Application of Leo G. Glasser, Robert J. Kanzler and Daniel J. Troy, 363 F.2d 449 (C.C.P.A. 1966)

Appellants here seek reversal of the decision of the Board of Appeals which affirmed the examiner's rejection of claims 5 through 9, the only remaining claims, of their application,¹  entitled "Method and Apparatus for Photometric Analysis," as unpatentable over the prior art.

The application relates to a photometric apparatus and method for such purposes as "the gaging of a transparent or translucent film, the measurement of the fluorescence of solids suspended in liquids, and the determination of the percentage concentration of a particular substance in a mixture with others without regard to whether the materials are in gaseous, liquid or solid phase." It discloses apparatus wherein a single beam of light including selected wave lengths is directed through a sample to a semitransparent mirror, splitting the transmitted radiation into two separate beams which are directed, through different filters having preselected filtering characteristics, to two separate phototubes serving as detectors. The outputs of the two detectors, described as varying exponentially, are impressed on separate channels of a dual-channel logarithmic amplifier forming part of a circuit designed to extract the logarithms of the output and perform a subtraction operation thereon, "thus effectively performing a ratio measurement." The output of the amplifier, which is said to vary linearly with concentration or thickness, is applied to an indicator or controller. It is further stated that some analytical determinations are such that the intensities of the radiations transmitted in the beam are linear functions and that a direct measurement of the linear ratio is made in those cases instead of measuring the logarithm of the ratio.

The application refers to photometric analysis as involving two variables. The first is described as the subject of analytical interest, while the second "is considered as referring to interference generally, which may be optically selective or non-selective, or to environmental changes having an effect on the radiation from the sample, such as random variations in the intensity of the light source, fluctuations in pressure or temperature of the sample, or the like." One of the two beams split from the beam transmitted through the sample, designated the "analytical" beam, is allocated primarily to the evaluation of the first variable and the other, or "compensation" beam, primarily to compensation for the effect of the second variable. The application further states:

* * * The transmitted radiation passed in each of the two beams is regulated by the interposition of appropriate filters in each of the beams preselected so that the minimum change in the first variable which it is desired to detect exerts a greater effect on the ratio of the intensity of radiation passed in a given one of the beams to the intensity of radiation passed in the other beam than is exerted by the anticipated maximum change in the second variable on this ratio within the range of the analysis. * *

The regulation effected by the filters in the analytical and compensation beams is such that the change in the ratio of the intensities of the radiation transmitted in the two beams is uniquely a measure of the sample characteristic under evaluation, independent of any non-selective interference or environmental changes or any anticipated selective changes which alter the analytical radiation from the sample. This results from the fact that the wave lengths of the beams are selected with the aid of absorption spectral data or routine empirical measurement so that the first variable affects the intensity of the analytical beam in a considerably greater proportion than it affects the intensity of the compensation beam, while at the same time the anticipated selectively-absorbing second variable affects the intensity of both beams in the same proportion within the sensitivity desired for the analysis of the first variable. It will be clear that non-selective interference by its very definition affects both beams in the same proportion and is therefore automatically canceled from the ratio, so that, after suitable calibration, the analysis is solely that of the isolated factor of interest.

As one example, the application describes use of the invention to measure the concentration of chlorine in dry air with a span of 0-1% chlorine at an accuracy corresponding to a maximum error of 0.01% chlorine in the presence of a concentration of nitrogen dioxide varying between 0 and 0.05%. It states that two wave lengths for the beam impressed on the sample are chosen "to give an appreciable difference in absorptivity towards chlorine and a difference in absorptivity towards nitrogen dioxide so small that the extreme variation in nitrogen dioxide content affects the absorbance less than the absorbance change equivalent to the desired sensitivity in terms of chlorine." The necessary conditions are said to be met by use of wave lengths of 365 mu. and 436 mu., to which the absorptivities of chlorine and nitrogen in liters/mol. — cm. are:

 365 mu. 436 mu. Chlorine 25.17 1.636 Nitrogen dioxide 318 327

One of the split beams is impressed on a detector through a 365 mu. filter, the other on the remaining detector through a 436 mu. filter.

Claims 5 and 7 are representative:

5. An analyzer for the photometric analysis of a sample on the basis of radiation transmitted from said sample wherein intensities of the several wave lengths in said radiation are singlevalued functions of a first variable and a second variable characteristic of said sample comprising in combination beam-splitting means adapted to receive and divide radiation from a specific region of said sample into two separate beams, one of which beams is passed through a first filter to a first radiation detector and the other of which beams is passed through a second filter to a second radiation detector, both detectors being disposed with respect to said beam-splitting means so that each said radiation detector views the identical area of said specific region, said first filter and said second filter being preselected to pass narrow bands of wave lengths in each of said beams so that the minimum change in said first variable desired to be detected exerts a greater effect on the ratio of the intensity of radiation passed in said first beam to the intensity of radiation passed in said second beam than is exerted by the anticipated maximum change in said second variable within the range of analysis, and a ratio-measuring electrical circuit connected to the output sides of said radiation detectors.

