmetric hydrocarbons. A number of suitable low-field MRI scanning platforms is becoming increasingly available: for example, Time Medical PICA and Hyperfine MRI scanner to name a few. Our ongoing efforts are currently focused on establishing the dedicated low-field MRI imaging facility with the goal of feasibility studies in phantoms, excised lungs and large animal model (e.g., sheep) hopefully in 2021–2023.

Conclusion

A consistent series of ¹H NMR spectroscopy experiments was introduced to compare the yields, polarization levels, build-up times and lifetimes of parahydrogen-induced polarization (PHIP) reaction products involving fluorinated molecules. The results show evidence that, depending on the F-substituted functional groups, fluorination can have no effect on the hyperpolarized states or nearly completely suppress them, as the partially fluorinated, hyperpolarized molecules studied here exhibit a broad range of polarization levels (with measured P_H maxima ranging from 0.1% to about 20%). The cross analysis of these results enable us to rationalize the following trends:

• For esters, F substitution on the carbon adjacent to the carbonyl group (α-carbon) is highly deleterious, both in terms of reaction efficiency (poor chemical conversion) and specificity (maximum polarization level < 0.1 %). However, F substitution within the –O–alkyl (alkoxy) group does not affect the efficiency and specificity of the reaction when the α-carbon is not bound to F atoms.

• For highly volatile ether such as diethyl ether, which represents a promising class of PHIP molecules to be used as inhalable anesthetics, the polarization levels are lowered by a factor of ~ 2.5 with the F substitution of the methyl group. The results are still deemed suitable for biomedical applications with highly desirable reduced flammability.

• A branched CF₃ group directly attached to the methine group of the CC double bond hosting the addition of parahydrogen seems promising for reaching high fluorine content, because the measured polarization levels are found significantly higher than in the non-fluorinated case. This effect could possibly mitigate the polarization decrease observed with linear, partially fluorinated ethers.

In fine, all the hydrogenation reactions studied here were intentionally catalyzed with the same Rh complex for comparison purposes; nevertheless each reaction could also benefit from the more rational design of a targeted catalyst.