A Guide for te Assessment and Mitigation of Bleed, Gloss Cange, and Mold in Inkjet Prints During Hig-umidity Conditions Jennifer Burger; University of Rocester and Daniel Burge; Image Permanence Institute, Rocester Institute of Tecnology; Rocester, NY, USA Abstract Te purpose of tis project was to define te absolute ceiling limits for time and relative umidity (RH) combinations at room temperature to prevent damage to inkjet printed materials in museums, libraries, and arcives wen tey are inadvertently exposed to sort-term ig-umidity conditions (under 8 ). Unintentional elevated umidity exposure can occur during HVAC malfunctions, transport, following water emergencies, and in uncontrolled storage or exibition areas. Previous researc as sown tat colorant bleed, gloss cange, and mold germination are te tree most common forms of inkjet deterioration during ig-umidity conditions. In order to provide collections care professionals wit te necessary information to mitigate all tree deterioration types, time limits for eac needed to be compiled into a single, concise guide. Data on ink bleed and mold germination limits were collected from previous researc, wile te gloss cange data required furter experimental investigation. Gloss cange experiments were performed wit dye on polymer-coated RC paper, as previous studies ave sown tis ink/paper combination to be particularly sensitive to gloss cange during exposure to elevated umidity. During te tests, samples were exposed to a series of time and RH variations. Te results sowed tat wile prints can be sensitive to gloss cange at elevated umidities, inkjet prints are even more sensitive to colorant bleed, wic is terefore te limiting factor. A guide for RH deterioration mitigation was developed and can now be used to predict ow prints ave or will respond to elevated umidity exposure for times less tan 8. ile all inkjet print types sould be safe at umidities at or below 65% for up to 8, relative umidity exposures above 8% sould be avoided at all costs as te most sensitive print types will likely be damaged witin ours. Te guide provides predictive times to damage for RH values between 65% and 8% tat can be interpolated to determine risk at tese intermediate conditions. Introduction Te purpose of tis project was to define te absolute ceiling limits for time and relative umidity (RH) combinations to prevent noticeable bleed, gloss cange, and mold germination in inkjet printed materials in collections wen tey are inadvertently exposed to sort-term ig-umidity conditions (under 8 ). Unintentional elevated umidity exposure can occur during HVAC malfunctions, transport, following water emergencies, and in uncontrolled storage or exibition areas. Previous researc as sown tat colorant bleed, gloss cange, and mold germination are te tree most common forms of inkjet deterioration during igumidity conditions []. In order to provide collections care professionals wit te necessary information to mitigate all tree deterioration types, time limits for eac needed to be compiled into a concise cart. Te cart is intended to assist institutions during, or in preparation for potential, adverse circumstances. Results may also benefit manufacturers of inkjet materials as well as artists and potograpers wo use tis medium. Inkjet printed materials ave included a wide variety of ink, coating, and support combinations since teir commercial popularization in te 98s. Tese variations in materiality ave a significant effect on ow inkjet prints will respond during exposure to adverse environmental conditions, wit certain combinations experiencing more severe degradation tan oters. A majority of inkjet printed materials in collections are not identified at te level necessary to know wic prints are most at risk for disfiguring decay. For tis reason, collection care protocols may need to be built around te most sensitive inkjet print type per deterioration force, in tis case ig umidity. Tis approac ensures tat all inkjet printed materials in collections, weter identified or not, would be protected during ig-umidity exposure. Materials and Metods Constructing te ig-umidity deterioration mitigation cart involved a four-step approac.. Coosing te parameters for temperatures, relative umidities, and times.. Collecting te previously defined limits for ink bleed and mold germination. 