Spectral performance of a composite single-crystal filtered thermal neutron beam for BNCT research at the University of Missouri.

Parameter studies, design calculations and initial neutronic performance measurements have been completed for a new thermal neutron beamline to be used for neutron capture therapy cell and small-animal radiobiology studies at the University of Missouri Research Reactor. The beamline features the use of single-crystal silicon and bismuth sections for neutron filtering and for reduction of incident gamma radiation. The calculated and measured thermal neutron fluxes produced at the irradiation location are 9.6 x 10(8) and 8.8 x 10(8)neutrons/cm(2)s, respectively. Calculated and measured cadmium ratios (Au foils) are 217 and 132. These results indicate a well-thermalized neutron spectrum with sufficient thermal neutron flux for a variety of small animal BNCT studies.

Synthesis, structure, and reactivity of closo-2,3,4,5,6,7,8,9,10,11-decahydroxy-1,12-bis(sulfonic acid)-1,12-dicarbadodecaborane(12).

The reaction of closo-1,12-bis(lithio)-1,12-dicarbadodecaborane(12) (1,12-bis(lithio)-p-carborane) with SO(2) formed closo-1,12-bis(lithiosulfinato)-p-carborane (10) in nearly quantitative yield. The latter was converted to closo-1,12-bis(sulfinic acid)-p-carborane (13) via H(+)-exchange. The corresponding 1,12-bis(sulfonic acid) derivative of p-carborane (12) was obtained in high yield by treating 10 with SO(2)Cl(2) and subsequent AlCl(3) mediated hydrolysis of the closo-1,12-bis(chlorosulfonyl)-p-carborane intermediate. The exhaustive oxidation of 12 in hot aqueous H(2)O(2) (30%) afforded B-decahydroxy-1,12-bis(sulfonic acid)-p-carborane (15) in 40% yield. As a byproduct, closo-B-decahydroxy-1-sulfonic acid-p-carborane (14) was formed. Both 14 and 15 were also obtained from the hydroxylation of 10 and 13. Compound 14 was obtained directly in 88% yield by heating 1-sulfinic acid-p-carborane (17) in H(2)O(2) (30%). Compound 17 was synthesized from diphenylmethylsilyl-protected p-carborane by using the method employed in the synthesis of 13. The X-ray structures of 15, its disodium salt, and its dipotassium salt are presented and discussed. Exhaustive methylation of 15 with methyl triflate furnishes closo-B-decamethoxy-1,12-bis(methyl sulfonate)-p-carborane (20). The characterization of closomer 20 also includes its crystal structure determination.

Synthesis and luminescence spectroscopy of a series of [eta(5)-CpFe(CO)2] complexes containing 1,12-dicarba-closo-dodecaboranyl and -ylene ligands.

Three new cyclopentadienyliron dicarbonyl compounds, 1-[eta(5)-CpFe(CO)(2)]-1,12-C(2)B(10)H(11), 1-[[eta(5)-CpFe(CO)(2)]-1,12-C(2)B(10)H(10)-12-yl](2)Hg, and 1,12-[eta(5)-CpFe(CO)(2)](2)-1,12-C(2)B(10)H(10), composed of 1,12-dicarba-closo-dodecaborane as a ligand precursor were synthesized and found to be luminescent. The uncoordinated 1,12-C(2)B(10)H(12) bridging ligand precursor is luminescent with a band maximum at 25180 cm(-1), while the iron complexes luminesce at lower energies in the range 13120-14210 cm(-1). The lowest energy excited electronic state in the iron complexes is assigned to a ligand field transition of the iron chromophore. Cyclic voltammetry of 1,12-[eta(5)-CpFe(CO)(2)](2)-1,12-C(2)B(10)H(10) displays two discrete one-electron oxidations, and the luminescence maximum is red shifted from that observed in 1-[eta(5)-CpFe(CO)(2)]-1,12-C(2)B(10)H(11). Both of these observations suggest that the iron-centered chromophores are weakly coupled. In contrast, the 1-[[eta(5)-CpFe(CO)(2)]-1,12-C(2)B(10)H(10)-12-yl](2)Hg complex is uncoupled as is evident from the single oxidation process observed with cyclic voltammetry. The extinction coefficient of 1,12-[eta(5)-CpFe(CO)(2)](2)-1,12-C(2)B(10)H(10) is six times that of 1-[eta(5)-CpFe(CO)(2)]-1,12-C(2)B(10)H(11), while the extinction coefficient of 1-[[eta(5)-CpFe(CO)(2)]-1,12-C(2)B(10)H(10)-12-yl](2)Hg is only twice that of 1-[eta(5)-CpFe(CO)(2)]-1,12-C(2)B(10)H(11). These spectroscopic properties are explained in terms of two coupled antiparallel transition dipole moments.

