US2444845A

US2444845A - ph-responsive glass electrode

Info

Publication number: US2444845A
Application number: US668393A
Authority: US
Inventors: George A Perley
Original assignee: Leeds and Northrup Co
Current assignee: Leeds and Northrup Co
Priority date: 1946-05-09
Filing date: 1946-05-09
Publication date: 1948-07-06
Anticipated expiration: 1965-07-06
Also published as: GB627442A

Description

July 6, 1948. G. A. PERLEY 2,444,845

PHRESPONSIVE GLASS ELECTRODE Filed lay 9, 1946 '7 Sheets-Sheet 1 4 1.a, .sf F2 Z f -2 INVENTOR.

WMODM# July 6, 1948. G. A. PERU-:Y 2,444,845

PH-RsPoHSIvE GLASS ELECTRODE Filed lay 9. 1946 7 Sheets-Sheet 2 /lf j {awww/fa) F/' /fawm/a) 6l ,4.515 /VPS/T/O/V Mal Pff cfvr Maga? Mm Amm/wf,

INVEN TOR.

@71 wd? me@ July 6 1948- G. A. PERLEY 2,444,845

Pri-RESPONSIVE GLASS ELEc'rRoDB Filed llay 9. 1946 7 Sheets-Sheet 3 7lf Z /75 Z ams.: onpas/77a Vs. 45cm/mz /PfflA/vrf IN VUV TOR.

July 6, 1948. G. A. PERLEY 2,444,845

Pfl-RESPONSIVE GLASS ELECTRODE Filed lay 9. 1946 7 Sheets-Sheet 4 INVENTOR. BY afd web Afm/WY July 6, 1948. G. A. PERLEY 2,444,845

PHRESPONSIVE GLASS ELECTRODE Filed lay 9, 1946 7 Sheets-Sheet 5 (ONT/NUMS' OPERATION /Al' C777 H4727? A7' 70 92C IN V EN TOR.

BY MP July 6, 1948. G. A. PERLEY 2,444,845

P11-RESPONSIVE GLASS ELECTRODE Filed lay 9. 1946 7 Sheets-Sheet 8 M525 6' F76 6 /fvfzafA/cf of mf P/PfsfA/cf of mms/v a/H595 INVENTOR.

A' TMF/Vir July 6, 194s. G. A. PERM-:Y 2,444,845

PH-RESPGNSIVE GLASS ELECTRODE Filed May 9, 1946 7 Sheets-Sheet '7 Aff/9465 IN VEN TOR.

Patenied July 6, i948 UNITED STATES PATENT OFFICE andN pany, Philadelphia, Fa., a

corporation of Pennsylvania Application May 9, i946, Serial No. 808,308

Claims. (Cl. 20d-195) My invention relates to pli-responsive glass electrodes, utilizable for example in determination of the pH, representative of the magnitudes of concentration of hydrogen-lons or hydroxylions and of acidity and alkalinity; more particularly my invention relates to the compositions, for the membranes of aforesaid electrodes, of the characteristics disclosed in my original application Serial No. 467.060, filed November 28. 1942` (now abandoned) of which this application is in part a continuation, and of the charatceristics hereinafter disclosed.

In general. glass electrodes of compositions known and used prior to my invention have been subject to large sodium error" whose magnitude was dependent upon the temperature and the pH of the solution, and upon the effect on the electrode glass by the sodium content of that solution, commonly known as the test solution: none of the aforesaid prior electrodes was suited for measurement of pH over long periods of time because of occurrence of changes in calibration, and impairment of speed of electrode response to changes in pH of the test solution; their sodium error, in terms of error in the pH measured by them, was great: none was suited for measurements of pH throughout the entire or substantially entire range from pH to 14 pH; in general they were unsuitable for measurement of pH of solutions of alkalinity greater than about 9.6 pH; and none of them was suitable for continuous measurement of pH oi solutions whose temperatures substantially exceeded 50 C.

In accordance with my invention operative and practical qualities oi' glass electrodes have been greatly improved: in respect to reduction of sodium error. from greater than 1 pH to about .5 pH and as low as .25 pH to about .10 pH, while still exhibiting very high qualities in respect to electrical volume-resistance of the glass membrane of the electrode: in respect to high stability or permanence in qualities and behavior over long periods of continuous use; in respect to measurement of pH of solutions having temperatures ranging from 0 C. to about 100 C.: in respect to procurement of consistently reproducible voltage/pH (potential/pH) characteristic which is rectilinear throughout substantially the entire range of pH measurements. from 0 to 14 pH, with a slope represented. for example, by about 60 millivolts per one pH: and in respect to desirable aspects and qualities herein more particularly referred to.

