Kinetic model of osmosis through semipermeable and solute‐permeable membranes

The gas analogy of the van't Hoff equation for osmotic pressure Δπ = RT/, where R is gas constant, T absolute temperature and mole volume of water, remained unexplained for a century because of a few misconceptions: (1) Use of supported membranes prevented the recognition that osmotic forces exert no effect on the solid membrane. During osmotic flow frictional force of solvent within membrane channels equals osmotic kinetic force π at the interface against the solution containing impermeant solute. (2) Retrograde diffusion of water is much less than osmotic flow even when dx in the gradient dc/dx approaches zero. (3) The gas analogy was thought to be accidental. Actually, the internal kinetic pressure is P = RT/, because intermolecular forces cancel out at the liquid interface, just as within a gas. The kinetic osmotic pressure is the difference in solvent pressure across the interface: π = RT/–(RT/)X₁ = (RT/)X₂, where X₁ and X₂ are the mole fractions of water and impermeant solute, respectively. Integration gives π = –(RT/)lnX₁, identical to the thermodynamic equation. This equation is correct up to 25 atmospheres, and up to 180 atmospheres by assuming that a sucrose molecule binds 4 and a glycerol molecule 2.5 water molecules. For solute‐permeable membranes, the reflection coefficient σ can be calculated by formulas proposed for ultrafiltration. Because the fraction (1–σ) of solute concentration behaves as solvent, osmosis may well proceed against the chemical potential gradient for water. The analogy to an ideal gas applies because π = –(RT/)lnX₁ is the small difference between enormous internal solvent pressures.     

Prevention of ischemic acute renal failure with impermeant solutes

Studies were performed to evaluate the effect of solute permeability and extracellular osmolality in protecting against ischemic acute renal failure. The functional protective effect of a 1-min intrarenal perfusion (prior to intrarenal norepinephrine 0.75 micrograms . kg-1 . min-1) of a hypertonic permeant solute (hypertonic saline, 1,400 mosmol/kg H2O) and an isotonic impermeant solute (isotonic mannitol, 280 mosmol/kg H2O or isotonic polyethylene glycol, IPEG, 300 mosmol/kg H2O) was evaluated. Three hours after ischemia, the glomerular filtration rate was significantly lower in hypertonic saline group vs. either the isotonic mannitol- or IPEG-treated animals (2.4 vs. 8.9 and 10.4 ml/min, respectively; both P less than 0.05). Mean renal blood flow was similar in all three groups. The effects of hypertonic saline and IPEG on glomerular filtration pressure and tubular obstruction were also evaluated. Stop-flow pressure, as an index of glomerular filtration pressure, was higher in the IPEG- vs. the hypertonic saline-treated animals (40 vs. 35 mmHg, P less than 0.001). Although proximal tubular pressure was increased in both groups, transglomerular hydrostatic pressure was higher in the IPEG vs. the hypertonic saline group (13 vs. 6 mmHg, P less than 0.01). Tubular microperfusion studies demonstrated increases in proximal tubular pressure in the hypertonic saline but not the IPEG studies. The present results indicate that isotonic, impermeant solutes provide functional protection against ischemic acute renal failure. The beneficial effect of impermeant solute is mediated, at least in part, by better maintenance of transglomerular hydrostatic pressure and prevention of secondary tubular obstruction.     

Osmotic reflextion coefficients of capillary walls to low molecular weight hydrophilic solutes measured in single perfused capillaries of the frog mesentery.