3. Determining te gloss cange limits troug experimental investigation.. Creating te final guide to prevent damage to inkjet collections during sort-term ig-umidity exposure. Cart Parameters Te cart parameters were determined, in part, from Salesin s previous bleed limit researc []. Tis investigation looked at colorant migration in inkjet printed materials at temperatures between 5 o and 35 o C, umidities between 6% and 9% RH, and times from to 8. Te relative umidity and time parameters remained te same for tis new guide, wile temperature was fixed at 5 o C. Tis temperature was cosen because prior bleed experiments sowed 5 o C produced te greatest bleed []. Table : Cart layout wit previously defined bleed and mold germination limits 5 o C 9% RH 8% RH % RH 6% RH Bleed no risk Gloss Cange Mold 3 no risk no risk
Table sows te initial limits collected from Salesin s bleed report []. Mold germination limits were determined using IPI s Dew Point Calculator, wic as an incorporated Mold Risk Factor predictor [3]. Te gloss cange limits would be filled in by te following experiments. Gloss Cange Metodology Te autors know of no existing standard or generally accepted metod for evaluating gloss cange in inkjet printed materials. ile ISO 89 (Imaging materials Processed silver-gelatin type black-and-wite films Specifications for stability) provides a test procedure for evaluating gloss cange in traditional potograpic negatives, motions picture films, etc. in direct contact wit eac oter, it does not take inkjet materiality into consideration nor contact wit enclosures []. It also uses te extremely ig temperature of o C. Te gloss cange metodology used in tis project was instead based on a previous study performed at IPI []. Tis study examined gloss cange in various inkjet printed materials tat were stacked wit a variety of interleaving, weigted, and incubated at te single, extreme condition 3 o C and 9% RH for seven. Results from tis study primarily informed sample selection and preparation for tis new project, but could not be used to fill in te above cart as tey used only one temperature, umidity, and time condition. Generally, gloss cange is caused by eiter abrasion or a combination of ig umidity, temperature, pressure, and time. Because te abrasion-induced gloss cange is a function of print type, abrader, and weigt, it is not considered in tis project, wic is solely focused on ig-umidity issues. Gloss cange by ig umidity, temperature, pressure, and time is often referred to as ferrotyping because te same forces are used in te latter. Te primary difference is tat ferrotyping is an intentional process and affects te entire print surface. Te gloss canges being considered in tis project are unintentional, non-uniform, and disfiguring forms of damage. Note tat tese gloss cange experiments are not accelerated aging tests, but are instead meant to replicate real world, worst-case scenarios, were inkjet printed materials may experience sort-term exposure to elevated umidity. Selection and Preparation In order to produce a guide tat would be useful for all collections of inkjet prints, researc focused on te print type most sensitive to tese forms of damage from te entire istory of inkjet printing and did not target current tecnologies and materials, wic may be significantly more stable tan tose previously available. Testing te most sensitive materials resulted in te production of conservative data necessary for mitigating inkjet print deterioration during tese conditions. Tis sensitive material is polymer-coated RC prints []. Figure : Target used during all gloss cange experiments. A printed target (Figure ) was cosen over unprinted samples in order to provide better visual comparison of canges in gloss as well as to insure tat te area most likely to experience gloss cange (printed or unprinted) was included. It consisted of black (Dmax), mid-tone gray, and wite (Dmin) areas. Inkjet samples were printed in-ouse following recommended printer settings and used CMYK inks for te two printed areas. samples were obtained at a local potograp laboratory. s were cut into x centimeter strips and labeled verso wit permanent marker. All samples were created in duplicate, including controls, and were preconditioned to o C and 5% RH for one week prior to experiment initiation. Initial gloss readings were recorded separately for black, gray, and wite areas for eac sample wit a Gardner micro-tri-gloss meter. Because