More particularly, my invention is characterized by inclusion in the glass membrane compositions. in addition to the usual components BiO-MO-XOz. of an oxygen modifier. OM, (one or more of the numerous oxides, including the rare earth metal oxides, hereinafter referred to) having the enect of increasing the ratio of oxygen to silicon in the glass, so securing or maintaining high the magnitude of that ratio when the mol per cent of SiO: is low, e. g., 59 to 66 per cent, thereby reducing sodium error and in addition greatly increasing stability and permanence of behavior of the electrode for long periods of time at normal and elevated temperatures. An outstanding example of oxygen modifier is one of the rare earth metal oxides, lanthanum oxide, LanOa. The oxygen modifier may be the oxide or a mixture of oxides of one or more of the hereinafter specified metals or elements, including rare earth metals. The mol percentage of oiu'gen modifier is in general small, for example from about 0.5 to about 6 mol per cent of the entire composition, as compared with the relatively greater mol per cent of the associated XO: component, such as SiO: which preferably constitutes the relatively low proportion of from about 59 to about 66 mol per cent of the entire glass composition; the mol percentages of the oxygen modiiier and of the associated S102 are complemental. in the sense that when the mol percentage of one is changed. within the approximate limits aforesaid, the mol percentage of the other, within its approximate limits aforesaid. changes in opposite sense and by approximately equal amounts.

My invention consists of a pli-responsive electrode of silicate glass yielding marked practical improvements and whose composition is represented by a prefuslon mixture of silica with alkali metal oxides and alkaline earth metal oxides, and having in conjunction therewith. to provide high ratio of oxygen to silicon, one or more of the oxides hereinafter for brevity designated as oxygen modiers.

My invention resides in the glass electrodes and compositions therefor hereinafter described and claimed.

For examples of my compositions and their characteristics, and for tables illustrative of various aspects and characteristics of my invention. reference is had to the accompanying drawings, in which:

Fig. 1, Table i, aii'ords numerous examples. including some from aforesaid original application, of results of extensive research respecting compositions and their operative properties of electrical volumetric-resistances and sodium errors in mag- 3 nitudes of pH, after periods of use for various lengths o! time;

Fig. 2. Table 2. is illustrative of eilects oi compositions upon aforesaid electrical resistance of the glass electrode;

lig. 3, Table 3, is illustrative of the adverse elects of some oi the components of glass compositions .upon'the magnitude of the sodium error of the resultant glass electrodes;

Fig. 4. Table 4, is illustrative of the iniluence or eilect of cesium on the sodium error;

Pig. 5, Table 5, is illustrative of the eiIect of the oxygen modifier and SiO: content on the stability of the electrode glass;

Pig. 5A, Table 5A, illustrates behavior of various glass electrodes in continuous operation at elevated temperature:

Mg. 6, Table 6. is illustrative of the influence of the presence, in certain oi' my glass compositions. of various oxygen modifiers;

Fig. 7. Table '1. is illustrative of the iniluence, on the sodium error. of the proportion of ianthanum oxide to silica ratio;

`Ii'lg. l, Table 8. is illustrative of the inuence on the electrode glass of the MO type oi' component: and Y '.Fig. 9 is a graph illustrating the voltage/pH characteristic of my glass compositions.

My glass compositions are the result of needto discover. as between components of diii'erent characters. or between individual oxides and their mol percentages or proportions, which shall well or best attain as many as possible, and to the greatest extent oi' each possible, of desirable aspects or characteristics as follows:

(o) The glass electrode should yield a rectilinear voitage/pH characteristic at least closely approximating that afforded by the ideal or perfect hydrogen gas electrode: thus for every change of one pH of the test solution, at 25 C.. the potential of the glass electrode should change 0.95912 volt. Actually a change of 0.003 volt for a change of 0.05 pH (60 miilivolts per one pH) is fairly representative.