1. Individual capillaries of the transilluminated frog mesentery have been perfused with suspensions of human red cells in frog Ringer solution containing 1-0 g albumin 100 ml.-1. The outer surface of the mesentery has been washed with normal frog Ringer solution and with frog Ringer solutions made hypertonic by addition of one of the following solutes: sodium chloride (100 m-mole. 1.-1); urea (100 m-mole.1.-1); sucrose (20-50 m-mole. 1.-1); cyanocobalamin (8-5 m-mole. 1.-1). The temperature of the mesentery was between 14 and 16 degrees C in all experiments. 2. Wtih the mesentery superfused with normal Ringer, the filtration coefficient was determined from measurements of the rate of fluid filtration across the capillary wall, at a series of known capillary pressures (Michel, Mason, Curry & Tooke, 1974). Filtration coefficient varied from 0-69 X 10(-3) to 4-45 X 10(-3) mum. sec-1 .cm H2O-1 with an average value of 1-87 X 10(-3) mum. sec-1. cm H2O-1. 3. When the superfusate was made hypertonic by the addition of a test solute, the osmotic reflextion coefficient (sigma) of the capillary wall to test solute was calculated from the additional rate of filtration, the concentration of test solute in the superfusate and the filtration coefficient. Average values for sigma were: sodium chloride, 0-068 +/- 0-03 (three capillaries); urea, 0-071 +/- 0.015 (four capillaries); sucrose, 0-115 +/- 0-023 (seven capillaries); cyanocobalamin, 0-100 +/- 0-03 (three capillaries). 4. In further experiments, the osmotic reflextion coefficients to sodium chloride, urea and sucrose were determined in the same capillary. Five technically acceptable experiments were carried out. Although there were differences in the value of sigma between different capillaries, in any one capillary values of sigma were of the same magnitude and there appeared to be no significant trend with the molecular size of the test solute. 5. Our findings are inconsistent with the hypothesis that there is a single pathway for water and small hydrophilic molecules across the capillary wall. 6. Our results may be interpreted in terms of an exclusive channel for water in parallel with a channel shared by both water and small hydrophilic molecules. It is suggested that the exclusive water channel may be the membranes and cytoplasm of the endothelial cells and the shared channel may be located in the intercellular junctions. 7. Our data suggest the exclusive water channel represents about 10% of the total filtration coefficient in frog mesenteric capillaries. The shared channel shows relatively little restriction to the molecules investigated. Estimates of the volume flow throught the two channels are made for conditions where hydrostatic pressure differences and osmotic pressure differences are the driving forces.     

Influence of perfusate oncotic pressure on the transcapillary clearance of albumin in maximally vasodilated rat skeletal muscle

An external detection technique was developed for repetitive and reliable measurements of clearance (Cl) of 99mTc-albumin during alterations of serum colloid osmotic pressure (pi p). Isolated, maximally vasodilated rat hindquarters were perfused with serum of two pi ps (20 mmHg and 45 mmHg) at four or five different filtration rates (Fv) for each pi p in each animal. The recorded accumulation rates of 99mTc-albumin (AR) were converted into dimensions of albumin clearance, setting the isogravimetric Cl at normal vascular pressures (pi p = 20 mmHg) at 0.0246 +/- 0.0012 ml min-1 per 100 g, which was obtained in defined muscle samples in 11 separate experiments. Serum perfusion with higher colloid osmotic pressure (45 mmHg) shifted the albumin clearance values upwards, without affecting the slope of the Cl vs. Fv relationship. Thus, the reflection coefficient (sigma) for albumin did not seem to be affected by the changes in pi p, while the isogravimetric albumin clearance was increased to roughly 0.058 ml min-1 per 100 g. Explicit two-pore equations were found to describe the experimental data fairly well, yielding an average sigma for albumin of 0.92 and a minor contribution of diffusion to overall transport even at low Fvs. Moreover, a coupling of macromolecular clearance to pi p may serve to minimize alterations in plasma protein concentration.     

Water permeability and pathways in the proximal tubule

The route of water transport in the proximal tubule could be either transjunctional or transcellular. A transjunctional route is supported by data showing high osmotic-to-diffusive water permeability ratios, the possible correlation of junctional leakiness to ions and nonelectrolytes with water permeability, and solvent drag of nonelectrolytes and ions. These data, however, are not convincing. A transcellular route of water transport is supported by data showing that the osmotic water permeability (Pf) for apical and/or basolateral cell membranes is sufficiently high to account for the transepithelial Pf, making a tentative conclusion for a transcellular route of water transport possible. In addition, measurements of Pf have yielded insights into the mechanism of solute-solvent coupling. Pf has been reported to be mostly between 0.1 and 0.3 cm/s. In the rabbit proximal straight and the Necturus proximal convoluted tubule, in which water transport rates are low, this range of Pf will account for volume absorption with only small osmotic gradients (less than 6 mosmol). Higher osmotic gradients are required in the rat and possibly the rabbit proximal convoluted tubule, where water transport rates are higher. Solute-solvent coupling in all species is probably due to both luminal hypotonicity and lateral intercellular space hypertonicity. These two processes are directly linked. Mass balance requires that generation of luminal hypotonicity also generates a hypertonic absorbate and, thus, some degree of lateral intercellular space hypertonicity. It is likely that, in the rabbit at least, effective osmotic pressure gradients due to differences in solute reflection coefficients play little role in solute-solvent coupling.     