7. A method of photometric analysis of a radiation-absorbing sample wherein intensities of the several wave lengths in the radiation transmitted from the sample are single-value functions of a first variable and a second variable characteristic of said sample comprising in sequence dividing radiation transmitted from said sample into a first beam of radiation and a second beam of radiation, filtering each said beam of radiation to transmit a different band of wave lengths for each said beam chosen so that the minimum change in said first variable desired to be detected exerts a greater effect on the ratio of the intensity of radiation transmitted in said first beam to the intensity of radiation transmitted in said second beam than is exerted by the maximum change in said second variable within the range of analysis, and electrically measuring the ratio of the intensity of the radiation transmitted in said first beam to the intensity of the radiation transmitted in said second beam together with a measurement of said first variable as a function of said ratio.

The issue is patentability over the prior art, the art of record being:

 Meyer 2,503,165 April 4, 1950 Wright et al. 2,737,591 March 6, 1956 Bliss 2,823,800 Feb. 18, 1958 Ator et al. 2,987,182 June 6, 1961 Fastie et al., Journal of the Optical Society of America, Vol. 37, No. 10, October 1947, pp. 763-768

The Wright et al. patent (Wright) relates to photometric analyzers for organic mixtures wherein a single beam of light is directed through the mixture in a sample cell and split into two separate beams upon emerging from the cell. The two resulting beams are then passed through separate filters to separate detectors which may be bolometers disposed in separate arms of an electrical bridge circuit. A potentiometer-type resistor interconnects the two bolometers with the adjustable contact on the resistor connected to one terminal of indicating means and normally so dividing the resistance between the two bolometer arms as to provide a balanced bridge. Variation in the substance in the sample cell which is being analyzed will cause unbalancing of the bridge and means are provided for automatically moving the adjustable contact to rebalance the bridge and recording the amount of adjustment required as a measure of the concentration of the substance. As an example, Wright discloses use of the apparatus to analyze a mixture of 1-butylene, 2-butylene, and butadiene. On the basis of a wave length of 6.25 mu. being strongly absorbed by butadiene but not by 1-butylene and 2-butylene, one of the split beams is passed through a filter which will prevent passage of the 6.25 mu. wave length, the other through a filter which will pass most of the 6.25 mu. wave length but is strongly absorptive of other wave lengths.

Meyer relates to photometric apparatus for absorption and emission analysis wherein a beam of radiation is split into two beams having the same characteristics. One is passed through a liquid containing a substance to be analyzed which absorbs part of the radiation, while the other is passed through the same liquid without the substance, the latter serving as a standard. The transmitted beams are alternately applied to a detector connected to circuitry producing a voltage "proportional to the logarithm of the ratio of the radiation intensities," which is a measure of the concentration of the substance to be analyzed.

The Fastie et al. publication (Fastie) discloses measurement of carbon dioxide through infrared radiation directed through a sample cell into two detectors. One detector differs from the other in that its input is filtered so that it does not receive any energy which carbon dioxide absorbs while both detectors are sensitive to other absorbing gases in the cell. The outputs of the two detectors are connected in opposition to an indicating device.

For reasons given later, the subject matter of Bliss and Ator et al. (Ator) need not be discussed.

The examiner rejected the appealed claims as unpatentable over Wright in view of any of Fastie, Bliss or Ator, and, optionally, over the same references further in view of Meyer. He held Bliss and Ator available as references because he considered an affidavit submitted to antedate them under Rule 131 inadequate.

In its original decision, the board held claims 5, 7 and 8 unpatentable over Wright alone, and claims 6 and 9 unpatentable over Wright in view of Meyer. It stated that Bliss and Ator were "cumulative, at best" and did not consider them further, thus not ruling on the sufficiency of the Rule 131 affidavit. The board did not mention Fastie. On reconsideration, the board modified its position as to the showing in Wright, but otherwise adhered to its original decision.

It appears uncontroverted that Wright discloses that part of claim 5 which precedes the recitation of the characteristics of "said first filter and said second filter * * *." However, that recitation and the following final recitation of a "ratio measuring electrical circuit * * *" are both in controversy.

Taking the last recitation first, as did the board the examiner conceded that Wright does not teach a "ratio-measuring circuit" as recited. Although the board originally stated that "there is no merit in appellants' argument that Wright et al. measures a difference and not a ratio," on reconsideration it stated:

Appellants are correct in stating that with bolometers as the detecting devices the ratio of intensities would not be obtained, except perhaps in a very limited range where the outputs of the two bolometers were about the same. Appellants lay down as a criterion that the detectors must have a linear current output with respect to radiation intensity.

The board also noted that Wright states "that thermopiles and other detectors may be used which produce an electric current in accordance with, or proportionate to, the absorbed light energy." However, the absence of any specific disclosure in Wright of circuits wherein those elements are used in a manner which measures the ratio of the intensities of the two beams leaves us unsatisfied that the reference suggests that a "ratio-measuring circuit" as required by the claims be used with the "thermopiles and other detectors."

The examiner conceded that Wright does not disclose measuring the required ratio.²  However, he noted that appellants admit that their logarithmic type measuring circuit per se "is old" and concluded that "only skill in the art would be involved in using a ratio measuring circuit in the Wright et al. device, if desired, since no unexpected results would follow." He further regarded Meyer's use of a ratio measuring electric circuit in photometric apparatus would make it obvious to use a ratio measuring circuit in Wright.