(b) The glass electrode should produce the same H-ion to electro-motive-force relationship ior both presence and absence of metallic ions in and from the test solution; thus the presence in the test solution of sodium. potassium. etc., ions. not permissible with glass electrodes known before my invention, should have no iniiuence upon the electro-motive-force/pH relationship.

(c) I'he glass of the electrode should have a long-time stability at temperatures of the range 0 C. to nearly 100 C.; which is to say, the composition oi' the glass should not be altered or affected by the usual test solutions in which it may be immersed, for pH measurement, even at elevated temperatures. .,(d) 'Ihe glass of the electrode should be of easily workable quality, to yield a glass membrane or membranous bulb of relatively robust characteristics. safely to withstand ordinary handling and use. The electrode composition should not be of character subject to devitriiication or opacity which connotes lack oi essential homogeneity.

.(e) The electrical volume-resistance oi' the composition for the glass membrane should be low, because the greater the resistance, the greater the problems of IR drop, leakage. pick-up and various other types of electrical disturbances. A bulb or membrane of the dimensions herein by way oi' example stated, having a volume-resistance substantialiv in excess of 800 megohms at any operating temperature. generally is deemed undesirable. In this connection it is recalled that such an electrode having at 0 C. a volumeresistance of 000 megohms, at 25 C. has a resistance of 45 megohms, and at 50 C. has a resistance of only 4 megohms. Similarly a glass electrode bulb or membrane having at 25 C. a volumeresisianoe ci' 500 megohms has, at'g50 C., a resistance of` 45 Imegolims. and at 90 C.. a resistance of 2 megohms.

' (l) 0f great practical importance is the need. for `practical and Vcommercial use. that the glass electrode continuously for long periods of use shall have sodium error of low magnitude. The sodium error. contemplated also by (c) above, is conveniently expressible in terms of the amount of error in. pH caused by the sodium content oi' the test solution. In the tables. aforesaid Figs. 1-8, sodium errors are expressed in terms of magnitudes of pH error at 25 C. for 2N sodium and at 12.8 to 13.0 pH.

Although the glasses or glass compositions here, ci' interest comprise complex silicates of certain metals, or homogeneous mixtures oi silicates of those metals with silica, their composition is in each case expressed in' terms of oxides, and. to use my conventional notation. each composition` is represented by an aforesaid general formula RzO/OM/MO/XO: in which. for purposes of in my terms dennlng my invention:

R isone or more of the metals lithium (Li), cesium (Cs). rubidium (Rb), potassium (K). and sodium (Na): ,V

M is one or more of the metals calcium (Ca) f strontium (Srhbarium (Ba). and magnesium (Mg) X, except as to small quantities ci' contaminants unavoidably present. is silicon (Si) 0M is one or more of the oxygen modiilers, consisting oi' the oxides and rare earth metal oxides herein referred to and producing the herein ,described beneficial eiIects accruing from increase in the ratio of oxygen to silicon. A In addition to aforesaid components aio-Mo-xoi there preferably is used one or more of aforesaid oxygen modiiiers, which for present purposes may be generically designated as component 0M. in which. for purposes oi defining my invention.- there may be employed: 1

One or more of the oxides, not easily attacked by alkaline solutions of sodium. potassium. or the like. of the following metals or elements. zirconium. ZrOz: titanium. T101; teilurium. TeOs; phosphorus, P205: columbium. CbzO; tantalum.- TazOs; thorium, ThOz; arsenic, AsnOs; uranium. U01: and of one or more of the rare earth metals cerium, CeOz: yttrlum, YzOs; lanthanum, LazOa; praseodymium, PrsOa: neodymlum, NdzOs: samarium. 811120:; and the like oxides of the further rare earth metals europium, .gadolinium, terbium; dysprosium, holmium, erbium, thulium, ytterbium and lutecium.

Substantially to eliminate the sodium error `in the measurement of pH, the composition oi' the glass of the electrode should not be such that any of its constituents is to appreciable extent exchangeable with sodium ions of the test solution; the silicon and oxygen of the silicates should be so chemically bound that there are not present to substantial extent unsatisfied valences productive at the surface of the glass of sodium4 silicate by reaction of sodium ions of the test solution, in which the eletrode glass is immersed.

with one or more constitutents of the glass. This end may be accomplished by selection of ratlos between R20, M0, and especially 0M, and XO: based on the considerations hereinafter discussed.