Relationship between the transfer rate and the gap in partial pressure of gas molecules at a heterogeneous interface

The kinetic energy of a gas molecule dissolved in an aqueous solution in the stationary state is inversely proportional to its solubility according to the distribution law of kinetic energy between the gas and liquid phases. From the difference in solubility, it is deduced that the gas molecule undergoes a change in energy in the passage across a heterogeneous interface, which is countered by a difference in energy level, or a gap in partial pressure, between the two sides. In order words, unless there is the gap in partial pressure, which has hitherto been disregarded in solving the diffusion equation, it is impossible to equalize the efflux of gas molecules with the influx. The relationship between the gap in partial pressure and the transfer rate is theoretically derived from the difference in solubility. The transfer rate is evaluated by multiplying the gap size by the transfer coefficient n, which in turn is expressed by an equation containing the unique value, k, for conductivity across the interface. Comparing the theoretical and experimental transfer coefficients obtained across the gas-liquid interface, the k value at which a gas traverses the liquid side per unit difference in partial pressure is estimated to be 6.58 x 10(-3) cm.s-1.Atm-1. Gas solubility in the RBC membrane is estimated to be ca. 22% of that in water, regardless of gas type, for O2, CO, and CO2. The increase in transfer coefficient across the RBC membrane by convection is explained by an additional change in kinetic energy in extracellular fluid attributable to an external force. The reduction in the transfer coefficient in the presence of increased protein is ascribed to an increase in kinetic energy of the membrane at the RBC boundary. The solubility in the membrane is linearly related to added protein concentration.     

Forces involved in transcapillary fluid movement in exercising cat skeletal muscle

Average capillary pressure (Pc) close to the venous end (fluid equilibrium point) of the exchange vessels (denoted Pc,v), arterial (PA) and venous pressure, and the rate of net transcapillary fluid flux were continuously recorded in sympathectomized muscle during 30 min of graded exercise and for 30 min in the post-exercise period. Regional changes in colloid osmotic pressure (pi pl) and total osmolality in plasma, the latter reflecting work-induced interstitial hyperosmolality, were measured at intervals. In the control state at rest with a Starling fluid equilibrium, Pc,v averaged 17.6 +/- 0.8 mmHg. Exercise caused a rapid transcapillary plasma fluid loss, the net driving pressure for which in the initial phase of heavy work was 58 mmHg (transcapillary fluid flux divided by the capillary filtration coefficient). This comprised an increase in Pc,v of 16 mmHg, a nonprotein osmotic force (Posm) related to exercise-induced tissue hyperosmolality corresponding to 46 mmHg and an opposing force established by a raised pi pl of 4 mmHg. A theoretical analysis indicated that the main fraction of the osmotic fluid loss passed through transcellular ultrapores and only a minor part through conventional small pores. In spite of the fact that Pc remained high throughout the exercise period, the outward fluid flux gradually declined and a Starling equilibrium was re-established 23 min after the commencement of heavy exercise. This was explained by a gradual decline of Posm and apparently also by a secondary increase in tissue pressure (Pif) and/or a decrease in interstitial colloid osmotic pressure (pi if). Net fluid absorption occurred in the post-exercise period as a result of a gradual decrease in Pc, reversed transcapillary Posm and also maintained high Pif and/or low pi if. Exercise (even light) abolished normal Pc autoregulation, implying that the filtration component of net transcapillary fluid flux becomes distinctly modulated if PA is altered.     