Octahedral coordination of halide ions (I-, Br-, Cl-) sandwich bonded with tridentate mercuracarborand-3 receptors.

The "anti-crown" B-hexamethyl 9-mercuracarborand-3 (1) was shown to complex halide ions (I-, Br-, Cl-) in an eta(3)-sandwich fashion. Symmetry-allowed interactions of the filled halide ion p-orbitals and the corresponding empty mercury p-orbitals result in three equivalent p(Hg)-p(halide)-p(Hg) three-center two-electron bonds and a sandwich structure. The molecular structures of [Li.(H(2)O)(4)][1(2).I].2CH(3)CN, MePPh(3)[1(2).Br].((CH(3))(2)CO)(2).(H(2)O)(2), and PPN[1(2).Cl] were determined by single-crystal X-ray diffraction studies. Compound [Li.(H(2)O)(4)][1(2).I].2CH(3)CN crystallized in the triclinic space group P-1, a = 13.312(8) A, b = 13.983(9) A, c = 13.996(9) A, alpha = 61.16(2) degrees, beta = 82.34(2) degrees, gamma = 86.58(2) degrees, V = 4365(2) A(3), Z = 1, R = 0.063, and R(w) = 0.171. Compound MePPh(3)[1(2).Br].((CH(3))(2)CO)(2).(H(2)O)(2) crystallized in the monoclinic space group C2/c, a = 24.671(8) A, b = 17.576(6) A, c = 26.079(8) A, beta = 106.424(6) degrees, V = 10847(6) A(3), Z = 8, R = 0.0607, and R(w) = 0.1506. Compound PPN[1(2).Cl] crystallized in the monoclinic space group C2/m, a = 37.27(2) A, b = 29.25(1) A, c = 10.990(4) A, beta = 100.659(7) degrees, V = 11774(8) A(3), Z = 4, R = 0.0911, and R(w) = 0.2369.

Dodecahydroxy-closo-dodecaborate(2-).

The cesium salt of the icosahedral borane anion dodecahydroxy-closo-dodecaborate(2-), Cs(2)[closo-B(12)(OH)(12)], Cs(2)1, was prepared by heating cesium dodecahydro-closo-dodecaborate(2-), Cs(2)[closo-B(12)H(12)], Cs(2)2, with 30% hydrogen peroxide. The other alkali metal salts A(2)1 (A = Li, Na, K, Rb) precipitated upon addition of ACl to warm aqueous solutions of Cs(2)1. The ammonium salt, [NH(4)](2)1, and the (mu-nitrido)bis(triphenylphosphonium) salt, [PPN](2)1, were obtained similarly. The [H(3)O](2)1 salt precipitated upon acidification of aqueous solutions of Cs(2)1 with hydrochloric acid. The solubility of these salts in water was determined by measuring the boron content of saturated aqueous solutions of A(2)1 (A = Li, Na, K, Rb, Cs), [H(3)O](2)1, and [NH(4)](2)1 using ICP-AES. Although these salts are derived from a dianion with twelve pendant hydroxyl groups, the alkali metal salts surprisingly displayed low water solubilities. Water solubility decreases with a decrease in the radius of A(+), except for the lithium salt, which is slightly more soluble than the potassium salt. The [H(3)O](2)1 and the [NH(4)](2)1 salts provide rare examples of water-insoluble hydronium and ammonium salts. The low water solubility of the A(2)1 salts is attributed to the dianion's pendant hydroxyl groups, which appear to function as cross-linking ligands. Four alkali metal salts, A(2)1 (A = Na, K, Rb, Cs), were characterized in the solid state by single-crystal X-ray crystallography. These data revealed intricate networks in which several anions are complexed through their hydroxyl groups to each alkali metal cation. In addition, the anions are engaged in hydrogen bonding with each other and, if present, with water of hydration. This cross-linking results in the precipitation of aggregated salts. Cation coordination numbers decrease with cation radius. Thus, cesium and rubidium are ten-coordinate, whereas potassium is seven-coordinate and sodium is six-coordinate. The geometry of anion 1(2)(-) is independent of cation identity; the B-B and B-O bond lengths of the various A(2)1 salts (A = Na, K, Rb, Cs) are identical.