For best chemical stability of the H2O component, I have found cesium may be used. In order to produce a glass whose electrical volumeresistivity is not excessively high at least a substantial or preponderant part of the R component should comprise lithium. Use of lithium as the sole metal of the R component is satisfactory when it constitutes. in terms of molecular percentage of the composition. not more than about 30%, to avoid undue decrease in durability or useful life of the electrode. The Re() component of my preferred glasses comprises both lithium and cesium in percentages herein indicated.

Rubidium may be used instead of oesium with about equal performance and with substantially similar limitations upon selection of its proportions with respect to other constituents of the glass.

Within narrow percentage limits, cesium may be replaced by potassium with retention of low sodium error characteristic: in general, increasing the KzO content of the R20 component of the glass effects relatively great increase in the electrical volume-resistivity while decreasing it results in impairment of stability of the glass in highly alkaline solutions.

In general, the limits for the oxides of cesiuml rubidium, potassium or sodium are from 0 to about 3 mol per cent of the total of the constituents; and of lithium from about 17 to 32 mol per cent.

The MO component may comprise CaO in combination with MgO (glass IA). Such glasses draw easily and are easily blown into clear thin membranous bulbs.

In the glasses (NB, M0 and MW) found best suited for applications requiring long continued exposure of the electrode to the test solution at elevated temperature, the M0 component of the glass preferably comprises both calcium and barium oxides.

strontium oxide may be substituted for barium oxide with some effect upon the electrical volumeresistance. but even less effect upon other properties. such as low sodium error, stability, and rapidity of response.

Concerning the selection and proportions of the X02 component of glasses of my invention, it is entirely SiOz.

In all cases, the molecular percentage of the X02 component should lie within the range oi' from about 59 to about 66% of the total composition of the glass: or from about 63 to about 65 mol per cent for my preferred glasses exemplified by the types following:

Liz0-Cs2O-BaO-LazOa-Si02, such as glass LizO-CszO-BaO-CaO-LazOa-SliOz, such as glass NB LisO-BaO-LazOa-SiOz, such as glass MT LinO-CszO-SrO-CaO-LazOa-Sion, such as glass LB Referring to Fig. 1: Table 1 comprises numerous compositions comprehended by my invention, and illustrations of influences by different kinds of, and variations of proportions of, individual constituents or combinations thereof.

The first 13 compositions, GR to IC. are exampies of those disclod in my aforesaid original application.

The compositions of Table 1 include those whose stability during continuous use as pH measuring electrodes for long periods, including periods from about six months to about two years, has been exceptionally good.

Of the compositions of Table 1 not disclosed in aforesaid original application, 66 are characterized by the significant content of SiO-z, to wit, less than 68 mol per cent, in conjunction with various oxygen modiers; and 58 of the compositions not disclosed in aforesaid original application, are characterized by various contents of CszO.

Aforesaid 13 compositions of my original application, are characterized in all but one case by SiO: content less than 66 mol per cent; each oi' 6 of aforesaid 13 compositions includes an oxygen modifier; and each of 11 of aforesaid 13 compositions includes cesium.

In Table 1, all of the oxygen modifiers are LaaOa. except as therein otherwise indicated.

The electrical volume-resistance of a pH responsive glass, in other respects suitable or practical for commercial or laboratory use, may pose a serious problem; the final choice of a glass composition is dependent upon or influenced in part by aforesaid resistance of the glass membrane or membranous bulb whose thickness may not, for purposes of reducing its aforesaid volumetric-resistance, be reduced below that magnitude which shall assure the membrane shall be sufficiently robust to withstand ordinary handling and use, particularly when the glass electrode as usual contains a solution.

A pH-responsive glass of low electrical volumeresistance-Table 2, Fig. 2, preferably less than megohms at 25 C.-is of advantage in that it is subject to less voltage error due to electrical leakage, as along the external surface of the stem of the glass electrode, adversely affecting the pH measurements; low resistance is also conducive to lessening of IR drop errors due to too high grid current of an electronic device for measuring the voltage developed by the pH measuring cell comprising the test solution, the measuring electrode of pH-responsive glass and the usual reference electrode; low electrical resistance is conducive also to less error due to electrostatic pick-up or electrical disturbances ambient to the glass electrode or the pH measuring cell.