Flow-dependent water permeability of the rabbit descending limb of Henle's loop

Because completely opposite results have been reported on the water permeability of the rabbit descending limbs of Henle's loop (DLH), we rigorously examined water permeability of the upper portion of the descending limb of the rabbit long-looped nephron. Even when the double-cannulation method was used in an attempt to reduce the resistance of tubular outflow, the collected fluid-to-perfusate inulin ratio was equal to or very close to the bathing fluid-to-perfusate osmolality ratio, indicating that osmotic equilibration occurred along the tubule by absorption of water. When perfusion rates were controlled by varying the height of the fluid reservoir connected to the perfusion pipette, osmotic (Pf) as well as diffusional (Pdw) water permeability was shown to be correlated with perfusion rate and/or perfusion pressure. Pf and Pdw at zero perfusion rate as determined from the values of the intercept of regression lines were 253 X 10(-3) and 4.54 X 10(-3) cm X s-1, respectively. The maximal values for Pf and Pdw were 737-1,098 X 10(-3) and 18.3 X 10(-3) cm X s-1, respectively. By changing the resistance to perfusion at the tubular outflow, it was shown that changes in Pf paralleled changes in perfusion rate rather than changes in perfusion pressure. Under stop-flow conditions the luminal fluid volume rapidly decreased after the osmolality of the bathing fluid was increased, suggesting that the segment is highly permeable to water even at zero flow rate. Reflection coefficients for urea and NaCl were 1.01 and 0.82, respectively. These data support the view that this segment is highly permeable to water and that increases in osmolality along the DLH in vivo may be accounted for mainly by abstraction of water rather than addition of solutes.     

Osmotic Pressures for Binary Solutions of Non-electrolytes

A comprehensive new theory of osmotic pressure is mechanistically derived to be applicable to incompressible binary solutions of non-electrolytes independent of solute concentrations, membrane selectivity, solution ideality, and without recourse to adjustable parameters. The approach employed relies on Ficks diffusion laws, and introduces the concept of uncoupled aggregate diffusion. The theory is verified using direct and indirect experimental data of sucrose solutions and then compared to conventional osmotic pressure laws. Direct consequences of the new theory entail novel a priori expressions for the activity, activity coefficient, and chemical potential of solvent. Applications to drug delivery and biomolecular separation nanotechnology are discussed.     

Saturation curve in gases of high atomic number at pressures up to 8 atm. I. Krypton and xenon.

The saturation curve has been studied in xenon and in krypton up to a pressure of 8 atm. An empirical formula has been found that describes the fraction of current collected over a wide range of voltages, pressures, ionization intensities, and electrode spacings. This is of practical value in the design of ionography chambers. For krypton the collection fraction fKr = (1 + 0.25eta-1.74)-1, and for xenon fXe = (1 + 0.16eta-1.88)-1, where eta = Fp-0.7Vd-2q-1/2 with F = 3.61 X 10(-7) and 2.50 X 10(-7) for krypton and xenon, respectively. The ranges of the variables covered in the experiments were p = 1-8 atm, V = 5-25000 V, d = 0.3-1.3 cm, and q = 4 X 10(-9)-6 X 10(-8) A/cm3.     

Influence of variations in hydration and in solute excretion on the effects of lysine-vasopressin infusion on urinary and renal tissue composition in the conscious rat

1. The changes in urinary and renal tissue composition induced by continuous, intravenous infusion of lysine-vasopressin (60 mu-u./min. 100 g body wt. until steady-state conditions prevailed) in normally hydrated, hydropaenic, saline-loaded (0.9%, w/v) and mannitol-loaded (15%, w/v) rats were determined and compared with those induced in water-loaded rats.2. Previous reports that the urinary responses to antidiuretic hormones vary both with hydration status and with concurrent solute excretion rate were confirmed.3. The data show that variations in urinary responses were accompanied by differences in the papillary responses to lysine-vasopressin.4. The results are discussed in terms of the effects of hydration and concurrent solute excretion on factors influencing (a) medullary accumulation of water and solute, (b) osmotic water reabsorption and (c) osmotic equilibration across the collecting duct; and of the effects of lysine-vasopressin on these factors.5. It is concluded that the effects of hydration and solute excretion on the antidiuretic responses to lysine-vasopressin may be interpreted by differences in (a) the medullary composition prevailing at the start and (b) any further changes in medullary composition that can be induced under the experimental circumstances.     

Cellular and extracellular dehydration, and angiotensin as stimuli to drinking in the common iguana Iguana iguana.