Dodecamethyl-closo-dodecaborate(2-).

Bis(tetraethylammonium) dodecamethyl-closo-dodecaborate(2-), [NEt(4)](2)[closo-B(12)Me(12)], [NEt(4)](2)2, was prepared employing modified Friedel-Crafts reaction conditions from [NEt(4)](2)[closo-B(12)H(12)], [NEt(4)](2)1, trimethylaluminum, and methyl iodide. The [NEt(4)](2)2 salt provides sufficient solubility in water to allow the synthesis of the important alkali metal salts A(2)2 (A = Li, Na, K, Rb, Cs) using cation-exchange procedures. The solid state structure of colorless [AsPh(4)](2)2 reveals a nearly perfect icosahedral B(12) cluster with B-B bonds ranging from 1.785(3) to 1.807(3) A and B-C bonds of 1.597(3)-1.625(3) A. In contrast, the crystal structure of dark-red [Py(2)CH(2)]2 (obtained from [NEt(4)](2)2 and [Py(2)CH(2)]Br(2)) contains a distorted icosahedral dianion [B-B = 1.740(13)-1.811(14) A, B-C = 1.591(13)-1.704(13) A]. In the [Py(2)CH(2)]2 salt, the dianion 2(2-) and its dipositive dipyridiniomethane counterion form a red charge-transfer complex. One-electron oxidation of 2(2)(-) by ceric(IV) ammonium nitrate affords the blue, air-stable radical [hypercloso-B(12)Me(12)](*-), dodecamethyl-hypercloso-dodecaborate(1-), 2(*-), isolated as the PPN salt. X-ray crystallography reveals that the geometries of the B(12) clusters observed in hypercloso-[PPN]2 and closo-[AsPh(4)](2)2 are identical and essentially undistorted icosahedra. The anion in the [PPN]2 structure contains B-B bonds ranging from 1.784(8) to 1.806(7) A and a range of B-C bonds from 1.596(7) to 1.616(7) A.

A Hydrogen-Bonded [(Mercuracarborand-Water)2 -Benzene] π-Sandwich Complex.

Two cyclic Lewis acidic receptors, namely B-hexamethyl-9-mecuracarborand-3, [9,12-(CH3 )2 -C2 B10 H8 Hg]3 , each host a single water molecule. A benzene molecule is situated between the two complexed water molecules in a sandwich fashion that suggests there is O-H⋅⋅⋅π hydrogen bonding.

Dodeca(carboranyl)-substituted closomers: toward unimolecular nanoparticles as delivery vehicles for BNCT.

The syntheses of the dodeca(carboranyl)-substituted closomers, dodeca[7-(2-methyl-1,2-dicarba-closo-dodecaboran-1-yl)heptanoate]-closo- dodecaborate(2-) and dodeca[7-(8-methyl-7,8-dicarba-nido-dodecaboran-7-yl)- heptanoate]-closo-dodecaborate(14-) are reported. These boron-rich, unimolecular nanospheres possess great potential for future use as drug-delivery platforms for boron neutron capture therapy (BNCT).

A selective optical sensor based on [9]mercuracarborand-3, a new type of ionophore with a chloride complexing cavity.

A highly selective optical sensor for chloride, based on the multidentate Lewis acid ionophore [9]mercuracarborand-3, is described herein. This sensor is constructed by embedding the mercuracarborand ionophore, a suitable pH-sensitive lipophilic dye, and lipophilic cationic sites in a plasticized polymeric membrane. The multiple complementary interactions offered by the preorganized complexing cavity of [9]mercuracarborand-3 is shown to control the anion selectivity pattern of the optical film. The film exhibits a significantly enhanced selectivity for chloride over a variety of lipophilic anions such as perchlorate, nitrate, salicylate, and thiocyanate. Furthermore, the optical selectivity coefficients obtained for chloride over other biologically relevant anions are shown to meet the selectivity requirements for the determination of chloride in physiological fluids, unlike previously reported chloride optical sensors. In addition, the optical film responds to chloride reversibly over a wide dynamic range (16 microM-136 mM) with fast response and recovery times.