bot polymer-coated RC and cromogenic samples measured as glossy before and after incubation, all targets were evaluated using a o angle of incidence ligt as per ISO 83- (Paints and varnises Determination of gloss value at degrees, 6 degrees, and 85 degrees) [8]. A measuring template was created to ensure consistency in recording. Polyester film was used as te surface in contact wit te prints. It was selected for two reasons. First, it was sown to induce gloss cange in previous work, and second, it is te recommended enclosure material for use in direct contact wit inkjet prints [, ]. Polyester is also known to block to cromogenic prints at te conditions used in tis experiment, so it acted as a positive control. Figure : Jig stacking order Table : Prints selected for investigation Paper Type Colorant Polymer RC- Dye Polymer RC- Dye 3 Dye Table lists te test samples. Two different polymer-coated RC paper brands were tested wit eac being printed on a different printer. Dye printers were used as tese were most often used wit tis paper type. prints ave been sown to undergo gloss cange during ig-umidity exposure and were meant to serve as a reference point for comparison. s were stacked in a steel specimen jig as sown in Figure. A strip of polyester was placed on top of eac print. Glass slides sandwiced te paper and interleaving stack. A weigt of 8gr/cm (.6 kpa) was placed on te stack to represent te average load experienced by potograps in an album at te bottom of a stack of albums []. All samples were tested in duplicate in separate jigs. 3
Experiment Conditions Test conditions were cosen to provide gloss cange data for te final cart and to reflect real-world reactions to sort-term ig-umidity exposures. Table 3: Experiment conditions Temperature Relative Humidity 5 o C 9% 5 o C 8% 5 o C % 5 o C 6% Table 3 lists te four test conditions. Te first set of samples were incubated at 9% RH, if any of te samples sowed noticeable gloss cange a new set of samples were incubated at 8% RH, and so on for % and 6% RH until no canges were seen for a particular umidity level. Eac ran for 8 wit duplicate jigs being removed after,,,,,, and 8. Jigs were placed in te center of an ESPEC LHL- Humidity Cabinet on a wire rack wit adequate room between jigs for air circulation. Duplicate jigs were removed at te end of teir incubation period and moved to a o C, 5% RH room for ours before disassembly and assessment. Unincubated control samples were stacked in te same manner as te test samples and kept at o C and 5% RH for te extent of te experiment. A second set of untested controls were created and left un-stacked in a o C/5% RH environment to provide a baseline to measure visual responses against. Results Visual Assessments In order to reduce variability, one person made te visual assessments. Tables, 5, and 6 list te visual assessments. Te rating scale indicates noticeable gloss cange wit Y and no noticeable gloss cange wit N. If eiter duplicate sample sowed noticeable gloss cange te incubation time was assigned a Y in order to provide te most conservative assessment. Similarly, if eiter brand of polymer-coated RC underwent noticeable gloss cange, te incubation time for polymer-coated RC was assigned a Y. Noticeable gloss cange was selected for te visual assessment, as it is a conservative approac and is more consistent and definable tan objectionable gloss cange. ile measureable gloss cange would ave been te most conservative parameter, it does not necessarily reflect a cange tat can be seen wit te naked eye and terefore is not te concerning condition. Furtermore, as will be discussed, quantitative gloss meter analysis did not provide consistent data wit wic to draw measureable gloss cange conclusions. Table : Noticeable gloss cange at 5 o C and 9% RH Polymer RC Y Y Y Y Y Y Y Y Y Y Y Y Y Y 8 Table lists noticeable gloss cange after incubation at 5 o C and 9% RH. All samples at tis condition sowed noticeable gloss cange after of incubation. Te degree of gloss cange was severe. Gloss cange was also irregular and blotcy and was only seen in te black (Dmax) areas. Te entirety of te black area experienced tis gloss cange. No noticeable gloss cange was seen in eiter te mid-tone gray or wite areas, up to and including 8 of incubation. Of additional note was a milkiness seen in te black areas of te polymer-coated RC samples after only of exposure at 9% RH. Tis