A high SiO: content (e. g., approximately 66 to 70 mol per cent) tends to procurement of low electrode resistance; but such high SiO: content results in undesirably high or prohibitive sodium error. A high LizO content (e. g., about 29 mol per cent or more) contributes to low electrode resistance; but such LizO content also results in undesirable or prohibitive magnitude of sodium error. The presence in the composition of a BaO constitutent results in higher electrical resistance than does CaO.

The choice of composition, with special regard to the electrical volume-resistance of the glass, is well illustrated by glasses NB and NV. Tables i and 2; thus, other things being the same. the NV glass is preferable to that of the NB type, though the sodium errors of both at room temperatures are substantially the same; but for use with solutions of elevated temperatures, the resistance of glass NB. though high at ordinary temperatures, is, at elevated temperatures, suitably low.

Referring to Fig. 3, Table 3: In general, the NaaO type of glass (such as PB, MA) has a lower resistance than an LizO type (such as PA, LZ) which in turn is of lower electrical volume-resistance than a KsO type (such as PC, MB). In general, however, the sodium errors of NaaO types are high.

Choice between LizO, and NazO, Kao, RbzO or CsaO types of compositions is simplified in part by the fact it is diilicult to procure a high (about 26 mol per cent) RbzO or CszO content, because of the high rates at which the oxide-producing salts of the two elements Rb and Cs decompose at the temperature of the glass making furnace. RbnO and CsaO readily decompose even at low temperature such as about 600 C., and the more rapidly at higher temperatures. The temperature range for fusion of my mixtures of oxide yielding salts varies from about 1100" C, to about 1300 C. The use of LizO in combination with CaO and high SiO: has been suggested (Zeitschrift fur Physikalische Chemie. A-lil, p. 366, 1932; B. I. Bsokolof and 0. H. Passynsky) I have found their glass is dimcuit to manufacture.

The sodium error of a LizO type of composition, consisting of a favorable mixture or combination with LizO of BaO with Lazo: and BiOz, or LizO with BaO, CaO, LaaC; and BiOz. is far lower than for the generally corresponding NaaO or X20 types of composition. Although the electrical resistance aforesaid of an equivalent NazO type of composition is lower than that for the Liao type. it is distinctly more advantageous or in some cases necessary to employ LizO in proper here disclosed proportions in order to secure freedom from large or prohibitive sodium error, i. e., to secure my new low order of sodium error.

Table 3 illustrates the relative merits of the LlzO, the NazO and the KzO types of glasses. It is predictable and observable from Table 3 that a LizO type of composition will have a smaller sodium error than a NazO type of composition, which in turn will have a smaller sodium error than a KzO type of glass. Respecting attack upon the electrode membrane by sodium salts in the solutions whose pH is under measurement, those salts will not materially alter my LizO-SiOz compositions, whereas sodium salts in the test solutions will greatly change or ail'ect a corresponding KzO-SiOz composition with resultant prohibitively high sodium error.

Referring to Fig. 4: Table 4 shows the presence of even small amounts (about l to about 2 mol per cent) of CszO as a constituent oi' the glass electrode composition highly signiilcantly reduces the sodium error, irrespective of whether the MO component of the glass comprises any one or more of the BeO, MgO compositions with either CaO or SrO in place of BaO.

While the influence of CszO (or RbsO) in my compositions may appear small because present in numerically small percentage. its eifect is in fact relatively great in reducing sodium error, herein always tabulated in terms of pH error. Nevertheless my broadly preferable LlzO- CsaO--BaO-LarOs-SiOn type of glasses are important even should aforesaid small additions of CszO be reduced in magnitude or entirely omitted because of the presence of LazOs or equivalent oxygen modifier.

Referring to Fig. 5, Table 5: A highly important result of my research is the discovery of the importance of the use of a relatively low SiO: content. to wit, of about 59 to about 66 mol per cent (which per se is in general significantly lower than ImOa. having the herein described beneficial effects) which is not easily attacked by either acid or alkaline test solutions of from about 0 pH to about 14 pH. The oxygen modider chosen should be of character not easily combining with the BiO: component thereby to form an undesirable stable compound. The effect and purpose of addition oi an oxygen modifier is to supply oxygen to and for the glass network structure. thereby securing high ratio of oxygen to silicon, which is novel. in eiIecting both signincant reduction of sodium error (due to low proportion of SiOz) and significant increase in the stability. permanence of behavior, of the electrode for significantly longer periods of use at normal and particularly (Fig. 5A) at elevated temperatures.