1. After water deprivation, the iguana promptly drank slightly more than enough water to restore the body fluids to isotonicity even under conditions of hypervolaemia. 2. In response to systemic injections of hypertonic solutions of NaCl and sucrose, the iguana drank and retained enough water to dilute the injected load to isotonicity irrespective of whether water was offered immediately or after 3 hr, and irrespective of whether the solute was administered I.V. or I.P. 3. Hypertonic solutions to glucose, urea, sorbitol and KCl caused little drinking. 4. The long latencies to drinking after hypertonic loads, which were not dependent on the nature of the solute, the route of administration or the dosage, were shown not to be a result of slow distribution of the solute throughout the body fluids. 5. Clearance of injected solutes via renal and extra-renal (nasal salt gland) routes was negligible during the 6 hr experimental period. 6. Measurements of plasma [Na], haematocrit, osmotic pressure and inulin space showed that the iguana drank, in response to cellular dehydration, enough water to restore the intracellular fluid volume to normal. 7. We conclude that, in response to substances which dehydrate cells, the iguana regulates its body osmolality precisely and efficiently provided it is able to do so by drinking. In this respect the responses of the iguana are similar to those of the nephrectomized rat since, in the short term, both rely exclusively on drinking to restore cellular water to normal. 8. The iguana also drinks in response to extracellular dehydration produced by hyperoncotic peritoneal dialysis, and in response to I.P. angiotensin II.     

Canalicular bile salt-independent bile formation: concepts and clues from electrolyte transport in rat liver

Studies on canalicular electrolyte transport are reviewed with reference to the concept that hepatocellular inorganic ion secretion may provide an osmotic drive for canalicular water flow. Cellular transport of electrolytes and of some nonelectrolytes appears directly or indirectly (cotransport or potential-sensitive transport) related to the activity of Na+-K+-ATPase of the sinusoidal cell membrane, but the role of the enzyme in regulating bile flow remains undetermined. Bile secretion of the isolated rat liver continues in the absence of either Na+, K+, Cl-, or HCO-3 when these ions are replaced in the perfusion medium by other permanent ions. Transepithelial salt concentration gradients, established experimentally, cause transient changes of bile flow and dissipate very quickly. Isotopic ion equilibration between sinusoids and bile proceeds faster than between sinusoids and liver cells. Both observations indicate extensive electrolyte diffusion through a paracellular shunt pathway. This pathway appears preferentially permeable to cations, and it restricts permeation of molecules of the size of sucrose (no apparent diffusion or effects of solvent drag) or bile acids (no backleak). In promoting canalicular osmotic water flow, transepithelial concentration gradients of NaCl are less effective than those of sucrose, revealing a reflection coefficient of NaCl of 0.3. By perfusion with hypertonic medium containing sucrose, bile flow is reduced. Bile production against this opposing osmotic gradient is accomplished by an increase in biliary organic anion concentration. Inorganic ion concentrations essentially conform to a Gibbs-Donnan distribution across the canalicular epithelium, established by the presence of impermeant anions in bile. Hence, the luminal electrical potential is expected to be negative with respect to the sinusoids. It is concluded that biliary secretion of endogenous organic anions is the major osmotic driving force for canalicular bile salt-independent bile flow and that transport of inorganic ions into bile results mainly from diffusion and solvent drag.     

Osmotic properties of the chromogranins and relation to osmotic pressure in catecholamine storage granules

The soluble proteins (chromogranins) of bovine chromaffin granules have been studied by micro-osmometry with semi-permeable membranes (UM2, PM10 and PM30 with cut-offs greater than 1, greater than 10 and greater than 30 kD, respectively) at 1 = 0.15 and pH 5-8 for protein concentrations up to 20 mg X ml-1. After lysis of chromaffin granules in phosphate buffer pH 6, the released chromogranins behaved as aggregating solutes, consistent with an inconspicuous osmotic pressure contribution from the chromogranins at the protein concentration of the intact granules. Thus, in the presence of phosphate about 90% of the molecules behaved as colloids with Mr = 30,300 at c = o. After lysis in phosphate-free buffers the chromogranins behaved as highly non-ideal solutes in a manner which was incompatible with isotonicity at the protein concentration of the intact granules. About two-thirds of the molecules in the lysates in Na-succinate pH 5-6 and K-acetate pH 6 exhibited Mr = 66,000 and 79,000, respectively. In dilute solutions (less than 12 mg protein X ml-1) and ATP/protein ratios corresponding to those in the intact granules, the UM2 pressures were markedly increased, indicating release of polypeptides with Mr 2000-3000 from aggregates. CaCl2 was without specific effect on the colloid osmotic pressures but reduced the ATP-dependent increase in pressure, suggesting release of molecules twice the size of those released by ATP alone. A model is presented for the contribution of the chromogranins to osmotic pressure regulation in the bovine adrenomedullary catecholamine-storing granules.     