milkiness increased in intensity as exposure time continued. It was most extreme in sections of te black areas tat did not appear to stick to te polyester. Areas tat experienced temporary sticking sowed an increased glossiness but no milkiness. Te milkiness was not seen in te gray or wite areas nor was it seen in any of te cromogenic samples. Tis effect was unexpected, but since tis was a real-time test, and not an accelerated test, te effect must be considered relevant and important and is terefore included in determining te limiting time to damage at te various test conditions. Table 5: Noticeable gloss cange at 5 o C and 8% RH Polymer RC Y Y Y Y Y Y Y N N N N N N N 8 Table 5 lists noticeable gloss cange after incubation at 5 o C and 8% RH. Polymer-coated RC samples underwent noticeable gloss cange after of incubation at tese conditions. Gloss cange in samples incubated at 8% RH was less severe tan samples incubated at 9% RH. ile gloss cange was also irregular and blotcy at 8% RH, it did not cover te entirety of te black area as it ad at 9% RH. Instead, gloss cange was found only along te edges of te samples. Te milkiness was also noticeable in te polymer-coated RC samples after of incubation at 8% RH. ile te degree of milkiness did increase wit time, it was not as noticeable as in samples exposed to 9% RH. Once again, milkiness was not seen in te gray or wite areas nor was it seen in te cromogenic samples. However, even toug te gloss cange was not seen until, te limiting time to damage is based on te to milkiness. Table 6: Noticeable gloss cange at 5 o C and % RH Polymer RC N N N N N N N N N N N N N N 8 Table 6 sows te visual assessments after incubation at 5 o C and % RH. No noticeable gloss cange was seen in eiter te polymer-coated RC or cromogenic samples up to and including tose incubated for 8. Tere was also no milkiness seen in any of te samples exposed to % RH. Because te % RH test sowed no canges to te samples, te 6% RH tests were dropped and RH conditions below % are considered safe for up to te maximum 8 period of te project. ile cromogenic samples were expected to undergo canges in gloss more readily and severely tan polymer-coated RC prints, results sow tat tey fair far better under all test conditions. Tis is an unexpected result and cannot be explained. In addition, and contrary to prior reports [], colorant density does appear to increase a tecnology s propensity to undergo surface modification. ile te wite and gray areas did not experience
gloss cange under any of te tested conditions, te black areas, were ink density is greatest, experienced severe gloss cange and milkiness. Colorant density sould be taken into consideration for future surface modification testing on inkjet printed materials. Gloss Meter Data Gloss unit measurements were taken before and after individual incubation periods in opes of providing a quantitative account of canges during ig-umidity exposure. Tables, 8 and 9 list te initial ( ) and final (-8 ) gloss units for incubated samples. ile gloss meter data for te cromogenic samples did follow a clear trend, te inkjet samples did not. At te 9% RH condition, te milkiness of te samples may ave ad a confounding effect on te gloss readings. For te 8% RH samples, te gloss canges were small and only along te edge of te samples. Tis made te gloss cange immeasurable by te instrument. Bot of tese issues made te gloss measurements too unreliable to draw conclusions. Te visual assessments ultimately provided a muc better assessment of te canges to te samples. Table : Gloss unit measurements at 9% RH Polymer RC- Polymer RC- Bl 6 5 53 56 55 8 6 Gr 35 36 36 36 3 3 8 58 5 5 8 5 9 9 5 Bl 6 5 6 6 6 5 8 35 Gr 6 58 5 9 9 5 6 63 58 59 58 5 5 5 Bl 6 66 56 6 6 55 55 Gr 8 6 53 58 6 55 9 9 68 65 63 58 6 6 6 Table 8: Gloss unit measurements at 8% RH Polymer RC- Polymer RC- Bl 6 5 9 8 5 8 9 Gr 38 35 3 39 36 36 3 58 53 5 5 5 53 53 5 Bl 6 6 5 5 5 53 55 56 Gr 6 6 6 59 58 5 59 58 6 6 6 63 6 58 63 6 Bl 69 6 65 66 66 6 6 Gr 8 65 6 6 6 66 68 9 6 6 66 66 6 66 8 8 Table 9: Gloss unit measurements at % RH Polymer RC- Polymer RC- Bl 6 59 59 58 5 5 58 55 Gr 38 39 38 39 3 58 55 55 5 53 5 55 5 Bl 6 6 66 65 6 6 6 63 Gr 6 6 65 6 6 6 6 6 6 66 6 6 65 6 3 6 Bl 5 5 Gr 8 5 5 3 3 3 9 3 6 3 3 3 Guide for RH deterioration mitigation Table : Inkjet print deterioration risk during sort-term exposure to elevated umidity 5 o C 8 9% RH 8% RH % RH 6% RH Bleed no risk Gloss Cange/ Milkiness no risk no risk Mold 3 no risk no risk Table combines data from te gloss cange experiments wit previous data collected on bleed and mold germination. Te cart sows ow long it will take te most susceptible inkjet printed materials to undergo a specific deterioration type during elevated umidity. Tis miner s canary approac provides te most conservative parameters to prevent disfiguring damage witin inkjet collections during sort-term adverse environmental conditions. Users can use te table to estimate weter teir prints ave been or will be put at risk. For example if a collection of inkjet prints as been exposed to 8% RH for tey would need to be immediately removed from te adverse environment and ten inspected for signs of colorant bleed, milkiness, or gloss cange. In real life situations, owever, umidities will rarely be exactly at one of te values listed in te cart above and interpolations between te RH values and corresponding times to failure may be needed. To interpolate between umidity values in te table, suc as for a % RH, select te next iger RH value in te cart to guide te decision making process. In tis example, te time to damage at 8% sould be assumed, as it will be te most conservative approac. However, because te differences between times to damage, or even no risk, can be extreme between two RH values in te cart, some estimating between values may be useful, especially in times of emergency, wen a wide variety of response and recovery activities need to be managed and prioritized. Since te limiting factor below % is ink bleed and te Salesin reported tat bleed did not occur below 65% RH, ten tat value may be used as te safe RH limit instead of 6% RH. 5
Conclusions ile prints can be sensitive to gloss cange at elevated umidities, inkjet prints are even more sensitive to colorant bleed, wic is terefore te limiting factor. Tese results are based on te most sensitive print types, oter print types, suc as pigment inks on matte surface paper, may be muc more resistant to bleed and/or gloss cange. Te guide for RH deterioration mitigation can be used to predict ow prints ave or will respond to sort-term elevated umidities less tan 8 Relative umidity exposures above 8% sould be avoided at all costs as te most sensitive print types will likely be damaged witin ours All inkjet print types sould be safe at umidities at or below 65% for up to 8 If materials are inadvertently exposed to ig umidity conditions, tey sould be removed to a safe environment as soon as possible and immediately assessed for bleed and gloss cange References [] D. Burge, IPI Guide to Preservation of Digitally-Printed Potograps, Rocester NY: Image Permanence Institute, 6. [] E. Salesin and D. Burge, Te Determination of Humidity Limits to Prevent Colorant Bleed in Inkjet Prints, NIP and Digital Fabrication Conference, vol. 3, pp. 9-83,. [3] Dew Point Calculator, ttp://www.dpcalc.org. [Accessed: -Jul- 6]. [] L. Rima and D. Burge, Tendency of Digitally Printed Materials to Ferrotype or Block, NIP and Digital Fabrication Conference, vol.5, pp. -5, 9. [5] D. Burge, A. Venosa, E. Salesin, P. Adelstein, and J. Reilly, Beyond Ligtfastness: Some Neglected Issues in Permanence of Digital Hardcopy, Proc. Intl. Symposium on Tecnologies for Digital Poto Fulfillment, vol., pp. 6-6,. [6] A. Robb, Te effect of relative umidity on ink jet prints, Preservation and Conservation Issues Related to Digital Printing, Conference Proceedings, Ruterford Conference Center, London, October 6-. London: Institute of Pysics, pp. 3-,. [] ISO 89 Imaging materials Processed silver-gelatin type blackand-wite films Specifications for stability. [8] ISO 83- Paints and varnises Determination of gloss value at degrees, 6 degrees and 85 degrees. [9] M. McCormick and H. ilelm, Te Influence of Relative Humidity on Sort-Term Color Drift in Inkjet Prints, Jour. Imaging Sci. and Tecnol., vol. 39, no., pp. 9-85,. Autor Biograpy Jennifer Burger received er M.A. degree in Potograpic Preservation and Collections Management from te University of Rocester and George Eastman Museum in 6. Se as been a researc assistant at te Image Permanence Institute for te past year. Her current researc interests include traditional and digital print identification, preservation, and conservation. Correspondence pertaining to tis paper sould be emailed to jennifer.d.burger@gmail.com