A study of the sequential acid-catalyzed hydroxylation of dodecahydro-closo-dodecaborate(2-).

The closo-[B12H12-n(OH)n]2- (n = 1-4) ions have been synthesized by the reaction of cesium dodecahydro-closo-dodecaborate(2-), Cs21, with aqueous sulfuric acid. Variation of the reaction temperature, time, and acid concentration results in the stepwise introduction of from one to four hydroxyl groups. Each individual hydroxylation step proceeds regioselectively, affording only one isomer per step. Further substitution of the hydroxylated cluster preferentially takes place at a B-H vertex meta to a B-OH vertex. The closo-[B12H12-n(OH)n]2- (n = 1-4) species, designated 2-5, respectively, are characterized by one- and two-dimensional 11B NMR spectroscopy, IR spectroscopy, and high-resolution fast atom bombardment (FAB) mass spectrometry. A rationale that qualitatively explains the influence of the hydroxyl group on the chemical shifts of the individual boron vertices is developed. Furthermore, the solid state structures of closo-[B12H11(OH)]2-, 2, and closo-1,7-[B12H10(OH)2]2-,3, are determined by X-ray diffraction. Crystallographic data are as follows: For [MePPh3](2)2, monoclinic, space group P2(1)/n, a = 890.1(5) pm, b = 1814(1) pm, c = 1270.5(7) pm, beta = 101.66(2) degrees, Z = 2, R = 0.055; for [MePPh3](2)3, monoclinic, space group P2(1)/n, a = 887.6(4) pm, b = 1847.2(8) pm, c = 1271.1(5) pm, beta = 101.17(1) degrees, Z = 2, R = 0.065. In addition, synthetic routes to O-derivatized species of the anions 2-5 such as closo-[B12H11(OTiCpCl2)]2-, 7, closo-1,7-[B12H10(OTiCpCl2)2]2-, 8, closo-1,7,9-[B12H9(OTiCpCl2)3]2-, 9, closo-[B12H11(OCONHPh)]2-, 10, and closo-1,7-[B12H10(OSO2Me)2]2-, 11, are described. The crystal structures of 7 and 11 are determined by single-crystal X-ray diffraction. Crystallographic data are as follows: For [MePPh3](2)7, monoclinic, space group Cc, a = 2530.5(2) pm, b = 1653.3(1) pm, c = 1281.3(1) pm, beta = 118.79(2) degrees, Z = 4, R = 0.085; for [HPy](2)11, monoclinic, space group P2(1)/n, a = 1550.9(8) pm, b = 993.1(5) pm, c = 1726.5(9) pm, beta = 112.36(2) degrees, Z = 4, R = 0.061.

The synthesis and characterization of bis-substituted derivatives of the [a2-B20H18]4- anion and the interconversion of isomers.

The apical-apical (a2) isomer of [Et4N]4[B20H18] reacts with oxalyl chloride in dichloromethane to produce a protonated bis-substituted carbonyl species, [Et4N][a2-B20H17(CO)2] ([Et4N][H1]), in 60% yield. Removal of the bridging hydrogen of the [H1]- anion in aprotic media results in rearrangement to form the equatorial-equatorial [e2-B20H16(CO)2]2- anion ([e2-1]2-). The reaction of [Et4N][H1] with sodium azide in acetonitrile produces [Et4N]3- [a2-B20H17(NCO)2] ([Et4N]3[H2]) in 53% yield, which subsequently reacts with isopropylamine in acetonitrile to provide the urea derivative [a2-B20H16(NH2C(O)NH(i-Pr))2]2- ([Et4N]2[4]) in 89% yield. The [H1]- ion is hydrolyzed in aqueous acetonitrile to give a protonated [a2-B20H17(CO2H)2]3- ion ([H3]3-) in 61% yield. The a2 isomers of the bis-substituted species [B20H16(CO2H)2]4- ([3]4-), [B20H16(NCO)2]4- ([2]4-), and [B20H16(NH2C(O)NH(i-Pr))2]2- ([4]2-), formed by the removal of the bridging proton from their protonated precursors, rearrange to form a mixture of ae isomers in solution.

Mercuracarborand "anti-crown ether"-based chloride-sensitive liquid/polymeric membrane electrodes.