, Fig. 6: Table 6 is illustrative of alternatives for the particular oxygen modiiler Lanos. In this table glass 0H is included for contrast. because it is not itself recommended because of its high volume-resistance and its unworlrability. The oxygen modifiers of glasses OW to PV, foot of Table 6, are included for contrast and general guidance; they are among those oi low orders of suitability or practicabiiity as compared with my more preferable compositions also having oxygen modifier components.

Referring to Fig. 7: Table 'l illustrates the iniluence, on sodium error, of the ratio of the LazOa content to the content of S102.

This table shows, as by glasses 0F, OE and MW, that with the LizO-CszO-CaO-BaO content held constant, or with the LizO-CszO content (glasses MH and NI) held constant, the decrease in mol per cent of SiO-i with a corresponding per cent increase in the oxygen modiiler, such as Lazos. produces a glass having significantly low sodium error.

Referring to Fig. 8: In Table 8 it is shot.: that, provided the Liso content does not exceed 25 mol per cent, it is a matter of indifference so far as concerns sodium error whether the M0 component be calcium. strontium or barium. or mixtures of two or more thereof, so long as there be present an oxygen modiiler, Lazo: or equivalent. and so long as the S102 content is below, or does not materially exceed, 68 mol per cent, even with Lazo: or other oxygen modiner present, the sodium error is practically unadected by choice of the MO component. However with higher LizO content, for example about 28 mol per cent, the use of BaO or Sr() is preferred over CaO.

In Fig. 9 is illustrated the voltage/pH characteristlc. B, of my hereinbefore described glass electrode compositions, for solutions of 2N sodium. The cell employed comprised, besides the glass electrode and the test solution, a reference electrode in a 'l pH solution. It will be noted the characteristic B is rectilinear throughout the range from about 0 to 14 pH; and has a slope. like that of the ideal, C. or perfect hydrogen pH measuring electrode, corresponding with about 60 milllvoits per one pH.

The voltage/pH characteristic of glass electrodes known and. used prior to my invention ir. substantially coincident with the aforesaid char acteristic B from about 0 pH to slightly above 8 pH, from which latter point onward, to about 14 pH. it is a branch A from characteristic B.

B and above i2 pH is substantially parallel to the axis of abscissae.

From the foregoing, it will be noted:

That a fundamental, in some aspects the most important fundamental o! my glass electrodes and compositions therefor, is that the B10: content shall be less than 66 mol per cent, and should have associated therewith a network oxygen modier;

That the oxygen modifiers, singly or in mixture, such as the rare earth oxides LazOa, Pr20s. NdzOa, CeOa, etc., and the oxides P205, TeOs. ZrOz, TiOz. ThOz, TazOr, CbaOs, AszOs give substantially equallygood results. In contrast with the foregoing, oxides W01, B201, A1203, U02, C0203, and V205 yield eifects similar in kind, but are less desirable. In general, oxide modifiers which are attacked by alkaline test solutions are not desirable:

That LinO, as R20 component, is highly important to low sodium error. However, addition of 0.20 or RbrO to the LizO results in favorable substantial further decrease ot the sodium error. Nago and KzO are in general undesirable in respect to low sodium error;

`That use yof CaO, Sr and BaO for one or m'ore constituents of the M0 component, is helpful when dealing with from about 22 to about 28 m'ol per cent LizO. `However, there is little choice, from the standpoint of their sodium error characteristics, whether SrO or BaO be used alone or in combination with the other sodium-error-reducing oxides, to Nit, Lino- Csa0-Lar0:-.Sl0z. Use of BeO and/or MgO is not desirable when low sodium error is in point;

l"That outstanding characteristics of those of my glass compositions comprising from about 50 to about 66 mol percent SiOa, plus about 6 to about 0.5 mol per cent of oiwgen modifier, reside not only' in their low sodium'error, but also in their outstanding performance when used ln test solutions at high temperatures, as exempliiled by stability or constancy of high performance in continuous operation for long periods in flowing water even at 90 C.;

And that with respect to my compositions including oxygen modifiers. the R20 component should constitute from about 22 to about 30 mol per cent of the entire composition; the OM component from about 6 to about 0.5 mol per cent of oxygen modifier; the XO: component should comprise from about 59 to about 86 mol percent, and the sum of the mol percentages of the 0M and X02 components should be about 'I0 moi per cent of the entire composition, the remainder of the composition consisting of the component M0.