Effect of acute lung injury on structure and function of pulmonary surfactant films

The structural and functional alterations in pulmonary surfactant that occur during acute lung injury were studied using rat lung surfactant large aggregates (LA) isolated from normal nonventilated lungs (N), and from standard ventilated (V) and injuriously ventilated (IV) excised lungs. N lungs inflated significantly better than IV lungs, with V lungs intermediate. Although IV LA phosphatidylcholine levels were unchanged, cholesterol and protein were elevated. V LA exhibited PC/cholesterol and PC/protein ratios intermediate between N and IV. In contrast to total cholesterol and protein levels, these ratios were not significantly different from IV LA. N and V LA, but not IV LA, adsorbed rapidly and were able to generate surface pressures (pi) near 70 mN/m during surface area reduction at 37 degrees C on a captive bubble tensiometer. Langmuir-Wilhelmy surface balance studies at 23 degrees C showed N LA films consistently attained pi approaching 70 mN/m during ten compression-expansion cycles. IV films were less effective and failed to achieve high pi consistently after the sixth cycle. V films were intermediate. Epifluorescence studies revealed compression of adsorbed N LA films formed well-defined liquid-condensed (LC) domains, but fewer, smaller domains were observed with IV films and, to a lesser extent, V films. Atomic force microscopy on Langmuir-Blodgett N films transferred at pi = 30 mN/m showed high, well-defined LC domains. IV films showed thinner, intermediate height, possibly fluid domains, which contain large numbers of small higher domains with heights corresponding to LC domains. V films were intermediate. We conclude that acute lung injury induced by hyperventilation, and to a lesser extent standard ventilation, of excised lungs alters surfactant surface activity and the ability of natural surfactant to form surface structures at the air-water interface.     

Osmometry with membrane-permeating solutes of low molecular weight

Experimental osmotic pressure (p)-time (t) curves for several low-molecular weight Solutes in toluene were studied by means of a recording automatic osmometer. The results for a monodisperse solute can be expressed by p = r π exp (?αt), with π the theoretical osmotic pressure. In the case of more than one solute the pressures are additive. for a given set of experimental conditions, the parameters r and α depend on the molecular weight of the solute, but to some extent also on its structure. It is shown how an approximate estimate of molecular weight and concentration of a polymer additive can be made from p versus t plots in the course of a molecular weight determination of the polymer. The results could not be brought in agreement with the equation of HOFFMANN and UNBEHEND . Also, the assumptions underlying the theory of LAIDLER and SHULER do not seem to be adequate in the general case to relate α to r.     

Permeability and model testing of heart capillaries by osmotic and optical methods

Rabbits were anesthetized with Nembutal; their hearts were removed and perfused with Ringers solution. Osmotic weight transients were then produced by test solute infusion into the perfusate. The rate constant for organ weight change was used to predict test solute venous concentration (CV) and the distribution volume (V) during an osmotic transient by use of the Johnson and Wilson (6) model for capillary tissue exchange. By use of Cr EDTA, strongly purple in solution, we compared the above predictions with a value of CV obtained directly by optical measurement of outflow venous concentration and values of V calculated from our measurements of organ weight and the known sucrose distribution volume. The close agreement between both sets of values with a permeability coefficient of 5 x 10(-5) cm/s and a volume of distribution of 3.2 ml/10 g heart lead us to conclude that the model used closly represents the conditions of the isolated perfused heart, and that both the osmotic transient and the extraction measurements provide good estimates of organ capillary permeability.     

Actions of glucose and potassium chloride on osmoreceptors slowing gastric emptying

1. Five subjects were given 373 test meals of 750 ml. water containing a range of concentrations of glucose or potassium chloride.2. The greater the concentration of solute in the meals, the greater was the volume of the test meal recovered from the stomach after a fixed time.3. When the concentrations of the solutes were expressed as m-osmole/l. corrected by osmotic coefficients based on vapour pressures at 37 degrees C, glucose and potassium chloride were indistinguishable in slowing gastric emptying.4. These results are consistent with the regulation of gastric emptying by a duodenal receptor responding to osmotic pressure.5. Potassium chloride was more nauseating than glucose on an osmolar basis.