Highly sensitive and selective chloride liquid/polymeric membrane electrodes are described that employ [9]-mercuracarborand-3 (MC3), a neutral preorganized macrocyclic Lewis acid, as the anion carrier. MC3-based chloride-sensitive membrane electrodes, doped with different mole percentages of cationic additives (5, 10, and 60 mol % tridodecylmethylammonium chloride) relative to the amount of the carrier, exhibit enhanced potentiometric selectivity for chloride over other anions, including more lipophilic anions such as perchlorate, nitrate, and thiocyanate. In addition, the selectivity coefficients obtained are shown to meet the requirement for clinical applications. The obtained selectivity pattern is shown to correlate very well with 199Hg NMR titrations of MC3 with various anions, performed in organic solvents. Optimized membrane electrodes show a near-Nernstian response toward chloride over a wide concentration range and have micromolar detection limits. MC3-based chloride sensors show a fast response time (in the order of few seconds), as well as short recovery time. The developed mercuracarborand-based sensors do not practically respond to pH changes over the pH range of 2.5-7.0. Response characteristics (e.g., detection limit, linear range, response slope, and selectivity) of the [9]mercuracarborand-3 based chloride sensors remain essentially the same over a period of approximately 2 months, reflecting remarkable stability and well-defined chemistry of the macrocyclic Lewis acid ionophore.

Toward a cancer therapy with boron-rich oligomeric phosphate diesters that target the cell nucleus.

The viability of boron neutron capture therapy depends on the development of tumor-targeting agents that contain large numbers of boron-10 (10B) atoms and are readily taken up by cells. Here we report on the selective uptake of homogeneous fluorescein-labeled nido-carboranyl oligomeric phosphate diesters (nido-OPDs) by the cell nucleus and their long-term retention after their delivery into the cytoplasm of TC7 cells by microinjection. All nido-OPDs accumulated in the cell nucleus within 2 h after microinjection. However, nido-OPDs in which the carborane cage was located on a side chain attached to the oligomeric backbone were redistributed between both the cytoplasm and nucleus after 24 h of incubation, whereas nido-OPDs in which the carborane cage was located along the oligomeric backbone remained primarily in the nucleus. Furthermore, cell-free incubation of digitonin-permeabilized TC7 cells with the nido-OPDs resulted in nuclear accumulation of the compounds, thus corroborating the microinjection studies. Our observation of fluorescence primarily located in the cell nucleus indicates that nuclear-specific uptake of sufficient amounts of 10B for effective boron neutron capture therapy ( approximately 10(8)-10(9) 10B atoms/tumor cell) via nido-OPDs is achievable.

Aromatic polyhedral hydroxyborates: bridging boron oxides and boron hydrides.

No explosion, but per-B-hydroxylation occurs if the icosahedral boron hydrides [closo-B12 H12 ](2-) (see picture), [closo-CB11 H12 ](-) , or closo-1,12-(CH2 OH)2 -1,12-C2 B10 H10 are refluxed in 30 % hydrogen peroxide. Thus, the three isoelectronic species [closo-B12 (OH)12 ](2-) , [closo-1-H-1-CB11 (OH)11 ](-) , and closo-1,12-H2 -1,12-C2 B10 (OH)10 were obtained. ○=BH, ○=BOH.

Homogeneous immunoconjugates for boron neutron-capture therapy: design, synthesis, and preliminary characterization.

The application of immunoprotein-based targeting strategies to the boron neutron-capture therapy of cancer poses an exceptional challenge, because viable boron neutron-capture therapy by this method will require the efficient delivery of 10(3) boron-10 atoms by each antigen-binding protein. Our recent investigations in this area have been focused on the development of efficient methods for the assembly of homogeneous immunoprotein conjugates containing the requisite boron load. In this regard, engineered immunoproteins fitted with unique, exposed cysteine residues provide attractive vehicles for site-specific modification. Additionally, homogeneous oligomeric boron-rich phosphodiesters (oligophosphates) have been identified as promising conjugation reagents. The coupling of two such boron-rich oligophosphates to sulfhydryls introduced to the CH2 domain of a chimeric IgG3 has been demonstrated. The resulting boron-rich immunoconjugates are formed efficiently, are readily purified, and have promising in vitro and in vivo characteristics. Encouragingly, these studies showed subtle differences in the properties of the conjugates derived from the two oligophosphate molecules studied, providing a basis for the application of rational design to future work. Such subtle details would not have been as readily discernible in heterogeneous conjugates, thus validating the rigorous experimental design employed here.