In manufacture of my glasses carbonatos, nitrates, or other salts or compounds of the R metals. the 0M metals and of the M metals are mixed with a suitable type of sand, such as potters sand or high grade quartz sand, to provide the X0: component, and the mixture fused in a crucible of platinum to avoid contamination of the melt by impurities usually or frequently present in refractory crucibles. The proportions by weight of the various ingredients are chosen to afford the above specified molecular percentages of oxides of the R metal, M metal, the 0M metal and silicon. The mol percentages, for practical purposes of composing the mixtures for fusion, may be taken as the same as the desired moi percentages existing after fusion.

During aforesaid heating process, requiring il@ temperatures of about i700 F. to 2400 F., the carbonates or other salts or compounds are decomposed, carbon dioxide is evolved, and interaction oi' the resultant oxides with the sand occurs. Y

In working the molten glass, it is subjected to a considerably lower temperature, about 1200 to i300 F., lower than its melting temperature, depending upon the particular glass. In forming the glass into a pH responsive member. as'by blowing it into bulb form, the flame temperature should be maintained as low as possible and the oxygen content of the name should not be ailowed to become too high; otherwise there may be abnormal lossof lithium or other element of the glass with resultant impairment of its value for pH measurements.

Membranes which are not clear, due to devitrlcation or other cause, should be discarded as usually unsuited for use in pH measurements.

The completed electrode may be in the general shape of a test tube all whose glass is of the same composition; or the electrode may be a bulb blown from one oi the glasses above described upon a tubular stem of glass of different composition. To distinguish between my glass electrodes having different ilelds of use or difierent characteristics, there may be used stems of diil'erently colored glasses, each identifying a particular glass bulb or membrane composition.

The stem must have low surface conductivity (low electrical leakage) and high volumetric.

(electrical) resistance to avoid secondary errors of measurement. In addition, the glass of the stem must have about the same thermal coemcient of expansion as that of the Glass to whichl attached and, of great importance, the glass of the stem must not appreciably diffuse into the glass of the electrode proper when joined. By way of example, t-he glasses known as Corning #001 or 01 lead glass are suitable for the stems: on the other hand some glasses, including soda lime glass, have been unsuitable because they diffused into the electrode glass and gave rise to appreciably large sodium errors when the electrode assembly was used for measurements in strongly alkaline test solutions.

The magnitudes of resistance listed in the tables are for bulbs of approximately B to 9 mm. outside diameter, about .2 mm. thick, weight of bulb about .l5 gram, joined to a stem of approximately 10 mm. internal diameter. The volumeresistances were obtained in each case by measurement of the resistance between conductors in solutions respectively disposed on opposite sides of and engaging all of the membrane oi the glass electrode. The resistances of the membranes at C. are approximately live thousandths of the resistances at 25 C.

The herein indicated sodium errors, in terms of error in pH, were measured under conditions of 2N sodium solution of 12.8 to 13.0 pH, at 25 C.

The mol percentages herein specified are in each case those of an entire glass composition.

As exemplary of measuring systems with which the glass electrodes comprehended by my invention may be used to advantage, reference is made to United States Letters Patent Nos. 2,108,294, Doyle et al.; 2,285,482, Wunsch: and 2,367,746, Albert J. Williams, Jr.

What I claim is:

1. A {JH-responsive glass electrode of composition having from about 20 to about 30 mol per cent LirO, from about 0.5 to about 6 mol per cent ammo Iaas, rom about 59 to about 86 mol per cent Sith, and the remainder consisting o! one or more of the oxides CaO. BrO and BaO.

2. A pH-responsive glass electrode o! composition having about 28 mol per cent LizO, about 2 mol per cent CsaO, about 5 mol per cent BaO, about 63 mol percent SiOz, and about 2 mol per cent Lazos.

3. A p11-responsive glass electrode oi composition having about 25 mol per cent LinO, about 2 mol per cent CsaO, about 2 mol per cent CaO, about 5 mol per cent BaO, about 63 mol per cent B102, and about 3 mol per cent LaaOa.