We do not think appellants' concession that their logarithmic type measuring circuit per se is old establishes that it would be obvious to substitute it for the measuring circuit of Wright, since the latter clearly measures and indicates a different function of the beam intensities than the logarithm of their ratio, and the examiner and board have not pointed out a reason why one skilled in the art would switch to determination of the latter function in Wright.

Although Meyer determines the logarithm of the ratio of the intensities of two beams in photometric apparatus, the comparison he makes is between one beam which passes through the material under investigation and another which passes through a reference material only, rather than between beams which have both passed through the test material as in Wright. The view, advanced by the examiner, that it would be obvious to use a logarithmic measuring circuit in Wright because Meyer shows such circuit to be "conventional" in the photometric analyzing art, is not convincing in view of that difference in the beams utilized. Meyer does point out that the exponential relationship involved in the absorption of radiation under the laws of Beer and Lambert results in the logarithmic function he measures being proportional to the concentration of the substance under investigation in his system, and the solicitor refers to the same laws in connection with a general discussion of the operation of the present device. However, no clear connection is established by the arguments here between those laws and the question of obviousness of modifying Wright to measure the ratio of radiation intensities. The point was not developed in the Patent Office prosecution of the case.

Moreover, the controlling issue here lies not only in the use of a ratio-measuring circuit but in whether it would be obvious to use such a circuit in combination with filters selected so that the minimum change in the variable desired to be detected provides a greater change in the ratio of radiation intensities than does the anticipated maximum change in the second variable. Concerning that limitation, the board stated:

* * * it is inherently disclosed in the Wright et al. patent as a result of the filter characteristics set forth in column 5, lines 57-63.

The portion of Wright which the board thought inherently disclosed the recited filter characteristics reads:

When the detectors 11 and 12 are bolometers, it is necessary that one of the light-filters 6 and 7 absorb a large proportion of light of the distinctive band of wave lengths absorbed by the component to be determined in the analysis and that it transmit light of other wave lengths.

Without further explanation, which the board did not provide, it is not apparent to us why that statement is thought to disclose that the two filters are "preselected to pass narrow bands of wave lengths in each of said beams so that the minimum change in said first variable desired to be detected exerts a greater effect on the ratio of the intensity of radiation passed in said first beam to the intensity of radiation passed in said second beam than is exerted by the anticipated maximum change in said second variable within the range of analysis."

The board also stated on reconsideration:

We find no error in our position relative to limitation (1) [the recitation of the filter characteristics], as stated in our decision. Furthermore, since the system of Wright et al. compensates for dirt on the windows of the sample cell and for other variations, such as variation of the light intensity of the source, which affects both beams equally, their system is such that the relation of the beam intensities, as set forth in the claims, is met, i. e., dirt on the windows of the sample cell, or variation in the intensity of the source light, will affect both beams to the same extent just as in appellants' device.

Of course, Wright, like appellants, does split a single beam after it passes through the material under analysis to cause both beams to be affected similarly by factors such as opaque material on windows of the cell and variation in intensity of the source of incident radiation. However, variations of that type, such as a reduction in intensities of the two beams to some proportion of their normal values, will change to different degrees the indicated output quantity of a measuring circuit which provides essentially a difference measurement while leaving a circuit which measures the ratio of intensities theoretically unaffected. Also, the recitation relative to the first and second filters, as in claim 5, leaves the clear implication that the result of having the minimum change in the first variable exert a greater effect on the ratio of the intensities of the two beams than the maximum change in the second variable is caused by the frequency characteristics of the filters. The correction which Wright obtains for factors which are nonselective as to wave length appears to arise from his splitting a single analytical beam in two and not from the characteristics of his filters.

As noted above, Fastie, relied on by the examiner but not mentioned by the board, connects the detectors for the two beams in opposition and thus does not measure the ratio of their intensities. It thus does not overcome the deficiency of Wright in that respect. In addition, we are not satisfied that Fastie would make it obvious to preselect filters having the characteristics called for in claim 5. The reference states that only one of the two detectors is responsive to the gas being measured while both respond equally to other gases. However, it does not deal with the problem of compensating for a second variable which is also absorptive of radiation of the same wave length as the measured gas.

We appreciate that the failure of a reference to disclose the recited relationship of the filters in the same words used by appellants would not be fatal to the rejection but, for the reasons set out, we are not satisfied on this record that the relationship required by claim 5 is taught or made obvious by the references relied on in the rejections we have specifically discussed. Since the same limitations are carried in the other appealed claims, we are obliged to reverse the rejection as to all claims.

The solicitor states in his brief that Bliss and Ator "needn't be considered." Under the particular circumstances, we agree with that concession. In both dismissing those patents as "cumulative, at best" and failing to rule on a contested holding by the examiner that appellants' affidavit under Rule 131 is insufficient to overcome the patent filing dates, the board has left us without any indication whether it considers the claims properly rejected on prior art which includes those patents.

The decision is reversed.

Reversed.