New horizons for therapy based on the boron neutron capture reaction.

Boron neutron capture therapy (BNCT) is currently undergoing clinical trials in the USA, Japan and The Netherlands with patients afflicted with deadly brain cancer (glioblastoma multiforme) or melanoma. This therapy relies on a binary process in which the capture of a slow neutron by a 10B nucleus leads to an energetic nuclear fission reaction, with the formation of 7Li3+ and 4He2+ and accompanied by about 2.4 MeV of energy. The fleeting 7Li3+ and 4He2+ travel a distance of only about the diameter of one cell, and they are deadly to any cell in which they have been produced. Research in progress is concerned with the development of advanced boron agents and neutron sources, other than nuclear reactors, for the treatment of a variety of cancer types using novel 10B delivery methods. Non-malignant diseases such as rheumatoid arthritis offer additional opportunities for BNCT. The entire BNCT area awaits commercialization.

Model studies directed toward the application of boron neutron capture therapy to rheumatoid arthritis: boron delivery by liposomes in rat collagen-induced arthritis.

The application of boron neutron capture therapy to rheumatoid arthritis requires the selective delivery of the boron-10 isotope to the synovitic tissue. The use of liposomes as a boron delivery method has been explored through the measurement of the time course biodistribution of boron in rats with collagen-induced arthritis (CIA). Small unilamellar vesicles were composed of a 1:1 mixture of distearoylphosphatidylcholine and cholesterol, incorporated K[nido-7-CH3(CH2)15-7,8-C2B9H11] as an addend in the lipid bilayer and encapsulated Na3[a2-B20H17NH2CH2CH2NH2] in the aqueous core. The tissue concentration of boron delivered by liposomes was determined by inductively coupled plasma-atomic emission spectroscopy after intravenous injection of liposome suspensions into Louvain rats with CIA. With the low injected doses of boron used [13-18 mg of boron per kg (body weight)], the peak boron concentration observed in arthritic synovium was 29 microg of boron per g of tissue. The highest synovium/blood boron ratio observed was 3.0, when the synovial boron concentration was 22 microg of boron per g of tissue. In an attempt to increase the synovium/blood boron ratio by lowering the blood boron concentration, a liposomal formulation characterized by a shorter blood clearance time was examined. Thus, the biodistribution of liposomes with additional K[nido-7-CH3(CH2)15-7, 8-C2B9H11] incorporated in the vesicle membrane not only demonstrated more rapid blood clearance and slightly higher synovium/blood boron ratios but also exhibited reduced boron uptake in synovial tissue. These studies with boron neutron capture therapy for CIA suggest that this form of therapy may be feasible in the treatment of rheumatoid arthritis.

Liposomes as drug delivery vehicles for boron agents.

The successful treatment of cancer by boron neutron capture therapy (BNCT) requires the selective concentration of boron-10 within malignant tumors. The potential of liposomes to deliver boron-rich compounds to tumors has been assessed by examination of the biodistribution of boron delivered by liposomes in tumor-bearing mice. Small unilamellar vesicles have been found to stably encapsulate high concentrations of water-soluble ionic boron compounds. Alternatively, lipophilic boron-containing species have been embedded within the phospholipid bilayer of liposomes, and both hydrophilic and lipophilic boron compounds have been incorporated within the same liposome formulation. The biodistribution of boron was determined at several time points over 48 hr after i.v. injection of liposomal suspensions in BALB/c mice bearing EMT6 tumors. The tumor-selective delivery of boron by the liposomes was demonstrated as tumor-boron concentrations increased for several hours post-injection. Even at the low injected doses employed (6-18 mg boron/kg body weight) therapeutic tumor boron concentrations were observed (> 30 micrograms boron/g tissue) and high tumor/blood ratios were achieved (> 5). The most favorable results were obtained with the polyhedral borane Na3[a2-B20H1-NH2CH2CH2NH2]. Liposomes encapsulating this species produced a tumor boron concentration of 45 micrograms/g tissue at 30 hr post-injection, at which time the tumor/blood boron ratio was 9.3.