4. A p11-responsive glass electrode oi composition -having from about to about 30 mol per cent LizO, 0 to about 3 mol percent CszO, about 59 to about 66 mol per cent SiO. about 8 to about 0.5 mol per cent LaaOa. and the remainder consisting oi the oxide BrO.

5. A glass electrode having a sodium error not exceeding 1 pH at 2N sodium at 12.6 pH at temperature C. to 35 C. whose voltage/pH characteristic is reproducible throughout the range 0 pI-I to 14 pH, and whose glass network structure is represented by the prefusion mixture consisting oi silica with LizO and at least one of the oxides oi `the group consisting oi CsO. RbaO. NaaO, KnO, and with at least one oi the oxides of the group consisting oi CaO, BaO. SrO, and high ratio or oxygen to silicon in said structure provided by inclusion in aforesaid mixture oi at least one o! the oxides of the group consisting of TeOn, P505, TlOz, ZrOz, AsaOs. CbzOt, Tano. ThOa. B20: and the rare earth metal oxides.

6. A glass electrode having a sodium error not exceeding 1 pH at 2N sodium at 12.6 pH at temperature 25 C. to 35 C., whose voltage/pH characteristics is reproducible throughout the range 0 pH to 14 pH. and oi' composition consisting oi. in mol percentages of the entire glass composition, from about 22 to about 28 per cent Liao with not exceeding about 3 percent additional of at least one of -g the oxides oi the group consbting of NaaO. CsaO. Bbz() and KzO. at least one oi.' the oxides of the group consisting of BaO, CaO, SrO totalling not more than from about 3 percent to about 9 percent. and Isilica from about 59 to about 86 percent with from about .5 percent to about l2 percentofatleastoneoxygenmodiiieroitho group consisting of TeOx. P10. TiOx. lr03. AlaOs. CbzQaTaeOaThOaBaOrandther-areearth metal oxides.

'1. A glass electrode having a sodium error not exceeding l pH at 2N sodium at 12.6 pH at temperature 25 C. to 35 C.. whose voltage/pH characteristic is reproducible throughout the range of 0 pH to 14 pH. and whose composition in mol percentages consists oi LizO lllrelathlg with at least one of the oxides oi the group consisting oi' C820, RbaO. NazO. KaO from about 25 percent to about 31 percent. from about 3 percent to about 9 percent of at least one of the oxides of the group consisting ot BaO. CaO, 8:0. a component of not more than about 6 percent of at least one ot the oxygen modiem of the group consisting oi TeOa. Pros, T101, 2x0. Assos. CbaOs. TazOs. T1101, B20: and the rare earth metal oxides. and a component o! silica not exceeding about 66 per cent.

8. A pli-responsive electrode o! glass having a sodium error not exceeding i pH at 2N sodium at 12.8 pI-I at temperature 25 C. to 35 C., whose voltage/pH characteristic is reproducible throughout the range 0 pH to 14 pH. and whose composition is represented by a prei'usion mixture consisting ai silica. lithia. at least one alkaline earth metal oxide, and at least one of the oiwgen modifiers of the group consisting of TeOs, P20, T101, ZrOa. AszOs, Cb10:.TaaOs, Thon. B10: and the rare earth metal oxides.

GEORGE A. PERLEY.

REFERENCES CITED The following references are of record in the ille of this patent:

UNITED STATES PATENTS Name Date Coleman Peb. 23, i943 Number Certiileate of Correction Patent No. 2,444,845.

July 6, 1948.

GEORGE A. PERLEY It is hereby certified that error appears in the printed specification of the above numbered patent re uiring correction as follows: Column 11, line 27, claim 5, for CSQ read 08,0; an that the said Letters Patent should be read with this correction thereln that the same may conform to the record of the oase in the Patent Office.

Signed and sealed this 16th day of November, A. D. 1948.

THOMAS F. MURPHY,

Assistant Commissioner of Patents.

ammo

Iaas, rom about 59 to about 86 mol per cent Sith, and the remainder consisting o! one or more of the oxides CaO. BrO and BaO.

GEORGE A. PERLEY.

July 6, 1948.

Signed and sealed this 16th day of November, A. D. 1948.

THOMAS F. MURPHY,

Assistant Commissioner of Patents.